MUTRFC2010 Full Version Proceeding

Green Transportation for
Future Generation
i
ii
Green Transportation for Future
Generation
Proceeding of Malaysian Universities Transportation Research Forum and
Conference 2010
Edited By
Engr. Mohd Sufian Abdul Karim

(Director of Alumni Relations Center, Universiti Tenaga Nasional)
Engr. Shuhairy Norhisham

(Lecturer, Civil Engineering Department, Universiti Tenaga Nasional)
Engr. Herda Yati Katman

(Lecturer, Civil Engineering Department, Universiti Tenaga Nasional)
Assisted By
Engr. Mohd Amiruddin Bin Zainol

(Research Engineer, Universiti Tenaga Nasional)
iii
Universiti Tenaga Nasional, 2010
All right reserved. No part of this publication may be reproduce, stored in a retrieval system, or
transmitted in any form or by any means electronic, mechanical, photocopying, recording or otherwise;
without permission in writing from the Co -Chairman of MUTRFC 2010.
Perpustakaan Negara Malaysia : Cataloguing-in-Publication Data
Malaysian Universities Transportation Research Forum and Conference (2010: Kajang, Selangor)
Green Transportation for Future Generation: Proceeding of the International Conference and
Building Technology / Editors Mohd Sufian Abdul Karim, Shuhairy Norhisham, Herda Yati
Katman.
ISBN 978-967-5770-08-1
1. Transportation--congress, 2.Highway--congress
I. Mohd Sufian Abdul Karim. II. Shuhairy Norhisham. III. Herda Yati Katman
This book is published on the understanding that the authors are solely responsible for the statements
made and opinions expressed in it and that its publication does not necessarily imply that such statements
and/or opinions are or reflect the views or opinions of the publishers. While every efforts has been made
to ensure that the statements made and opinions expressed in these publications provide a safe and
accurate guide, no liability or responsibility can be accepted in the respect by the authors or publishers.
Cover Design : Fathoni Usman

Typeface : Times
Type size : 10-12
Printed and bound in Malaysia by Universiti Tenaga Nasional

Universiti Tenaga Nasional, Jalan IKRAM-UNITEN, 43000 Kajang, Selangor, Malaysia
Tel: +603- 8921 2020 Fax: +603-8921 2116 e-mail:shuhairy@uniten.edu.my
iv
PREFACE
Malaysian Universities Transportation Research Forum and Conference (MUTRFC) is jointly
organized by the Department of Civil Engineering, UNITEN, in collaboration with the Transportation
Science Society of Malaysia (TSSM) and supported by the Institution of Engineers, Malaysia (IEM)
Transportation and Highway Technical Division. This conference provides an interdisciplinary forum for
the presentation of the current and latest in research, development and commercialization of any
transportation products including consultation activities.
Objectives:
1) To give a platform to academician, researcher and practicing engineers to present research works in
transportation engineering.
2) To gather all researchers in transportation engineering and related fields in one fraternity to present
their research outcomes.
3) To promote a knowledge and research environment in transportation engineering
The topics of interest include, but are not limited to Transportation energy issues, Green Transportation,
Intelligent Transport Systems, Road and Rail Transport Modes, Public Transportation System,
Motorcycles and Non-motorized Transport, Maritime Transport and Ports , Airports and Aviation, Low
Cost Carrier (LCC), Transportation Demand and Modelling, Traffic Control System and Policy,
Transport Demand Management, Transportation Information System and GIS, Physical Distribution and
Logistics, Transport and the Environment, Traffic Accident and Road Safety, Transportation land use and
Spatial Analysis, Privatization and Financing of Transport, Issues in Transport Economics, Roads and
Highway, Road Infrastructure and Maintenance, Highway Planning, Pavement Engineering and Materials
and other related topics.
The keynote speakers for the forum were delivered by prominent scholars and professionals from well-
established organizations and institutions.
The response from scientists, researchers, academics, professionals and practising engineers was
overwhelming. As spirits of gathering of transportation related fields in Malaysia, 50 papers had been
accepted for this conference after a series of reviewing by the reviewer.
Universiti Tenaga Nasional Mohd Sufian Abdul Karim

Putrajaya Campus Co-Chairman
21 December 2010 MUTRFC 2010
v
CONFERENCE PROGRAM
Date : 21 December 2010 (Tuesday)
Time : 8:00am 6:30pm Venue
Venue : College of Information Technology, Universiti Tenaga Nasional
21 December 2010
8.00am 8.30am Registration
BW-G-R28, College of Information Technology, UNITEN
8.30am 9.30am Opening Ceremony
Negaraku and UNITEN Song
Doa Recitation
1.Conference Chairman Salutation
2. TSSM President Salutation
3. Salutation and Conference Officiating by Vice Chancellor
9.30am 10.30am Photo Session
Refreshments
10.30am 12.30pm Keynote speakers
1) Professor Dr Ahmad Farhan Mohd Sadullah,

Director General, Malaysian Institute of Road Safety Research
Transportation Research in Malaysia: Are we making the difference?
2) Prof. Ir. Dr. Ruslan Hassan

Professor/Specialist Consultant, Faculty of Civil Engineering, UiTM
Transportation Sustainability in a Climate-Constrained World
3) Ir. Looi Hip Peu

Deputy President , Malaysia Green Building Confederation Board
12.30pm 1.45pm Lunch

Room A Room B Room C Room D
Session 1 Session 1 Session 1 Session 1
BW-2-L06 BW-2-L14 BW-2-L15 BW-2-L16
1.45pm 2.00pm 040 G 035 PG 002 G 014 UG
2.00pm 2.15pm 041 G 024 PG 003 G 016 UG
2.15pm 2.30pm 046 G 026 PG 007 G 020 UG
2.30pm 2.45pm 050 G 028 PG 018 G 021 UG
2.455pm 3.00pm 012 PG 029 PG 019 PG 023 UG
3.00pm 3.15pm 013 PG 030 PG 027 G 042 UG
3.15pm 3.30pm 017 PG 031 PG 037 PG 051 PG
3.30pm 3.40pm Q&A Q&A Q&A Q&A
3.40pm 4.00pm Refreshments
Room A Room B Room C Room B
Session 2 Session 2 Session 2 Session 2
BW-2-L06 BW-2-L14 BW-2-L15 BW-2-L16
4.00pm 4.15pm 032 PG 039 PG 001 PG 038 PG
4.15pm 4.30pm 033 PG 043 PG 004 PG 049 PG
4.30pm 4.45pm 034 PG 044 PG 005 PG 011 PG
4.45pm 5.00pm 022 PG 045 PG 006 PG 010 PG
5.15pm 5.30pm 036 PG 047 PG 009 PG 048 PG
5.30pm 5.45pm Q&A Q&A Q&A Q&A
5.45pm-6.30pm Closing Ceremony
vi
ORGANIZING COMMITTEE
Patron: Prof. Dato Dr. Ir. Mashkuri Yaakob, Vice Chancellor of UNITEN
Advisory Committee:
Prof. Dr. Mohd. Zamri Yusoff (UNITEN)
Prof. Ir. Mohamed Rehan Karim (UM)
Dr. Muhammad Zaly Shah Muhammad Hussein (UTM)
Assoc. Prof. Dr. Sumiani Yusof (UM)
Prof. Engr. Dr. Amiruddin Ismail (SUTRA, UKM)
Prof. Dr. Jamilah Mohamad (UM)
Assoc. Prof. Dr. Norlida Abdul Hamid (UiTM)
Assoc. Prof. Dr. Madzlan Napiah (UTP)
Conference Committee:
Co-Chairs
Engr. Mohd Sufian Abdul Karim (Director of Alumni Relations Center, UNITEN)
Assoc. Prof Dr. Ir. Marlinda Abdul Malek (Head, Department of Civil Engineering, UNITEN)
Secretariat
Engr. Siti Aliyyah Masjuki
Ir. Norlela Ismail
Assoc. Prof. Dr. Engr. Lee Choon Yong
Engr. Dr. Fathoni Usman
Engr. Dr. Ho Chee Kit
Engr. Megat Asyraf Megat Ibnu Rashad
Engr. Ean Lee Woen
Proceeding Committee:
Editor
Engr. Mohd Sufian Abdul Karim
Reviewer
Dr. Muhammad Zaly Shah Muhammad Hussein
Assoc. Prof. Dr. Sumiani Yusof
Prof. Engr. Dr. Amiruddin Ismail
Prof. Dr. Jamilah Mohamad
Assoc. Prof. Dr. Norlida Abdul Hamid
Assoc. Prof. Dr. Madzlan Napiah
Assoc. Prof. Dr. Engr. Lee Choon Yong
Ir. Norlela Ismail
Engr. Dr. Ho Chee Kit
vii
Key note Speaker:
Professor Dr Ahmad Farhan Mohd Sadullah, Director General,

Malaysian Institute of Road Safety Research Professor Farhan has nearly 17
years working experience in transport and traffic engineering as well as road
safety. He has been lecturing at the School of Civil Engineering, Universiti
Sains Malaysia, since 1995, before resuming his current duty as the Director
General of the Malaysian Institute of Road Safety Research (MIROS) in
August 2008.
Professor Farhan is presently the Chair of the World Road Association

(PIARC) Technical Committee C.2 on Safer Road Operations, a council member of the Road Engineering
Association of Malaysia (REAM) and the Chair of the REAM Standing Committee on Road Safety. He is
also a board member of the Road Traffic Injury Research Network (RTIRN). He has been involved in
numerous international and national research and consultancy projects and has made a multitude amount
of contribution to published research works. His areas of research interest include sustainable transport
issues, travel behaviour, traffic engineering, and road safety. Presently, at MIROS, Farhan leads a young
and dedicated team to provide the evidence-based research support to combat the road safety menace in
Malaysia and the world. Research work in MIROS covers the user, vehicle and the roads perspectives, as
well as to support the safe-system approach. Farhan obtained his Bachelor of Civil Engineering from the
Catholic University of America, Washington DC, before pursuing his MSc in Transport from the Imperial
College London and subsequently Ph.D in Transport from the University of London, at University
College London.
Synopsis Presentation: Transportation Research in Malaysia: Are we making the difference?
Transportation research in Malaysia has had a slow start as only recently we have seen growing interest in
transportation-related research focuses. The Malaysian Universities Transportation Research Forum and
Conference (MUTRFC) have provided the platform for many researchers to share their research findings
and interests. The late start in research interest for transport may be contributed to the fact that in the past
transport in Malaysia has been taken for granted, and was always considered to be available for our use
for our daily undertakings. The scenario changed when the country began to see problems associated with
transport. This includes safety, inefficiencies, high cost, congestion and environmental issues. We as a
nation began to see the potential of research towards solving the transportation issues, as we began to
realize that the transport-associated problems has begun to cost the nations lives, lost in productivity,
image issues as well as environmental impacts. Having analyzed the array of researches currently carried
out in the country, a question came to my mind whether we, who are the advocates of better transport for
the country, are making that needed difference? Are our research findings being used by policy makers
and practitioners to help solve or alleviate the problems we are facing? Conversely, are our research
findings merely repeating and duplicating others at addressing same or similar problem statements of
which there are no offered solutions to the real world. Are we just going through the research exercise for
the sake of our degrees, our KPIs or are we there to make the difference? This presentation shall address
this issue with the hope that it will eliminate less impactful research which does not have any impact to
the problems we face in the transportation area. It is a moral obligation for us researchers to use our
resources and expertise to solve real problems and in doing so we hope to make the difference?
viii
Key note Speaker:
Ir. Dr. Ruslan Hassan

Professor/Specialist Consultant
T1-A19-3C,Science & Technology Complex

FACULTY OF CIVIL ENGINEERING
Universiti Teknologi Mara (UiTM)
e-mail: ruslan290@salam.uitm.edu.my
Dr. Ruslan Hassan is the Professor (Civil and Environmental Eng.) and also the Research Fellow at the
Malaysian Institute of Transport (MITRANS).He has been on the Faculty of Civil Engineering since
1980. He was the Hon. Research Fellow, Birmingham University, UK (1991). Dr. Ruslan Hassan taught
graduate level courses in Environmental Assessment, Air Pollution and Environmental Management,
among others. He is the author of Environmental Pollution: Assessment and Management (UNIVISION
Press) and more than 100 papers and technical reports. Dr. Ruslans current fieldS of active research
include Sustainability Design of Basins, Buildings and Transportation. He is an EIA Registered
Consultant. He is a Fellow Member of I.E.M., T.A.M. and M.S.A. Dr. Ruslan is currently the President
of C.O.S.T.A.M., and Vice-President of I.E.M. He was the Past President of T.A.M. (2003-07) and has
served as a Board Member of Malaysia (BEM)(2004-09). He was the recipient of IEM Tan Sri Yusuf
(2000) and British High Commissioners Awards (1991). He was appointed as one of the Trustees of the
Construction Research Institute Malaysia (CREAM-CIDB) since 2006.
Synopsis Presentation: Transportation Sustainability in a Climate-Constrained World
In this presentation, the transportation eras of Infrastructure, Systems and as Complex Systems with
Socio-Technical issues will be discussed. The comparisons will be made between passengers with freight
(logistics) transportation with respect to Greenhouse Gases emission. Several Solutions for attaining
sustainability in Environment, Society and Economics in the Logistics sectors will be forwarded. An
innovative solution such as Policy measures for Greenhouse Gas mitigation as a result of coupling air
pollution model with behavioral model will also be discussed. Finally, adaptive measures for further
actions will be recommended.
ix
Key note Speaker:
Ir. Looi Hip Peu
Vice President, Malaysian Green Building Confederation (2010/11)

Hon. Secretary, Association of Consulting Engineer Malaysia (2010/11)
Member, GBI Accreditation Panel
Managing Director, Mektricon Sdn Bhd (Consulting Engineers)
B.Eng (Hons), P.Eng, Jurutera Gas, Int. P.Eng (My), F.IEM, M.ACEM
hplooi@unifi.my; mektricn@unifi.my, www.hplooi.blogspot.com
Ir. Looi Hip Peu is a practicing engineer with more than 30 years experience. He is a founding director of
Mektricon Sdn Bhd (established since 1983), a firm of consulting engineers involved in a wide repertoire
of projects from large industrial plants and townships to commercial developments and small bungalows.
The firm has also commissioned projects within the ASEAN/ASIAN region. Ir. Looi currently specialises
in plant design.
Ir. H.P. Looi has been active in the engineering community, currently holding the post of Vice President
of MGBC (2010/11), Hon Secretary of ACEM (2010/11) and member of the GBI Accreditation Panel. He
has been involved in the sustainability agenda since 2000, being member / contributing author at national
level to various technical committees and workgroups (a very brief summary list) such as (a) work group
on MS1525; (b) Workgroup Guidelines on E.E. in Building, Motor & Lighting published by PTM; (c)
workgroup Thermal Guidelines for Industry, Co-Gen published by PTHM; (d) workgroup on GBI NRNC,
NREB, Factory and Township tool; (e) chairman Electrical Paper Committee, Professional Competency
Exam (BEM-ACEM).
Synopsis Presentation:
x
CONTENTS
Preface v
Organizing Committee vii
Proceeding Committee vii
Invited Keynote Speakers viii
Technical Papers
001 PG TRIP PRODUCTIVITY EVALUATION OF BUS SERVICE: MEDAN KIDD 1

BUS STATION
Madzlan Napiah, Amirah Suriati Ahmad Farid and Suwardo
002 G PASS-BY NOISE OF MALAYSIAN ARMED FORCES (MAF) VEHICLES 11

Shamsul Akmar Ab Aziz, Adam Gani and Abdul Hamid Hassan
003 G AIRCRAFT NOISE LEVEL IMPACT AT AIRPORT AND ITS 19

SURROUNDINGS
Sakti Adji Adisasmita and Achmad Zubair
004 PG INCREASED USE OF OWN CAR IN TRIPOLI LIBYA 33

Adel Ettaieb Elmloshi and Amiruddin Ismail
005 PG INFLUENCE OF BLENDING INTERACTION OF CRUMB RUBBER 43

MODIFIED BITUMEN ON PAVEMENT PROPERTIES
Nuha S. Mashaan, Mohd Rehan Karim, Asim Hassan Ali and Mahrez Abdul Aziz
006 PG PRELIMINARY ASSESSMENT ON RELIABILITY OF PUBLIC BUS 49

SERVICE IN KOTA BHARU
Madzlan Napiah and Noorfakhriah Yaakub
007 G APPLICATION OF PROBABILISTIC DYNAMIC PROGRAMMING IN 59

FORMULATING AN AIRCRAFT ACQUISITION DECISION MODEL
Teoh Lay Eng and Khoo Hooi Ling
009 PG USING THE BEHAVIORAL SCIENCES THEORY AND STRUCTURAL 67

EQUATION MODEL (SEM) IN BEHAVIORAL INTERVENTION:
HELMET USE
Kamarudin Ambak, Rozmi Ismail, Riza Atiq Abdullah and Muhamad Nazri Borhan
010 PG EFFECT OF THERMOPLASTIC COPOLYMERS ON 79

MICROSTRUCTURE AND VISCOELASTIC BEHAVIOUR OF BITUMEN
Noor Zainab Habib, Ibrahim Kamaruddin, Madzlan Napiah and Isa Mohd Tan
011 PG THE EFFECT OF FINE AGGREGATE PROPERTIES ON THE RUTTING 89
BEHAVIOR OF THE CONVENTIONAL AND POLYMER MODIFIED
BITUMINOUS MIXTURES USING TWO TYPES OF SAND AS FINE
AGGREGATE
Ibrahim Kamaruddin, Madzlan B. Napiah and Yasreen Gasm ALkhalig
012 PG REGIONAL FREIGHT TRANSPORTATION PLANNING: AN OVERVIEW 99

OF DEVELOPING METHODOLOGY
Noor Mahmudah, Danang Parikesit, Siti Malkhamah, Sigit Priyanto, and Mark
Zuidgeest
013 PG EFFECT OF RAINFALL ON MICROSCOPIC TRAFFIC FLOW 117

PARAMETERS
Hashim Mohammed Alhassan and Johnnie Ben-Edigbe
014 UG DESIGN OF BATTERY PACK FOR ELECTRIC VEHICLE BASED ON 127

LITHIUM-ION TECHNOLOGY
Z.Y Phuan, G.D Gan, C.K Leong, B.K Cheah, M.K Yoong, Y.H Gan and K.W
Chew
016 UG STUDIES OF ELECTRIC MOTORS FOR LIGHT-WEIGHT ELECTRIC 135

VEHICLE
Gan Guo Dong, Gan Yu Han, Leong Chee Ken, Phuan Zheng Yi , Vincent Cheah
Beng Keat, Yoong Mun Kiat and Chew Kuew Wai
017 PG QUICKBIRD IMAGE-BASED ESTIMATION ON NUMBER OF 149

BUILDINGS IN EACH ZONE FOR TRIP GENERATION MODELLING
Qadriathi Dg. Bau, Danang Parikesit , Hartono and Totok Gunawan
018 G TEST PROTOCOL FOR MALAYSIAN ARMED FORCES (MAF) 161

VEHICLES - HEALTH HAZARD ASSESSMENT (HHA) AND VEHICLE
PERFORMANCE
Adam Gani, Shamsul Akmar Ab Aziz and Abdul Hamid bin Hassan
019 PG IMPROVISED ENGINEERING SPECIFICATION DESIGN FOR ROAD 169

HUMP: A CASE STUDY IN RESIDENTIAL STREET SHAH ALAM
Nor Izzah Zainuddin, Muhammad Akram Adnan, Mohd Yusof Abd. Rahman and
Jezan Md Diah
020 UG CASE STUDY OF EV CONTROLLER AND POWER MANAGEMENT 175

SYSTEM
Vincent Cheah Beng Keat, Gan Yu Han, Leong Chee Ken, Phuan Zheng Yi, Gan
Guo Dong, Yoong Mun Kiat and Chew Kuew Wai
021 PG A STUDY ON ELECTRIC CAR CHASSIS AND DESIGN PRINCIPLE 191
Gan Yu Han, Gan Guo Dong, Leong Chee Ken, Phuan Zheng Yi, Vincent Cheah
Beng Keat, Yoong Mun Kiat and Chew Kuew Wai
022 PG EXPLORATION OF 85th PERCENTILE OPERATING SPEED MODEL ON 203

CREST VERTICAL CURVE TWO-LANE RURAL HIGHWAYS
Syed Khairi Syed Abbas, Muhammad Akram Adnan and Intan Rohani Endut
023 UG BUS TRANSIT SYSTEM MODELING USING PARAMICS: A CASE 213

STUDY OF KLANG VALLEY NETWORK
Ho Yun Li and Toh Boon Leong
024 PG KEEPING CITIES MOVING: A LITERATURE REVIEW ON GREEN 221

CAMPUS
Siti Hafnida Binti Razali
026 PG E-ASSIST: A TRANSPORT DEMAND MANAGEMENT EXPERT 229

ADVISORY SYSTEM
Resdiansyah Mansyur, Riza Atiq O.K. Rahmat, Amiruddin Ismail and Mohamad
Raduan Kabit
027 G TIMES SPENT IN DRIVING TAXI A CASE STUDY IN PUTRAJAYA 241

SENTRAL
Chuah Pooi Yee, Sulistyo Arintono and Hussain Hamid
028 PG ESTIMATION OF DISPERSION OF CARBON MONOXIDE (CO), 249

NITROGEN DIOXIDES (NO2), AND CARBON DIOXIDE (CO2) FROM
PORT KLANG - KLIA ROAD: PRELIMINARY FINDINGS
Nur Bazla Mohmed Yusof, Mohamad Firdaus Ibrahim and Ruslan Hassan
029 PG INSTITUTIONAL RELATIONSHIP IN PORT CITY, PENANG MALAYSIA 259

Mazuin Dollah and Zakaria Ahmad
030 PG THE PHYSICAL DEVELOPMENT OF PORT CITY 269

Raja Noor Hafizah Raja Amir Shah, Jamalunlaili Abdullah and Zakaria Ahmad
031 PG THE HEADWAY PATTERNS AND POTENTIAL PARAMETERS OF BUS 279

TRANSPORTATION IN PENANG
Suria Haron, Sarina Muhamad Noor, Ahmad Farhan Mohd Sadullah and Leong Lee
Vien
032 PG COMPARISON OF SPACE MEAN SPEED USING CAR CHASING 291
METHOD AND MOVING VEHICLE METHOD AT TWO-LANE
HIGHWAYS IN MALAYSIA
Mohd Azahar bin Awang and Leong Lee Vien
033 PG EXPERT SYSTEM FOR PAVEMENT CONDITION RATING 303

Norlela Ismail, Amiruddin Ismail and Riza Atiq O.K. Rahmat
034 PG FREE-FLOW SPEED AT TWO-LANE HIGHWAYS IN MALAYSIA 311

Leong Lee Vien, Mustaqin Alpi and Hazwaruaida Muhammad
035 PG TRAFFIC ACCIDENTS ANALYSIS ALONG EAST COAST 323

EXPRESSWAY
Intan Suhana M.R., Shahril Faiz M.N. and Azlina I.
036 PG PROPERTIES OF BITUMINOUS BINDER MODIFIED WITH WASTE 333

POLYETHYLENE TEREPHTHALATE
Zahra Niloofar Kalantar, Abdelaziz Mahrez and Mohamed Rehan Karim
037 PG FINITE ELEMENT ANALYSIS OF FLEXIBLE PAVEMENT SUBJECTED 345

TO HEAVY AXLE LOAD AT MALAYSIAN FEDERAL ROAD ROUTE
ONE
Osama Mahmoud Yassenn, Bahardin Bin Baharom, Intan Rohani Endut, Mohamed
Ahmed Hafez and Md Yunus Ab Wahab
038 PG THE ARRANGEMENT OF SHUTTLE TRAIN TRANSPORTATION 359

MODEL AS AN INTEGRATED TRANSFER SERVICE MODA OF
ADISUTJIPTO AIRPORT, YOGYAKARTA AND ADISUMARMO
AIRPORT, SURAKARTA, CENTRAL JAVA
Lilik Wachid Budi Susilo and Alvin Nur Muhammad
039 PG DEVELOPMENT OF ONELANE ONRAMP MODEL FOR 369

EXPRESSWAYS IN MALAYSIA
Leong Lee Vien and Tuti Azmalia Azai
040 G CUBE DYNASIM AND AASIDRA COMPARISONS IN ANALYZING A 381

SIGNALIZED INTERSECTION
Azlina I. and Intan Suhana M.R.
041 G A THEORETICAL APPROACH IN EVALUATING THE POTENTIAL 391

IMPACT OF WOMEN TELECOMMUTERS ON TRAVEL DEMAND
Mootaz Munjid Mustafa and Abdul Azeez Kadar Hamsa
042 UG PALM OIL CLINKERS IN VARIOUS COMPOSITIONS AS COMPOSITE 401
MATERIALS IN MARSHAL MIX DESIGN
Shuhairy Norhisham, N. Ganesh Rao, Herda Yati Katman, Siti Aliyyah Masjuki,
Fathoni Usman, Megat Asyraff Megat Ibnu Rashad and Ean Lee Woen
043 PG EVALUATION OF THE LABORATORY COMPACTING EFFORT ON 411

STONE MASTIC ASPHALT MIXTURES
Pourtahmasb,M.S. and Karim M.R.
044 PG TYPES OF DAMAGES ON FLEXIBLE PAVEMENT FOR MALAYSIAN 421

FEDERAL ROAD
Nurul Elma Kordi, Intan Rohani Endut and Bahardin Baharom
045 PG THE EFFECTIVENESS OF A ROAD TRANSPORTATION NETWORK 433

SYSTEM IN A PORT CITY: TOWARDS GREEN LOGISTICS IN
MALAYSIA
Noraini Anor and Zakaria Ahmad
046 G ECONOMIC IMPACTS OF THE SHIFTING OF FUEL CONSUMPTION IN 445

INDONESIA
Heru Sutomo and Joewono Soemardjito
047 PG AN ALTERNATIVE METHOD IN EVALUATION APPROACH OF 459

AIRPORT PAVEMENT PERFORMANCE
Ervina Ahyudanari, Nasir Shafiq and Ibrahim Kamaruddin
048 PG POPULATION BASED VEHICLE ROUTING PROBLEM - CASE STUDY : 469

SCHOOL BUS FOR NEEDY STUDENT IN SURABAYA, INDONESIA
Daniel Hary Prasetyo, Jamilah Muhamad and Rosmadi Fauzi
049 PG PERFORMANCE MEASUREMENT OF MALAYSIAN AIRPORTS USING 487

DEA METHOD
Syamsul Rizal Razali and Muhammad Zaly Shah
050 G GREEN SHIPPING IN MARITIME HUMAN CAPITAL DEVELOPMENT 493

Muhammad Zaly Shah and Lee Ghim Teck
051 PG CASE STUDY BASED ON VISUAL INSPECTION TOWARDS BRIDGE 501

DISTRESS AT PLUS EXPRESSWAY CONNECTING SUNGAI MERAB
KAWASAN PERINDUSTRIAN BANGI
Siti Aliyyah Masjuki, Shuhairy Norhisham, Herda Yati Katman, Fathoni Usman,
Megat Asyraff Megat Ibnu Rashad
Proceeding of Malaysian Universities Transportation Research Forum and Conferences 2010
(MUTRFC2010), 21 December 2010, Universiti Tenaga Nasional. ISBN 978-967-5770-08-1
TRIP PRODUCTIVITY EVALUATION OF BUS SERVICE: MEDAN KIDD BUS STATION
Madzlan Napiah 1*, Amirah Suriati Ahmad Farid 2 and Suwardo 3
ABSTRACT:
Bus scheduling is an important factor to improve public transportation service demand. The more
punctual bus operates, the more passengers using this service. From earlier observation, current bus
system has punctuality problem where passengers waiting time become longer at bus stop before bus
arrive. Growing traffic congestion, the need to preserve the environment and the problems of road safety
are the main reasons to be considered for the improvement in public transportation services. To provide a
good public transport services, travel schedule of each bus should be provided and adhered to by all
drivers. In addition, the bus scheduling should be displayed at each terminal and the bus stop for
passengers guide. Therefore, the passenger will be more interested in using public transport because of
accessibility and reliability of this facility. The objective of this study is to evaluate trip productivity of
city bus services. The quality of service of city bus operation will be evaluated by level of services. For
application, a case study based on Ipoh city bus was conducted. Three routes were chosen to assess bus
operation for the whole day in a week per route. The survey consists of tracking location along bus route,
bus headway, service frequency and bus travel time, and boarding and alighting of passenger. The result
shows the frequency of bus is very low compared to Bank World standard. Even the level of service
(LOS) of this operation is in good level (LOS C), the passengers still need to wait for a long time to ride
the bus. The improvement of the service needs to be considered in terms of condition of bus and the
competition factors with other bus operators. Thus, trip productivity of bus services will be increase.
Keywords: Bus Scheduling, Trip Productivity, City Bus Services, Public Transportation Facilities
1
Senior Lecturer, Universiti Teknologi Petronas, Bandar Seri Iskandar, 31750 Tronoh, Perak Darul Ridzuan, MALAYSIA
2
M.Sc. Student, Universiti Teknologi Petronas, Bandar Seri Iskandar, 31750 Tronoh, Perak Darul Ridzuan, MALAYSIA
3
PhD. Student, Universiti Teknologi Petronas, Bandar Seri Iskandar, 31750 Tronoh, Perak Darul Ridzuan, MALAYSIA
*Correspondence Author: Madzlan Napiah, Universiti Teknologi Petronas, Malaysia. Tel: +6053687281 Fax:
+6053654086. E-mail: madzlan_napiah@petronas.com.my
1
INTRODUCTION
Bus scheduling is an important factor to increase public transportation service demand. The more
punctual bus operates, the more passengers using this service. Growing traffic congestion, the need to
preserve the environment and the problems of road safety are the main reasons to be considered for the
improvement in public transportation services. To provide a good public transport services, travel
schedule of each bus should be provided and adhered to by all drivers.
From earlier observation, current bus system has punctuality problem where passengers waiting
time become longer at bus stop before bus arrive. Therefore, private transportation stills the main
transportation mode in this country. As a result, its bring many problems to the society and environment
such as congestion, high accident rate, shortage of parking spaces, low air quality and high maintenance
cost to maintain transportation infrastructure and facilities.
In Malaysia, public transport planning has always trailed behind rapid urban development when it
should be vice versa. With the expectation of few well-planned housing areas, most residents would find
themselves landlocked from any routes that are viable to be served by public transport, especially if the
surrounding areas are consisted of different conflicting land zones.
For application, a case study based on Ipoh city bus was conducted. This study aims to evaluate
trip productivity of city bus services. With focus on Ipoh city bus (Medan Kidd), fieldwork survey was
conducted covering primary data survey and secondary data collection. The primary data consisted of bus
service operation and passenger boarding and alighting.
LITERATURE REVIEW
Description of Study Area

Ipoh's traffic is not as congested as other cities such as Kuala Lumpur. However, during heavy
traffic such as the morning drive to work (around 7-8am),afternoon break (12-2pm) and late evening
return from work (around 5-6pm), minor traffic jams may occur. Note that the Ipoh town centre has many
one-way streets. It would affect the punctuality of bus services.
The city bus terminal is located at Ipoh city centre nearby Ipoh Railway Station (formerly in
Medan Kidd, Old Town). Medan Kidd is the main terminal for over 20 years accommodating people in
and around Ipoh that have 643 kilometer square wide. The station is viewed as the main caterer for
transportation for nearly 300,000 people surrounding the Ipoh city. Many bus companies are operating at
this terminal. Ipoh Omnibus Co. is one of the companies that shuttle between Medan Kidd, Ipoh city
centre and around Ipoh. The bus fare is between RM1 and RM2, depending on distance of travel.
Standard of Bus Characteristic Performance

In practice, there are some standard of bus operation. The World Bank standard may be used in
this study as general guidance to assess the performance of bus services. As cited by Sulistyorini (1997),
World Bank standard (Wright and Thiriez, 1987) can be used to evaluate the performance of urban public
transport system (Table 1). The standard was needed to evaluate performance such as travel distance per
bus per day, number of passengers per bus per day and load factor, etc. In addition, as in Table 2, Vuchic
(1981); Khisty and Lall (2003) stated guidance for evaluating the regular bas performance. The other
standard that can be used is Transit Capacity and Quality of Service Manual (TCQSM 2003) to evaluate
the service frequency, regularity and level of service (Transport Research Board of National Academics,
2003). Based on the TCQSM 2003 manual, the level of service is determined by referring to Table 3 and
Table 4.
2
Table 1: Table of World Bank Standard

No Parameters (unit) Standard
1 Headway (minutes) 1-12
2 Travel distance (km/bus/day) 210-260
3 No. of passenger (pass/bus/day) 440-525
4 Load factor (%) 70
5 Availability (%) 80-90
Source: Wright and Thiriez (1987); Sulistyorini (1997)
Table 2: Table of World Bank Standard (for regular bas)

No Parameters (unit) Standard
1 Vehicle capacity (pass/bus) 40-120
2 Frequency (bus/h) 60-180
3 Passenger capacity of route (pass/h) 2400-8000
4 Operating speed (km/h) 15-25
5 Lane width (m) 3.00-3.65
6 Vehicle control man/vis
7 Reliability low-med
8 Safety Med
9 Station spacing (m) 200-500
Note: man = manual, vis = visual, med = medium
Source: Standard adapted for regular bus by Vuchic (1981) as stated in Khisty and Lall (2003)
Table 3: Passengers Loading LOS Thresholds

LOS Passengers per Seat Comments
A 0.00-0.50 No passenger need sit next to another
B 0.51-0.75 Passengers can choose where to sit
C 0.76-1.00 All passengers can sit
D 1.01-1.25* Comfortable standee load for urban transit
E 1.26-1.50* Maximum schedule load for urban transit
F >1.50* Crush load
*Approximate values for comparison
Source: Transport Research Board of National Academics TCRP Report 100 (TCQSM 2003)
Table 4: Service Frequency LOS Thresholds

Average headway Frequency
LOS Comments
(min) (buses/hour)
A <10 >6 Passengers do not need schedules
B 10-14 5-6 Frequent service, passengers consult schedules
C 15-20 3-4 Maximum desirable time to wait if bus missed
D 21-30 2 Service unattractive to choice riders
E 31-60 1 Service available during the hour
F >60 <1 Service unattractive to all riders
Source: Transport Research Board of National Academics TCRP Report 100 (TCQSM 2003)
3
Quality of Service
Quality of service reflects the passengers perception of transit performance. In general, the
performance measures defined as the process whereby an organization establishes the parameters within
which programs, investments, and acquisitions are reaching the desired results. In term of this study, the
performance measures used to describe the perception of bus users about the quality of bus service based
on reliability (Transport Research Board of National Academics, TCQSM 2003).
Quality of service depends to a great extent on the operating decisions made by a transit system
and measures how successful an agency is in providing service to its customers. The manual states that
on-time performance is the most widely used reliability measure that users can relate to. Table 5 shows
various measures to evaluate quality of bus service suggested in TCQSM.
Table 5: Table of Quality of Service Framework

Service Measures
Transit Stops Route Segments/Corridors System
Availability Frequency Hours of service Service coverage
Comfort & Convenience Passenger load Reliability Transit-auto travel time
- On-time performance
- Headway adherence
Source: Transport Research Board of National Academics (2003)
These measures are closely dependent upon each other. The availability of transit service is
important for determining whether transit service is even an option for a particular trip. When service is
not available, other aspects of service quality do not matter for a given trip. If transit service is available,
then transit becomes an option for a given trip. At this point, passengers weigh the comfort and
convenience of transit against competing modes. The availability measures, along with the travel time
measures, are particularly suited to short and long-term planning efforts, while the remaining comfort and
convenience measures in each framework are well suited for ongoing service delivery monitoring. Among
which, reliability is one of the most important measures in determining bus service level from the
viewpoint of users as well as operators (Transport Research Board of National Academics, TCQSM
2003).
Scheduling
Public bus is usually run to a fixed route and schedule, serving specified bus stops or bus stations.
Some services may be semi-flexible in that they can vary where they stop on a fixed route by operating in
a hail and ride way. The route and schedule of some services may also be flexible to some degree.
Researcher from TCQSM studied the more frequent the service, the shorter the wait time when a bus is
missed or when the exact schedule is not known, and the greater the flexibility that customers have in
selecting travel times (Transport Research Board of National Academics, 2003).
Reliability
Reliability is one of the most important measures in determining bus service level from the
viewpoint of users (Seung et al., 2005). According to TCQSM, reliability affects the amount of time
passengers must wait at a bus stop for a bus to arrive, as well as the consistency of a passengers arrival
time at a destination from day-to-day (Transport Research Board of National Academics, 2003). It also
affects a passengers total trip time where if persons believe a bus may depart early, they may arrive
earlier than they would otherwise to make sure not missing the bus.
4
Reliability can be evaluated by the data collected or by on-time performance and headway
adherence. The following are factors that influence reliability:
a. Traffic conditions
b. Road construction and track maintenance
c. Vehicle and maintenance quality
d. Vehicle and staff availability
e. Transit preferential treatments
f. Schedule achievability
g. Evenness of passenger demand
h. Differences in operator driving skills
i. Wheelchair lift and ramp usage
j. Route length and the number of stops
k. Operation control strategies
METHODOLOGY
For this study, Taman Meru, Taman Buntong and Taman Rapat routes were chosen to assess bus
operation. These areas are located around Ipoh. Figure 1 and Table 6 shows information of the selected
routes.
Figure 1: Route of Bus Operation

Table 6: Table of Route Information
No. Particulars No. of Buses Distance (km)
1 Medan Kidd-Taman Buntong 4 15
2 Medan Kidd-Taman Meru 2 21
3 Medan Kidd-Taman Rapat 3 14
Data for analysis were collected through a series of field survey. The survey consists of tracking
location along bus route, bus headway, service frequency and bus travel time (on board survey), boarding
and alighting of passenger (on board survey). All primary data were collected from Ipoh city bus, Ipoh
Omnibus Co. Sdn Bhd. The secondary data were collected from the relevant resources, such as related
project report, RSN Perak 2010, and Ipoh Omnibus Co. Sdn Bhd administration.
5
Figure 2: Equipments for Field Survey
By using handheld GPS (Garmin), the name and distance of every bus stop or others point of stop
location can be identified. The points are located between two main stop which are start at bus terminal
(Medan Kidd) and ends at bus terminal. Along two main stop, all points indicate the potential passenger
and the location of passenger boarding and alighting.
The surveyor or observer placed inside the bus records the number of passenger boarding and
alighting at or between points. The surveyor also recorded point/location code into GPS where passengers
get in or out along the route from start terminal until end terminal. The number of boarding and alighting
passenger were collected during period, 7.00-19.00 (full one day) for a week per route. This data was
used to analyze passengers loading profile.
Schedule of bus operation was collected from Ipoh Omnibus Co. Sdn. Bhd administration. By
using this schedule, the actual arrival and departure time of bus at bus terminal can also be determined.
This schedule contains date of operation, bus ID, destination, scheduled time for bus to depart from bus
terminal everyday and the actual and departure time recorded by operator for each bus.
RESULT ANALYSIS
Table 7 shows the results of the analysis in comparison with the values of World Bank Standard.
This standard applies particularly to city buses which of the same category with this study, but the results
are very much different compared to the values of World Bank Standard.
The frequency of bus is only six to seven buses per hour which is very low compared to standard
(Table 8). The headway of 10 to 15 minutes is higher than the World Bank Standard (1-12 minutes),
while the travel distance in kilometer per bus per day is an average (100 to 386 km/bus/day). The higher
headway reflect the higher waiting time, thus makes the current bus system unattractive to passengers.
The number of passengers (pass/bus/day) of 301 is lower than its standard. The load factor of 85% is
higher compared to normal condition of 70% and the availability of 100% is satisfactory although the
number of passengers is lower. Other parameters such as reliability, safety and station spacing are not
available for this study.
6
Table 7: Bus service characteristics based on World Bank Standard and survey results
No Parameters (unit) Standard Results
1 Headway (minutes) 1-12 10~15
2 Travel distance (km/bus/day) 210-260 100-386
3 No. of passenger (pass/bus/day) 440-525 301
4 Load factor (%) 70 85
5 Availability (%) 80-90 100
Table 8: Bus service characteristics based on Standard (regular bas) by Vuchic (1981) and survey results
No Parameters (unit) Standard Results
1 Vehicle capacity (pass/bus) 40-120 30~48
2 Frequency (bus/h) 60-180 6~7
3 Passenger capacity of route (pass/h) 2400-8000 210~336 (per bus)
4 Operating speed (km/h) 15-25 40~50
5 Lane width (m) 3.00-3.65 3.65
6 Vehicle control man/vis Man/vis
7 Reliability low-med na
8 Safety Med na
9 Station spacing (m) 200-500 na
Note: man: manual, vis: visual, med: medium, na: not available
Figure 3 shows (a) the number of passengers and (b) load factor during one week of field survey.
The field survey covered three routes namely Taman Meru, Buntong and Gunung Rapat. On Sunday, no
bus operate for Buntong Route.
600
500
400
Meru
300
Buntong
200 Gunung Rapat
100
0
Mon Tue Wed Thu Fri Sat Sun
(a)
7
160
140
120
100
Meru
80
Buntong
60
40 Gunung Rapat
20
0
Mon Tue Wed Thu Fri Sat Sun
(b)
Figure 3: (a) Number of Passengers Boarding per Bus and (b) Load Factor during One Week Field
Survey
The number of passengers boarding and alighting is recorded and takes the average value in
passenger per bus per day to compare to its LOS. As shown in Table 9, all routes have Level of Service
(LOS) C which means all passengers can sit when the bus operated (see Table 3). This is because the
frequency of bus at every route is more than six per hour.
Table 9: Service Frequency Analysis

Passenger in Passenger out On bus LF
Route (pass/bus/day) (pass/bus/day) (pass/bus/day) (pass/seat) LOS
Tmn Meru 51 51 41 0.85 C
Buntong 43 45 37 0.77 C
Gunung Rapat 52 54 28 0.92 C
From the analysis, there are several factors affecting the changeable passengers demand to other
bus operators; journey time, accessibility or frequency of the service, waiting time, convenience and
comfort. Passengers will choose the best service to make their trip. For example, passengers usually have
their own purpose of their journey. They must choose the service that provides on-time service, accessible
and easy to get. Thus, they can arrive at their destination on time.
CONCLUSION
In conclusion, trip productivity of bus service can be improved at optimum level if bus operators
at Medan Kidd merge their operation and reorganize their operation in terms of scheduling of bus trip,
frequency of bus per route per hour, ticket fare and condition of bus for convenience people. Therefore,
the user of public bus may increase and many traffic problems could be reduced.
The quality of service of public transport system in Medan Kidd needs to be improved even
though the level of service indicates good level. This is because, according the service frequency analysis,
the LOS is C which mean all passengers can sit when the bus operated but passengers still need to wait
for a long time to ride the bus because of long headway.
8
The improvement of the service needs to be considered in terms of condition of the bus and
competition factors with other bus operators. This is important to provide the convenience and comfort
service for user, at the same time to attract people using public bus. Thus, trip productivity of bus services
will be increase.
REFERENCES
Jabatan Perancangan Bandar dan Desa Negeri Perak, (2001) Draf Rancangan Struktur Negeri Perak 2020.
Kazushi Sano, Wisinee Wisetjindawat, Yumiko Suga & Pairoj Raohthanachonkun, (2007) A Study on the
Benefits of Improving Local Bus Service Punctuality, Journal of the Eastern Asia Society for
Transportation Studies, Vol. 7.
Khisty, C. J. and Lall, B. K. (2003) Transportation Engineering: An Introduction, 3rd ed. Prentice-Hall,
Pearson Education, Inc., Upper Saddle River, New Jersey 07458.
Macafee, K., (2008) Bus Punctuality Statistics GB: 2007, Department for Transport.
Seung-young Kho Young-Ho Kim, Jun-Sik park & Eun-Ho Kim, (2005) A Development of Punctuality
Index for Bus Operation, Eastern Asia Society for Transportation Studies, Vol. 6, pp 492-504.
Sulistyorini, R. (1997) The Performance of Urban Public Transport System in Yogyakarta Special
Region, Time Series Analysis 1988-1997, Indonesia: Department of Civil Engineering, Gadjah Mada
University.
Transport Research Board of the National Academics, (2003), TRCP Report 100: Transit Capacity and
Quality of Service Manual-2nd Edition (TCQSM 2003), Washington D.C.
Yuko J. Nakanishi, (1997) Bus Performance Indicators: On-Time Performance and Service Regularity.
Transportation Research Record 1571, pp 3-13.
Wright, A. A. and Thiriez, S., (1987) Bus Service: Reducing Costs, Raising Standards, Technical Paper
Number 68: Urban Transport Series, World Bank, Washington, D.C. 20433, U.S.A.
9
10
PASS-BY NOISE OF MALAYSIAN ARMED FORCES (MAF) VEHICLES
Shamsul Akmar Ab Aziz 1*, Adam Gani 2 & Abdul Hamid Hassan 3
ABSTRACT:
Noise, Vibration and Harshness (NVH) is an important engineering study attribute and focus for
every vehicle in research and development. Pass-by noise is one of the NVH parameters that need
to be considered in order to improve noise in a vehicle as a whole. Hence, this paper will
demonstrate the opportunities that pass-by noise test can provide in a new development of
Malaysian Armed Forces (MAF) vehicles. This paper will consider all the requirements and
advantages that pass-by noise can accomplish in a NVH study on MAF vehicles. In this study,
pass-by noise test is performed on MAF vehicles by the Vehicle Branch, Science Technology
Research Institute for Defence (STRIDE). The main objective of this study is to ensure that MAF
vehicles meet the Malaysian Governmental Regulation (Environmental Quality Act 1974
Environmental Quality (Motor Vehicle Noise) Regulations 1987), and consequently improve
overall noise quality by demonstrating procedures that will accomplish the targets of NVH. In
order to perform this study, the method derived from the Environmental Quality Act 1974
Environmental Quality (Motor Vehicle Noise) Regulations 1987 is referred to as a guideline to
fulfil the procedures and requirements. The measurement methods must be exactly the same as
stated in Environmental Quality Act 1974 Environmental Quality (Motor Vehicle Noise)
Regulations 1987. This is a very important step to ensure the results are valid and can be accepted
by the international society. The result obtained will be compared to the maximum permitted
sound level given by the Environmental Quality Act 1974.
Keywords: Pass-by noise; Noise, Vibration and Harshness (NVH); Malaysian Armed Forces
(MAF).
1
Research Officer, Science Technology Research Institute for Defence (STRIDE), Kajang, Selangor, MALAYSIA
2
Research Officer, Technical Acceptance Unit, Procument Division, Kuala Lumpur, MALAYSIA
3
Senior Research Officer, Science Technology Research Institute for Defence (STRIDE), Kajang, Selangor,
MALAYSIA
*Correspondence Author: Ir. Shamsul Akmar Ab Aziz, STRIDE, Ministry of Defence, Malaysia. Tel: +603
87324400 ext. 4510, Fax: +60387336219. E-mail: shamsulakmar.abaziz@stride.gov.my
11
INTRODUCTION
Sound pollution has become an important issue in present day society, with a large
number of scientific papers emphasizing on the analysis of Noise, Vibration and Harshness
(NVH) generated by vehicles. When a vehicle is passing in front of us or surrounding us, the
sound emitted by the vehicle is called pass-by noise. Pass-by noise generated by Malaysian
Armed Forces (MAF) vehicles is one of the relevant environmental issues that affect the comfort
reaction of MAF personnel.
MAF vehicle is a complex vehicle in which every component is a potential source of

noise. The noises emitted by MAF vehicles may be from the engine, exhaust, tyre path or from
other particular components, or a combination of these sources. To monitor these noises, the
method described in the Environmental Quality Act 1974 is currently used. In most cases, the
vehicle manufacturers not only want to check that the vehicle conforms to regulations, they also
desire the ability to analyze the results within the framework of research and development
studies. Usually, the equipment setup often requires installation and dismantling which takes most
of the time needed to perform a test.
One of the problems in performing an outdoor measurement is the weather condition.

Adverse weather conditions is totally difficult and no possible measurement can be done.
Consequently, the measurement system should not only be resistant to difficult outdoor weather
conditions, but also be very compact, easy to use and fast to setup (Cousin & Mauss, 2003).
In this paper, the investigation of difference types of MAF vehicles is conducted in order
to determine if they comply with the pass-by noise requirements indicated in the Malaysian
Governmental Regulation (Environmental Quality Act 1974 Environmental Quality (Motor
Vehicle Noise) Regulations 1987). If the MAF vehicles pass-by noise complies with this
standard, it means that the MAF vehicle is safe and comfortable to be used by MAF personnel.
At the same time, one of the noise pollution sources can be reduced, providing a significant
tactical during military operations. As one of the government body responsible in monitoring
human comfort of MAF vehicles, the Vehicle Branch, Science Technology Research Institute for
Defence (STRIDE) is very concerned about noise pollution of MAF vehicles. Therefore, it is
hoped that this control method will help reduce the pass-by noise of MAF vehicles.
PASS BY NOISE ASSESSMENT
Noise-induced hearing loss, which is increasing slowly over the years, is caused by high
levels of noise regularly exceeding a daily average of 85 dB(A) (Fernandes et al., 2005). One of
the main causes of hearing loss is occupational noise exposure. In this study, the levels of daily
personal noise exposure for MAF personnel is calculated using the following standard expression:
(1)
where L A(EP,d) is an 8-hour daily sound average and the L A(eq) is the A-weighted equivalent
continuous sound level for an interval of time T corresponding to the type of noise k that the
person is exposed to during T hours per day (Fernandes et al., 2005). A-weighting is used with
noise measuring equipment to give approximately the same response to different frequencies as
the human ear (Powell & Forrest, 1988). Most of the measurements related to vehicle noise use
A-weighting which is already set up in the equipment. The unit dB(A) is defined as decibel A-
weighted, being a unit of measurement of sound level obtained when using a sound level meter
12
corrected to the A weight scale using a reference pressure of 20 micropascals (Environmental

Quality Act 1974, 1987).
Vehicle noise can cause hearing loss, annoyance, discomfort and interfere with personal
performance if it exceeds the permissible limit. Table 1 shows the maximum sound level
permitted for motor vehicles having more than three wheels. It is divided into seven categories
depending on usage of vehicle, vehicle permissible weight and engine capacity.
Table 1: Maximum Sound Level Permitted For Motor Vehicles Having More Than Three Wheels
(Source: Environmental Quality Act 1974 (1987))
No. Maximum Sound

Category of Vehicle Level Permitted
(dB (A))
1. Used for the carriage of passengers and comprising 80
not more than 9 seats (including the driver's seat)
more than 9 seats. Permitted maximum weight
does not exceed 3.5 tonnes
3. Used for the carriage of goods. Permitted 81
maximum weight does not exceed 3.5 tonnes.
Engine is less than 200 h.p
exceeds 3.5 tonnes. Engine is less than 200 h.p
exceeds 3.5 tonnes. Engine is 200 h .p. or more
maximum weight exceeds 3.5 tonnes. Engine is
less than 200 h.p
maximum weight exceeds 3.5 tonnes. Engine is
200 h.p. or more
METHODOLOGY
Types of Vehicles
Five types of vehicles are used in this study; 3-Ton Truck Mercedes Benz 4X4 911B,
Vamtac S3, Land Rover Defender 130 TDI 4x4 LRV, High Mobility Load Carrier (HMLC) Tatra
7 Ton 4x4 and Tatra 10 Ton 6x6. All these vehicles are wheeled vehicles, and shall confirm to
Malaysian Governmental Regulation (Environmental Quality Act 1974 Environmental Quality
(Motor Vehicle Noise) Regulations 1987).
13
Table 2: Vehicles Specification (Source: STRIDE (2004, 2005, 2007a, 2007b, 2007c)).
Vehicles 3-Ton Truck Vamtac S3, Land Rover Tatra 7 Ton Tatra 10 Ton
Mercedes Spain Defender 130 4x4 6x6
Benz TDi
Type OM 352, Turbo 2495 cc 4 stroke with 4 stroke with
Diesel 4 Intercooler, 6 6 cylinder in- 6 cylinder in-
Stroke 6 line 188 hp, line line
Cylinder, 3200 cc
130hp
Max Torque 363 Nm @ 410 Nm @ 265 Nm @ 950 Nm @ 1966 Nm @
2000 rpm 2000 rpm 1800 rpm 1700 rpm 1200 rpm
Vehicle use Goods carrier Goods carrier Passenger Passenger and Passenger and
carrier cargo carrier cargo carrier
Maximum 3000 2000 3500 7000 10000
pay load (kg)
Max sound 81 81 80 85 passenger 85 passenger
level carrier) carrier)
permitted 88 (good 88 (good
(dB(A)) carrier) carrier)
Methods
Pass-by test on MAF vehicles is measured using a Sound Level Meter (SLM) Bruel &
Kjaer Type 2500 or Type 2238. The equipment should meet the requirements of a Type 1
instrument according to the International Electrotechnical Commission (IEC) Publication 651.
The measurements shall be made by using the frequency weighting A and time frequency F
(Environmental Quality Act 1974 (1987)).
An SLM is positioned at each side of the vehicle (left position and right position). Each
SLM is positioned 7.5 m from reference line CC (Figure 1) on the test track. The height of the
SLM shall be 1.2 m above ground level. In order to comply with the standard, the measurement
must be carried out under certain conditions. Background and wind noise interference will give
effect to the reading. The weather shall be clear, and the wind speed shall be not more than 5 m/s.
The surroundings must be clear of high vegetation or buildings. The background noise (including
wind noise) shall be at least 10 dB less than levels measured during the test. Measurements shall
not be made in adverse weather conditions (Environmental Quality Act 1974 (1987)).
The track must have certain characteristics and comply with certain rules. Standard ISO
10844 or SAE1470 describe the characteristics of the track. The test site shall be substantially
levelled. The surface of the test track shall be dry. The test track and the surface of the site shall
be dry and free from absorbing materials such as long grass, or big trees (Environmental Quality
Act 1974 (1987)). By considering all the parameters involved in this type of measurement, a
complete system shall be able be handle, control and adhere to the measurement process.
At least two measurements shall be made on each side (right and left). The vehicle shall
approach line AA with constant speed 50 km/h. When the front of the vehicle reaches line AA,
the throttle shall be fully opened as rapidly as practicably possible and held fully opened until the
rear of the vehicle reaches line BB, where the throttle shall be decelerated as rapidly as possible.
The measurement shall be made on each side of vehicle, with 2 readings for each side of vehicle.
14
The results are valid if the differences between 2 readings do not exceed 2 dB, and the highest
value given by the measurement shall be adopted as the representative value of the noise
generated from running vehicles on roads (Environmental Quality Act 1974 (1987)).
Figure 1: Microphone positions for measurements

(Source: Environmental Quality Act, 1974)
RESULTS & DISCUSSION
3-Ton Truck Mercedes Benz 4X4 911B (Figure 2)
This trial was carried out at Kahang Air Strip, Johor (STRIDE, 2004). This location was
selected because it has a long path road, and is located far away from main road. The wind speed
and environmental noise levels were recorded to ensure these parameters do not exceed the given
limits.
Reading SPL dB(A)

No. Left Right
1 84.0 81.0
2 84.1 81.6
3 84.3 82.6
* Environment parameters:
Wind Speed : 1.2 1.5m/s
Background Noise : 33.5 dB(A)
Outside Temperature: 38oC
Figure 2: 3-Ton Truck Mercedes Benz Table 3: 3-Ton Truck Mercedes Benz pass-by noise
Comparisons were made between the measurement readings and the standard values
(Schedule of Standard C). From the vehicle specification given in Table 2, this vehicle falls into
Category 3 (Table 1), for which the permitted sound level is 81 dB(A). From the results (Table 3),
the highest reading obtained for this vehicle was 84.3 dB (A) at the left position. We can
conclude that the overall sound produced by this vehicle exceeds the permitted limit.
15
Vamtac S3 (Figure 3)
URO VAMTAC (Vehculo de Alta Movilidad Tctico) is a four-wheel drive military

vehicle manufactured by the Spanish Company, Urovesa (STRIDE, 2007c). This vehicle was
physically evaluated by the MAF trial team on February 2007. The location for the test is at
Kahang Air Strip, Johor.
Reading SPL dB(A)

No. Left Right
1 96.0 100.3
2 97.6 100.5
3 96.8 101.1
Figure 3: Vamtac S3. Table 4: Vamtac S3 pass-by noise.
the highest reading obtained for this vehicle was 101.1 dB(A) at right position. We can conclude
that the overall sound produced by this vehicle exceeds the permitted limit.
Land Rover Defender 130 TDI 4x4 (Figure 4)
The Land Rover Defender is a British four-wheel drive off-road utility vehicle (STRIDE,
2004). In Malaysia, this vehicle is a common utility vehicle that is used to carry armed forces
personnel and light goods.
Reading SPL dB(A)

No. Left Right
1 79.0 78.9
2 79.2 78.4
3 79.1 78.7
Figure 4: Land Rover Defender 130 TDI 4x4. Table 5: Land Rover Defender 130 TDI
pass-by noise.
the highest reading obtained from this vehicle was 79.2 dB(A) at the left position. We can
conclude that the overall sound produced by this vehicle does not exceed the permitted limit.
16
High Mobility Load Carrier (HMLC) Tatra 7 Ton 4x4 (Figure 5)
The MAF trial team conducted the physical evaluation for this truck, which was
manufactured by the Tatra Company, Czech Republic (STRIDE, 2007a). This truck is used for
transporting personal, logistics and goods.
Reading SPL dB(A)

No. Left Right
1 87.0 88.1
2 87.4 88.5
3 87.2 88.7
Wind Speed : < 5 m/s
Temperature : 36oC
Figure 5: High Mobility Load Carrier (HMLC)
Tatra 7 Ton 4x4. Table 6: Tatra 7 Ton 4x4 pass-by noise.
Category 5 (Table 1) if used as a passenger carrier, and Category 7 (Table 1) if used as a cargo
carrier. The permitted sound level is 85 dB(A) and 88 dB(A) respectively. From the results (Table
6), the highest reading obtained for this vehicle was 88.7 dB(A) at the right position. We can
conclude that the overall sound produced by this vehicle does not exceed the permitted limit if
used as a passenger carrier, but exceeds the permitted limit if used as a cargo carrier.
High Mobility Load Carrier (HMLC) Tatra 10 Ton 6x6 ( Figure 6)
The Tatra 10 Ton vehicle has the same functions as the Tatra 7 Ton vehicle, but has a different
axle system, which was upgraded to the 6x6 type (STRIDE, 2007b).
Reading SPL dB(A)

No. Left Right
1 83.5 84.6
2 83.8 84.9
3 83.5 84.0
Wind Speed : < 5 m/s
Temperature : 36oC
Figure 6: High Mobility Load Carrier (HMLC)

Tatra 10 Ton 6x6. Table 7: Tatra 10 Ton 4x4 pass-by noise.
(Schedule of Standard C). From the vehicle specification data given in table 2, this vehicle falls
into Category 5 (Table 1) if used as a passenger carrier, and Category 7 (Table 1) if used as a
cargo carrier. The permitted sound level is 85 dB(A) and 88 dB(A) respectively. From the results
(Table 7), the highest reading obtained from this vehicle was 84.9 dB(A) at right position. We can
17
conclude that the overall sound produced by this vehicle exceeds the permitted limit for the both
its functions, as a passenger carrier and as a cargo carrier.
CONCLUSION
Pass-by noise measurement was conducted to determine the maximum overall sound
pressure level emitted by an MAF vehicle while accelerating between a set of two SLMs. As high
sound pressure levels can cause damage to the human hearing, a necessary reduction of exposure
in sound levels is required for acoustic comfort. From the results above, three vehicles were
identified as failing to fulfil the standard requirement as stated in Malaysian Governmental
Regulation (Environmental Quality Act 1974 Environmental Quality (Motor Vehicle Noise)
Regulations 1987). They were 3-Ton Truck Mercedes Benz 4X4 911B, Vamtac S3 and High
Mobility Load Carrier (HMLC) Tatra 10 Ton 6x6. Only one vehicle passed the standard
requirements; the Land Rover Defender 130 TDI 4x4. Meanwhile, the High Mobility Load
Carrier (HMLC) Tatra 7 Ton 4x4 passed the standard requirement as a passenger carrier but is not
recommended as a cargo carrier.
To overcome the noise pollution of MAF vehicles, regular maintenance and early repairs
as well as replacement of defective items should be conducted on the MAF vehicles. Currently
the exterior noise level of the accelerated pass-by noise measurement is the only legal limit for
vehicle acoustics. Within the scope of further noise emission reduction efforts, a tightening of
limits and the procedure can be expected. Pass-by noise measurements can support the
development activities of car manufactures in producing environmental friendly MAF vehicle.
We hope that the enforcement of the rules and regulations will be more stringent to ensure the
noise level emitted by MAF vehicle is under permitted level limit.
REFERENCES
Environmental Quality Act 1974 (Environmental Quality (Motor Vehicle Noise) Regulations)
(1987). Federal Subsidiary Legislation.
Fry.J. & Jennings.P. (2003). Using multi-layer perceptions to predict vehicle pass-by noise.
University of Warwick, United Kingdom.
Cousin, G. & Mauss, P. (2003). An overview of vehicle pass-by noise. OROS. Virginia, USA.
Fernandes.S.J.C, Carvalho.A.P.O, Gallas.M, Vaz.P. & Matos.A. (2005). Noise Levels in
Dental Schools. University of Porto, Portugal.
Powell, R.F. & Forrest, M.R. (1988). Noise in the Military Environment. Shrivenham, United
Kingdom.
STRIDE, (2004). Laporan STRIDE Ujinilai Land Rover Defender 130 TDI 4x4 LRV. Science &
Technology Institute for Defence (STRIDE), Ministry of Defence, Malaysia.
STRIDE, (2005). Laporan STRIDE Ujinilai Technology Demonstrator Trak 3 Ton Mercedes
Benz 4X4 911B/911C Yang Telah Di Baikpulih. Science & Technology Institute for
Defence (STRIDE), Ministry of Defence, Malaysia.
STRIDE, (2007a). Laporan STRIDE Ujinilai High Mobility Load Carrier (HMLC) Tatra 7 Ton
4x4. Science & Technology Institute for Defence (STRIDE), Ministry of Defence,
Malaysia.
STRIDE, (2007b). Laporan STRIDE Ujinilai High Mobility Load Carrier (HMLC) Tatra 10 Ton
6x6. Science & Technology Institute for Defence (STRIDE), Ministry of Defence,
Malaysia.
STRIDE, (2007c). Laporan STRIDE Ujinilai Vamtac S3 dari Spain. Science & Technology
Institute for Defence (STRIDE), Ministry of Defence, Malaysia.
18
AIRCRAFT NOISE LEVEL IMPACT AT AIRPORT AND ITS SURROUNDINGS

(Case Study: Djalaluddin Airport, Gorontalo, Indonesia)
Sakti Adji Adisasmita*, Achmad Zubair
ABSTRACT:
The study is conducted to share information by collecting information/data from Djalaluddin Airport
Gorontalo in order to measure, prevent, minimize and evaluate the environments in airport and its
surrounding. The study outputs are: to know the noise level and to evaluate the noise impact in
Djalaluddin Airport Gorontalo such that the noise impact can be manage and monitor effectively and
efficiently not only in the current situation but also in the future condistions.
The purpose of this study is to measure and evaluate noise impact on airport activities and also manage
the airport noise to prevent, minimize and improve positive impact suct that Djlaluddin Airport activity
can give optimum benefits to the people.
The data used are primary and secondary data. The primary data is based on field survey at Djalaluddin
Airport Gorontalo, interview and discussion with official government, and people who live in surrounding
airport, while secondary data is based on data collected from various agencies, namely the Directorate
General of Transport Department, the Statistic Central Bureau, PT. Angkasa Pura 1, and so on.
Analytical tools used to measure the noise level in Djalaluddin Airport Gorontalo was WECPNL
(Weighted Equivalent Continuous Perceived Noise Level) Method which focused on departure and arrival
aircrafts at certain points and times.
The findings and recommendations of the study is based on the information of distinctive standard,
technology, procedures methods and tools regarding to the noise measures especially on airport
construction and operation derived from questionnaires and field survey. In general, the findings and
recommendations of this study can be summarized as follows: based on the condition in Djalaluddin
Airport and its surrounding shows that noise WECPNL index level at all measure points, the results is in
the range of noise standards. Neverthelles, for airport operational, noise also come from vehicles traffic
activities not only to/from airport but also from traffic flows in airport surrounding, cause housing area
located next to the airport and traffic frequency at roads in airport surrounding also influence the airport
development activities.
Keywords: Noise, Airport
__________________________________________________________________
Lecturer, Faculty of Engineering, Hasanuddin University, Jl. Perintis Kemerdekaan KM10, Tamalanrea 90245, Makassar, INDONESIA
Lecturer, Faculty of Engineering, Hasanuddin University, Jl. Perintis Kemerdekaan KM 10, Tamalanrea 90245, Makassar, INDONESIA
* Correspondence Author: Sakti Adji Adisasmita, Hasanuddin University, Indonesia. Tel: +62411 587 636, Fax: +62411 580
505. E-mail: adjiadisasmita@yahoo.com
19
BACKGROUND
Gorontalo Province located at Integrated Economic Zones (IEZ) Batui (Central Sulawesi
Province and Manado, North Sulawesi). Location is strategic make it a transit area of all commodities.
Transportation to and from this region become crowded, air transport is the main entrance, so that the
airport is expected to provide excellent service both in terms of technical, environmental, safety and
comfort.
Airport operational activities are expected to have an impact on the surrounding environment,
both effects are positive or negative, so the impact of airport operations to be managed either by
increasing the positive impacts and minimize negative impacts, it is as mandated in Law No. 23 in the
year of 1997 on Environmental Management.
Djalaluddin Airport, Gorontalo continues to experience development in fulfilling the need for
airport service excellence. For example the construction of airport facilities for aviation safety, namely
Instrument Landing System (ILS) is expected to flow front arrival (landing) and departure (take-off)
aircraft that will enter and exit the airport more secure.
The purpose of this study was investigates the impact of noise arising from the activities of the
airport, handling negative impacts, planning and implementation of action to prevent, reduce or
overcome the negative impact and increase the positive impact, such that the activities will be able to
provide optimum benefit.
When the research study conducted, on going activities at Djalaluddin Airport, Gorontalo
was extension of runway 27 from 2,250 m to 2,500 m and widening runway from 30 m to 45 m that
began in 2008. Djalaluddin Airport development plans is divided into three stages of development and
construction of facilities, namely Phase 1 in 2010, phase two in 2025 and the ultimate stage. An
estimated number of aircraft arrival-departure as many as 5,710 aircrafts with the number of passenger
arrival-departure-transit as many as 264,320 people in the plan of Phase I in 2010. Meanwhile, plans for
Phase II, with the prediction arrival-departure 7,640 aircrafts and 696,517 passengers.
Operational activity is a routine activity that is part of the operation of runway, taxiway and
apron. This activity is the movement of aircraft which includes take-off, landing and the activity on the
apron including refueling aircraft. Based on data from October 2008 at the Airport's air traffic at
Djalaluddin Airport, Gorontalo serve 80 planes arrived and departed of 89 aircraft per month. Djalaluddin
Airport, Gorontalo serving passenger and cargo and based on data from passengers who arrived in
October 2008 were 9,411 people and departed were 3,519 people per month
Operation of aircraft activity, especially landings and take-offs would cause noise in the area
around the airport. Noise level in the area are measured by a method and regulations. Based on the noise
level can be determined on an area of land use with reference to the method based on WECPNL Method
and regulations from Director General of Civil Aviation No. 109 of 2000.
Noise is considered the most important environmental concern for many people who live or
work near to an airport. As one would expect, in general, the greatest noise levels are experienced close to
the airport, however, disturbance can also occur many miles away under aircraft approach and departure
routes. Although individual planes have become quieter, any benefit to residents has been offset by the
increase in the volume of aircraft traffic.
20
Noise disturbance around an airport is caused by aircraft in the air; reverse thrust used by
aircraft to slow down after landing; aircraft on the ground, including taxiing, engine testing and running
on-board electrical generators; departing aircraft that stray from the Preferred Noise Routes (PNRs); road
traffic to and from the airport; and construction activity. Aircraft noise disturbance is commonly
measured as a function of the number of aircraft and the noise levels made by each, this provides the
noise climate.
The effect of noise in the airport area itself and the surroundings has caused damage to
buildings. Such conditions as seen in buildings in the village of SDN 04 Isimu at South Isimu Village
whose roof damaged by jet blast (vibration). The SDN 04located 200m from the end of runway and
included in the noise level areas 1. Thus, the existence of these schools are planned to be moved by the
local government. Similarly with the existence of these villagers who numbered as many as 200 heads of
households.
OVERVIEW OF RESEARCH STUDY AREA
Djalaluddin Airport is located in the city of Gorontalo, Gorontalo Province. The position of this
province is located in the northern part of Sulawesi Island that is directly adjacent to the North Sulawesi
province in eastern and Central Sulawesi Province in the west, while its northern faced the Celebes Sea
and in the south to the Gulf of Tomini.
Gorontalo region is very strategic as it considers economically, because it is located on the

central axis of economic growth areas, namely between 2 (two) Integrated Economic Zones (IEZ) Batui,
Central Sulawesi and Manado, Bitung in North Sulawesi Province. This strategic location can be used as
a transit area of all commodities from and to both these IEZ. As a result of activities flow of goods in the
IEZ area, then have a positive impact on improving economic activity in the region of North Sulawesi,
Central Sulawesi and Southeast Sulawesi and even the South. In addition, Gorontalo are also on the
countries overlooking the Pacific Ocean facing the countries of Korea, Japan and Latin America. Excess
this position can provide a good opportunity in trade development.
Overall Gorontalo province has an area of 12,215.44 km2. Compared with other parts of
Indonesia, this province is an area of only 0.64 percent. Gorontalo province consists of 5 (five) districts
and 1 (one) city, namely: Boalemo Regency, Gorontalo Regency, Pohuwato Regency, Bone Bolango
Regency and Gorontalo City. Area of each Regency/City are Boalemo Regency: 2,248.24 km2 (18.4%),
Gorontalo Regency: 3,426.98 km2 (28.05%), Pohuwato Regency: 4,491.03 km2 (36.77%), Bone Bolango
Regency: 1,984.40 km2 (16%), and Gorontalo City: 64.79 km2 (0, 53%). The map of Gorontalo Province
and Gorontalo City can be seen in Figure 1.
21
Gorontalo Province Gorontalo City
Figure 1 Maps of Gorontalo Province and Gorontalo City
The layout of airport facilities which consists of airside and landside areas in detail can be seen
in Tables 1 and 2 below.
Table 1 Airside Facility
NO FACILITY DIMENSIONS
1 Runway 2500x45m
2 Taxiway A 115x23m
Taxiway B 115x23m
3 Apron 230x80m
4 Strip 2500x150m
5 Shoulder 2500x60m
Source: Djalaluddin Airport Data, 2009
Table 2 Landside Facility
NO FACILITY DIMENSIONS
1 Departure Terminal
a. Check-in counter
b. Check-in area 160.5
c. Waiting room 163.5
d. Toilet 100
e. Praying room 9
f. Lounge 90
2 Arrival Terminal
a. Arrival room 163
b. Toilet 27.5
c. Praying room 36
d. Curb 12
22
e. Concourse 12
3 Cargo terminal 550
4 Supporting Facility
a. Administration office 1,300
b. Operational building 480
c. ATC 180
d. Fire building + Parking area 416.8
e. NDB building 36
f. DVOR/DME building 96
g. Power house building 375
h. CCR building 63
i. ILS building -
j. AMSC building -
k. Meteorology building -
l. Workshop building 280
5 Air Navigation Facility
a. Non Directional Beacon 1 unit
(NDB) 1 unit
b. Very high frequency 1 unit
c. Omni directional radio range
6 Precision Landing Aids Facility
a. Instrument Landing System 1 unit
(ILS) 1 unit
b. Middle maker 1 unit
c. Loclizer 1 unit
d. Glide path
Source: Djalaluddin Airport Data, 2009
Development in the field of air transportation plays a very important and strategic role in
supporting, encouraging all aspects of life in the economic, social, cultural, politic, defense and security.
Air Transportation become increasingly important, due to the vast territory of the unitary Republic of
Indonesia which are separated by islands, and can be a tool that can connect in a fast travel time.
Structuring a reliable air transportation system, integrated, and directed, must be supported by improving
the quality of human resources, as well as the fulfillment of air transportation statistics. The construction
of Djalaluddin Airport is expected to support the need for air transportation facilities that can provide
support for the development of Gorontalo region and surrounding areas. Increasing the number of
passengers and flights will affect the level of regional income and welfare of the community. Air traffic
data from year 2000 until the year 2008 can be seen in Table 3 below.
Table 3 Air Traffic Passenger Flows through Djalaluddin Airport, Gorontalo
Aircraft Passenger
Year Depart Arrive Depart Arrive Transit
2000 177 178 1,547 1,627 637
2001 377 377 4,626 3,514 3,599
2002 511 514 14,520 11,769 3,836
2003 900 901 30,059 29,691 267
2004 1,229 514 63,026 56,638 509
23
2005 1,062 377 75,767 70,068 1,141

2006 1,035 178 83,599 79,021 876
2007 1,170 1,174 93,802 91,577 2,492
2008 968 970 95,791 96,043 206
Source: Gorontalo in Figures 2008
NOISE LIMITATION AREA
Regions Noise Limit is a limit of a certain area around the airport whose are affected by aircraft
engine sound waves and can interfere with the environment. Noise Limit Zone is divided into three
categories that can be seen in Table 4 as follows.
Table 4 Noise Limitation Area
CATEGORY DESCRIPTION INDEX VALUE

AREA WECPNL
Region I Can be used for various activities, except to build new

schools and hospitals, while for building schools and
hospitals are advised to existing fitted with silencers o 70WECPNL<75
sound-proof trap.
Region II Can be used for various activities, except for building

schools, hospitals, and new homes, while schools, hospitals
and existing homes are advised to trap equipped with 75WECPNL<80
silencers or soundproof.
Region III Can be used to construct buildings or facilities in the airpor

which is specifically used for the operation of the airport is
equipped with a silencer or sound-proof trap. And can be WECPNL80
used as green belt or agricultural facilities and environmenta
controls that did not attract the birds coming.
Source: . Directorate General of Air Transport Regulation/109/VI/2000
METHODOLOGY
To determine the level of noise therefore the measurement is done directly in the airport area. It is
based on the likely distribution of noise that will be generated by movement of the aircraft and all activity
at the airport itself. Noise measurements were performed on 4 (four) locations as a point of
measurement and measurement results are obtained through calculation WECPNL (Weighted
Equivalent Continuous Perceived Noise Level). Thus, the value of the noise level at the airport
area can only be obtained through a study on the computation WECPNL (Weighted Equivalent
Continuous Perceived Noise Level). The equipment used to measure the noise level shown in
Figure 2. As for determining noise levels are calculated using the following WECPNL formula.
WECPNL = dB (A) + 10 Log N 27

Where:
dB (A) = average noise level within 1 (one) day, measurement (24 hours).
24
N = N1 + 3 N2 + 10 ( N3 + N 4 )
Where:
N = Number of events (arrival and departure of aircraft) in 1 (one) day (24 hours)
N1 = Number of events in the period in hours 07:00 to 19:00 local time
N2 = Number of events in the period in hours 19:00 to 22:00 local time
N3 = Number of events in the period 22:00 hrs - 24.00 local time
N4 = Number of events in the period 24.00 hours - 07.00 local time
Noise Monitoring Terminal
(Out side) Microphone
Figure 2 Noise Equipment
Note book P/C
(Inside the office)
Laser Printr
Security SERVER
XY Plottrr
Stabilizer
RESULTS AND DISCUSSIONS
NOISE LEVEL
Results of Noise survey around Djalaluddin Airport, Gorontalo:
Point 1, location at TNI Air Force complex, south of runway located at noise limitation area level 1
with a noise average level value of WECPNL 54.
25
Point 2, location at runway end (runway) 27, located at noise limitation aea level 2 with noise average
level value of WECPNL 78.
Point 3, location at arrival room arrival terminal, located at noise limitation area level 1 with noise
average level value of WECPNL 74.
Poin 4, location at runway edge (runway) 09, located at noise limitation area level 1 with noise average
level value of WECPNL 62.
Based on initial environmental conditions (exsisting), the WECPNL calculation results presented in Table
5 below.
Table 5 Calculation Result of WECPNL

TIME Point 1 Point 2 Point 3 Point 4
8/6/2009 53 75 72 56
8/7/2009 59 80 78 77
8/8/2009 52 77 72 52
Average 54 78 74 62
Source: Survey Results
Description:
Point 1 = E 12251269 N 003759.62
Point 2 = E 12252.417 N 00385.62
Point 3 = E 122512.98 N 003820.43
Point 4 = E 122506.98 N 003822.52
While the distribution of noise in the area of the airport on WECPNL index value can be seen
in Table 4 in the previous page.
From these noise measurements can be made noise contour map. Furthermore, from noise
contour map can be seen that for the points 1,3 and 4 which is an area of noise level 1, which has noise of
WECPNL index below 75, i.e. 74 points for point 3, 62 for point 4 and 54 for points 1. While for point 2
the noise level is 2 with WECPNL index of 78.
Under these conditions showed that the noise level as an index of WECPNL at all points of
measurement are, within the provisions of the noise region, even for points 1 and 4 may even lie below
the amount of noise index of WECPNL.
However, in the operational activities of these noise sources can also be derived from vehicle
traffic activity, both in airport or around the airport area.
Consideration was given that residential areas were around the airport and the frequency of
traffic on the roads around the airport was also affected by development activities at the Djalaludin
Airport. Therefore, it is felt the need to assess the level of noise caused by traffic around the airport and
the region. The road network layout around airport shown in Figure 3 and the results of this study will be
used in evaluating for noise impact area surrounding the airport. The study of noise in the form of an
estimate using equation (Rau Wooten) as follows:
26
Figure 3 Road Network Layout around Djalaluddin Airport
Description:
Li : Noise caused by vehicles (i) (dB (A)

Leq (i) : Vehicle noise level (dB (A)
n : Amount of traffic per type of vehicle
ti : Noise exposure time (minutes)
r : affected by noisy distance (m)
Lt (total) : Noise generated by all vehicles dB (A)
Using the formula above, where the exposure time (ti = 3 minutes), the distance
affected by noise (r = 10m to 50m). The 50m distance is the normal distance between the
housing areas and roads. Thus, the noise can be estimated by traffic generated by each type of
vehicle at this time. The results shows that noise level generated by each vehicle is 80 dB (A)
for motorcycles, 70 dB (A) for cars, 85 dB (A) for trucks and 82 dB (A) for bus. The calculation
is presented in Table 6 below.
Tabel 6 Noise Predictions

Distance Noise
A1 A2 A3 A4 A5
(m) dB (A) dB (A) dB (A) dB (A) dB (A)
10 68.18 63.31 65.19 65.43 65.37
20 62.16 57.29 59.17 59.41 59.35
30 58.64 53.77 55.65 55.88 55.82
40 56.14 51.27 53.15 53.38 53.33
50 54.20 49.33 51.21 51.45 51.39
Source: Analysis Results
27
Description:
A1 = Road exit airport enter
A2 = Isimju Rd Bangomeme Rd
A3 = Isimu Rd Kwandang Rd
A4 = Isimu Rd Paguyaman Rd
A5 = Isimu Rd Limboto Rd
Based on the above calculation shows the total noise value of various sources of noise at all
observation locations in the area of the airport and the streets surrounding, it can be in a safe condition
began to distance 40m. At that distance the highest total noise on the path Isimu, it shows at 54.45 dB
(A). Such conditions still show that the noise value is below the noise level standard for residential areas
that is equal to 55 dB (A) according to Indonesian Environmental Regulation No. 48, 1996. Thus, the
noise impact of traffic in the airport and surrounding area is still in a fair and does not potentially interfere
with public health and comfort.
Construction Activity
Increased levels of noise which occurs due to the noise sources from which its existence is very
significant both in terms of numbers and movements at the airport and surrounding area, as for the
sources of vehicles and heavy equipment operating in land clearing work, and the runway pavement.
Similarly, the effects felt by residents about noise living near existing roads around the airport as the
traffic lane of vehicles.
The traffic noise based on studies showing that the normal noise level for residential areas to a
distance of about 50m from the source but high noise sources are derived from large trucks like dump
trucks that is equal to 84 dB (A) for each truck. Thus, the noise impact will be felt for workers in the
surrounding airport area, especially people who live adjacent to the route passing vehicle project.
Operational Activities
Sources of noise impacts from aircraft movements such as landing and take-off on the runway,
taxiway and apron. The impact of these activities is felt in the airport area and also in the surrounding area
and not infrequently also caused little damage to buildings, such conditions as seen in buildings in the
village of SDN 04 Isimu where roof damaged by jet blast plane
The presence of SDN 04 is about 200m from the end of runway 09 and was included in the
noise level of a region. Thus, the existence of these schools are planned to be moved by the local
government. Similarly with the existence of these villagers who numbered as many as 200 heads of
households. However, as has been previously known that the noise level caused by the movement of
aircraft still in the range and even below the noise index for classification of regions according to
WECPNL, Director General of Air Transport Regulation/109/VI/2000.
Noise Impact Evaluation
One of the important negative impacts from airport activities is noise. In the work of
construction activities from the 2,250 m extension to 2,500 m and widening the runway from 30 m to 45
m, sources of noise can be derived from the vehicles, heavy equipment and generators used for work
purposes. While the impact of the noise itself will be felt by workers in the the airport surrounding,
especially people who live adjacent to the route passing vehicle project. For that necessary arrangements
28
relating to the management of project work in determining the timing of the work until the volume of
vehicles in use.
Meanwhile, in addition to the operational activities from the movement of aircraft, the source of
noise is also derived from the density of traffic in the area and around the airport. For noise from aircraft,
nothing else comes from the aircraft engine itself which is when the plane landing and take-off and
during operation in the adjacent time.
One effort to muffle the noise is planting trees that can reduce noise, one of them is the
Japanese Tree Bamboo (Bambusa japonica). Besides the need for regular medical check-up for
communities such that the effect of noise can be directly addressed. In addition, other noise sources are
vehicular traffic around the airport and outside the airport. This condition is expected to increase from
year to year.
In the construction of the runway extension and construction of landside facilities, the noise
source can be derived from work activities and occupations of land resettlement, runway, apron pavement
and construction of buildings. From the activity causing increased presence of vehicles in the area
surrounding the airport and roads, which is used as a mobility vehicle lane project and heavy equipment,
so the noise impact will be felt by workers in the airport surrounding area, especially people who live
adjacent to the route passing vehicle project. Therefore they need for regular medical check-up such that
the effect of noise can be directly addressed
MANAGEMENT OF NOISE IMPACT
Noise generated during construction activities based on mobillisasion of heavy equipment and
materials, including the movement of heavy equipment such as excavators and bulldozers on the project
activities. The recommended to engage in activities that potentially cause such noise between 07:00 am to
05.00 pm. Use ear-plug for the workers to avoid the noise of heavy equipment usage.
Operational Activity
Sources of noise caused of the increasing of traffic vehicles and aircraft movements. The
purpose of environmental management plan is to minimize the occurrence of noise caused by vehicle
traffic to and from the airport terminal for passengers, aircraft movements, loading and unloading of
goods. Technological approach is done by planting crops with plants that did not attract birds and does
not interfere with flight operations, has the function of ecological and aesthetic, and also able to reduce
the noise level in the area of the airport locations, such as Japanese Bamboo (Bambusa japonica);
implement Regions Noise Limit, which determined by the government by conducting field investigations
on the use of land around the airport; and implementing regulations concerning noise limits on
commercial aircraft (noise standards for aircraft to land and air).
MONITORING THE IMPACT OF NOISE
Noise generated during construction activities based on mobillisation of heavy equipment and
materials, including the movement of heavy equipment such as excavators and bulldozers on the project
29
activities. Collecting data to determine the level of noise was done by measuring the noise level, ie using
a Sound Level Meter and by conducting field observations. The analytical method used is to analyze data
from field measurements using the formula of Noise Monitoring Terminals WECPNL (Weighted
EquivalentContinous Perceived Noise Level). Location of noise monitoring was done along the transport
route and airport development plan such as apron, aircraft parking, passenger terminal, runway, and so on.
Map of environmental management and monitoring can be seen in Attachment 1.
Operational Activity
Noise generated during operations sourced from aircraft take-off and landing, the increase in
number of passenger traffic, movement of vehicles and aircraft in line with widening the operation of
runway. Collecting data to determine the level of noise was done by measuring the noise level, i.e. using
a Sound Level Meter as well as field observations. The analytical method used isto analyze data from
field measurements by using a formula based on the Noise Level and Standard Formulas WECPNL
(Weighted Equivalent Continuous Perceived Noise Level). Location of noise monitoring component
conducted in the vicinity of the airport development were four sampling locations, as has been stated in
the previous page.
CONCLUSION AND RECOMMENDATION
- The study shows that the noise level as an index of WECPNL at all points of measurement are within
the provisions of the noise index WECPNL region. Noise source also comes from vehicle traffic
activity, both in airport area and its surrounding due to residential areas and traffic on the roads around
the airport was also affected by development activities at the Djalaluddin Airport.
- Noise occurs during construction and operational activities of the airport. For that we need to evaluate,
manage and monitor, especially in the aspects of noise at airports in accordance with the standard noise
limit area.
- Tree planting (landscaping) which functions to reduce, i.e. plant the Japanese Tree Bamboo (Bambusa
japonica), installation of silencers on the terminal, the use of ear protectors for workers/technicians.
Also need for routine medical check-up for communities around airport such that the effect of noise can
be reduced.
- Setting the operational management at the airport and its surroundings which should be integrated with
spatial planning systems so that airports can function effectively and efficiently in the present time and
future.
- The utilization of green aircraft in the near future, which is quiter and cleaner, i.e. SAX 40 (Silent
Aircraft eXperiment, <25% quiter and less energy consumption).
- The use of quiter and low polluting ground vehicles for reducing noise and energy
consumption.
REFERENCES
Adisasmita, S.A. (2005), Air Travel Demand Forecasting and Passenger Flow Distributions at Airport
Terminal Building, Ph.D. Thesis, the University of Newcastle, Australia, 2005.
30
(2007), Design and Development of Hasanuddin International Airport based on Eco Airport System,
Tribun Timur Newspaper, Indonesia, 2007.
(2007), Hasanuddin Airport Management in Environmental Perspectives. Environmental Agency

Magazine, Makassar, Indonesia, 2007.
(2007), Team Leader, Consultant Report. AJAT-3 : ASEAN-Japan Airport Study Project. PT.
Konsultasi Pembangunan Atmadasa. Department of Transport, Directorate General
Air Tranport, Republic of Indonesia, 2007.
(2009), Team Leader, Consultant Report. Study on Management and Monitoring of Environment at
Gorontalo and Tarakan Airports. Ministry of Transportation, Directorate General of
Air Transportation, Republic of Indonesia, 2009.
Anonim. (2003), Eco-Airport Guideline, Aviation Bureau Ministry of Land, Infrastructure and
Transportation, 2003.
Ashford, N., & Wright, P.H. (1992), Airport Engineering (3rd ed). New York: A Wiley-Interscience
Publication, 1992.
David, C. (1995), The Impact of New Aircraft Development on the Design and Conctruction of Civil
Airports, Proc. Instn. Civ. Engrs. Transp., vol. 111, (1995), pp. 59, 63.
De Neufville, R. (1995), Designing Airport Passenger Buildings for the 21st Century, Proceeding of
the Institution of Civil Engineers, Transport, vol. 111:2, (1995), pp. 97, 101.
Ghobrial, A., and Kanafani, A. (1995), Future of Airline Hubbed Networks: Some Policy
Implications, Journal of Transportation Engineering, vol. 121:2, (1995), pp.
Website Narita Airport (http://www.narita-airport.jp).
Raguraman, K. (2001), Key Concepts and Issues in Airport Capacity Planning and Management,
Journal of Aviation Management, (2001), pp. 1, 131.
31
MAP OF ENVIRONMENTAL MANAGEMENT AND MONITORING
B C

A
3

2
2

E
1

3
I

H
OL
R S 1

PQ

2
M J K N
1

4 4

= Management & Monitoring

Vegetation to prevent noise

4
INCREASED USE OF OWN CAR IN TRIPOLI - LIBYA
Adel Ettaieb Elmloshi 1* & Amiruddin Ismail 2
ABSTRACT:
The own car is the first most common mode of transportation using in Tripoli. About 66 % Tripolis
resident prefer and used own car for all trips propose and 34 % whos prefer to use the private
transport such as private taxi and private buses namely (micro bus with capacity from 8 to 16 person)
in Tripoli city. The rapid increase in the use of own car model transport with the poor public transport
system in Tripoli as caused acute traffic jam, accidents, environmental impacts such as (air pollution
and noise) and land use damage for building more new car parks. In this paper, policy aimed at
discouraging the use own car and private transport in the Tripoli city. Also this paper research to
identify factors that prevent own car and private transport users from used public transport so that
logical policies could be formulated to support greater use of public transport. A survey was carried
out on users of own cars and private transport namely (private taxi and private buses) (n= 900).
Descriptive and analytical transportation issue is the main research methodology in this paper. The
most important variables found likely to support the use of public transport were reduced travel time;
reduce the distance from home to public transport stop and suitable fares. In conclusion for the
traveller to change from private vehicles to public transport, proper incentives need to be provided for
better performance.
Keywords: Own car, Private transport, poor public transport, Descriptive and analytical
1
* Engr. Adel Ettaieb Elmloshi, MSc. PhD Student, Department of Civil and Structural Engineering, Faculty of
Engineering and Built Environment, Universiti Kebangsaan Malaysia (U K M), adelelmloshi@yahoo.com,
adel0467@vlsi.eng.ukm.my Phone No. 0060173741296 Malaysia, Phone No. 00218913785296 Libya
2
Prof. Engr. Dr. Amiruddin Ismail, Head Sustainable Urban Transport Research Centre, Department of Civil and Structural
Engineering, Faculty of Engineering and Built Environment Structural Engineering, Universiti Kebangsaan Malaysia
(UKM), abim@vlsi.eng.ukm.my, abimismail@gmail.com, Phone No.0060389216203 / 12 Malaysia
33
INTRODUCTION
LIBYA LOCATION
Libya is located on the northern coast of Africa, bordering the Mediterranean Sea. It has
borders with Egypt to the east, Sudan to the southeast, Chad and Niger to the south, Algeria to the
west and Tunisia to the northwest. Libya has an area of approximately 1,759,540 sq km and
population of about 6,036,914. The religion of the people in Libya is Islam and the official language
is Arabic, for business purposes, some Berber language, English and Italian are also used. It has a
coastline of 1,770 kilometres. Dominated by the Sahara desert, most of the land is barren and
relatively flat. The capital city of Libya named Tripoli Figure 1 show the Libya location and Tripoli
position on Libyan coastal pelt (Libya country Profile 2007, 2008).
Figure 1: The location of Tripoli - Source (http://www.libyana.org/cities/index.htm)
PRIVATE TRANSPORT IN LIBYA
Private transport in Libya simply means shared taxis, buses and coaches. Private transport is
always assigned together with the required escort for those who arrive without transport. Currently
there are no trains in Libya. Libya has no railways since 1965, where the government was the
dismantling of this system, which was created in order to restore the occupation established. Bus and
taxi there are few bus services, mainly luxury, air-conditioned coaches between Tripoli and Benghazi
and other major cities. A minibus service operates between cities. Long distance taxi fares can be
quite expensive and should be agreed in advance. Taxis are available in large cities, and are often used
to share works under the supervision own union and majority of taxis are cheap and lease negotiation,
and prices specified in the fixed lines. Private taxis in the capital operate like other private taxis in
other countries. Private taxis are found only in large cities as well as the capital Tripoli.
The increasing numbers of motor vehicles in Libya, all kinds from private cars, private
transport, taxis and trucks and heavy equipments in the period between 1998 to 2008 an indication of
the severity of congestion on major roads between main and interior cities, which reached up at the
end of 2008 to more than of 2,000,000 vehicles. The vehicles total number is 2052679 except that of
government vehicles and the international organizations. The traffic statistics of the number of
vehicles registered in Libya until the end of 1998 (203575) reached up in the car to (2052679) car
until the end of 2008 or increased to about ten times as it was in 1998 and is distributed as following
the private transport of (367824), own car (1525429), taxi (77320), trailer and lorry (61124),
Motorcycles (873), heavy equipment (20109), (data as of 2008).
34
PRIVATE TRANSPORT IN TRIPOLI
In many capital cities in Arab countries including Tripoli travel demand management has
been receiving increasing attention for its potential to improve urban transport problems. Managing
urban travel demand is not only planning the transport system but also to get solution the issues of
traffic congestion, accidents and environmental pollution from the increasing use of vehicles.
Dominant mode of transport in Tripoli is the taxis, white and yellow, and that contact the city to other
cities as (Zuwara - Sabratha - Sorman Tajora - Gheryan,....) also taxi black and white, which operate
within the city and operated by the private sector and overseen by the Association taxi General, where
the taxi go and stop wherever referred to passengers without the presence of stations and parking on
the road. Tripoli also does not have bus lines in the city at present, although they were operating on a
regular basis period of the eighties through the public transport company. But in the latest of the
eighties when to dissolve the company stopped all the bus service in city.
METHODOLOGY
Descriptive and analytical transportation issue is the main research methodology in this study,
field survey collection of data from the Tripoli organizations which means references and previous
studies. Data collection and survey procedure, traffic counts were carried out on 5 entrance and major
streets in Tripoli city in October 2009. Traffic counts were collected for 10 hours per work day time
starting at 08 am till 18 pm in order to have most of the day traffic volume fluctuation. Eight complete
trips were made for five main entrances into Tripoli city to measure the travel time and speed for both
directions at the morning and afternoon peak period.
A questionnaires study was carried out in selected urban areas of the Tripoli city to determine
and investigate why people use private car instead taxi(for example cab and micro or mini bus) in
Tripoli city, suggest and support public transportation use in Tripoli. A total of 900 questionnaires
were collected over period of five months from (21 July to 15 December). This questionnaire was
categorized into 3 parts as: (1) Travellers personal information, (2) Trip characteristics and
respondents current which travel modes prefer, (3) Public transport option. Used Excel and SPSS
software in this study to descriptive and analysis data.
POPULARITY OF THE OWNCAR
Car is the most available one, figure 1 show the popularity of the own car. Which represent 88
percent of Libyan people prefer use his or her own car. Each Libyan person has own car or more. He
can transfer from place to place in safety way it takes a few times to transport from place to other.
Most the people like the car because they can feel freely to do, what are ever they want? So there are
many reasons why the people prefer to use own car and why the car is more popular than another
current modes use in Libya. It is reliable, comfortable, safe and convenient way to travel. The own a
car has become more popular and dominant than other current modes use in Libya, because it is
usually available when required, such as takes the user from door to door, and can reach to several
destination. Also the own a car represent the power status and prestige. The increasing number of own
car use has increased road congestion, environmental pollution and traffic accidents, table 2 show
latest statistic in 2008 of different vehicles in Libya.
35
Table 1: Show latest vehicles statistic in 2008 (in Libya)
Type of vehicle Libyan citizen Foreigner Total

citizen
Own car 1465776 59653 1525429
Private Transport 355145 12679 367824
taxis 77320 0 77320
motorcycles 808 65 873
Lorry 33446 3344 36790
Trailer 21365 2969 24334
Tractor 2743 240 2983
Heavy equipment 13206 3920 17126
Total 1969809 82870 2052679
Source: General People's Committee of General Security (data as of 2008)
8%
4% Car
Taxi
88%
Micro or mini
private bus
Figure 2: Popularity of own car
TRAFFIC ACCIDENT IN LIBYA
Traffic accidents have increased dramatically in Libya and the deaths will reach up to 25 cases
a day in the next ten years if there is no real measures as duplication and separate roads and the roads
and lighting control on the quality of vehicles and the introduction of high import. Although the men
are more likely to suffer a fatal injury than women (men accounted for two thirds of the total number
of deaths due to injury in 2008), injuries are a leading cause of death for both sexes and in all age
groups. The age groups affected in the road deaths accidents were from the age of 25 years until 35
years and who are the pillar, hope and the future support of the country.
Libya is one of the most affected countries because of traffic accidents damage and human
deaths, the bodies, and social and physical property, it is clear to us that lead to the death of 5 persons
per day according to the (Secretariat of the Libyan justice) and this is a very large number in a
population that does not exceed seven million (Office of traffic and licensing Tripoli (data as 2008).
and through the traffic statistics for the proportion of road traffic accidents and the resulting that there
is a direct relation between the proportion of the increasing numbers of vehicles and the increase of
the proportion of road traffic accidents, who is also reached up at the end of 2008 to 126536 accident,
resulted about of 18946 people dead, about 52915 people major injuries, 66707 people in minor
injuries, reconciliation and damage accidents is 42592 accidents, by total cost is 189665666 Libyan
Dinar.(1.0 U.S. dollar equivalent 1.3 of LD).
36
Table 2. Represent the road traffic accident in Libya between years (1998 2008). Although
table 3 shows the men are more likely to suffer a fatal injury than women (men accounted for two
thirds of the total number of deaths due to injury in 2008), injuries are a leading cause of death for
both sexes and in all age groups.
Table 2: Road Traffic Accidents in Libya (1998 2008)
Year Number of Vehicles involved in deaths Hared Simple Damages

Accident road Accidents injured injured accidents
1998 9393 10531 1224 3573 4770 3414
1999 9370 10574 1204 3515 4879 3267
2000 10667 12271 1504 4155 5462 3609
2001 10895 12528 1598 4141 5892 3505
2002 12017 13952 1751 4483 6575 3965
2003 12154 13938 1744 4479 6023 4266
2004 11643 14171 1785 5150 5596 3895
2005 11898 14620 1800 5380 6161 3954
2006 11982 16919 1866 5663 6501 3894
2007 13165 17859 2138 5950 7547 4553
2008 13352 18662 2332 6424 7301 4270
Source. General People's Committee of General Security (2009)
Table 3: Deaths resulting from traffic accidents (cases) by sex and age groups in 2008
Age Gender Drivers Passengers Pedestrian Total Age Gender Drivers Passengers Pedestrian Total
Groups Groups
(-)- Male 0 7 19 26 35 - 44 Male 137 25 58 224
04
Female 0 8 10 18 Female 12 40 17 69
05 Mal 0 17 47 65 45 - 54 Male 76 20 49 148
09
Female 0 14 21 35 Female 08 30 27 65
10 - 14 Male 0 17 42 61 55 -64 Male 44 10 33 87
Female 0 14 10 21 Female 0 22 19 41
15 Male 26 29 52 115 65 - 74 Male 27 06 33 66
19
Female 03 15 14 32 Female 0 02 05 07
20 Male 162 87 75 344 75 ( Male 04 02 16 22
24 +)
Female 13 38 20 71 Female 01 04 02 07
25 - Male 399 166 214 804 Total Male 875 386 638 1962
34
Female 38 109 42 189 Female 75 293 187 555
Source. Secretariat of the Libyan justice (2009)
TRIPOLI CASE OF STUDY
Tripoli is Libya's largest city and port, and is the country's capital. Meeting-place of the
People's Congress, it is known in Arabic as Tarabalus Al-Gharb, or Tripoli of the West. Tripoli is
situated on the shore of the Mediterranean Sea, north of the Equator at latitude (32.560), and longitude
(13.100), east of Greenwich. The importance of its location made Tripoli one of the best economic and
37
commercial centres in North Africa (Libya Nation Atlas 1978) as shown in Figure 1. Tripoli
Population is 1,682,000 people with Area of 400 square kilometres and population density 2207.32
people / sq km (Libya country Profile 2007, 2008). More than twenty years ago there were only a
handful of cars in the Tripoli, today the streets are bumper to bumper with modern vehicles, own cars,
taxis and different types of micro buses. The continuing growth in vehicle ownership and lack of road
spaces and public transport, many Tripoli streets and roads network started to suffer from various
traffic problems.
The number of vehicles increased approximately to 705025 private cars, 43740 taxis, and
lorry from one to three tons 120599 vehicles, vehicles heavy load 29343 vehicles. Then the total
vehicles use the road of Tripoli every day are 898707 vehicle, at peak period from 7 am to 7pm, there
is continual congestion, traffic congestion is frequently a problem through weekdays.(data as of
2008). Tripoli is growing very fast for the last ten years, both geographically and administratively.
More and more policies and regulations are being developed at Tripoli level, particularly in the field
of traffic and transport. The free movement of people, goods and services is a key priority of the
Tripoli city. The survey was carried in Tripoli due to the high number of private car ownership and
poor public transport services together with high congestion at selected area as shown in Figure 2. 3T
Figure 3: Shows the study area

Source: Google Earth 2009
An enormous influx of the Libyan populace into the Tripoli resulted in a five-fold increase in
Tripoli's population during the seventies and eighties. To accommodate this huge increase, many new
suburbs sprang up all around, and a lack of planning resulted in a sprawling metropolitan area with
severe traffic congestion at peak times. Traffic congestion occurs when a volume of traffic or modal 2T
split choice generates demand for space greater than the available road capacity. There are a number
2T
of specific circumstances which cause or aggravate congestion; most of them reduce the capacity of a
road at a given point or over a certain length, or increase the number of vehicles required for a given
throughput of people or goods.
SELECTION OF THE STUDY AREA
The survey was carried in Tripoli because it has high number of ownership cars, private
transport vehicles and poor the public transport, and because the congestion of this selected area seem
to come from all over Libyan cities. For the reasons stated above Tripoli was expected to be the best
case study representing Libyan cities. Table 4 shows the number of registered own cars and several
vehicles to 31 \ 10 \ 2009 in Tripoli city.
38
Table 4: Show latest vehicles statistic untill the thirty first of October 2009 (in Tripoli)
Type of vehicle Libyan Foreigner Total

citizen citizen
Private 713258 30538 743796
Private Transport 118256 5301 123557
taxis 47031 0 47031
motorcycles 543 55 598
Lorry 11944 1589 13533
Trailer 7509 791 8300
Tractor 476 119 595
Heavy equipment 6170 1678 7848
Total 905187 40071 945258
Source: Traffic office and licensing of Tripoli, data as of 2009
Tripoli has over more three times the number of own cars than second Libyan capital city at
eastern called (Ben Ghazi). And three times the number of private transport namely (taxi and micro
buses), table 5 shows the number of own car and private transport in some Libyan cities. The study
was carried out in the main of five Entrances into the city, such as the southern entrance for each of
the Al-Gabis Entrance, Airport Road Entrance and the Al-Hadaba (Salah Al-Din) Entrance, the Al-
Giran Entrance in the west side and the Tajora (Al-Bive) Entrance at east of city. And flowing
through several downtown roads and main streets in the city, such as Omar AL-Mukhtar Street two
direction and Mizran street and the First of September street in the city centre and road two way along
the Tripoli city port named Al-Fatah street, and the old and modern streets in city centre, figure 2
above shows the main routes in study area.
Table 5: Registered own car and private transport numbers in Libya cities (2009)
City Own car Private transport ( City Own Private transport (

taxi & micro buses) car taxi & micro
buses)
Al beida 29394 711 Darnah 22815 3262
Misrath 65712 1339 Algoba 2462 61
Zalitan 59652 1532 Tobruk 15773 759
Zawiah 65728 2695 Ajdabiya 13604 474
Tripoli 1 713258 47031 Almarj 14066 226
Alkhums 33263 1347 Aljabal 3164 94
alakhdar
Tarahuna 17432 869 Nalut 9009 263
Surt 24394 407 Gahrian 49805 1088
Benghazi 196668 12420 Ghafrin 12973 238
2
Sabrata 35132 1410 Sahel Algafara 41600 959

Surman 18971 771 Tajora 15689 1685
Source: General People's Committee of General Security (2009)
39
REASONS FOR CHOICE TRIPOLI AS CASE OF STUDY AREA
Tripoli is taken as an illustrative case study for several reasons for Tripolis traffic problems
such as
Tripoli is the capital city of Libya and lively greater area.

Tripoli has a very high natural population growth.
Approximately 40 % of the total Libyan population resides in Tripoli.
One of the main causes of congestion is rapid growth for the number of private cars.
Poor public transport infrastructure such as (no train, light rail, metro, public buses and car
parking).
The dramatically increase in the number of taxis and micro buses in Tripoli centre.
The road system is another cause of problem and the roads are poor developed.
The port that handles the first largest volume of cargo in Tripoli city.
NUMBER OF CAR OWNERSHIP PER HOUSEHOLD
Car ownership by the household is present 100 % the best mode in Tripoli especially and in
Libya generally and majority of Libyan people depend on the own car in travel. Figure 4 shows the 49
% had one to two car in their house, 31% had 2 to 3 cars in their house and 11 % had from 3 to 4 cars
in house, also had responded by 9% non car and used other transport choice such as taxi and micro
bus.
11% 9% 1 to 2 cars
49% 2 to 3 cars
31% 3 to 4 cars
non car
Figure 4: Household Car ownership
Total Traffic Volume Using Own Cars on main entrances:
Traffic volumes along peak hours reflect that the behaviour traffic volume occurred between
8 am and 18 pm, and reached 187366 vehicles entered Tripoli from five main entrances. It should be
noted that during the traffic count survey, ministry of general security prevented the trucks to use the
Airport Road entrance. Thus the percent of trucks presented in this study are lower than the expected.
Traffic Speed Survey and Level of Service
Eight complete trips were made for from Tripoli city centre (Green Square) to Al-gabis Road
Entrance, Al-Fath University, Al-Giran Road and Umm- Aitiqah Airport. To measurement the travel
time, delay and speed for both directions at the morning and afternoon peaks. The floating car method
was used. The results of this survey would help in determining the level of service on each street in
this study and could be seen in table (6). Table 6 show average running speed during morning and
afternoon peak hour in 8 trips started from Green Square to each access point for and ended in Green
Square.
40
Table 6: Average Running Speed during Morning and Afternoon Peak Hour
Distance Direction Running Speed (km/hr) Average Running speed

Street (km) From, To (km/hr)
Morning Afternoon Morning Afternoon
period Period period Period
10.200 From 30 23 26.33 19
Green Al-Gabis 28 19
Square 21 15
10.200 To 35 22 28.67 20.33
Al-Gabis 29 21
22 18
7.100 From 18 20 18 17.67
Green AL-Fath 20 18
Square university 16 15
7.100 To 16 18 17.33 16.33
AL-Fath 18 15
university 18 16
13.900 From 20 29 19 31.67
Green Al-Giran 18 31
Square 19 35
13.100 To 25 30 24.67 29.33
Al-Giran 24 29
25 29
08.00 From Umm 45 39 42 39.67
Green Aitiqah 43 40
Square 38 40
08.00 To Umm 40 35 43.33 36.67
Aitiqah 45 35
45 40
Table (6) shows the average running speed for Al-gabis Road Entrance, Al-Fath University,
Al-Giran Road and Umm- Aitiqah Airport in both direction during morning and afternoon peak hours.
There are number of factors such as safety, freedom to maneuver, number of interruptions,
comfort, convenience, economy, and speed which can illustrate the level of service on any section of
roads, therefore; only two features have been incorporated in this research in order to estimate the
existing level of service :
Operating speed
Ratio of traffic volume to capacity
It is worth mentioning that both of these two features were determined during the peak
periods. In fact, this will help in the determination of the level of congestion during the peak hours for
Tripoli streets.
41
CONCLUSION
Understanding travel behaviour and the reasons for choosing one transport mode over another
is an essential issue. However, travel behaviour is more complex. For each trip, commuters have the
choice between different modes of transportation. Each mode is having specific characteristics, such
as advantages and disadvantages depended on travel time and travel cost. There are some people that
use private transport namely (micro bus and private taxi) and own cars to reach their works or studies.
The factors behind conducting this research include the estimation of traffic problems and congestion
of traffic are and to establish suitable models to reasonably describe commuters attitude in Tripoli.
The own car has become more popular and dominant than other current modes use in Tripoli,
because it is usually available when required, such as takes the user from door to door, and can reach
to several destination. Also the own a car represent the power status and comfort and safety. The
increasing number of private cars use in city caused increased road congestion, environmental
pollution and traffic accidents.
REFERENCES
Alrabotti Fathi Bashir (2006) Traffic Lights, part one (Tripoli)
Alrabotti Fathi Bashir (2007) Traffic Lights, part two (Tripoli)
General People's Committee of General Security (data as of 2009)
Libya country profile, Published: 13 November 2007, last updated: 21 September 2008
Secretariat of the Libyan justice, general traffic department - office and licensing of Tripoli,
(Data as of 2008)
Secretariat of the Libyan justice, general Traffic department - office and licensing of Tripoli,
(Data as of 2009)
Tripoli location - Source (http://www.libyana.org/cities/index.htm)
42
INFLUENCE OF BLENDING INTERACTION OF CRUMB RUBBER MODIFIED BITUMEN

ON PAVEMENT PROPERTIES
Nuha S. Mashaan 1*, Mohd Rehan Karim 2,

Asim Hassan Ali 3 and Mahrez Abdul Aziz 4
ABSTRACT:
The application of crumb rubber modifier in asphalt mixture is intended to improve the properties of
asphalt binder by reducing the binders inherent temperature susceptibility. Therefore, during the
interaction with asphalt binder, the CRM particles absorb a portion of the oils in asphalt binder and the
particles swell; lead to increase the viscosity and stiffness of the CRM binder. Yet, the performance
properties of CRM binders in hot mix asphalt pavement are unclear due to various interaction factors of
CRM with asphalt binders; the effect of CRM percentage, and the blending mix conditions. In this study,
a laboratory investigations was conducted on the asphalt binder properties as a function of various
blending treatments (temperatures and times) ,and the CRM percentages. A total of ten CRM binders ( 2
CRM blending times , 1 CRM blending temperatures, 5 CRM percentage by weight of binder) were
produce through an accelerated un aged binder process by using the three binder tests ( Penetration test
,Viscosity test, Softening point test). The result from this study showed that (1) blending interaction times
and blending interaction temperatures for CRM binders were observed to have effect on the binder
properties; (2) The longer time and high temperature for interaction of CRM binders results in an
increase in viscosity, and softening point. (3) The CRM percentage lead to decreased the penetration,
which would enhance CRM binder stiffness, and asphalt binder consistency and resistance to flow , which
led to improve rutting resistance .However, this study found that control binder of 80/100 grad
penetration had little change of the binder properties as a function of interaction blending conditions.
Keywords: Bitumen Binder, CRM binder, Viscosity, Blending Effect, Rutting.
1
Student, Universiti Malaya, Jalan Petalling Jaya, 50603 Kuala Lumpur, MALAYSIA
2
Lecturer, Universiti Malaya, Jalan Petalling Jaya, 50603 kuala lumpur , MALAYSIA
3
Student , Universiti Malaya, Jalan Petalling Jaya, 50603 Kuala Lumpur , MALAYSIA
4
Lecturer, Universiti Malaya, Jalan Petalling Jaya, 50603 Kuala Lumpur, MALAYSIA
*Correspondence Author: Nuha S. Mashaan,, Universiti Malaya,, Malaysia.

. E-mail: nuhaasim@yahoo.com
43
INTRODUCTION
There are several distresses of road pavement related to bitumen properties. Since bitumen is viscous/
elastic material, its rheological properties are very sensitive to temperature as well as rate of loading. With
respect to temperature, the most frequent problems of road pavement are rutting and fatigue cracking
.Asphalt rubber binder use in road materials applications is getting increasing interest from the paving
industry since the 1960s.
Asphalt rubber blend produced according to the wet process, are consider as a good alternative to pure
bitumen and were investigated. The used of recycled rubber particles in bitumen binder applications is an
environmentally efficient valorisation of the reclaimed material. Unlike using tire rubber as fuel for
cement kilns, asphalt rubber binder take advantage of the exceptional elastic properties of rubber while
allowing further recycling of the asphalt mixture at the end of lifespan.
LITERATURE REVIEW
Researches and applications of CRM asphalt binders in United states showed that the CRM asphalt
binder has many advantages characteristics like improved resistance to rutting due to high viscosity, high
softening point and better resilience, improved resistance to surface initiated, reduce fatigue/ reflection
cracking, reduce temperature susceptibility, improved durability and lower pavement maintenance costs,
and saving in energy and natural resource by using waste products (Shtang et al.,2009). The increasing
use of CRM in asphalt pavement requires better understanding of its effects on the physical, chemical,
and performance properties of CRM binders. In general, the addition of CRM to asphalt binder is
intended to improve in binders properties such as reducing the inherent temperature susceptibility of the
binder. The improvement of the properties of CRM binders likely depends on the interaction between
CRM binders and asphalt binders where CRM particles swell in the binders to form a viscous gel;
resulting in an increase in the viscosity of the CRM binders (Aiery et al., 2003). However, the properties
of CRM binders at a wide range of temperature are consider to be some what unclear due to the various
interaction effects of CRM with asphalt binder, largely depending on the chemistry of the asphalt binder
,the CRM percentage, particle size, texture service of CRM and the blending interaction temperature and
time (blending conditions mix).
From previous literature, it was concluded that swelling of the rubber particles due to the absorption of
the light fractions into rubber particles, and stiffening of the residual binder phase are the main
mechanism of the interaction (Aiery et al., 2003; Shen and Amirkhanian, 2005). The control binders of
grade 80/100 penetration were mixed at the same blending conditions as well as ,the five CRM
percentages binders, without adding the crumb rubber.
METHODOLOGY
Materials and Productions
One asphalt binder grade 80/100 penetration was used in this study. This asphalt binder has wide use in
different areas especially in Malaysia. Table 1 shows the properties of the base binder included in this
study.
The CRM produced by mechanical shredding at ambient temperature was obtained from one source, size
30 # (0.6 mm) and used with gradation as showed in table 2., which is widely used to produce the CRM
mixture in the United States. To ensure that the consistency of the CRM was maintained throughout the
study, only one batch of crumb rubber was used in this study. The binder mixing used in this study was
the wet process, in which the CRM is added to the base asphalt binder before introducing it to the
aggregate. The CRM binder was manufactured in the laboratory by an open blade mixer at a blending
speed 200 rpm. This mixing speed matches the field practices used to produce field mixture. The crumb
rubber was added to the base binder using two interaction blending times (30min and 60 min.), and
44
interaction blending temperatures (160C ) ,and five CRM percentages (0,4,8,12,16,and 20% by binder
weight). The two blending interaction times and the blending interaction temperatures were decided as
different agencies use different CRM percentages. To evaluate the property difference between CRM
binders and control binders, the control binder (bitumen grade 80/100 penetration) was also mixed at the
same interaction blending times and temperatures without adding the crumb rubber modifier.
Table 1: Table of the properties of control binder in this study
Test properties Test result
Viscosity @135 C (pas) 0.65

G*/ (sin ) @ 64C (kpa) 1.35
Ductility @ 25 C 100
Softening point @ 25 C 47
Penetration @ 25 C 88
RESULT ANALYSIS
Effect of CRM Content
Binder Results:
The results of all three binder tests including penetration test , softening point test, and Brookfield
viscosity show in figures 1- 3 as function of both blending time and crumb rubber content.
160c,30min. 160C,60min.
Penetration (mm)
100
80
60
40
20
0
0 4 8 12 16 20
CRM percentage %
Figurer 1: Penetration of CRM binders as function of blending time and rubber content.
The penetration was deceased as crumb rubber increase and blending time. The effect of blending time
has been less on control binder and more effective on crumb rubber content 16% and 20%.
45
C.M.160C,30min C.M.160C,60min
Sotening Point (C)

80
60
40
20
0
0 4 8 12 16 20
CRM percentage %
Figure 2: Softening point results of CRM binders as function of blending time and rubber content.
The softening point results have similarity to penetration, which increased as crumb rubber content
increased with longer blending time as shown in figure 2. This increase in softening point has effect on
rubber binder behavior under high temperatures, which mean binder with higher softening point the less
susceptibility to temperature change during service life.
160C,30m in. 160C,60m in.
5000
viscosity m.pass
4000
3000
2000
1000
0
0 4 8 12 16 20
% of CRM
Figure3: Brookfield Viscosity results of CRM binders as function of blending time and rubber content
at 135C.
According to Brookfield viscosity the increase in crumb rubber content lead to significance increases in
viscosity and at 20%. Specifications include a maximum viscosity limit (3 pas) for an unaged
binder at 135C( The asphalt institute, 2003). Thus increase in viscosity can explain that rubber
binder has more swelling particles due to absorb some of oil of bitumen in interaction process between
bitumen and crumb rubber particles. Figure 3 shows the effect of rubber content on viscosity been more
obvious than the blending time function. This results give a good indicator of increase binder ability to
resist rutting, as be less susceptibility to temperatures and more stiffness and more flexible at high
temperatures service in field.
46
CONCLUSION
In conclusion, to investigate the characteristics of binders due to the interaction with the crumb rubber,
CRM binders were produced in the laboratory, using two blending times, , and five CRM percentages. In
addition, a control binder of PG 80/100 grade was mixed using the same blending conditions without
adding the rubber. A series of Brookfield viscometer, softening point, and penetration were conducted.
From these results, the following conclusions were drawn:
(1) The control binder of PG 80/100 penetration grade was found to have little change of the viscosity at
135 C , penetration at 25 C ,and softening point at 25 C.
(2) In general, the longer blending time and higher content of CRM modified binder seemed to lead to an
increase in the viscosity at 135 C, which is related to the increase in the rubber mass through binder
absorption, and improve in binder stiffness through low in penetration and high in softening point.
REFERENCES
Shtang, l.,Weidong, C. , Jianguo, F. ,Shujie, S. (2009).Variance analysis and performance evaluation of

different crumb rubber( CRM) asphalt. Construction and Building Materials. 23. 2701-2708.
Airey GD, Rahman MM, Collop AC.(2003). Absorption of bitumen into crumb rubber using the basket
drainage method. Int J Pavement Eng. Vol. 4 (2): 105-19.
Shen, J. ,Amirkhanian, S.(2005).The influence of crumb rubber modifier (CRM) microstructure on the
high temperature properties of CRM binder. Int. J. Pavement Eng. Vol.6 (4):265-71
The Asphalt Institute. (2003). Perfomance graded asphalt binder specification and testing,SP-1.
Lexington(KY).
ASTM 2000, Standard Test Methode for Penetration of Bitumminous Materials, D5-97, Annual Book of
Standards, Vol.04.03,Eston,MD,USA,2000.
ASTM 2000, Test Method for Softening Point of Bitumen (Ring and- Ball Apparatus) , D36-06,Annual
Book of ASTM Standards,Vol.04.04,Easton, MD,USA,2000.
ASTM 2000, Test Method for Viscosity Determinations of Unfilled Asphalts Using the Brookfield
Thermosel
47
48
PRELIMINARY ASSESSMENT ON RELIABILITY OF PUBLIC BUS SERVICE IN KOTA

BHARU
Madzlan Napiah1, Noorfakhriah Yaakub2
ABSTRACT:
Reliability of public bus service is a crucial factor in determining the effectiveness of the
operation. Reliability can be measured in many ways; one of them is the Punctuality Index where bus
performance is measured whether the actual departure of certain bus is according to the scheduled
departure time. The use of public transport will help reduce traffic congestion, save money and time, as
well as saving the environment as less pollution is emitted to the air. In Stockholm, Sweden, the citizens
in 10 km radius tend to cycle to their office. Personal cars are kept in garage because the community
enjoyed using bicycle or rather use public transportation. The issue here is that the three parties involved
in public transportation are arguing about the state of service and performance of the public bus in
Malaysia. The parties are the authority, the bus operator, and most importantly the users. This paper
investigates the Punctuality Index of public bus service in Kota Bharu, Kelantan where the objective is to
assess the state of reliability of the public bus service. In Kota Bharu, there are no commuters or train
services that provide full coverage of transportation such as in Kuala Lumpur and Klang Valley.
Therefore, the public bus demand is very crucial in this city. The research will be conducted using on
board method where surveyor will ride on buses and record the required data. During the ride, the
surveyor can observe passenger profile and reasons of delay as well. This paper only presents the
preliminary of conceptual framework from the study, namely the routes, schedule, travel time, and
observations of the public bus service in Kota Bharu. The actual results will be used to suggest plans of
improvement so that the public can experience better service in Kota Bharu.
Keywords: Public Bus Performance, Punctuality Index, Public Bus Service, Reliability
1
Lecturer, Universiti Teknologi PETRONAS, Bandar Seri Iskandar, 31750 Tronoh, Perak, MALAYSIA
2
Student, Universiti Teknologi PETRONAS, Bandar Seri Iskandar, 31750 Tronoh, Perak, MALAYSIA
*Correspondence Author: Madzlan Napiah, Universiti Teknologi PETRONAS, Malaysia. Tel: +605 368 7281
Fax: +605 365 6716 E-mail: madzlan_napiah@petronas.com.my
49
INTRODUCTION
Due to the rapid growth of urbanization in Malaysia, there existed an increase in economic activities
which led to the increased trips made and longer distance. Therefore, many users choose to use private
transportation in order to improve their access to urban services, and before long, existing road capacity is
pushed to its limit resulting in severe traffic congestion.
The most popular public transport around Malaysia is the public bus. In Kota Bharu, Kelantan,
and Penang, where the train commuter service is not available, the public bus is the major alternative for
its population to move, and to avoid stresses due to traffic congestion during peak hour in the morning,
during lunch break, and after office hour. Moreover, the public bus has wide destination coverage and the
routes seem well-planned and organized. Rapid KL and Rapid Penang are examples of good public bus
service in Malaysia.
Problem Statement
Public bus performance is claimed by the bus operator as satisfactory, while the government stated that
the transport infrastructure in the country is ready to meet the needs of the 21st century expectation. On
the contrary, the public is frustrated with the present situation of the public bus system in Malaysia.
There is a need to determine whether the public bus performance is up to standard or below it so
that the service can be improvised to meet the requirements and satisfy all parties involved. The output
can be viewed as a mechanism for producing improvements desired by the society (Garber, 2002). Lack
of reliability results in discomfort for public bus user, thus reduce its competitiveness to personal vehicle
use. This paper investigates some of the measures that can be used to assess reliability, namely the
punctuality index PI as well as the on-time performance.
The area chosen for assessment is Kota Bharu, the capital city of Kelantan which is rapidly
developing as a main attraction for investment and tourism as the city denoted on maintaining the
traditional culture while asserting modern attributes. The monopoly of bus operator (Cityliner) also
signifies that service reliability and integration is easier to achieve compared to other urban municipalities
which usually has more than one private bus operator.
Objectives of Study
This paper is a conceptual framework and a preliminary survey to establish parameters and study
area. This paper does not present analysis and results yet. The objectives that can be extracted from this
study are:
1. To determine the punctuality index for Kota Bharu city bus service
2. To determine user comfort and problems arise in the Kota Bharu city bus service
3. To conclude the state of reliability for Kota Bharu city bus service
4. To recommend solutions for arising problem in the Kota Bharu city bus service
5. To advise new route or necessary schedule for the Kota Bharu city bus service
Scope of Study
This study is mainly conducted in Kota Bharu as the centre of population density in Kelantan. On board
method is done instead of off board because the reasons for delayed buses or missed trips can be
determined. The on board method will be done on buses serving in Kota Bharu City only which will
narrow down the scope to six (6) bus routes.
50
LITERATURE REVIEW
Current state of Malaysias public transport especially public bus is perceived of unable to meet citizens
needs satisfactorily. Many claim that if the public transport is being improved, they are willing to use it
instead of paying for outrageous petrol prices and tolls and ending up stuck in traffic jams (Kambala,
2007).
Before the rapid growth in the number of private vehicles, the percentage of public bus user in
Klang Valley is higher; 34% in 1985, and later 20% in 1997. By January 2010, the percentage decreased
to 10%. The Malaysian Government introduced the Government Transformation Programme (GTP) in
accordance with the principles of 1Malaysia, People First, Performance Now to improve the public
transport in certain populated areas. In June 2010, the percentage of public bus user in Klang Valley has
increased to 12%. The governments target is to achieve 13% modal share by 2010 and 25% by 2012
(GTP Roadmap, 2010). Figure 1 shows the Public Transport Users Modal Share in Klang Valley until
2010, and its targeted modal share by 2012.
Figure 1: The percentage (modal share) of public transport user in Klang Valley by year. Broken line
indicates the target to be achieved by end of 2010 (13%) and 2012(25%).
Public Bus Service in Kota Bharu
Kota Bharu is the capital city for Kelantan state located in east side of West Malaysia (see Figure 2). The
city expands as 403 square km, consisting of 65 sections with population of 500,000 citizens (Statistics,
2010). In Kota Bharu, the public transports available are buses, taxis, and the bicycle trishaws. Train
services are only available in smaller towns at the other side of Kelantan River
(http://www.mpkbbri.gov.my/).
CitylinerKelantan Operations mainly focused on stage bus operations, the services which allow
passengers to board and alight buses at a number of designated stops along a route. In Kota Bharu, stage
bus services utilize two types of buses; one type is the normal bus with 44 seats, and the second type is
the smaller-sized 25-seater mini-buses (Jamilah Mohamad, 2005).
Prior to Konsortium Transnasional Berhad (KTB), the CitylinerKelantan Operations were run by
Syarikat Kenderaan Melayu Kelantan (SKMK), a subsidiary of Kumpulan Kenderaan Malaysia Berhad
(KKMB). Being the only bus operator, CitylinerKelantan Operations has a sole the responsibility of
upgrading its services for the local population because the citizens totally rely on the company for public
bus service integration and performance (http://www.cityliner.com.my/), especially after 2005 when the
status of the Kota Bharu Municipal Council (MPKB) was upgraded to Bandaraya Islam (MPKB-BRI).
51
Figure 2: Kota Bharu as the capital city of Kelantan state (Department of Survey and Mapping Malaysia
JUPEM, 2005)
Reliability of Public Bus Service
Existing studies has been done using various measures, and factors influencing the reliability of bus
service had also been widely analyzed. Reliability has been defined in a variety of ways. In normal
language, reliability means something we can depend on where it will always give the same outcome. For
example, an equipment, machine, or system is considered reliable when it performs its intended or
required function or mission, on demand without degradation or failure (http://www.
businessdictionary.com/definition/reliability).
Reliability of transportation system can be defined as the ability of the system to adhere to a
schedule or maintain regular headways and a consistent travel time; in other words, as the on-
time/punctuality performance and headway evenness (Chen, 2009). For Malaysian Ministry of
Transportation, the reliability of public transport is defined as the percentage of journey times completed
within 60 minutes during the morning peak period (GTP Roadmap, 2010). Reliability is well attested for
its importance in attributing the level of service quality for transportation (Bates, 2009).
Liu (2008) presented factors affecting reliability classification into the following groups:
1. Traffic Characteristics, such as traffic composition in travel demand and traffic congestion levels.
2. Route Characteristics, such as the length of route, number of lanes, location of the bus stops, and
provision of bus lanes.
3. Passenger Characteristics, which include variability passenger volumes at stops, passenger route
choice and their arrival distribution.
4. Bus Operational Characteristics, such as scheduling system, number of staffs, fleet availability and
maintenance, and ticketing system.
52
Chen (2009) has used performance parameters consisting of Punctuality Index based on routes
(PIR), Deviation Index based on stops (DIS), and Evenness Index based on stops (EIS). Sampling survey
techniques were used to estimate service reliability at the stop, route, and network levels in a case study in
Beijing. Effects of route length, headway, distance from a stop to the origin terminal, and the use of
exclusive bus lanes were considered on service reliability.
Gershenson (2009) used equal headway instability as the performance parameter and published
the two causes of losing equal headways; vehicles go faster than expected, and the other one is that
vehicles go slower than expected. In principle, when vehicles go faster, the problem can be rectified by
forcing them to wait at stations until their expected departure time comes. As for slow vehicles, the issue
is more complicated as most cases are external, including heavy traffic and passenger behaviors.
Friman (2004) concluded that the people satisfaction when using public transport service is
influenced by quality improvement or reliability of the service. Punctuality of the public transportation is
an important factor which has major impact on satisfaction of users.
Punctuality Index Indication
Nakanishi (1997) proposed reliability measures in terms of on-time performance (OTP) indicator as well
as service regularity in New York City Transit (NYCT). The NYCT defined OTP as the percentage of
trips departing from all scheduled time points, not including terminals, between 0 and 5 minutes after
their scheduled departing time.
In this paper, the Punctuality Index of a bus service is a factor indicating the time gap between
scheduled and actual arriving and departing time for a bus, or also called the headway adherence. The
Traffic Capacity and Quality of Service Manual (TCQSM, 2003) suggests that both OTP and headway
adherence can be used as a measure in reliability indication.
Therefore the formulation suggested for Punctuality Index is as shown in Equation 1 and
Equation 2 (Kho, 2005):
2
S
P= (1)
h
where P is punctuality index,

h is scheduled headways,
S is the variables with following formula
1 I
S2 = (t i i )2 (2)
I i =1
where I is the number of operations,

t 1 is the actual arrival time of i-th bus operation,
1 is scheduled arrival time of i-th bus operation
The Punctuality Index value will then be converted into percentage value as in Equation 3:
P% = (1 P ) 100 (3)
53
METHODOLOGY
The project started with selection of topic in interested area of Transportation and Highway Engineering.
The area chosen is Public Transportation. After thorough background studies and problem statement on
the proposed area, the title of research is then determined using selected bus operator. In this case, the bus
operator chosen is the Cityliner Kelantan Operation which is the only public bus service provider in
Kelantan State.
Next is literature review where documents such as journals, thesis, published papers, and manuals
are accessed to understand the topic better. Secondary data are gathered from the bus operator
management as well as the information board at bus stations. Secondary data here mean the schedule of
bus trips and the map of journey of each city bus trip ID.
From there, data collection is done using On Board Method, also known as the Ridership or Ride
Check Method. Questionnaire surveys are also distributed to bus users and observers who might not use
public bus for certain reasons. Analysis will then be done according to collected data, and results of the
research will be used to recommend methods so that the public bus service in Kota Bharu can be
improved.
On Board Method
Survey is done using on board method instead of off board method. The difference between on
board and off board is that during off board method, the surveyor will do observation of arriving and
departing buses at the station or one of the stops, while for on board method; the surveyor will do
observation on arrival and departure time at stops from the bus ride. Off board method will record more
bus trips per period of time compared to on board but using on board method, the surveyor can observe
the passenger profile and reasons of delay.
The required data are:
i) Bus schedule: Start time and end time are acceptable

ii) Number of trips for each route
iii) Frequency of bus or headways
iv) Distance from bus station to each stops and end of the route
v) Arrival and Departure time from station and bus stops
vi) Number of designated bus stops and undesignated stops where commuters frequently alight and
aboard.
vii) Traffic, route, passenger, as well as bus and driver characteristics.
Questionnaire
Questionnaires are distributed to the bus passengers to get the overall passenger view of the public bus
service in Kota Bharu. The question should include users personal information as well for profiling
purposes.
Use of GPS Instrument
Global Positioning System method will be used to determine the distance from terminals to each stops
and destination.
54
PROGRESS
Secondary Data
There are six bus routes on city bus trips in Kota Bharu. Figure 3 shows the Stesen Hilir Pasar, the main
station that caters for all bus trips in the state of Kelantan and Figure 4 shows Stesen Buluh Kubu, the
previous main station that is now being used to cater for three bus trips only, which are bus 10, 61A, and
61B. Table 1 shows the available routes for city bus service in Kota Bharu.
Figure 3: Hilir Pasar bus station (main station). Figure 4: Buluh Kubu bus station
Table 1: Bus routes in Kota Bharu

No. Station Trip ID Bus Route Route Type
1 4 Kota Bharu - Kompleks Sekolah - Padang Tembak Direct
2 Hilir Pasar 8 Kota Bharu - Kemumin - Sabak Direct
3 9 Kota Bharu - Pengkalan Chepa - Sabak Direct
4 10 Kota Bharu - Penambang - Pantai Cahaya Bulan Direct
Buluh
5 61A Kota Bharu - Kubang Kerian - Pasir Hor - Kg. Sireh Loop
Kubu
6 61B Kota Bharu - Kg. Sireh - Pasir Hor - Kubang Kerian Loop
Figure 5: Direct route and loop route.
55
From all six bus routes, all are direct route type except for bus number 61A and 61B. The
difference between direct route type and loop route is shown in Figure 5. Bus 61A and 61B have the same
looping route, except that the two buses travel in opposite direction.
The buses details are shown in Table 2. Column 3 shows the number of trips each bus make for
each day. For example, for bus ID 4, the number of trips per day is seven from Kota Bharu station to
Padang Tembak station, and seven from Padang Tembak to Kota Bharu, fourteen (14) trips for each bus
wherein there are four buses with ID 4. Therefore there are actually 56 trips to and from padang Tembak
in a day in Kota Bharu. Travel time for Route 4 is 30 minutes from Kota Bharu to Padang Tembak and
another 30 minutes from Padang Tembak to Kota Bharu. As for Route 10, the travel time is longer from
Pantai Cahaya Bulan to Kota Bharu because the bus will enter housing areas in Taman Buaya to board
passengers.
Table 2: Trip details
No. of Trips No. of Travel Time/Trip

No. Trip ID
per day per bus buses (minutes)
1 4 7+7 4 30+30
2 8 7+6 3 35+35
3 9 7+6 3 30+30
4 10 8+7 3 25+35
5 61A 9 1 65
6 61B 9 1 65
Figure 6 shows the map of Kota Bharu and the route of city buses. The city buses do not cover
the whole Kota Bharu because there are buses connecting passenger to other towns that cover some of the
routes.
Figure 6: Kota Bharu e- map (JUPEM, 2004) with the city bus routes.
56
Observation
The Passenger
The public bus users in Kota Bharu complained that buses always late and waiting time was very long for
certain bus routes. They claimed that due to many disabled buses, the company was having bus shortage
problem. Some routes have fewer bus allocations thus resulting in bigger headway difference. As a result,
users were shifting to other mode of transportation such as taxis and personal vehicles to fulfill their
travelling needs. Decreasing of bus passengers was very serious in Kota Bharu for the past year.
However, for some routes, passengers were very satisfied due to punctuality and frequency of the service.
The Schedule
From early observation, the buses depart according to schedule, however due to bus shortage, headway
between buses departure is still a big gap. Some buses might depart earlier than scheduled due to users
demand. Causes of bus delay could be due to heavy loading of school children at station or stops. School
children represent more than half of the bus users, which explain why the bus schedules are according to
their schooling or extra classes timetable.
The Drivers
At bus station, some of the drivers departed earlier than scheduled when there were adequate numbers of
passengers in the bus. This subsequently caused long passenger waiting time for the next trip, which
resulting in higher loading. Higher passenger loading on the second trip caused the travel speed to be
lower from frequent loading and unloading activities.
For instance, bus trip that supposed to depart at 12.30 pm departed 30 minutes earlier since the
number of passengers already large. Those who arrived after 12.00 pm will have to wait for the next bus
schedule at 2.00 pm. This group of passenger will have to wait longer and add up to the passenger for
2.00 pm trip. In some instances, passenger cant afford to wait that long will shift to other mode of
transportation, making public bus unattractive in the long run.
The Bus Condition
The bus condition is comfortable with Grade B bus (27 seats and 7 standing) as well as air-conditioned.
There are two automated doors located at the front and middle of the bus to ease passengers loading and
unloading at peak hours. The ticketing system use Personal Digital Assistant (PDA) equipment which
only prints out small piece of paper, indicating environmental friendliness.
The Company
The Cityliner Kelantan Operation admitted that a number of buses were disabled due to malfunctioned
air-conditioner and automated doors. To prevent the problems from recurring, the company has imposed
that the bus drivers must help take care of their bus. It is now forbidden to leave the bus with the engine
and air-conditioner on during the headway gap at the station. Besides, the company is now in the process
of buying new buses to satisfy users demand.
FUTURE DIRECTION
This paper is a part of an ongoing research whereby only the conceptual framework is presented. This
study is focused on determining the reliability of public bus on Kota Bharu in terms of Punctuality Index
57
analysis. The expected outcome will be the analysis of data gathered during survey which will determine
the reliability and performance of the bus service. From the analysis, discussion and inference can be
carried out to enhance this study, then recommendation can be advised to improve the service or
introduce new feasible route. Such studies should be conducted in other places to improve the public bus
service in many parts of Malaysia. Thus the people will have more option for better transportation.
REFERENCES
Abdalla Nurdeen Kambala, Riza Atiq, O.K. Rahmat, Amiruddin Ismail. (2007). Why Do People Use
Their Cars: A Case Study In Malaysia. Journal of Social Sciences 3 (3): 117 122.
Basic Population Characteristics by Administrative Districts June 2010. (2010). Department of Statistics.
Malaysia. Retrieved from http://www.statistics.gov.my/~Itemid=53&lang=en/
Carlos Gershenson, Luis A. Pineda. (2009). Why Does Public Transport Not Arrive on Time? The
Pervasiveness of Equal Headway Instability. Plos One open access, Vol. 4, Issue 10.
E-Map, Kelantan Electronic Map. (2004). Department of Survey and Mapping (JUPEM). Kelantan,
Malaysia.
Government Tranportation Programme Roadmap. (2010). Land Public Transport Commission (SPAD).
Retrieved from http://www.pemandu.gov.my/~gid=41&lang=en/
Jamilah Mohamad dan Nik Ibtisamiah Ibrahim. (2005). Characteristics of Bus Service Provision in Kota
Bharu, Kelantan. Journal of Transportation Science Society of Malaysia 1 13-25.
John Bates. (2009). An Agenda for Research on Reliability. Association for European Transport and
Contributors.
Margareta Friman. (2004). Implementing Quality Improvement in Public Transport. Journal of Public
Transportation. Vol. 7, No. 4.
Nicholas J. Garber & Lester A. Hoel (2002). Traffic and Highway Engineering, 3rd Edition. Toronto,
Canada. Thomson Learning. p. 571.
Ronghui Liu and Shalini Sinha. (2008). Modelling Urban Bus Service and Passenger Reliability.
Seung-Young Kho, Jun-Sik Park, Young-Ho Kim, Eun-Ho Kim. (2005). A Development of Punctuality
Index for Bus Operation. Journal of the eastern Asia Society for Transportation Studies. Vol. 6.
p 492 504.
Transit Capacity and Quality of Service Manual. (2003). 2nd Edition. Retrieved from
http://trb.org/~/part%201.pdf
Xumei Chen, Leio Yu, Yushi Zhang, Jifu Guo. (2009). Analyzing Urban Bus Service Reliability at The
Stop, Route, and Network Levels. Journal of transportation Research Part A 43.
Yuko J. Nakanishi. (1997). Bus Performance Indicator: On-Time Performance and Service Regularity.
Transportation Research Record 1571 3-12.
58
APPLICATION OF PROBABILISTIC DYNAMIC PROGRAMMING IN FORMULATING

AN AIRCRAFT ACQUISITION DECISION MODEL
Teoh Lay Eng 1*, Khoo Hooi Ling 2
ABSTRACT:
In the past, some studies had been conducted to inspect deterministic travellers demand of air
transportation system. However, these literatures might not be practicable to capture the real
occurrence of travellers demand as it remains unknown until the day of real operation. Motivated by
this fact, this study focuses on how to develop an optimization model in order to capture the
stochastic demand in a cost-effective manner. With the objective to maximize the expected profit of
the airlines companies, this study aims to determine the decision policy to acquire new aircrafts in
order to account for the stochastic demand as well as to replace the ageing aircrafts. In other words,
the developed optimization model is solved to determine the number and the types of aircrafts that
should be acquired during the planning horizon. In doing this, probabilistic dynamic programming is
applied due to its capability to comprehend the element of travellers demand stochastically. It is
anticipated that this study may reveal useful insight for air transportations management authorities as
well as the decision makers in making profitable managerial decision in providing a better service in
such a competitive airline industry.
Keywords: Stochastic Demand, Aircrafts Acquisition, Dynamic Programming, Probabilistic Dynamic

Programming, Linear Programming, Non-linear Programming
1
Lecturer, Universiti Tunku Abdul Rahman, Setapak Campus, Off Jalan Genting Kelang, Setapak, 53300 Kuala Lumpur,
MALAYSIA
2
Assistant Professor, Universiti Tunku Abdul Rahman, Setapak Campus, Off Jalan Genting Kelang, Setapak, 53300 Kuala
Lumpur, MALAYSIA
*Corresponding Author: Teoh Lay Eng, Universiti Tunku Abdul Rahman, Malaysia. Tel: +6003 4107 9802, fax:
+60341079803. E-mail: teohle@utar.edu.my
59
INTRODUCTION
At present, undoubtedly the world including airline industry is changing with many
unexpected events such as the economy crisis and the outbreaks of flu diseases which could affect the
demand of air travellers as well as the operational performance and the profit of airline companies.
To account for unpredictable stochastic demand, a proper managerial strategy is certainly needed by
airlines companies in order to gain a profitable return. One of the utmost operational decisions that
should be highlighted is the decision making to acquire new aircrafts to comprehend the
inconsistency of demand. In other words, airline companies have to decide strategically on the fleet
size of their aircraft based on the anticipated level of demand. Undeniably, an optimal strategy or
decision policy to determine the number and the types of the aircraft to be acquired in meeting the
uncertain level of demand is extremely important. Therefore, this study aims to provide answers to
various critical questions: What are the impacts of stochastic demand in purchasing new aircrafts?
How to make a desirable and profitable decision to purchase new aircrafts by taking stochastic
demand into consideration? How to improve the decision making for air transportation system by
managing travellers demand under uncertainties? In response to these critical issues, an optimization
model is developed by adopting probabilistic dynamic programming with the goal to maximize the
profit earned by the airline companies. Probabilistic dynamic programming is applied as it is able to
tackle uncertainties stochastically. Concisely, this study deals with an developed optimization model
to determine the number and the types of aircrafts to be acquired during a planning horizon in order
to capture the uncertainty of demand in a better manner.
LITERATURE REVIEW
Earlier literatures revealed that it is important to consider the quantity of travellers demand
in order to obtain the optimal solution for the aircraft acquisition model (New (1975), Wei and
Hansen (2005)). In particular, Listes et al. (2005) mentioned that the obtained solution is more robust
and closer to practical implementation by considering stochasticity. Apart from this, to account for
stochastic demand in airline operational planning, Pitfield et al. (2009) adopted simultaneous
equations approach to evaluate the real airline data. They found that aircraft size and the service
frequency could be affected by the level of demand elasticity. Additionally, in order to solve the
robust optimization problem, uncertainty of future demands was inspected by List et al. (2003) by
using a partial moment measure of risk. Then, Listes et al. (2005) applied scenario aggregation-based
approach to inspect the best fleet composition i.e. choice of aircrafts for the most profitable operation
subject to the airlines planned schedule. They looked into an airline fleet configuration that accounts
the robustness explicitly for short-term stochastic demand fluctuations. In the study, they found that
the stochastic approach is pertinent and viable in capturing the robustness of larger set of realistic
data. However, the limitation of the study is that only short term planning horizon is considered.
Besides, there are some studies in fleet assignment and scheduling problem that relevant to stochastic
demand (Feldman (2002), Yan et al. (2008)).
Observably there are limited studies for aircraft acquisition decision model that consider
stochastic demand. This is essential as in reality the airline company has to consider uncertain
demand in planning aircraft acquisition as a long term planning (more than 1 year) problem. As such,
the approach and models developed for short term planning might not be functional and effective in
such circumstances. To remedy this limitation, this study proposes an aircraft acquisition decision
model by taking stochastic demand into consideration. By using the probabilistic dynamic
programming, an optimization model is developed to maximize the earned profit of airline
companies. This mathematical approach is selected as it is capable to decompose the proposed model
into a chain of simpler single-period sub-problems during the planning horizon. More importantly,
this approach considers states (i.e. decision variables) and the corresponding profits which are
probabilistic (not deterministic) at each stage in order to comprehend the level of demand that is
uncertain until the day of real operation. In this study, to capture the demand uncertainty, it is
60
assumed that the travel demand could be described by some probabilistic distributions while the
decision variables of the acquisition model are the number and the types of aircrafts that need to be
purchased in order to maximize the operational profit. Besides, the probable phenomena are defined
in response to the targeted confidence level as the state variables are probabilistic due to the uncertain
demand. This component is necessary not only to capture the uncertainty of state variables properly
but also to ensure that the targeted level of service by the airline company could be met satisfactorily
to gain a profitable return. It is then shown that the probabilistic dynamic programming model could
be transformed into a linear or non-linear programming model based on the linearity of the objective
function and the interested constraints with respect to the decision variables. Apart from this, it is
important to note that the number of aircrafts considered for the acquisition decision could affect the
complexity of the formulation model and its optimal solutions. As reported by Gu et al. (1994), the
problem might become an NP-hard problem which can only be solved by meta-heuristic methods if
more than two aircrafts types are considered.
METHODOLOGIES
In order to tackle demand uncertainty, this section demonstrates the application of

probabilistic dynamic programming in formulating an optimization model for the decision making to
acquire new aircrafts.
Nomenclature
Following are the notations used in this study:
Parameters
T Horizon length for the planning period.
MAX budget (t ) Budget constraint allocated for the acquisition of new aircrafts at the beginning
of the operating period t.
DtS Random variable for stochastic demand which correspond to phenomenon S for
the operating period t.
ORDER t Total number of aircrafts that could be purchased in the market for the operating
period t.
PARK t Area of hangar for the period of t (as geometry limitation).
rt 1
Discount rate for the period of t for which the discount factor is .
(1 + rt )t
Significance level of demand constraint.
E ( faret ) Expected value of flight fare per passenger for the operating period t.
E ( cos tt ) Expected value of flight cost per passenger for the operating period t.
ps Probability to own I t as the initial number of aircrafts at phenomenon S.
Functions
P (t ) Function of discounted profit earned for the operating period t.
(
f Dt , Atn ) Function of number of flights in terms of stochastic demand, Dtn , and the total
of aircraft owned, Atn for n types of aircrafts during the operating period t.
(
hgf Dt , Atn ) Maintenance cost function as a composite function in terms of the function of
total mileage travelled, g, and the function of number of flights, f, for n types of
aircrafts during the operating period t.
61
Sets
X t = ( xt1 , xt 2 , , xtn ) Number of n types of aircrafts to be purchased at the
beginning of the operating period t.
I t = ( Int1 y , Int 2 y ,..., Intny ) Initial number of n types of aircrafts at age y at the
beginning of operating period t.
U t = ( ut1 , ut 2 , , utn ) Setup cost for the acquisition of n types of aircrafts at the
beginning of operating period t .
S = ( s1 , s2 , , sk ) Phenomenon of owning I t as the initial number of
aircrafts.
PURCt = ( purct1 , purct 2 , , purctn ) Purchase cost for n types of aircrafts at the beginning of
the operating period t.
SEATn = ( seat1 , seat2 ,, seatn ) Number of seats for n types of aircrafts owned.
SOLDt = ( soldt1 y , soldt 2 y , , sold tny ) Number of n types of aircrafts sold at age y at the
beginning of operating period of t.
RESALEt = ( resalet1 y , , resaletny ) Resale price for n types of aircrafts (at age y) for the
operating period of t.
DEPt = ( dept1 y , dept 2 y , , deptny ) Depreciation values for n types of aircrafts (at age y) for
the operating period of t.
SIZE = ( size1 , size2 , , sizen ) Size of n types of aircrafts.
Problem Formulation
By assuming that there is a choice of n types of aircraft that could be acquired and operated
for a given origin-destination (OD) pair, this study aims to determine the number and the types of
aircrafts that should be acquired with the objetive to maximize the profit of the airline company. For
the mentioned OD pair, the travellers demand is assumed to be stochastic and could be expressed by
some random distributions. To deal with stochastic elements, the aircrafts acquisition decision model
is formulated as a probabilistic dynamic programming problem by considering numerous practical
constraints which are relatively significant in the operational planning.
Probabilistic Dynamic Programming Model
Stage, State Variables and Optimal Decision

For a probabilistic dynamic programming model, the stage of the model is the planning
horizon of the aircraft acquisition period for which the planning period, t , in terms of years is the
stage variable of the model. The state variable at each stage t consisted of various interrelated
variables, namely the number of aircrafts to be purchased as the main decision variable for this study
as well as the initial number of aircrafts owned, number of aircrafts to be sold and stochastic demand.
State variables at any stage yields the desired elements which subsequently results in the optimal
decision (which is also known as alternatives at each stage) to acquire new aircrafts in order to meet
stochastic demand while making decision to sell ageing aircrafts with the aim to maximize the profit
earned.
Constraints
To ensure the efficiency of the operational planning of airline companies, some constraints
that need to be considered are explained as follows:
(i) Budget constraint

This is the most practical constraint to ascertain that the solution obtained is financially feasible for
the airline companies. Accordingly, the total purchase cost of the aircrafts which should not be more
than the allocated budget can be expressed as follows:
62
purc x
i =1
ti ti MAX budget ( t ) (1)
(ii) Demand constraint

The stochastic demand of air travellers can be represented by some probability distributions. In order
to meet the stochastic demand at a targeted level of service, the demand constraint can be formulated
as the following expression.
n
( (
P ( SEATi ) f Dt , Ati DtS 1
i =1
)) (2)
where P is the probability of the occurrence of the desired level of service while 1 is the
confidence level (i.e. targeted level). If the travellers demand is assumed to follow the normal
distribution with mean and standard deviation , the demand constraint could be expressed by,
( SEAT ) ( f ( D , A ) ) F
n
1
i t
i
t (1 ) + (3)
i =1
where F 1 (1 ) is the inverse cumulative probability of 1- .
Specifically, by considering the projected level of demand which is normally distributed for
each operating period of t during the planning horizon, the following expression signifies the demand
constraint for t = 1, 2, , T .
( SEAT ) ( f ( D , A ) ) F
n
1
i t
i
t (1 ) t + t (4)
i =1
where Dt ~ N ( t , t ) . Generally, the projected level of demand could be determined with an

appropriate forecasting approach.
(iii) Parking constraint

Sometimes, the choice of the aircraft to be acquired would be constrained by the specification or the
geometry design of the airports. For instance, the jumbo plane A380 could only be used at certain
airports due to its large body size and heavy weight. Besides, the aircrafts have to be grounded at the
hanger of the airport when they are off-duty. As such, it is feasible constraint to be considered and
this constraint can be expressed as follow:
n m
( ln
=i 1 =
y 0
tiy + xti ) ( SIZE ) PARK .
i t (5)
(iv) Sales of aircrafts constraint

For some airlines, ageing aircrafts which are less cost-effective might be considered to be sold at the
beginning of certain operating period t when the airlines make the decision to acquire new aircrafts.
Nevertheless, to sustain a certain level of operational efficiency, the number of aircraft sold should
not be more than the aircrafts owned by the aircrafts companies. It is formulated as follows:
soldtjy In(t 1) j=
( y 1) for t 1,=
2,..., T , j 1,=
2,..., n, y 1, 2,..., m . (6)
(v) Order delivery constraint

Sometimes, there might be delay in fleet supply due to the delivery of the new aircrafts which is
relatively depended on the efficiency and productivity of the manufacturing company. As such, the
aircrafts that could be purchased should not be more than the number of aircrafts available in the
market. This constraint is expressed as follows:
n
x ORDER .
i =1
i t (7)
63
Objective Function
The objective of this study is to maximize the expected profit of the airline companies which
could be resulted by the subtraction of the total operating cost from the total revenue obtained. For an
airline company, on the whole the total revenue comes from the operational income (i.e. the sales of
flight ticket) and the sales of ageing aircrafts. On the other hand, the total operating cost includes the
total purchasing cost of new aircrafts, the total operational cost of aircrafts owned, the total
maintenance cost and the total depreciation expenses of aircrafts owned.
For the operating period t, the total revenue, TR ( I t ) , is expressed as follows:

n m
TR ( I t ) E ( faret ) DtS +
= sold
=i 1 =
y 1
tiy resaletiy (8)
where the first term of the right hand side of eqn. (8) indicates the expected income obtained from the
sale of flight tickets by considering the stochastic demand DtS for which DtS F 1 (1 ) t + t .
The second term indicates the revenue obtained from the sales of ageing aircrafts.
The total operating cost for the operating period t, TC ( I t ) is expressed by,
( ) ( In )( dep )
n n n m

TC ( I t ) =uti + purcti ( xti ) +E ( cos tt ) DtS +
=i 1 =i 1
hgf Dt , Ati +
=i 1 =
y 1
tiy tiy (9)
where the first, second, third, fourth and the last term of the right hand side of eqn. (9) indicates the
setup cost for the acquisition of n types of aircraft, the purchasing cost of the new aircrafts, the
expected operating cost, the maintenance cost and the total depreciation expenses respectively.
The Probable Phenomena, s1 ,..., sk

The actual value of stochastic demand remains unknown until the day of operation. As a
consequence, the state variables are probabilistic and hence the probable phenomenon for which the
likely state variables to be occurred should be defined appropriately in order to capture the
uncertainty of state variables properly. In doing this, the airline company ought to consider the
possible level of service (i.e. actual level of demand) in order to plan their profitable operations
strategically. Only by recognizing the probable phenomenon, the targeted level of service by airline
companies could be met satisfactorily. This signifies that possible phenomenon to be happened
correlates very much with the level of service achieved by airline companies. In general, let s1 ,..., sk
be k possible phenomenon to meet the level of service (i.e. actual level of demand) at a targeted
confidence level. In spite of this, the phenomena, s1 ,..., sk differ in terms of the attainment of the
targeted level of service. In other words, subject to the targeted confidence level, s1 ,..., sk is the
concerned phenomena for which the actual level of demand is achieved differently.
Indeed, component of s1 ,..., sk in the developed modal is extremely significant as it turns out
to be an essential indicator to imply the possession of aircrafts in order to capture the actual
occurrence in reality. Only with this indicator, the actual operation under uncertainties will then be
monitored closely with the developed optimization model. Correspondingly, the possibility for the
phenomena s1 ,..., sk to be happened has to be considered and hence ps1 ,..., psk (i.e. the possibility for
s1 ,..., sk to be occurred) is included necessarily in the developed optimization model. In reality, the
phenomenon s1 ,..., sk and the corresponding probability ps1 ,..., psk ought to be treated strategically
by considering the companys decision policy, qualitative judgement from experts or consultants, the
past operational performance as well as the feedbacks from air transportation users (i.e. air
travellers).
64
The Optimization Model

With the aim to maximize the expected profit earned by acquiring new aircrafts to meet the
travellers demand under uncertainty, the formulation of the optimization model can be phrased as
follows:
For t = 1, 2, , T
n m n

E ( faret
s1
) Dt
s1

+ sold tiy resaletiy uti + purcti ( xti )

=i 1 = y 1 =i 1
ps1 + ... +
( ) )( )
n n m
t t
E cos t s1 D s1 hgf ( D , Ai )
t t ( Intiy dep tiy

1 = = =
P ( I t ) = max
i 1 i 1 y 1 (10)

X
(1 + rt ) t

n m n

E ( faret ) Dt +
sk sk

soldtiy resaletiy uti + purcti ( xti )
p =i 1 = y 1 i =1
+ P ( I )
sk t +1
( ) )( )
n n m

E cos ttsk Dtsk hgf ( Dt , Ati ) ( Intiy dep tiy

=i 1 =i 1 = y 1
subject to (1) and (4)-(7) for which DtS , Xt , It , SOLDt Z + {0} . Note
1
that, DtS F 1 (1 ) t + t . The term, is needed in order to obtain the discounted value
(1 + rt )
t
across the period of time while k indicates the k-th possible phenomenon for owning I t as the initial
number of aircrafts. It is important to note that the optimization model is formulated by assuming that
the developed model drives operational decision of airline companies particularly from the
perspective of flights frequency and its scheduling to meet stochastic demand. In other words, the
aircrafts acquisition decision will consequently leads to the optimal operational decision of fleet
routing at a desired level of service.
SOLUTION METHOD
The proposed probabilistic dynamic programming can be solved by decomposing it into a

series of simpler sub-problems. With the working backward (i.e. from the last period to the first
period), the solution method commences by solving the sub-problem at the period of T for which T is
the last period of the planning horizon. The current optimal solutions found for the decision variables
(i.e. the states of dynamic programming) at current stage leads to the problem solving at the period of
T 1 , T 2, ..., 1 . This procedure continues until all the sub-problems have been solved optimally
so that the decision policy to acquire new aircrafts can be determined eventually. For the developed
optimization model (10), the types of solution method i.e. linear programming problem or non-linear
programming problem can be identified particularly based on the key components as follows:
function of the number of flights, f Dt , Atn ( )
function of the maintenance cost, hgf Dt , Atn ( )
constraints (1), and (4)-(7)
If the abovementioned components are linear in nature, the proposed model (10) could be
converted to a linear programming model otherwise it could be converted and solved as non-linear
programming model. In reality, the linearity of these components is based on the real data for a
particular airline company. It shall then be validated by conducting the regression test with the aid of
some mathematical software. Empirically, the function of the number of flights, f Dt , Atn and the ( )
function of the maintenance costs, hgf Dt , Atn ( ) seems to emerge as polynomial function which
appears to be either increasing linear function or upward quadratic function. These functions show
that both response variables (i.e. number of flights and maintenance cost) are increasing constantly
65
when the explanatory variables (i.e. stochastic demand and total of aircrafts owned) are on the rise.
This shows that the response variables correlate positively with the interested explanatory variables.
It is anticipated that the developed model able to produce viable optimal solutions. Nonetheless, the
complexity of probabilistic dynamic programming makes it not practicable to develop a general
computer code and this explains the absence of commercial software for dynamic programming.
Therefore, an Excel-based algorithm is proposed in this study to perform the solution method. This
spreadsheet is straightforwardly implemented and more importantly, fully utilization of this
spreadsheet can solve the optimization model directly in an understandable manner.
CONCLUSIONS
This study formulated the aircraft acquisition decision model for which the traveller demand
appears to be stochastic in nature. The probabilistic dynamic programming approach is adopted in
order to incorporate the stochastic demand which is assumed to follow the normal distribution. By
taking some practical constraint into consideration, the developed model aims to maximize the
airlines companies profit i.e. the model is solved to determine the number and the types of new
aircrafts that should be purchased at every time period in order to produce a profitable return. It is
anticipated that the proposed methodology is viable and this study are able to steer the relevant
authorities at management level as well as the decision makers in making a wise profitable
operational decision to perform better in such a competitive airline industry. For the future work, the
proposed model will be tested with a set of real data collected from the airline company.
REFERENCES
Feldman, J. M. (2002). Matching planes to people. Air Transport World, 39, 3133.
Gu, Z., Johnson, E. L., Nemhauser, G. L., & Wang, Y. (1994). Some properties of the fleet
assignment problem. Operations Res. Lett., 15, 5971.
List, G. F., Wood, B., Nozick, L. K., Turnquist, M. A., Jones, D. A., Kjeldgaard, E. A., & Lawton, C.
R. (2003). Robust optimization for fleet planning under uncertainty. Transportation Research
E, 39, 209227.
Listes, O., & Dekker, R. (2005). A scenario aggregationbased approach for determining a robust
airline fleet composition for dynamic capacity allocation. Transportation Science, 39, 367-
382.
New, C. C. (1975). Transport fleet planning for multi-period operations. Operational Research
Quarterly, 26, 151-166.
Pitfield, D. E., Caves, R. E., & Quddus, M. A. (2009). Airline strategies for aircraft size and airline
frequency with changing demand and competition: a simultaneous-equations approach for
traffic on the North Atlantic. Journal of Air Transport Management, 16, 18.
Wei, W., & Hansen, M. (2005). Impact of aircraft size and seat availability on airlines demand and
market share in duopoly markets. Transportation Research Part E, 41, 315 327.
Yan, S., Tang, C. H., & Fu, T. C. (2008). An airline scheduling model and solution algorithms under
stochastic demands. European Journal of Operational Research, 190, 2239.
66
USING THE BEHAVIORAL SCIENCES THEORY AND STRUCTURAL EQUATION

MODEL (SEM) IN BEHAVIORAL INTERVENTION: HELMET USE
Kamarudin Ambak 1*, Rozmi Ismail 2,

Riza Atiq Abdullah 3 and Muhamad Nazri Borhan 4
ABSTRACT:
Nowadays, health and safety behaviors becoming a great concerned among researchers and practitioners.
Behavioral sciences theories and models have potential in enhancing efforts to reduce unintentional
injuries. In Malaysia, road accidents data statistics showed that motorcycle users contributes more than
50% of fatalities in traffic accidents, and the major cause due to head injuries. One of strategy can be used
to reduce the severity of head injuries by proper helmet usage. Although the safety helmet is the best
protective equipment to prevent or reduce severity of head injury, majority motorcycle user did not use or
did not fasten properly. To have better understanding on this problem, the behavioral sciences theory and
engineering aspect are necessary to help create a full picture of the often fragmented injury puzzle to
provide comprehensive insights into solutions. The Theory Planned Behavior (TPB) and Health Belief
Model (HBM) were used in predicting the behavioral intention toward proper helmet usage among
motorcyclist. While, a new intervention approach were used in Technology Acceptance Model (TAM)
that based on the perception of a conceptual system called Safety Helmet Reminder System (SHR).
Preliminary analysis shows that the constructs variables are reliable and statistically significant with the
exogenous and endogenous variables. Full structured models are proposed and to be tested so that, the
significant predictor can be identified. In future works, a multivariate analysis technique, known as
Structural Equation Model (SEM) will be used in modeling exercise. Finally, the good-of-fit models will
be used in interpreting the implication of intervention strategy toward injury prevention program.
Keywords: Behavioral Intervention, Structural Equation Model, Theory of Planned Behavior, Health
Belief Model, Technology Acceptance Model, Helmet use
1
Lecturer, Universiti Tun Hussein Onn Malaysia, Parit Raja, 84600 Batu Pahat, Johor, MALAYSIA
Ph.D Candidate, Universiti Kebangsaan Malaysia, Bangi, 43600 UKM Bangi, Selangor, MALAYSIA
2
Assoc. Prof., Universiti Kebangsaan Malaysia, Bangi, 43600 UKM Bangi, Selangor, MALAYSIA
3
Professor, Universiti Kebangsaan Malaysia, Bangi, 43600 UKM Bangi, Selangor, MALAYSIA
4
Lecturer, Universiti Kebangsaan Malaysia, Bangi, 43600 UKM Bangi, Selangor, MALAYSIA
*Correspondence Author: Kamarudin Ambak,, Universiti Tun Hussein Onn Malaysia. Tel: +6012 7662895
Fax: +6074536070. E-mail:kamardin@uthm.edu.my
67
INTRODUCTION
The motorcycle is one of popular mode of transport. In Malaysia, for instance, the registration
number of vehicles increased by 84% to 16.8 million from 1998 to 2000, and motorcycle consists of 50%
(PDRM, 2009). As a result, many road accidents occurred in the transportation system that leading to
casualties and deaths. According to Royal Malaysian Police (PDRM), more than 6000 people killed in
road accidents annually. In 2007 alone, PDRM reported that 6282 of fatalities were recorded in road
crashes. Of this figure, motorcycle users accounted for 50% (or 3197) of all road fatalities. Since,
motorcycle has become a common and popular in many developing countries (Conrad et. al 1996), its
also known as vulnerable road user in term of safety-risky exposure and instability compared to other
vehicles. Many researches indicated that the major cause of death involving motorcycle users due to head
injuries (Barbara et al. 1995; Kulanthayan et al. 2000; Radin et al. 2005; Shoa, 2005 and Zamani et al.
2009). Facts from road accidents statistics showed that the most part of body injured lead to fatality is
head by 65% (PDRM, 2009). Safety helmet is the best equipment that can be used to protect motorcycle
users from head injuries (NSTHA, 2009; Radin et al. 2005 and Shuaieb et al. 2002). Many studies shown
that the helmet is effective in preventing and reducing the severity of head injuries by 37% to 72%
(David, 2007; Li et al. 2008) or deaths by 20% to 24% (Masao et al. 2003; Thomas, 2009).
Despite the fact that the safety helmet is the best protective equipment to prevent or reduce
severity of head injury, majority motorcycle user did not use or did not fasten properly. There are several
studies in developing countries found that the percentage of proper usage of helmet among motorcycle
users is consider low (Conrad et al. 1996; Ichikawa et al. 2003; Hung et al. 2006; Li-Ping et al. 2008;
Kulanthayan et al. 2000; Zamani et al. 2009). However, Radin et al. (2005) highlighted that Malaysian
government has taken role of safety concern regarding helmet issue by implementing series initiatives as
early seventies. Beginning with Introduction of Motorcycle Safety Helmet Standard MS1: 1969,
Implementation of Helmet Law in 1973, Targeted safety helmet campaign in 1997 to date and newly
intervention program is the Community Based Program in 2007 to date. Then, the effectiveness of helmet
initiatives has been evaluated with a few studies. Radin et al. (2005) reported that since 1995, 1998 and
2000, the rates of proper usage of safety helmet were increased by 33%, 41% and 54% respectively. It is
seem positively improve but the percentage rate was saturated at 66%. Furthermore, this figure is
represents in average for both areas in urban and rural. But, the compliance rate in rural area was
considered low at 33% (Kulanthayan et al. 2001) and needs to do extra effort regarding their safety
concerned. Li et al. (2008) suggested that there is a need to implement new interventions to increase
helmet use. Therefore, the aim of this study is to apply a behavioral sciences theory or model in
predicting intention toward proper usage of helmet and to determine a significant predictor that contribute
to the behavioral intention of safety helmet usage.
BEHAVIORAL SCIENCES THEORIES
Behavioral and social sciences theories and models have the potential to enhance efforts to reduce
unintentional injuries (Trifiletti et al. 2005). The behavioral sciences or social psychological theories such
as Theory of Planned Behavior (TPB; Ajzen, 1991), Health Belief Model (HBM; Rosenstock, 1966) and
Technology Acceptance Model (TAM; Davis, 1989) provide a potentially fruitful framework to
understand in prediction of behavioral intention. For instance, Lajunen & Rasanen (2004) were adopted
the TPB and HBM in their study to understand why cyclist are so unwilling to use bicycle helmets.
Warner et al. (2006) used the TPB as a conceptual framework in prediction of drivers decision to speed.
Simsekoglu & Lajunen (2008) found that the social psychological theories provide potentially useful yet
rarely used tools for explaining how attitudes, beliefs, and values influence seat belt use. Chen et al.
(2007) proposed the integration of TAM and TPB to understanding on the critical antecedents of
motorists intention of electronic toll collection (ETC) service adoption. Thus, behavioral science is an
integral part of a comprehensive injury prevention strategy and the need to improve behavioral
68
interventions by using better empirical data about determinants of behavior as well as theories and
frameworks pertaining to change in health behavior (Geilen & Sleet, 2003).
Theory of Planned Behavior (TPB)
The Theory of Planned Behavior (TPB) is an extension of the Theory of Reasoned Action (TRA).
Ajzen (1991) extended his earlier work with Fishbein (Fishbein & Ajzen, 1975) to include an explanation
of all behaviors, not simply those under voluntary control by including measures of perceived behavioral
control (Johnson & Hall, 2005; Letirand & Delhomme, 2005). According to the theory of planned
behavior peoples attitude towards the behavior, their subjective norm, and their perceived behavioral
control determine their behavior indirectly via their intentions (Warner and Abreg, 2006). Attitudes are a
persons overall evaluations of a behavior while subjective norm consists of the persons beliefs about
whether significant others think he/she should engage in that behavior (Ajzen, 1985, 1991; Conner &
Sparks, 1996). In addition, perceived behavioral control has both direct and mediated effects (by
behavioral intention) on behavior and refers to the persons perception of control on engaging in that
behavior (Ajzen, 1985, 1991; Conner & Sparks, 1996). Letirand & Delhomme (2005) mentioned that
behavioral intention, is determined by the combination of attitude toward the behavior, subjective norm
(perceived social pressure from important others to perform or not to perform a given behavior), and
perceived behavioral control. According to these (TRA & TPB) models, behavioral intention is
influenced by a persons attitude toward performing a behavior, and by beliefs about whether individuals
who are important to the person approve or disapprove of the behavior (subjective norm) and perceived
behavioral control is construct has to do with peoples beliefs that they can control a particular behavior
(National Institutes of Health, 2005).
Health Belief Model (HBM)
The HBM was developed in the early 1950s by a group of social psychologists at the U.S. Public
Health Service in an attempt to understand the widespread failure of people to accept disease preventives
or screening tests for the early detection of asymptomatic disease (Janz et al. 1984). The basic
components of the HBM are derived from a well-established body of psychological and behavioral theory
(Becker et al. 1977). The concept of the HBM focuses on two aspects of health behavior: threat
perception and behavioral evaluation (Rosenstock, 1966, 1974). Threat perception refers to a perceived
susceptibility to illness and a perceived severity of the consequences of such an illness, whilst behavioral
evaluation concerns the perceived benefit and the perceived barriers to enacting behaviour (Sheeran &
Abraham, 1996). Additionally, the HBM proposes cue to action and health motivation as two other
cognitive components were included to the model (Sheeran & Abraham, 1996). According to
McClenahan et al. (2007) the HBM is a health-specific model, which suggests that health behaviors are a
result of a set of core beliefs and it has been used to predict many health behaviors. The HBM suggests
that the core beliefs should be used to predict the likelihood that a behavior will occur but recently it has
been suggested that intention should be included as a mediator between beliefs and health behavior
(Quine et al. 1998).
Technology Acceptance Model (TAM)
The technology acceptance model (TAM) is grounded in both TRA and TPB (Davis, 1989; Davis
et al. 1989). The TAM is perhaps the most widely applied and most notable is the application of the
Technology Acceptance Model (TAM) to the prediction and explanation of end-user reactions to health
IT (Holden & Karsh, 2010; Yi et al. 2006). TAM was specifically tailored for modeling user acceptance
of an information system with the aim of explaining the behavioral intention to use the system (Chen et
al. 2007). He added, this model, perceived usefulness and perceived ease of use were considered as two
predecessors affecting attitude toward a technology, which affects behavioral intention to use that
69
technology. The TAM model states that an individuals system usage is determined by behavioral
intention, which is, in turn, determined by two beliefs: perceived usefulness, the extent to which a person
believes that using the system will improve his or her job performance, and perceived ease of use, the
extent to which a person believes that using the system will be free of effort (Yi et al. 2006). The TAM
has been tested by many researchers with different populations of users and IT innovations (Polancic et
al. 2010). Besides this, Hong et al. (2006) concluded that TAM is the most simple and generic model that
can be used to study both initial and continued IT adoption.
Structural Equation Model (SEM)
The development of structural equation modeling (SEM) methods and software has proceeded
rapidly since the 1970s (MacCalum & Austin, 2000). An SEM is an extremely flexible linear-in-
parameters multivariate statistical modeling technique and it has been used in modeling travel behavior
and values since about 1980s (Golob, 2003). Structural equation modeling (SEM) is a family of statistical
techniques permitting researchers to test such models and as a hybrid of factor analysis and path analysis
that researchers can test hypothesized relationships between constructs (Weston, 2006). Also, SEM is a
technique used for specifying and estimating models of linear relationships among variables. Variables in
a model may include both measured variables (MVs) and latent variables (LVs). LVs are hypothetical
constructs that cannot be directly measured (MacCalum & Austin, 2000). SEM is a relatively new method
and most applications have been in psychology, sociology, the biological sciences, educational research,
political science, market research and travel behavior (Golob, 2003).
An SEM has two primary components: the measurement model and the structural model. The
measurement model describes the relationships between observed variables (e.g., instruments) and the
construct or latent variables are hypothesized to measure. In contrast, the structural model describes
interrelationships among constructs. When the measurement model and the structural model are
considered together, the model may be called the composite or full structural model (Weston, 2006).
Figure 1 shows a basic example of component in structural equation model.
Figure 1: A basic example of SEM component (Lee et al. 2008)
METHODOLOGY
Since the previous studies show the compliance rate of proper helmet usage in rural area and
outside-town area was low, so the collections of data were carried out within outside-town centre
including countrysides, housing estates and residential areas. Selangor state was chosen as location of
study due to this state recorded highest road accidents statistical report (PDRM, 2009) and Bangi
was represents as typical suburban in the state. There are two methods were used in this study. The
first method, observation on helmet usage among motorcyclists those using their motorcycles in daily
70
activity such as to sundry shop, working, send children to school, etc. The observation activities were
carried-out at six station locations, three in countrysides and three in housing estate or residential areas.
A week of observed were carried out within half-an hour as early 7.30 a.m. 8.00 a.m. and rush hour at
5.30 p.m. 6.00 p.m. This period is normal peak hour that been used in observational activity in the
selected stations.
Second method is using face-to-face approached and self-administered questionnaire were used to
motorcyclists who not wearing a helmet or untied their helmet. If they refuse to do so, another respondent
were approached and prior to giving the questionnaire, the way they using a helmet were noted (either
unfastened properly or with-out helmet) and recorded separately. Locations of data collection were
divided into six zones. Three zones consist of a group number of section in housing estate and three zones
in countryside respectively. The questionnaire is consists of five sections: background, riding experience,
knowledge and attitude, behavioral sciences model (Theory Planned Behavior, Health Belief Model and
Technology Acceptance Model) and feedback. A pre-tested questionnaire session was carried-out with 20
respondents and after analyzed the outcomes, the reliability analysis was carried out to improve the
questionnaire and to meet respondent acceptance level. Three hundred (300) respondents were chosen as
sample size to represent their general characteristics and the surveyed was achieved with response rate of
57% (out of 533 respondents were approached). However, eight cases were dropped out for further
analysis due to incomplete. This sample size is reasonably enough to analyze descriptive statistics,
multivariate analysis and structural equation model. There are several studies using less than 300 of
sample size, such as seatbelt use (N=277) by Simsekoglu and Lajunen (2008), motorcyclists intention to
speed (N=110) by Elliot (2010), drivers decision speed (N=250) by Warner and Abreg. (2006) and truck
driver behavior (N=232) by Poulter et al. (2008). The data were analyzed using the Statistical Package for
Social Sciences Software (SPSS) version 18.
RESULT ANALYSIS
Based on the observation of helmet usage of the 1150 motorcyclists, results show that only 46.9%
used helmets properly and 10.8% untied helmet and 42.3% did not use helmet at all (see Table 1). In
term of proper usage of safety helmet, housing estate were recorded slightly higher (54% to 64%)
compared to countryside (28.5% to 39%) by. Results significantly (p<0.05) show that those who are
riding in housing estate or residential area more likely to used helmet properly compared to those in
countryside (see Table 2). In term of gender, Table 3 shows that female motorcyclists are more likely to
used helmet (64%) compared to male motorcyclist (56%). This result is significantly associated at 95%
confidence level (p=0.037).
Table 1: Helmet usage among motorcyclist at location of study

Helmeted
Location Unhelmeted (%) Untied (%) (%) Total
Properly
Countryside
Zone 1 116 (58.9) 25 (12.7) 33 (28.5) 197

Zone 2 85 (51.5) 18 (10.9) 37 (37.6) 165
Zone 3 94 (51.6) 17 (9.3) 55 (39.0) 182
Zone 4 73 (33.9) 22 (12.1) 98 (54.0) 215
Housing
Estate
Zone 5 65 (32.0) 17 (11.8) 95 (56.2) 203

Zone 6 54 (28.8) 11 (7.0) 107 (64.3) 188
Total 487 (42.3) 124 (10.8) 540 (46.9) 1150
71
Table 4: Helmet usage by location

Helmeted Unhelmeted/
Location Total
Properly Improperly
Countryside 189 (34.7%) 355 (65.3%) 544 (100%)
Housing estate 351 (57.9%) 255 (42.1%) 606 (100%)
(2 = 61.831, df = 1, p=0.000< 0.05)
Table 5: Helmet usage by gender

Gender No helmet Used helmet Total
Male 408 (43.8%) 523 (56.2%) 931(100%)
Female 79 (36.1%) 140 (63.9%) 606 (100%)
(2 = 4.363, df = 1, p=0.037< 0.05)
Reliability and Correlation Analysis
The reliability analysis was conducted on specific questionnaire for TPB (with seven items),
HBM (with eleven items) and TAM (with seven items) respectively. An Alpha Cronbach () was adopted
to evaluate the reliability of these items that used in the instruments. The acceptable for Alpha Cronbach
value is when > 0.7 (Bland & Atlman 1997). The results show that the Cronbachs ( ), for TPB is
0.738, HBM is 0.778 and TAM is 0.913 respectively, and those indicate the items were used in the
variables is reliable.
Table 6, Table 7 and Table 8 show the means, standard deviations and pearson-correlations for
each measure for TPB, HBM and TAM models respectively. Thus, all the variables in TPB and HBM
model are viewed as distinct but highly correlated and were found to have significant positive correlations
with intentions and behavior toward proper usage of helmet. As well as TAM model, the bivariate
correlation shows the variables was significantly correlated with behavioral intention to use SHR (Safety
Helmet Reminder System). In order to examine the relationships between independent variables and
dependent variable, a further examination of hypotheses testing is needed.
Table 6: Means, Standard Deviations and Correlations for Variables in TPB (N=292)
1 2 3 4 5 Mean Std. D
1 1 5.53 1.56
2 .562** 1 6.03 1.20
3 .282** .456** 1 18.86 2.73
4 .266** .403** .598** 1 12.28 2.08
5 .160** .221** .394** .380** 1 10.68 2.64
Note: 1- Intention, 2-Behavioral, 3-Attitude, 4-Subjective Norm, 5-Perceived Behavioral Control
** Correlation is significant at .01 level (2-tailed)
72
Table 7: Means, Standard Deviations and Correlations for Variables in HBM (N=292)
1 2 3 4 5 6 7 8 Mean Std. D
1 1 5.53 1.56
2 .562** 1 6.03 1.20
3 .148* .314** 1 6.29 1.14
4 .347** .358** .267** 1 5.55 1.40
5 .267** .361** .415** .464** 1 12.34 1.40
6 .162** .192** .259** .122* .378** 1 11.66 2.53
7 .156** .233** .378** .224** .360** .157** 1 11.12 3.06
8 .185** .218** .391** .163** .415** .217** .577** 1 12.26 2.32
Note: 1- Intention, 2-Behavioral, 3-Health Value, 4-Cue to Action, 5-Perceived Benefit, 6-Perceived Susceptibility,
7-Perceived Barrier, 8- Perceived Severity; * Correlation is significant at .05 level (2-tailed)
Table 8: Means, Standard Deviations and Correlations for Variables in TAM (N=292)
1 2 3 4 Mean Std. D
1 1 5.52 1.55
2 .212** 1 11.4 2.58
3 .220** .659** 1 17.68 3.25
4 .187** .709** .712** 1 11.52 2.29
Note: 1- Behavioral, 2-Intention to use SHR, 3- Perceived Usefulness, 4- Perceived Ease of Use
For further analysis and modeling exercise (SEM) in future work, the TPB, HBM and TAM
models have been adopted to testing the relationship constructs variables between exogenous (intention)
and endogenous (behavior) variables based on review of previous studies. These proposed models have
been adapted from Lajunen & Rasanen (2004), Simsekoglu & Lajunen (2008), McClenahan et al. (2007)
and Polancic et al. (2010) those successful in predicting behavioral intention in their studied.
ATT1
ATT2 Attitude
ATT3
SN1 Subjective Proper

Intention
Norm helmet use
SN2
PBC1 Perceived
Behavioral
PBC2 Control
Figure 2: Proposed TPB model for predicting intention toward proper helmet usage
73
HV Health Value
CUA1
Cues to Action
CUA2
PBen1
Perceived
Benefit
PBen2
Proper
Intention helmet use
PBar1
Perceived Barrier
PBar2
PSus1 Perceived
Susceptibility
PSus2
PSer1 Perceived
Severity
PSer2
Figure 3: Proposed HBM model for predicting intention toward proper helmet usage
PU1
PU2 Perceived
usefulness
PU3
Intention to Proper
Use SHR helmet use
PEOU1
Perceived
ease of use
PEOU2
Figure 4: Proposed TAM model for predicting intention to use SHR toward proper helmet usage
DISCUSSION AND CONCLUSION
Motorcycle crashes cannot be totally prevented but resultant head injuries and their severity can
be avoid or minimized by protective equipment like safety helmet (Kulanthayan et al. 2000). However,
this study found that the compliance rates of proper helmet use among motorcyclists were considered low
(47% in average) and alarmingly, in some zone as low as 29%. This finding similar in Kulanthayan et al
(2001) studied found that 54% of motorcyclists were used safety helmet properly. In fact, their result
showed at outside-town centre was only at 33%. Other studies, in Indonesia, Conrad et al. (1996) reported
that only 55% of the riders wore helmets correctly for urban area and would predicted that compliance
would be considerably lower on byways and backroads and in the villages. In China, based on
74
observational study shows, that the rate of proper helmet use was low, with less than one-third (32.3%) of
riders (Li et al. 2008). Hence, present study indicates that the lower rate of compliance of safety helmet
issue still occurs, even though Malaysian government concerned on this problem seriously. It seems that
helmet initiative programs namely helmet law enforcement, safety helmet campaign and Community
Based Program are insufficient to overcome the problem. Therefore, a new approach is needed to be
introduced to mitigate current issue as recommended by Li et al. (2008). They suggested the need to
implement new interventions to increase helmet use.
In fact, an interdisciplinary approach that involves behavioral sciences, injury prevention and
engineering aspect all together would create a proper solution. Behavioral science when combined with
engineering, epidemiology and other disciplines creates a full picture of the often fragmented injury
puzzle and informs comprehensive solutions (Winston & Jacobsohn, 2010). Trifiletti et al. (2005) stated
that the behavioral and social sciences theories and models have the potential to enhance efforts to reduce
unintentional injuries. The applications of such behavior sciences theories or models (TPB, HBM, and
TAM) are able to predict and explain the significant predictor (e.g attitude toward behavior). The
relationship between construct variable and target behavior also can be determined. Then, the implication
of the model analysis would suggest some strategies to be taken onto intervention program.
For instance, Simsekoglu and Lajunen (2008) found that the TPB results emphasize the important
role of attitudes and subjective norms in developing intentions to use a seat belt. Also, they suggested that
seat belt campaigns should be aimed at forming and strengthening positive attitudes towards seat belt use.
While, Lajunen & Rasanen (2004) mentioned that both the HBM (perceived barriers) and TPB
(subjective norm) results emphasize the role of parents and peers in a teenagers intention to use a bicycle
helmet. According to the results of their study, bicycle helmet campaigns should aim mainly at changing
peers and parents attitudes. Finally, with regards to these significant approaches, there is a need to come
out with such models (TPB, HBM &TAM) to be proposed with a new intervention measure regarding
motorcycle safety program, particularly on head injury prevention.
FUTUREWORK
This paper presents part of result for on going research to develop a behavioral intention
model toward proper helmet usage among motorcyclist by using Theory Planned Behavior
(TPB), Health Belief Model (HBM) and Technology Acceptance Model (TAM). In TAM
modeling analysis, a conceptualize system called (Safety Helmet Reminder System) will be used
in order to predict that such system could influence motorcyclists behavior and their acceptance.
Kamarudin et al. (2009) suggested possibility to adapt and apply a seat belt reminder system into
motorcycle as helmet reminder system. Similarly, Li et al. (2008) suggested that there is a need
to implement new interventions to increase helmet use. Hopefully, this research may contribute a
significant finding and method that be able to mitigate current problem in the country regarding
motorcycle safety program.
ACKNOWLEDGEMENT
The author would like to thank Universiti Tun Hussein Onn Malaysia (UTHM) and
Ministry of Higher Education (MOHE) of Malaysia who provide a scholarship for Ph.D studies
and sponsoring this work.
75
REFERENCES
Ajzen, I. (1985). From intentions to actions: A theory of planned behavior. In J. Kuhl & J. Beckmann
(Eds.), Action control: From cognition to behavior. Berlin, Heidelber, New York: Springer-Verlag.
Ajzen, I. (1991). The Theory of Planned Behavior. Organizational Behavior and Human Decision
Processes, 50, 179211.
Barbara, G., Khaty, L.R., Richard, E.H., Magdalena, C., Lorann, S. (1995). Relationship of helmet use
and head injuries among motorcycle crash victims in El Paso County, Colorado, 1989-1990, Accident
Analysis and Prevention, 27(3), 363-369.
Becker, M.H., Maiman, L.A., Kirscht, J.P., Haefner, D.P., Drachman, R.H. (1977). The Health Belief
Model and prediction of dietary compliance: a field experiment. Journal of Health and Social
Behavior, 18, 4, 348-366.
Bland, J.M. and Atlman, D.G. (1997). Statistics Notes: Cronbachs alpha. British Medical Journal, 314,
572.
Chen, C.D., Fan, Y.W. and Farn, C.K. (2007). Predicting electronic toll collection service adoption: An
integration of the technology acceptance model and the theory of planned behavior. Tranportation
Research Part C. 15, 300-311.
Conner, M. and Sparks, P. (1996). The Theory of Planned Behaviour and health behaviours. In M.
Conner, & P. Norman (Eds.), Predicting health behaviour , 121 162. Buckingham: Open University
Press.
Conrad, P., Bradshaw, Y.S., Lamsudin, R., Kasniyah, N., Costello, C. (1996). Helmets, Injuries and
cultural definitions: motorcycle injury in urban Indonesia. Accident Analysis and Prevention, 28 (2),
193-200.
Davis, F.D., (1989). Perceived Usefulness, Perceived ease of use and user acceptance of Information
Technology. MIS Quarterly 13 (3), 319340.
Davis, F.D., Bagozzi, R.P., Warshaw, P.R. (1989). User acceptance of computer technology: A
comparison of two theoretical models. Management Science, 35 (8), 9821003.
David J. Houston. (2007). Are helmet laws protecting young motorcyclists?. Journal of Safety Research
38, 329-339.
Elliot, M.A. (2010). Predicting Motorcyclists intentions to speed: effects of selected cognitions from the
Theory of Planned Behaviour, Self-Identity And Social Identity. Accident Analysis and Prevention,
42, 718-725.
Gielen, A.C., Sleet, D. (2003). Application of behavior-change theories and method to injury prevention.
Epidemiologic Reviews, 25, 65-76.
Golob, T.F. (2003). Structural equation modeling for travel behavior research. Transportation Research
Part B, 37, 1-25.
Holden, R.J., Karsh, B.T. (2010).The Technology Acceptance Model: Its past and its future in
health care. Journal of Biomedical Informatics, 43, 159172.
Hong, S.J., Thong, J.Y.L. and Tam, K.Y. (2006). Understanding continued information technology usage
behavior: a comparison of three models in the context of mobile internet. Decision Support Systems
(3), 18191834.
Hung, D.V., Stevenson, M.R., Ivers, R.Q. (2006). Prevalence of helmet use among motorcycle riders in
vietnam, Injury Prevention, 12: 409-413.
76
Janz,N.K.,Becker, M.H. (1984). The Health Belief Model: A decade later. Health Education
Quarterly, 11(1), 1-47.
Johnson, S.E., Hall, A. (2005). The prediction of safe lifting behavior: An application of The Theory of
Planned Behavior. Journal of Safety Research, 36, 63 73.
Ichikawa, M., Chadbunchachai,W.,Marui, E. 2003. Effect of the helmet act for motorcyclists in Thailand.
Accident Analysis and Prevention, 35 (2), 183189.
Kamarudin, A., Riza Atiq A., Rozmi, I. (2009). Intelligent Transports System for motorcycle safety and
issues, European Journal of Scientific Research, 28 (4) 600-611.
Kulanthayan, S., Umar, R.S., Hariza, H.A., Nasir, M.T., Harwant, S. (2000). Compliance of proper safety
helmet usage in motorcyclists. Medical Journal of Malaysia 55 (1), 4044.
Kulanthayan, S., Umar, R.S., Hariza, H.A., Nasir, M.T. 2001. Modeling of compliance of motorcyclist to
proper usage of safety helmet in Malaysia. Journal of Crash Prevention and Injury Control, 2 (3),
239-246.
Lajunen, T and Rasanen, M. (2004). Can social psychological models be used to promote bicycle helmet
use among teenagers? A comparison of the Health Belief Model, Theory of Planned Behavior and
the Locus of Control. Journal of Safety Research, 35, 115-123.
Lee, JY., Chung, JH., Son, B. (2008). Analysis of traffic accident size for Korean Highway using
Structural Equation Models. Accident Analysis and Prevention, 40, 19551963.
Letirand, F., Delhomme, P. (2005). Speed Behavior as a choice between observing and exceeding the
speed limit. Transportation Research Part F, 8, 481492.
Li-Ping, L., Gong-Li, L., Qi-En, C., Anthony, L.Z., Sing, K.L. (2008). Improper motorcycle helmet use in
provincial areas of a developing country. Accident Analysis and Prevention 40, 1937-1942.
Li, GL., Li, LP., Cai, QE. (2008). Motorcycle helmet use in Southern China: An Observational Study,
Traffic Injury Prevention 9, 125-128.
Masao, I., Witaya, C., Eiji, M. (2003). Effect of the helmet act for motorcyclists in Thailand, Accident
Analysis and Prevention 35, 183189.
McClenahan,C., Shevlin, M., Adamson,G., Bennett, C., ONeill, B. (2007). Testicular self-
examination: A test of the Health Belief Model and The Theory of Planned Behavior. Health
Education Research, 22(2), 272-284.
Maccallum, R.C., Austin, J.T. (2000). Applications of Structural Equation Modeling in
psychological research. Annual Reviews Psychology, 51, 201226.
National Highway Traffic Safety Administration (NHTSA) 2009. An analysis of motorcycle helmet use
in fatal crashes, Annals of Emergency Medicine, 53, 501.
National Institutes of Health. (2005) A Guide for Health Promotion Practice (2nd Edition). U.S.
Department of Health and Human Services: Rimer, B., Glanz, K.
Polancic, G., Hericko,M., Rozman, I. (2010) An empirical examination of application frameworks
success based on Technology Acceptance Model. The Journal of Systems and Software, 83, 574-
584.
Poulter,R.D., Chapman, P., Bibby,P.A, Clarke, D.D., Crundall, D. (2008). An application of the Theory
of Planned Behaviour to truck driving behaviour and compliance with regulations. Accident Analysis
and Prevention, 40, 2058-2064.
Quine L, Rutter, D.R., Arnold L. (1998). Predicting and understanding safety helmet use among
schoolboy cyclists: a comparison of The Theory of Planned Behaviour and The Health Belief Model.
Psychological Health, 13, 25169.
77
Radin Umar, R.S., Kulanthayan, K.C., Law, T.H., Ahmad, H., Musa, A.H., Mohd, N.M.T. (2005).
Helmet initiative program in Malaysia, Pertanika Journal of Science and Tech. Supplement 13(1),
29-40.
Rosenstock, I. M. (1966). Why people use health services? Milbank Memorial Fund Quarterly, 44, 94
124.
Royal Malaysian Police (PDRM), (2009). Statistical Report on Road Accidents in Malaysia 2007. Traffic
Branch, Bukit Aman, Kuala Lumpur.
Sheeran, P., Abraham, C. (1996). The Health Belief Model. In: Connor M, Norman P (eds). Predicting
Health Behavior: Research and Practice with Social Cognition Models. Buckingham, UK: Open
University Press.
Shao-H, K. (2005). Helmet use and motorcycle fatalities in Taiwan. Accident Analysis and Prevention,
37, 349355.
Simsekoglu, O. and Lajunen, T. (2008). Social psychology of seat belt use: a comparison of Theory of
Planned Behavior and Health Belief Model, Transportation Research Part F 11,181191.
Thomas, S.D. (2009). Motorcycle helmets and traffic safety. Journal of Health Economics, 28, 398412.
Trifiletti, L.B, Gielen, A.C., Sleet, D.A., Hopkins, K. (2005). Behavioral and social sciences theories and
models: Are they used in unintentional injury prevention research?. Health Education Research,
20(3), 298-307.
Warner, H.W. and Abreg, L. (2006). Drivers decision to speed: a study inspired by The Theory of
Planned Behaviour, Transportation Research Part F, 9,427433.
Weston, R., Gore (Jr), P.A. (2006). A brief guide to Structural Equation Modeling. The
Counseling Psychologist, 34: 719-751.
Winston, F.K. and Jacobsohn, L. (2010). A practical approach for applying best prevention practices in
behavioural interventions to injury. Injury Prevention, 16, 107-112.
Yi, M.Y., Jackson, J.D., Park, J.S., Probst, J.C. (2006). Understanding Information Technology
acceptance by individual professionals: Toward an integrative view. Information
Management, 43, 350-63.
Zamani, F.A., Nikami, S., Mohammadi, E., Ali, M., Ghofranipour, F., Ahmadi. F., Hejazi, B.S. (2009).
Motorcyclists reaction to safety helmet law: A qualitative study, BMC Public Health, 9: 393.
78
EFFECT OF THERMOPLASTIC COPOLYMERS ON MICROSTRUCTURE AND

VISCOELASTIC BEHAVIOUR OF BITUMEN
Noor Zainab Habib 1*, Ibrahim Kamaruddin 2,

Madzlan Napiah 3 and Isa Mohd Tan 4
ABSTRACT:
This paper presents a part of research to study the effect of the thermoplastic copolymer namely linear
low density polyethylene (LLDPE), High density polyethylene (HDPE) and Polypropylene (PP) modified
on 80 pen bitumen, using the rheological and morphological parameter. The modification of bitumen by
the use of polymers enhances its performance characteristics but significantly affects its rheological
properties. The tools used for rheological study of polymer modified bitumen (PMB) were penetration,
ring & ball softening point and viscosity test. The morphological analysis was done to study the
compatibility of bitumen with polymer, through field emission scanning electron microscopy (FESEM).
The results were then related to the changes in the rheological properties of polymer modified bitumen. It
was observed that thermoplastic copolymer shows profound effect on penetration rather than softening
point. From the viscosity results it is revealed that the viscoelastic behaviour of PMB depend on the
concentration of polymer in the mix, mixing temperature, mixing technique, solvating power of base
bitumen and molecular structure of polymer used. PP offer better homogeneous blend in comparison to
LLDPE and HDPE. The viscosity of base bitumen was enhanced with the addition of polymer and
pseudoplastic behavior was also observed for LLDPE and HDPE. Oxidation during blending at 160C
may be one of the causes of hardness of PMB. The best results were obtained when polymer
concentration was kept below 3% for HDPE and LLDPE. PP shows least variation in all the test results.
Keywords: Polymer modified bitumen, Rheology, Linear low density polyethylene, High density
polyethylene, Polypropylene, Morphology.
1
Ph.D Student, Universiti Teknologi PETRONAS, Tronoh 31750, Perak, MALAYSIA
2
Associate Professor, Civil Engg. Dept. Universiti Teknologi PETRONAS, Tronoh 31750, Perak, MALAYSIA
3
Associate Professor, Civil Engg. Dept. Universiti Teknologi PETRONAS, Tronoh 31750, Perak, MALAYSIA
4
Associate Professor, Chemical Engg. Dept. Universiti Teknologi PETRONAS, Tronoh 31750, Perak, MALAYSIA
*Correspondence Author:Noor Zainab Habib, Universiti Teknologi PETRONAS, Malaysia. Tel: +600136059264
Fax: +6053656716. E-mail:noorzainab47@hotmail.com
79
INTRODUCTION
Thermoplastic copolymers have been widely used for the modification of bituminous pavement.
The objective of this paper is to investigate the effect of the thermoplastic copolymer on bitumen as the
efficient use of these polymers can only be made if its viscoelastic behaviour is completely understood.
Rheological and morphological analysis is used as it gives the insight picture of the phase system of the
modified blend and its effect on the viscoelastic behaviour.
The selected thermoplastic copolymers LLDPE, HDPE and PP when blended with the bitumen
show different compatibility due to difference in molecular weight and structure of base bitumen. The
results from the empirical test and morphological test revealed that the concentration of the polymer in the
blend and its compatibility with the base bitumen are the governing factor of the viscoelastic behaviour of
the polymer blend.
LITERATURE REVIEW
Continuous increase of wheel loads, tyre pressure and change in climatic condition severely affect
the performance of bituminous mix pavements. The main modes of deterioration encountered usually are
rutting, fatigue and temperature cracking. In order to overcome this problem, polymer modification is
considered as best solution as the modification of virgin bitumen with synthetic polymer binder provides
a way to overcome new technical demands (Lesueur 2009). The most commonly used polymer globally
include approximately 75% elastomeric modified binder, 15% plastomeric and remaining 10% belongs to
either rubber or other modification(Gonzalez et al 2004).
This study focus on the use of thermoplastic copolymers namely, linear low density polyethylene
(LLDPE), high density polyethylene (HDPE) and polypropylene (PP) as modifier for the 80pen base
bitumen. Concentrations of all polymers were kept till 5% by weight of the bitumen. As most polymers
when blended remains insoluble to certain extent in bitumen thus phase separation may occur which may
be due to the formation of a multiphase system, a phase rich in asphaltenes not absorbed by the polymer
(Bardesi A, et al. 1999) which enhances the viscosity by the formation of more complex internal structure
(Hadidy Al, Qui Tan Yi 2009). Thus to study this multiphase system, beside empirical test field emission
electron microscopy FESEM was also used, to have an insight view which helps in understanding the
morphology and phase system existing in the polymer blend. While mixing either using mechanical or
chemical method the differences in molecular weight and polarity of base bitumen and polymer has
critical affect on compatibility (J.Stastna et al. 2002).Thermoplastic does not swell well when blended
with bitumen but increases the viscosity, even this minor volumetric change was because of the structural
change in the PMB blend. Polyethylene which belongs to plastomer gives rigidity to the binder and
reduces the deformation under load (Hadidy Al, Qui Tan Yi 2009).The affect of property become more
significant when the concentration of polyethylene was kept below 1% by weight of the base bitumen (
Yousefi AA, Kadi A AIT 2000). For LLDPE the concentration up to 2.5% shows better results in terms of
Marshall Stability, resilient modulus, water susceptibility and fatigue life of the modified binder (Hadidy
Al, Qui Tan Yi 2009).Thermoplastic when mixed with bitumen even at ambient temperature increases the
viscosity and thus stiffness at service temperature but unfortunately do not show any significant elastic
behavior (Lu.X and Isacsson1997). Polyethylene morphology is strongly affected under stress and
deformation as sliding of chains with respects to entanglements occurs at the nodes (Sybilski.D (1997).
Addition of bitumen improves deformation resistance as the viscosity of blend enhanced tremendously
which is observed with increase in softening point and decrease in penetration values (Hadidy Al, Qui
Tan Yi 2009). PMB blend also shows a two-phase morphology, depending on the polymer type,
concentration, base bitumen composition and the aromaticity (lower molecular weight polar naphthenic),
80
the system can have an asphalt-rich phase with a homogeneous dispersion of polymer spheres or
continuous polymer-rich phase with dispersion of asphaltene globules in it(DrozdovAleksey.D, Yuan Q
2003). When a polymer and compatible base bitumen are mixed, the polymer strands absorb part of the
low molecular weight oil fraction of the base bitumen and become swollen (Sybilski.D 1997). When the
polymer-rich phase becomes the continuous phase due to the relatively higher fraction of swollen
polymer, the swollen strands connect together at nodes and form a three dimensional network. This
network significantly affects the mechanical properties of the binders and ultimately the bituminous
binder mixes (Sybilski.D 1997). Thus the rheology of the multi-phase systems is very complex due to
instability of the mix as bitumen and polymer having different density and viscosity at 25C and 160C,
which ultimately influence morphology (Wekumbura et al. 2007).
In short to know about the internal structure of polymer modified bitumen, rheological and
morphological studies of the PMBs is the only answer. The development of new polymer blends of
specific properties using different mixing technique with varying ratio of polymer is the most difficult
task as it is related to the behaviour of the material which in turn depend upon its chemical structure and
internal phase formation which ultimately controls the rheology of the PMB and its mechanical
properties. The objective of this paper is thus to investigate the characteristics of the polymer modified
bitumen PMB using empirical test, morphological analysis and relating its rheolgical characteristics to its
mechanical properties at the later part of this study.
METHODOLOGY
Materials used in this study includes

80/100 penetration grade base bitumen obtained from PETRONAS refinery at Malacca,
Malaysia.
The polymer used for modification supplied by Titan Polyethylene Malaysia, were Linear low
density polyethylene LLDPE, High density polyethylene HDPE and Polypropylene Resin PP.
LLDPE and HDPE were in pallet form while Polypropylene was in powder form.
1) Sample Preparation
Samples were prepared using melt blending technique. The bitumen about 400gm was heated in oven till
fluid condition and polymer was slowly added, while the speed of the mixer was maintained at 120 rpm
and temperature was kept between 160C and170C. The concentration of PP, LLDPE and HDPE used,
were 0.5, 1.0, 1.5, 2.0, 2.5, 3 and 5% by weight of blend. Mixing was continued for 1 hr to produce
homogenous mixtures. The modified bitumen was then sealed in containers covered with aluminium foil
and stored for further testing. Empirical test such as penetration, softening point and viscosity were then
conducted on the prepared samples.
2) Laboratory testing
Test on the prepared samples was conducted according to ASTM (ASTM Standard 2000) method to
characterize the properties of PMBs. The different percentage of polymer concentration provides a wider
range of results which helps in analysing each type of the polymer blend at particular concentration. The
rheological test includes penetration at 25C, softening point and viscosity test. The test results of base
bitumen are presented in Table 1.
81
Table 1: Conventional properties of base bitumen
Grade Pen @ 25C (dmm) Soft. pt (C) PI Viscosity@135C

(Pa s)
80/100 84 53 0.5 0.4
Penetration (ASTM D-5)
The standard penetration test at 25C was performed on base bitumen and PMB with the concentration of
polymer varying between 1%- 5% by weight of the bitumen. The results of the test are shown in Table 2.
Table 2: Penetration and softening point results

Polymer HDPE LLDPE PP
% 1% 2% 3% 5% 1% 2% 3% 5% 1% 2% 3% 5%
Polymer
Pen. 32 22 23 23 35 30 25 13 34 30 28 15
(dmm)
Soft. 51 58 67 79 53 53 60 67 54 55 55 59
Point C
Softening Point (ASTM D-36)
Ring and ball is the standard test to determine the consistency of the bitumen, which represent the
temperature at which a change of phase from solid to liquid occurs. The results are shown in Table 2.
Viscosity (ASTM D-4402)
Viscosity test was conducted using Brook Field viscometer on neat bitumen and polymer modified
bitumen. The dynamic viscosity measurement were made at 135C .The test was carried out on all
polymers up to the concentration of 3% and viscosities measurements were made on increasing shear
rates( shear stresses).
RESULT ANALYSIS
1) Penetration Results
From Table 1 it was observed that there is a sharp decrease in the penetration value of base bitumen from
84dmm to 36dmm for PMB at 1% concentration of polymer shows that increase in the hardness of the
PMB was because of the use of the high molecular weight polymer. It is obvious from the observation
that thermoplastics influence more on the penetration with the increase in the viscosity of the bitumen
(Whiteoak 2003).Since the melting temperature of all polymers are below160C, thus at mixing
temperature of 160C it absorb some oil and release low molecular weight fraction into the bitumen
which increases the viscosity of the PMB (Yousefi AA, Kadi A AIT 2000). Thus by the end of mixing
process, viscosity increases and by the time it cools harden mixture was formed. The hardening of the
bitumen can be beneficial as it increases the stiffness of the material, thus the load spreading capabilities
82
of the structure but also can lead to fretting or cracking (Whiteoak 2003). This can be evaluated at later
stage of this study by performance tests.
2) Softening Point Results
The results obtained from the test are shown in Table 1. It shows that there is no great difference in the
softening temperature for PMBs up to 2% concentration of polymer in blend as compared to base
bitumen, as thermoplastics modification does not significantly affect the softening point as compared to
the penetration (Whiteoak 2003). One of the cause was the internal structure formed by polymer seems to
be thermodynamically stable which do not significantly affect the softening point of the PMB, but as the
concentration of polymer increases HDPE and LLDPE shows rapid increase in softening point in
comparison to PP. This can be considered as positive sign and can be further checked with the
performance test at the later stage of this study, whether it enhance the pavement performance
characteristics in terms of rutting, fatigue and temperature susceptibility. From Table 2 it can also be
observed that PP offer lesser variation in softening point which may be due to homogeneity achieved
during blending of PP with base bitumen as PP is in powder form, which reflects that it may offer better
rutting resistance at higher temperature when tested later.
3) Viscosity of Bitumen and PMB blend
The flow behaviour of the material described in terms of viscosity, exhibits Newtonian and non-
Newtonian characteristics depending on the composition and source of the crude. Temperature and
loading also affect the behaviour describing the viscoelastic properties of the material. The internal
structure of the base bitumen also play key role (Sybilski.D 1997). Bitumen with viscosity of 0.4Pas at
135C shows increase in viscosity with the increase in polymer concentration. The non- Newtonian
characteristic with the decrease in viscosity with the increase in shear rate was observed for all
concentration of polymer in blend. These non-Newtonian phenomena are dependent on the shear rate and
it influences the internal structure of the PMB (Sybilski.D 1997). From Figure 1- 3, shows the non-
Newtonian behaviour with the decrease in viscosity with increase in shear rate. The fluctuation of
viscosity was observed more for HDPE and LLDPE till 3% concentration of polymer as shown in Figure
1-3. This mixed behaviour with the decrease in viscosity with increase in shear rate and again increase in
viscosity do not truly represent non-Newtonian behaviour but it represent thixotropic effect. The
thixotropic behaviour of modified bitumen may be due to reversible breakdown of structure which is
commonly found in multiphase system of polymer modified bitumen (Bhattacharya Sati N 1997). It is
very difficult to say that polymer modified bitumen purely exhibit shear thinning phenomenon with the
decrease in viscosity with the increase in shear rate as shear thickening phenomenon was also observed.
This pseudo plastic behaviour of PMB may be attributed due to breakdown of structure, or polymer
particles existing in equilibrium state being more aligned offering lesser resistance to flow but with the
increase in shear rate these offer higher resistance due to agglomeration, aggregation or flocculation of
particle in multiphase system as the interparticle forces like Brownian, van der Walls forces become
prominent (Bhattacharya Sati N 1997). HDPE and LLDPE as both are not fully dispersed in bitumen
because of higher molecular weight exhibit both thixotropy and viscoelasticity at all concentration of
polymer in the blend. PP offer less fluctuation as can be observed for all concentration of polymer in the
blend as shown in Figure 1-3. One of the main cause of this instability is the tendency of PMB to
morphological phase separation which is due to Brownanian coalescence followed by gravitational
flocculation and later on end up with creaming (Gonzalez et al (2004).As this viscometer did not allow
measurement of viscosity at both higher and lower shear rate, true statement about the shear thickening or
shear thinning cannot be made. It was also observed that for HDPE it shows shear thickening behaviour at
shear rate around 2000 1/sec to 3000 1/sec., which may be due to agglomeration of HDPE particle at
higher temperature of 135C, in which the chain realign and offer more resistance to higher shear stress.
The phenomenon which was observed for HDPE and LLDPE may offer better resistance to rutting when
83
analysed at later stage of this study. More expectation are made for LLDPE and PP as rearrangement of
molecules were observed more for these two types of polymer by linear decrease in viscosity with the
increase in shear rate but actual behaviour can only be obtained when tested for fatigue and rutting at later
stage of this study.
Figure1: Viscosity of 1% PMB
Figure 2: Viscosity of 2% PMB
Lu.X and Isacsson(1997
Figure 3: Viscosity of 3% PMB
4) Morphological Analysis:
Field emission scanning electron microscopy (FESEM) was used to study the internal morphology of the
PMB. The analyzed sample describes the extent of the continuous phase in the blend and compatibility of
the bitumenpolymer within the blend. The compatibility of the bitumen with the polymer depends on
molecular weight, as higher molecular weight polymers are incompatible with lower molecular weight
bitumen (Whiteoak 2003). From Fig.4 till Fig. 6 the phase separation was not very prominent although
the viscosity of the blend was increased as being proved by penetration results, PP and LLDPE seems to
be better blended forming homogenous mix in comparison to HDPE which can be confirmed by the
scanned samples till 3%. HDPE phase separation becomes more prominent when the temperature lowers
84
showing two distinct phases, polymer floating on highly viscous mix which was being produced by some
of the solvation of polymer at high temperature with bitumen increasing the viscosity of the mix. From
Fig.6 incompatible phase system was observed more for LLDPE and HDPE which may be due to higher
molecular weight polymer, as higher the difference in molecular weight between polymer and bitumen,
and the presence of asphaltene which was produced due to blending PMB may be one of the cause which
disturb the phase equilibrium of the system and soon phase separation occurs (Whiteoak 2003). The
solvency of the HDPE was almost same up to 2% but the creaming effect was more observed for PP in
Fig 6, reflecting an increase in viscosity. It may be because Polyethylene at temperature greater than
160C releases lower molecular weight fraction, which enhances the viscosity (Hadidy Al, Qui Tan Yi
2009). Thus solvency of PP and LLDPE found to be better as compared to HDPE, because lower
molecular weight and less branched polymer offers less resistance while blending.
Figure 4: (a) 1% PP (b)1% HDPE (c)1% LLDPE
Figure 5: (a) 2% PP (b) 2% HDPE (c)2% LLDPE
Figure 6: (a) 3% PP (b) 3% HDPE (c)3% LLDPE
CONCLUSION
It was observed that binder rheology was strongly affected by modification with polymers.
LLDPE & PP modified bitumen shows least variation in penetration and softening point in comparison to
HDPE modified bitumen, when the concentration of polymer is kept below 3%. It was found that
thermoplastic modified bitumen was more affected by change in softening point in comparison to
penetration, which may be due to the formation of thermodynamically stable structure which may offer
85
resistance in terms of rutting, fatigue and temperature susceptibility. The sharp decrease in the penetration
value of virgin bitumen after modification, even at 1% polymer concentration shows rapid enhancement
of the hardness of the PMB .One the cause of it may be due to the formation of dispersed polymer
network that may offer better resistance to pavement deformation. Thus the viscoelastic behaviour of
virgin bitumen was strongly influenced by polymer concentration, temperature and penetration grade of
the bitumen. The Newtonian behaviour of virgin bitumen is strongly affected by the addition of polymer
in blend as they exhibited non-Newtonian behaviour. Pseudoplastic behaviour was more prominent for
concentration at 1% and 2% for HDPE and LLDPE, as pallet form of both polymers were not fully
absorbed and was dispersed in the bitumen even when mixed at 160C. Viscosity at 3% for all polymer
blends showed a linear decrease in viscosity with the increase in shear rate when the temperature was kept
constant at 135C.Polypropylene with linear structure added in powder form to the base bitumen was
easier to blend to produce homogenous blend as observed by the linear decrease in penetration, increase
in softening point and increase in viscosity, thus presenting the continuous formation of stable internal
structure. It was found that HDPE and LLDPE exhibited both thixotropic and viscoelastic behavior at all
concentration of polymer in the blend. HDPE showed shear thickening behaviour at shear rate around
2000 1/sec to 3000 1/sec., which may be due to agglomeration of HDPE particle at higher temperature of
135C and may offer better resistance to rutting. FESEM of PMBs up to 3% shows better blended mix,
which reflects that lower the concentration of polymer in mix, better the blends can be achieved.
REFERENCES
Whiteoak CD, (2003) The shell bitumen hand book. Surrey, UK.
Yousefi AA, Kadi A AIT (2000) Composite asphalt binders: Effects of modified RPE on asphalt
Journal of Materials in civil Engineering, pp113-123
Didier Lesueur (2009) The colloidal structure of bitumen: Consequences on the rheology and on the
mechanism of bitumen modification. Advances in colloid and Interface Science.145, pp 42-82.
Gonzalez et al (2004) Rheology and stability of bitumen/EVA blends European Polymer Journal,
2004Vol 40, pp 2365-2372
Bardesi A, et al. (1999) Use of modified bituminous binders, special bitumens with bitumen additives
in pavement applications. Technical Committee Flexible Roads (C8) World Road Association
(PIARC) 1999
Hadidy Al, Qui Tan Yi (2009) Effect of Polyethylene on life of flexible pavements Construction and
Building Materials 23, 2009, pp 1456-146
J.Stastna et al. (2002) Viscosity functions in polymer modified asphalts Journal of Colloid and
Interface Science, 2002, vol 259, pp.200-20
Lu.X and Isacsson(1997) Compatibilty and storage stability of styrene-butadiene-styrene copolymer

modified bitumens Journal of Materials and Structure,1997, Vol.30, pg.618-626
Isacsson and Lu (1998) Testing and appraisal of polymer modified road bitumens RILEM Report 17,
1998, pp 1-38
86
DrozdovAleksey.D, Yuan Q (2003) The viscoelastic and viscoplastic behavior of low density
polyethylene International Journal of Solid and Structure, 2003, vol. 40, pp2321-2342.
Sybilski.D (1997) New simplified equation for the computation of absolute viscosity of polymer-
bitumens Materials and Structure vol.30, April 1997, pp182-187
Wekumbura et al. (2007), Destruction and Recovery of Internal Structure in Polymer-Modified

Asphalts Journal of Materials Civil Engineering, 2007, pp 227-232
Chen Jian-Shiuh et al (2002). Asphalt Modified by Styrene-Butadiene Styrene Triblock Copolymer:

Morphology and Modal Journal of Materials in Civil Engineering May/June 2002, pp 224-229
Traxler(1961) Ashaplt its composition, properties and uses Reinhold publishing corporation.Chabman
& Hall.Ltd., London.1961
Bhattacharya Sati N, (1997) Rheology Fundamentals and Measurement chapter 1, Royal Melbourne
Institute of Technology. Australia 1997
American society for Testing and Materials (ASTM), standard Specification, Section 4, vol. 04-03; 2000
87
88
THE EFFECT OF FINE AGGREGATE PROPERTIES ON THE RUTTING BEHAVIOR OF

THE CONVENTIONAL AND POLYMER MODIFIED BITUMINOUS MIXTURES USING
TWO TYPES OF SAND AS FINE AGGREGATE
Ibrahim Kamaruddin 1, Madzlan B. Napiah 2, Yasreen Gasm ALkhalig 3*
ABSTRACT:
Permanent deformation (rutting) continue to be the main challenges in improving the performance of
bituminous mixture pavements. The purpose of this paper is to look at some aspects of the effects of
fine aggregate properties on the rutting behavior of hot mix asphalt. Two types of sand (quarry and
mining sand) with two conventional bitumens (PEN 50/60 & PEN 80/100) and four polymer
modified bitumens PMB (PM1_82, PM1_76, PM2_82 and PM2_76) were used. Physical, chemical
and mechanical tests were performed on the sands to determine their effect when incorporated with a
bituminous mixture. According to the wheel tracking results, quarry sand that has more angularity,
rougher, higher shear strength and a higher percentage of Aluminum oxide presented higher
resistance to rutting. Also a PMB mixture gives better rutting results than conventional mixtures, this
is due to the PMB having better viscosity property than that of the conventional bitumen.
Keywords: fine aggregate, physical property, chemical property, mechanical property, Marshall Test,
wheel tracking test.
1
Lecturer, Universiti Teknologi PETRONAS, Bandar Seri Iskandar, 31750 Tronoh, Perak, Malaysia.
2
Lecturer, Universiti Teknologi PETRONAS, Bandar Seri Iskandar, 31750 Tronoh, Perak Malaysia.
3
Student, Universiti Teknologi PETRONAS, Bandar Seri Iskandar, 31750 Tronoh, Perak Malaysia.
*Correspondence Author: Yasreen Gasm ALkhalig, Universiti Teknologi Petronas, Malaysia. Tel:
+60176291665. E-mail: neersay@yahoo.com, neersay@gmail.com
89
INTRODUCTION
Hot mix asphalt (HMA) pavement has been found as an alternative, to prevent premature failure.
However the use of HMA mixture in pavement construction has been associated with some
performance problems which become the main focus of present day research. The frequent problems
associated with road pavements like rutting, abrasion, fatigue cracking, thermal cracking, aging and
stripping cause the roads to wear away or fail (Navarro et al, 2002 ; Lu and Isacsson, 1997). Among
these road related problems, rutting is considered to be the main problems in highway pavements (Lu
et al, 1998). Rutting occurs only on flexible pavements as indicated by the permanent deformation or
rut depth along the wheel paths. Earlier studies have identified that rutting occurs as a result of
accumulated plastic deformation due to high traffic loads and high temperatures (Navarro et al, 2007).
Another study was considered rutting as a longitudinal subsidence localized in the wheel tracks
caused by vehicles loads. It occurs when road does not have sufficient stability of the asphalt material
at the surface, insufficient compaction of the pavement and insufficient pavement strength (Robinson
& Thagesen, 2004). Many studies have found that the properties of fine aggregates namely its
physical, chemical and mechanical properties played a significant role in improving the bituminous
mixtures rutting performance.
Several researches have been carried out on mixtures using polymer modified bitumen to
increase its resistance towards permanent deformation (rutting).
LITERATURE REVIEW
Fine aggregate is a primary constituent in asphalt mixtures. For that reason, the properties of fine
aggregates namely its physical, chemical and mechanical properties played a significant role in
determining the characteristics of the resulting bituminous mixtures.
The physical characteristics of fine aggregate (shape and surface texture) have been found to
affect the asphalt mixture properties and its performance (Topal & Sengoz, 2005; Eyad et al., 2001;
Chapuis & Legare, 1992). The amount of chemical component of aggregate was found to effect on the
bonding strength and on the adhesion performance between aggregate and binder and also on the
hardness of the bituminous mixture (Abo Qudais & Al Shweily, 2007; Wu et al., 2007). Other factors
that can influence the performance of HMA mixture are the mechanical properties of the fine
aggregate (Topal & Sengoz, 2006; Das, 1998; Park & Lee, 2002). It was found that HMA mixtures
containing river gravel fines and natural sand fines shows higher rut depth, while the mixture
containing granite and limestone fines shows lower rut depth (Park & Lee 2002). Fine aggregate
surface texture plays an important role in HMA rutting resistance. A good correlation between
frictional angle and rut depth was found, therefore HMA mixtures containing river fines with a
measured friction angle in direct shear test of 40.30 has the highest rut depth compared with mixtures
containing granite fines with a friction angle of 45.20 (Park & Lee, 2002). Aggregate gradation
(distribution of particle sizes) is one of the most important factors to resist pavement distress (Shen et
al, 2005).
Some researchers investigated the possibility to overcome the rutting phenomena by using
different types of polymers. Tayfur et al. (2007) used five types of polymer to modify asphalt mixture
namely amorphous polyalphaolefin, cellulose fibers, polyolefin, bituminous cellulose fibers and
styrene butadiene styrene; while Ahmedzade and Yilmaz (2007) used polyester resin (PR). Another
study by Sirin, et al. (2006) found that the high viscosity of SBS modified binder was the main reason
for the high rutting resistance. All of these polymers were found to increase the resistance to
permanent deformation of asphalt concrete mixture.
90
METHODOLOGY
In this study several tests were carried out to determine the following properties: specific gravity
of coarse aggregate, fine aggregate and filler, physical, chemical and mechanical properties of fine
aggregate, physical properties of polymer modified bitumen and conventional bitumen, Marshall Test
and wheel tracking test.
The conventional bitumen were PEN 50/60 and PEN 80/100, while the polymer modified
bitumen were designated as PM1 (PM1_82& PM1_76) and PM2 (PM2_82& PM2_76). PM1
modified bitumen is made up by adding styrene butadiene styrene (SBS) to bitumen penetration 82
and bitumen penetration 76 to come up with PM1_82 and PM1_76. PM2 modified bitumen is made
up by addition one of the plastomer polymer to bitumen penetration 82 and bitumen penetration 76 to
come up with PM2_82 and PM2_76. Both of the conventional bitumen are manufactured from refined
crude oil and were obtained from Bellamy Precision and Berne Science respectively. Polymer
modified bitumen was manufactured by adding polymers to the bitumen and were obtained from
PPMS Technologies supplier in Malaysia.
Table 1 shows the physical properties of polymer modified bitumen and base bitumen. The
consistency tests proved that bitumen PEN 50/60 has higher consistency (harder) than bitumen PEN
80/100. Because PEN 50/60 has lower penetration and ductility, and higher softening point compared
to the 80/100. It can be also noticed that polymer modified bitumen has lower penetration and higher
softening point compared to the conventional bitumen. This means the stiffness for polymer modified
bitumen is higher than the conventional bitumen, thus exhibit better resistance to rutting. PM1 has
lower penetration, higher softening point and lower density and higher ductility values as compared to
PM2. This is because PM1 polymer modified bitumen consists of SBS polymer and the SBS which is
considering one of the elastomer polymers consists of two elements: hard polystyrene end blocks and
rubbery midblock. The hard polystyrene end blocks give high tensile strength and flow resistance at
high temperature, whereas the rubbery midblocks are responsible for the elasticity, fatigue resistance
and flexibility at low temperature (Ahmedzade et al., 2007).
Table 1: Properties of binder

Penetration Specific
Softening Ductility
Type of @ 250C gravity at
point 0C (cm)
bitumen (0.1mm) 25 0C
ASTM D36 ASTM D113
ASTM D5 ASTM D04
PEN 50/60 49 52 36 1.032
PEN 80/100 90 49 125.6 1.030
PM1_82 39 70 59.1 1.028
PM1_76 47 67 56.4 1.027
PM2_82 48 65 35.5 1.032
PM2_76 48 53 29.6 1.035
Table 2 shows the specific gravity and source of coarse aggregate, fine aggregate and filler.
Table 2: Specific gravity of aggregate

Types of the aggregate Source Specific gravity(g/cm3)
Coarse aggregate (granite) Quarry 2.655
Fine aggregate (Mining sand) Pond 2.695
Fine aggregate (Quarry sand) Quarry 2.690
Filler (Portland cement) Factory 3.135
91
Physical Property
The physical property of sand can be determined by using the fine aggregate angularity (FAA)
test. Table 3 shows the fine aggregate angularity results. The fine aggregate angularity is measured by
determining the percentage of voids in the sand, the higher the percentage of voids the more angular
the fine aggregate. It can be noticed from the table, that quarry sand has more angularity compared to
the mining sand. It was found that fine aggregate with higher values of FAA produce more angular
particles and greater rough surface texture, resulting in a larger interlock between the particles
consequently resulting in higher shear strength, and this is enhanced the rutting resistance (Fernandes
& Gouveia, 2003).
Table 3: Fine aggregate angularity for the sand used

Fine aggregate
Types of the aggregate
angularity (FAA) %
Fine aggregate (Quarry sand) 46.91
Fine aggregate (Mining sand) 42.58
Chemical Property
The chemical composition of sand can be determined by the X-ray fluorescence (XRF) test. The
XRF test result in Table 4 showed that mining sand has the higher percentage of SiO 2 (80%),
compared to the quarry sand (66.1%). It was found that large amount of SiO 2 can cause stripping of
HMA pavements because silica reduces the bond strength between the aggregate and binder (Abo-
Qudais & Al-Shweily, 2007). Quarry sand has the higher total percentage of alumina (Al 2 O 3 ) i.e.
12.6% compared to the mining sand which contains 8.98%. The Al 2 O 3 content is related to the
hardness of the material, this can be seen from the Table 5. Quarry sand also has the higher total
percentage of hematite (Fe 2 O 3 ) i.e. 5.30% compared to the mining sand which contains 2.2% of the
hematite. The Fe 2 O 3 content which has the smallest particle size compared to the other elements as
shown in the Table 5 increases the density of the bituminous mixture. The oil absorption property is
needed to absorb the extensive oils in the bituminous mixture; this property could decrease the rutting
behavior (Wu et al, 2007). Al 2 O 3 also has the highest oils absorption value (25-225 g/100g) followed
by Fe 2 O 3 and SiO 2 respectively. This indicates that quarry sand showed the highest ability to absorb
the extensive oils in bituminous mixture, because it has got higher Al 2 O 3 and Fe 2 O 3 whilst mining
sand exhibited the lower ability to absorb the extensive oils in bituminous mixture.
Table 4: Chemical composition results for fine aggregate
Fine CaO SiO 2 Al 2 O 3 Fe 2 O 3 K2O Na 2 O

aggregate (%) (%) (%) (%) (%) (%)
type
Quarry sand 7.19 66.1 12.6 5.30 10.5 1.95
Mining sand 3.6 80.0 8.98 2.2 2.2 0
Table 5: Characteristics of Compounds (Wypych, 1999)

Solubility in Hardness Particle Size Oil Absorption
Components
Water (Mohr) (m) (g/100g)
CaO Hydrophilic - - -
SiO 2 Hydrophobic 7 2-19 17-20
Al 2 O 3 Hydrophobic 9 0.8-10 25-225
Fe 2 O 3 Hydrophobic 3.8-5.1 0.0130-105 10-35
K2O Hydrophilic - - -
92
Mechanical Property
The mechanical Property of sand can be determined by using shear box test. The angle of internal
friction () is an indication of particle interlocking and hence particle shape and surface texture. The
direct shear test results in Figure 1 showed that quarry sand has the higher value (450) compare to
the mining sand (34.90). It was found that rounded smooth textured aggregate particles tend to slide
past one another producing a HMA mixture with relatively low shear strength (Park & Lee 2002).
These results explain why crushed aggregate like quarry sand has higher shear resistance compared to
natural aggregate samples (mining sand). Because natural sand have more rounded particles with
smooth surface textured resulting in less interlock between particles (Fernandes & Gouveia, 2003).
The high shear resistance is an indicator of resistance to mixture deformation (Park & Lee 2002;
Topal & Sengoz, 2006). Study confirmed that the higher internal friction angle of fine aggregate
indicates better interlocking mechanism results in a more resistant granular structure (Fernandes &
Gouveia, 2003).
Figure 1: values for different types of sand.
Marshall Test
Marshall Test determination of optimum asphalt content by weight of total mix is done by using
Marshall Mix design method (Aksoy et al, 2005). The procedure involves 75 blows per face; three
specimens of each asphalt cement contents (3.5 %, 4.0%, 4.5%, 5.0%, 5.5%, 6.0%, 6.5%, and 7%) are
prepared. A total of 288 specimens have been tested with the stability, flow, air voids, unit weight,
voids filled binder and voids in mineral aggregate (Ahmedzade & Yilmaz, 2007; Ahmedzade et
al,2007). The optimum binder content is calculated as the average of asphalt contents that meet the
maximum stability, maximum unit weight, minimum voids in mineral aggregate and 5.0% air voids.
The results of optimum bitumen content are tabulated in Table 6. It can be seen that mining sand has
higher OBC compared to the quarry sand, this is because of the higher surface area of the mining sand
which need higher amount of binder to cover the particles.
Table 6: OBC (%) for all types of binder & sands
Type of binder Quarry sand Mining sand

PM1_82 4.19 6.33
PM1_76 4.53 6.06
PM2_82 4.48 6.40
PM2_76 4.69 6.61
PEN 50/60 4.59 5.86
PEN 80/100 4.80 6.14
93
Wheel tracking test
Wheel tracking test was used to measure the rutting behavior on the asphalt mixture and it was
conducted at 40 0C. An actual wheel of 200 mm diameter and 50 mm width with a total wheel load of
520 N was applied on the square specimen as shown in Figure 2. The wheel was run backward and
forward across the centre of the bituminous specimen at the frequency of 42 times per minute for 46
minutes loading. The total rut depth was determined and recorded by the Wessex software that came
together with the testing machine. The testing procedures conform to the specification of BS 598 110:
1998. Figure 3 shows the rut depth for the two sand samples having different stiffness values.
Figure 2: Wessex wheel trackers
High rut depth for mining Low rut depth for quarry
sand mixture sand mixture
Figure 3: Wheel tracking specimen with different rut depth
RESULT ANALYSIS
Wheel Tracking Results:
It was found that wheel tracking test has been used by many researchers because it can be carried
out in many shape like Georgia loaded wheel test (GLWT) and Asphalt Pavement Analyzer (APA)
where the sample can also be fabricated in a shape of a beam or cylindrical (Collop et al, 1995). The
choice of sample can also be taken from core extracted from road structure. Therefore the laboratory
results can be compared with the actual performance of the road structure, because the test simulates
the actual road conditions. The test machine was set to operate for 46 minutes at 42 cycles per minute.
The rut depth was therefore obtained after 1932 cycles. The wheel tracking test results are presented
in Table 7.
94
Table 7: Wheel tracking results for the different sands
Type of fine aggregate/ Rut depth (mm)

Type of binder
quarry sand mining sand
PM1_82 0.04 1.8
PM1_76 0.6 2.5
PM2_82 0.3 2.8
PM2_76 0.5 2.7
PEN 50/60 2.1 3.7
PEN 80/100 3.4 4.6
From Table 7, it can be noted that quarry sand mixture exhibit higher resistance to rutting compared
to the mining sand. This is because quarry sand mixture has the least rut depth as compared to the
mining sand. This behavior has similar trend as physical, chemical and mechanical properties of fine
aggregate and as other researchers have obtained that physical, chemical and mechanical properties of
fine aggregate could improve the rutting resistance (Park & Lee 2002; Abo-Qudais & Al-Shweily,
2007; Topal & Sengoz, 2006; Lee et al, 1999; Eyad et al, 2001). The mixtures containing bitumen
PEN 80/100 exhibits higher rate of deformation compared to the mixture containing bitumen PEN
50/60 and polymer modified bitumen. This trend was also observed in two types of sand. Bitumen
PEN 80/100 has lower viscosity and stiffness, therefore it can deform easily at long time of loading
and high temperature. It is also observed that all rut depth values are affected by the use of polymer
modified bitumen, in contrast to conventional mixtures. The mixture containing polymer modified
bitumen (PM1& PM2) has lower rate of deformation compared to the mixture containing
conventional bitumen and this agrees with earlier study (Tayfur et al, 2007). This observation can also
be explained by the viscosity of the PMB, which is higher in PMB than conventional binders. Earlier
study found that more viscous binder has better rutting resistance (Sirin et al, 2006).
CONCLUSIONS
Based on the experimental results the following conclusions can be drawn:
The penetration and softening point results have demonstrated that the use of polymer modifier
increased the stiffness of the binder at high pavement service temperature. This has the potential to
improve the rutting resistance of the polymer modified bituminous mixture. At the same time, the
stiffness of the binder at low pavement service temperatures is expected to decrease therefore
reducing the mix potential to brittleness and cracking.
The fine aggregate angularity (FAA) was measured using a simple laboratory test. The higher the
values of fine aggregate angularity, the more angular are the particles with rougher surface texture.
This resulted in a better interlocking mechanism between the particles and thus offering better shear
strength. Quarry sand was found to have the higher FAA values compared to the mining sand.
The content of alumina (Al 2 O 3 ) in fine aggregate affects the quality of bituminous pavement by
increasing the hardness of the mix and hence the pavement becomes more resistant to permanent
deformation. However, increase in hardness may also cause the pavement to become brittle faster,
promoting fatigue cracking to occur early in the pavement life. From the results obtained, it can be
concluded that quarry sand has the higher value of hardness compared to the mining sand.
The high shear strength is a good indicator of bituminous mixture resistance to rutting. Quarry
sand portrayed the higher shear strength value compared to the mining sand. A good correlation has
been found between the angle of shear resistance () and rut depth. The was obtained from the
direct shear test and the rut depth was obtained from the wheel tracking test. Therefore a fine
95
aggregate mixtures containing sand with the highest frictional resistance angle exhibited lowest rut
depth.
Rutting results showed that fine aggregate with more angular particles, rougher surface texture,
higher shear strength, higher oil absorption and higher hardness values contribute to the increase in
rutting resistance of hot mixture asphalt. Mixture containing quarry sand exhibits higher rutting
resistance compared to mining sand. The conventional mixtures have maximum deformation of
about 4.6 mm while the modified mixtures have deformations of 2.8 mm indicating that modified
mixture is more resistant to permanent deformation.
REFERENCE
Abo-Qudais, S., & Al-Shweily, H. (2007). Effect of Aggregate Properties on Asphalt Mixtures
Stripping and Creep Behavior. Science Direct Journal of Construction and Building
Materials, 21, 1886-1898.
Ahmedzade, P., & Yilmaz, M. (2007). Effect of Polyester Resin Additive on the Properties of Asphalt
Binders and Mixtures. Science Direct Journal of Construction and Building Materials, 22, 481-
486.
Ahmedzade, P., Tigdemir, M . , & Kalyoncuoglu, S. F. (2007). Laboratory Investigation of the

Properties of Asphalt Concrete Mixtures Modified with TOP-SBS. Science Direct Journal of
Construction and Building Materials, 21, 626-633.
Aksoy, A., Samlioglu, K., Tayfur, S., & Ozen, H. (2005). Effects of Various Additives on the
Moisture Damage Sensitivity of Asphalt Mixtures. Science Direct Journal of Construction and
Building Materials, 19, 11-18.
British Standard Part 598-110. (1998). Method of Test For The Determination of Wheel-Tracking
Rate and Depth. Sampling and Examination of Bituminous Mixtures for Roads and Other
Paved Areas.
Chapuis, R. P., & Legare, P.P. (1992). A Simple Method for Determining the Surface Area of Fine
Aggregate and Filler in Bituminous Mixtures, Effect of Aggregates and Mineral Filler on
Asphalt Mixture Performance. ASTM STP 1147. American Society for Testing and Materials,
Philadelphia.
Collop, A. C. , Cebon, D., & Hardy, M. S. A. (1995). Viscoelastic Approach to Rutting in Filexible
Pavements. Journal of Transportation Engineering, 82-91.
Das, B.M. (1998). Principles of Geotechnical Engineering. PWS publishing company, Fourth Edition
Copyright.
Eyad, M., Dana, O., Thomas, W., & Laith, T. (2001). Correlation of Fine Aggregate Imaging Shape
Indices with Asphalt Mixture Performance. Journal of Transportation Research Record, 148-
156.
Fernandes Jr, J. L. , & Gouveia, L. T. D. (2003). Limitation of the Fine Aggregate Angularity (FAA)
Test to Predict the Behavior of Asphalt Mixtures. Department of Transportation, University of
Sao Paulo, Brazil, 1-9.
Lee, C. J. , White, T. D., & West, T. R. (1999) The Effect of Fine Aggregate Angularity on Asphalt
Mixture Performance. Final Report, Transportation Research Program, Purdue University.
96
Lu, X., & Isacsson, U. (1997). Rheological Characterization of Styrene-Butadiene -Styrene

Copolymer Modified Bitumens. Science Direct Journal of Construction and Building
Materials, 11, 23-32.
Lu, X., Isacsson, U., & Ekblad, J. (1998). Low-Temperature Properties of Styrene-Butadiene-Styrene
Polymer Modified Bitumens. Science Direct Journal of Construction and Building Materials,
12, 405-414.
Navarro, F. J., Partal, P., Garca, M. M., Martinez, B. F. J., & Gallegos, C. (2007). Bitumen
Modification With A low-Molecular-Weight Reactive Isocyanate-Terminated Polymer.
Science Direct Journal of Fuel, 86, 2291-2299.
Navarro, F. J., Partal, P., Martinez, B. F. , Valencia, C ., & Gallegos, C. (2002). Rheological
Characteristics of Ground Tire Rubber-Modified Bitumens. Chemical Engineering Journal, 89,
5361.
Park, D.W., & Lee, H. S. (2002). Test Methods for Fine Aggregate Angularity Considering
Resistance of Rutting. KSCE Journal of Civil Engineering, 6, 421-427.
Robinson, R . , & Thagesen, B. (2004). Road Engineering for Development. Published by Spon Press,
Second Edition.
Shen, D. H., Kuo, M . F. , & Du, J. C. (2005). Properties of Gap- Aggregate Gradation Asphalt
Mixture and Permanent Deformation. Science Direct Journal of Construction and Building
Materials, 19, 147-153.
Sirin, O. , Kim, H. J., Tia, M., & Choubane, B.(2008). Comparison of Rutting Resistance of
Unmodified and SBS-Modified Superpave Mixtures by Accelerated Pavement Testing.
Science Direct Journal of Construction and Building Materials, 22, 286-294.
Tayfur, S., Ozen, H., & Aksoy, A. (2007). Investigation of Rutting Performance of Asphalt Mixtures
Containing Polymer Modifiers. Science Direct Journal of Construction and Building
Materials, 21, 328-337.
Topal, A. , & Sengoz, B. (2006). Evaluations of Compacted Aggregate Resistant Test Compared with
the Fine Aggregate Angularity Standards. Science Direct Journal of Construction and Building
Materials, 22, 993-998.
Topal, A., & Sengoz, B. (2005). Determination of Fine Aggregate Angularity in Relation with the
Resistance to Rutting of Hot-Mix Asphalt. Science Direct Journal of Construction and
Building Materials, 19, 155-163.
Wu, S., Xue, Y., Ye, Q. & Chen, Y. (2007). Utilization of Steel Slag Aggregates for Stone Mastic
Asphalt (SMA) Mixtures. Science Direct Journal of Building and Environment, 42, 2580-
2585.
Wypych .G. (1999). Handbook of Fillers. Toronto: Chem Tec Publishing.
97
98
REGIONAL FREIGHT TRANSPORTATION PLANNING:

AN OVERVIEW OF DEVELOPING METHODOLOGY
Noor Mahmudah 1*, Danang Parikesit 2,

Siti Malkhamah 3, Sigit Priyanto 4, and Mark Zuidgeest 5
ABSTRACT:
There is a parallel evolution of a political economy and the transportation system either in developed or
developing countries. The growth of economic will increase the number of ton-km of freight
transportation. As flows of freight and their impacts increase, transportation planners, managers, and
operators have a more concern to develop a better methods for tracking and monitoring the flows of
commodities and analyzing the impacts of these flows to the transportation infrastructures. Current
estimation of freight flows and analysis methods have several problems, which are related to the data and
the method itself.
This paper attempts to elaborate the relationship of the economics activities of commodities toward export
- specifically the production of goods and services that generate freight transportation, the spatial data of
commodities related to the economics activities (resources, factories, and outlets), the transportation data
(infrastructures, modes, and services), and government regulations, in formulating the method for freight
transportation planning to produce an efficient freight movement to enhance sustainable mobility and
economic development in the regional scale.
Keywords: methodology; planning; regional freight transportation
1
Ph.D Candidate, Dept. of Civil and Environmental Engineering, UGM, Jl. Grafika No. 2, Yogyakarta, 55281, INDONESIA
2
Prof. Dr., Dept. of Civil and Environmental Engineering, UGM, Jl. Grafika No. 2, Yogyakarta, 55281, INDONESIA
3
4
5
Assistant Professor, Dr., ITC, University of Twente, Hengelosestraat 99, 7514 AE, Enschede, THE NETHERLANDS
*Correspondence Author: Noor Mahmudah, Jl. Grafika No. 2, Yogyakarta, 55281, INDONESIA. Tel: +62 274 6497249,
Fax: +62274631178. E-mail: mahmudah@hotmail.com
99
1. INTRODUCTION
There is a parallel evolution of a political economy and the transportation system either in developed or
developing countries. The growth of economic will also increase the number of ton-km of freight
transportation. The case of European Union and Indonesia showed that across all freight transport modes,
road freight transportation remains dominant (Stead and Banister, 2006; Lubis et al, 2005). On the other
hand, empirical studies also showed that the domination of road freight transportation caused economic
loss, safety and environmental problems. Road deterioration and traffic accident due to overloaded truck,
and adverse impact to the environment due to gas emissions and noise of truck, are the main problems
raised from road freight transportation (Kreutzberger et al, 2003).
As flows of freight and their impacts increase, transportation planners, managers, and operators have a
more concern to develop a better methods for tracking and monitoring the flows of commodities and
analyzing the impacts of these flows to the transportation infrastructures. Current estimation of freight
flows and analysis methods have several problems, which are related to the data and the method itself.
The existing freight transportation planning methods, for instance, are generally adopted from the
conventional four-stage model (Houlguin-Veras et al, 2001; Southworth, 2002; De Jong et al, 2004).
To formulate an effective strategy for regional transportation planning, the planners or policy makers
need to know the location of resources, processing or factory locations, and the final outlet (ports) to send
their goods to the outside world. To do this, various characteristics of the different commodities and the
available networks that can be used, need to be known. An understanding of how the relationships
between commodities activities and transportation are linked by spatial structure is essential. Therefore,
this study attempts to elaborate the relationship of the economics activities of commodities toward export
- specifically the production of goods and services that generate freight transportation, the spatial data of
commodities related to the economics activities (resources, factories, and outlets), the transportation data
(infrastructures, modes, and services), and government regulations, in formulating the method for freight
transportation planning by considering intermodality concept, to produce an efficient freight movement to
enhance sustainable mobility and economic development in the regional scale.
In general, this paper aims to formulate the methodology for a regional freight transportation planning by
employing the concept of inter-modality and spatial models in order to achieve the efficient transportation
system. In more specific, the objectives of this paper are: a) to summarize the conceptual framework of
the current models of freight transportation planning; and b) to elaborate the developing method of
regional freight transportation planning.
2. LITERATURE REVIEW
There is a strong relationship between transportation and regional economic development. Moreover, the
availability and performance of transport infrastructures are very important in supporting freight transportation.
According to Lem (2002), the relationship between transport infrastructure and economic development
have, at least, three important implications for the goods movement and trade. Firstly, the relationship
between transport infrastructure and regional productivity implied that the improvements of transport
infrastructure are very important aspects of economic development strategy to increase regional
productivity. Secondly, transport infrastructure that adequately facilitates the goods movement of
industries within the region can reduce the production costs. Finally, transport infrastructure affects the
location of industries. It is implied that a development strategy involving attraction of new industries to a
region must ensure that regional transportation infrastructure is adequate for the targeted industries. The
access of industries to supply routes must be evaluated in accordance with existing transportation
infrastructure and the potential location of factories (plants) and distribution.
100
To enhance efficiency and quality in transportation development, it is necessary to improve the overall
cost effectiveness and functioning of transportation operations and infrastructures. Nowadays, inter-
modality is more and more realized to provide a better performing of transportation system (Kreutzberger
et al, 2006b) because of environmental concerns, reasons of efficiency and the benefits of co-ordination
of modes to cope with growing transportation flows (OECD, 1997 in Bontekoning, 2000; Bontekoning et
al, 2004). In addition, inter-modality is now widely used to enhance the freight transportation that is
seamless, efficient, and sustainable to achieve specific goals (Rodrigue, et al, 2004).
The existing models of freight transportation planning, in general, are adopting from the ones used for
passenger transportation like the four-stage-model (Houlguin-Veras et al, 2001; Southworth, 2002; De
Jong et al, 2004). Southworth (2002) introduced a model of freight transportation, which is known as a
multi-step freight planning model. The conceptual framework of this model is originally consisted of: (a)
freight generation/attraction; (b) flow modelling/trip distribution; (c) modal split; and (d) traffic route
assignment respectively. When the planning process is intended to translate the commodity flows into
vehicle movements so the fifth step is required, this is called modelling of vehicle load factors. This
vehicle/fleet loading may occur as step four in the modelling process, which is conducted after mode split
model or alternatively, it may occur at trip generation stage, producing truck trip forecasts that are
suitable for direct application to the subsequent traffic route assignment stage. Where truck traffic is
concerned, it is usual to carry out mixed freight-passenger travel assignments to capture the effects of
traffic congestion on shipment times and hence freight delivery costs. In theory, these congestion
inclusive costs can then be fed back through the freight flow modelling, modal split, and vehicle loading
steps, and iterated until the system of model equations stabilizes on a set of transportation costs and flows.
Variations on such process have been used to analyze corridor-specific, metropolitan area wide and state-
wide freight movement systems (Holguin-Veras and Thorson, 2000; 2003a; 2003b).
Holguin-Veras and Thorson (2000) and Holguin-Veras et al (2001) have promoted the models of freight
transportation planning which are known as commodity-based and trip-based models. Both models have
been identified as the appropriate models for the New York Metropolitan Region that has specific
location and transportation demands. In general, these two freight transportation models have been built
on either of two platforms: vehicle (trip)-based or commodity-based modelling. These platforms model
vehicle trips and commodity type typically by size and weight respectively. The common approaches of
both models are trip generation, trip distribution and traffic assignment.
Commodity-based models focus on modelling the amount of freight measured in tons, or any other unit of
weight. It is accepted that the focus on the cargoes enables commodity based models to capture more
accurately the fundamental economic mechanisms driving freight movements, which are largely
determined by the cargoes attributes (e.g. shape, unit weight). Commodity based modelling is comprised
of the process: (a) generation models that are used to estimate the total number of tons produced and
attracted by each zone in the study area; (b) distribution models, where the tonnage moving between each
origin-destination pair is estimated using gravity models and other forms of spatial interaction model; (c)
mode split model that is intended to estimate the number of tons moved by the various modes and is done
by applying discrete choice models and/or panel data from focus groups of business representatives; and
(d) finally, in the traffic assignment phase of commodity-based models, a combination of vehicle loading
models and complementary models that capture empty trips, applied to origin-destination matrices by
mode, are used to assign vehicle trips to the networks. Trip-based models focus on modelling vehicle-trip,
which presupposes that the mode selection and the vehicle selections were already done. Therefore, trip-
based models do not need mode split or vehicle loading models. The advantages and disadvantages of
each model can be summarized as depicted in Table 1.
101
TABLE 1. The advantages and disadvantages of commodity-based and trip-based models

Model type Advantages Disadvantages
A. Commodity-based Freight Model 1. Focus on modelling the amount Unable to capture empty
of freight (measured in tons) trips
2. Capture the economic
mechanism driving freight
movement
B. Trip-based Freight Model 1. Focus on modelling vehicle- 1. Cannot capture mode
trip choice process
2. Able to track the vehicle 2. Difficult to identify
movement economic and behavioural
3. Capture empty-trip vehicle mechanism of freight
demand
3. METHODOLOGY
The conceptual framework used in developing methodology for regional freight transportation planning
is as illustrated in Appendix 1. The regional economic activities, especially related to the production of
goods and services, is considered as the freight driven. The goods movement and transportation services
will be analyzed further through spatial location and transportation modelling. The spatial location model
will be simulating and then mapping of the optimum access/location of factories and outlets of selected
commodities in term of distance. Furthermore, transportation model analyze the freight flows originated
from the resources to the factories and from the factories to the desired outlets (ports) by considering the
optimum generalized cost as the indicator of transportation performance. The flows of these commodities
are supposed to be transported by specific transportation modes through particular route of available
transportation networks. Moreover, the vehicle ownership status will also be considered as the
attributes/factors influencing the generalized cost of these flows. The type and location of transportation
infrastructures (networks and ports) are applied as strategy (scenario) in the model to achieve the efficient
transportation system (insfrastructure and service). Hence, the expected valuable output of these models
are the optimum generalized costs resulted from the optimum location of resources, factories, and
transportation infrastructures to enhance freight mobility either for production or distribution activities in
the regional scale.
The proposed methodology for regional freight transportation planning is as depicted in Appendix 2. In
general, this method consists of two parts. The first part is desk study on socio-economic development of
particular region and selection of important commodities in the region using Regional Income (RI) model.
In this part, all data used in analysis is already available (given) in form of secondary data. The second
part is the process of freight transportation planning, in which spatial location and transportation models
are part of it. The data that will be used in model building should be obtained from the research. This part
consist of qualitative survey on goods production and movement (flow) of selected commodities,
mapping of (optimum) spatial location of selected commodities using ArcGIS, and modelling of freight
transportation using FlowMap and ArcGIS to determine the generalized cost (GC), which is used as an
indicator of efficient transportation system. Identification of type and location of infrastructures are then
considered as strategies or scenarios that are applied in transportation modelling in order to achieve an
optimum model, which is reflected into optimum generalized cost. The transportation planning processes
can be illustrated as follows:
a) The first stage is conducting desk-study on socio economic condition and regional development in
study area by analyzing the available secondary data. Selecting of specific commodities that is
influencing the regional economic growth in particular region is conducted by analyzing secondary
economics data using Regional Income (RI) approach. Afterward literature study on process of
goods production and goods transportation activities of selected commodities is conducted. This
102
aims to prepare the good quality of questioner to be used as a tool in qualitative survey on goods
production and goods movement.
b) The second stage is conducting qualitative survey on goods production and transportation of
selected commodities in study area. It is primarily directed to the shippers. However, it may also
need to survey on carriers to obtain more precise models representation of the actual conditions.
The data obtained from the survey are then refined and selected as the attributes considered in
transportation model either using ArcGIS or FlowMap. In general, the expected data from this
survey are: (1) economic activities of selected commodities, especially the activities related to
goods production and services (annual production, economic value, logistics chain, technology);
(2) spatial data related to commodity and infrastructures (e.g. location of resources, factory, outlet
of produced commodity, and transportation networks); (3) transportation data such as transport
infrastructures characteristics (road and river networks, river and sea ports), vehicle characteristics
(type, capacity, speed, ownership), and transportation services (transport cost/fare, travel time, and
reliability); and (4) policy such as government regulations and regional master plan.
c) The third stage is mapping of spatial location of selected commodities using ArcGIS. The data
required as the input are spatial data of resources (raw materials), outlet for sending the produced
goods, transportation networks (road and river), and available policy/regulations. The location of
factories is then simulated by considering the accessibility of such factories to the available
transportation networks. Regional master plan will be considered as the constraints to the model.
Shortest route (in term of distance) between particular factory to certain networks (road, river, or
both road and river) is considered as the optimum spatial location of selected commodities.
d) The fourth, fifth, sixth, and seventh stages are actually attached with transportation modelling as
well as spatial location of selected commodities. Modelling of freight transportation using ArcGIS
and FlowMap is considered as the fourth stage. The data required in this stage are spatial data of
selected commodities (location of resources, factories, and outlets), transportation availability
(networks, ports, and vehicle characteristics), and policy/government regulations. In addition, to
capture the shippers or carriers preference and intention to specific attributes of transportation such
as modes, destinations, or service quality, so the quantitative survey (the fifth stage) is required to
conduct. The outputs of freight transportation model are utility and generalized cost models.
Generalized cost is also used as indicator of transportation system performance. To improve
transportation performance, identification of type and location of transport infrastructures (sixth
stage) are then applied as strategy or scenario in the model. The transportation system performance
is checked again by comparing the value of generalized cost before and after applying such
strategy. This procedure is conducted iteratively until the satisfied model - with optimum
generalized cost, is achieved. Model calibration and validation (seventh stage) will be done by
using statistical analysis (significance test or t-test, goodness of fit statistics or R2, and root mean
square error or RMSE) and by testing the models to other areas or commodities. The valid model
obtained is accepted as the optimum model.
The freight transportation planning models, which are promoted by Southworth (2002), Holguin-Veras
and Thorson (2000) and Holguin-Veras et al (2001), give more attention to the transportation modelling
with several stages to be conducted. The freight movements, which are determined from certain
economics activities in particular location, are then converted to the tonnage of goods (flows) and
transported from certain origin to particular destination by using trip distribution, modal split and traffic
route assignment models respectively. The locations of industries or economic activities applied in the
transportation modelling are using a given data. In fact, the spatial location of these activities may differ
one to another or change rapidly and this will have significant influence to the transportation model
obtained. As a new innovation in freight transportation planning, this new method does not only consider
103
the transportation modelling but also involving the spatial location modelling/mapping, especially in
determination of factories/industries location, and considers the government regulations as the constraint,
in formulating the freight transportation planning for regional scale. Thus, it is expected that this method
can produce the more realistic/practical transportation model because of the location of economics
activities that generate freight transportation is considered in the model and can be changed dynamically
over time. This new method also offers a more efficient stage of transportation modelling where the
distribution, modal split, and route assignment models are conducted in the same time. All simulation
models will be built on the basis of up-to-date GIS software that is now widely used in the spatial location
modelling (and mapping) and transportation modelling. However, this method faces with the great
challenges in data collection (by using disaggregate data) and database construction because it requires
large data to be formatted and converted into GIS database.
4. FREIGHT TRANSPORTATION MODEL
The field of freight transportation modelling has developed very fast into different directions all over the
world. As noted by Tavasszy (2006), since the commencing of transportation modelling, freight demand
modelling has gone through a number of major development stages. In general, freight transport demand
modelling can be distinguished by considering the decision problem (e.g. production and consumption,
trade, logistics services, network and routing), typical modelling challenges (e.g. trip generation and
facility location, mode choice, intermodal transportation, and the relationship of freight economy), and
typical techniques employed (e.g. discrete choice models, network assignment, and simulation).
This research is intended to utilize the freight network simulation, especially using ArcGIS and FlowMap
softwares by considering spatial and transportation factors as the input, and government regulation as the
constrains, to modelling intermodal freight transportation in regional scale by considering the generalized
costs as efficiency indicator of optimum transportation system. It is important to be noted that spatial
location and transportation models that are then constructed in this modelling are part of proposed
methodology for regional freight transportation planning as the general objectives of this research. The
considered factors in building the models among others: the type of commodities; transportation services
(travel cost, travel time, and climate); the type of transportation modes (truck, barge/ship, truck-barge
combination); and status of vehicle ownership (own-vehicle and hired vehicle). The expected output of
the research is the optimum generalized cost resulted from the optimum location of resources, factories,
and outlets, and the type and location of transportation infrastructures to support the freight flow either for
production or distribution activities of specified commodities in the regional scale. The framework
considered to model freight transportation for the regional scale is as depicted in Figure 1.
To address the objectives as stated earlier, however, this research is limited and is only considering: (a)
commodity to be used in the model is a primary commodity toward export and is selected by considering
the added value. Other derivations of this commodity are not considered in this model; (b) The model is
only considered the simple logistics chain that consists of the resources, factories (including storage), and
ports (sea and river ports) as the outlets; (c) The specific area considered in the model is that it should be
consisted of road and river networks that can be used for freight transportation. Due to the river networks
are highly dependent on climate thus their reliability will also be considered in the model; (d)
Transportation cost (in term of economic value), travel time, and climate are the attributes/factors of
utility considered in determining the probability of choosing transportation modes (truck, barge, and
combination of truck-barge) in the first analysis of transportation model (namely transportation cost 1 or
TC1); (e) Transportation cost (in term of economic value), travel time, climate, and status of vehicle
ownership (own or hired vehicle) are the attributes/factors of utility considered in determining the
probability of choosing transportation modes (own-truck, hired-truck, hired-barge, and combination of
hired-truck-barge) in more detailed analysis of transportation model (namely transportation cost 2 or
TC2); and (f) Generalized cost, which is a combination of transportation cost and travel time cost, is used
104
as the indicator of an efficient transportation system in the shippers point of view. The optimum
generalized cost is the optimum of total cost spent from origin to the final destinations.
Transport Costs-1 (TC1)

Utility function: Type of
Transportation
Modes
Generalized Cost function:
Transport Costs-2 (TC2)

Utility function:
Type of Vehicle
Ownership
Generalized cost function:
Type of
Infrastructures
Identification of Type and Location of (road and river
Transport Infrastructures networks, river
and sea ports)
FIGURE 1. Framework of regional freight transportation model
The specification of freight transportation model can be distinguished as: 1) Mode split model at the first
layer and so called transportation costs 1 (TC-1); and 2) Mode split model at the second layer and so
called transportation costs 2 (TC-2).
1) At the first layer, transportation model is constructed by considering the mode split of truck, barge,
and combination of truck-barge. The attributes, which are represented utility, are total travel time,
vehicle operation cost, and climate. The probability in choosing specific transportation mode is
calculated using Multinomial Logit Model. The framework of mode split as illustrated in Figure 2.
Mode Split
Truck (T) Barge (B) Truck- Barge(TB)
FIGURE 2. Framework of modal split model at first layer
105
Utility functions is using linear form that is expressed as follows:

(1)
Generalized Cost is expressed as follows:
(2)
Where
Vm : utility function for mode type m
Cij : generalized cost (Rp)
: fare from i to j on link r for mode type m (Rp), measured in economic cost
: value of time (Rp/hr)
: total travel time spent from i to j on link r for mode type m (hr)
0, 1, 2 , 3 : Constants to be obtained from the model
Climate : Number of hours per year the road/river to each destination are restricted the
circulation of particular mode
Multinomial logit model utilized in the model are:
(3)
(4)
(5)
2) In the second layer, transportation model is constructed by considering the differentiation in

transportation modes (truck, barge, and combination of truck-barge) and vehicle ownership
(own/hired vehicle). Firstly, the mode split of truck, barge, and truck-barge, will be established in
the lower level. Moreover, in the upper level, these transportation modes are then grouped on the
basis of their ownership (own or hired vehicle). Nested Multinomial Logit model will be used in
analysis of mode split. The attributes considered are total travel time, travel cost, climate and
vehicle ownership. The group of transport mode is as illustrated in Figure 3.
Calculation of utility function, generalized cost, and probability at lower level is as same as the
calculation at first layer analysis. While at upper level analysis, utility function is using linear form
and is expressed as follows:
(6)
(7)
Generalized Cost is calculated as the followings:

(8)
(9)
106
Where:
Vm : utility function for transportation mode type m
Cij : generalized cost (Rp)
: value of time (Rp/hr)
: vehicle operation cost (Rp/Ton)
: total travel time spent from i to j on link r for transport mode type m (hr)
0, 1, 2 , 3 : Constants to be obtained from the model
Climate : Number of hours per year the road/river to each destination are restricted the
circulation of particular mode
Transport mode used

from i to j
Upper
Level
Own Hired
Vehicle Vehicle
Hired Hired
Lower Truck Barge
Level Own
Truck
Hired Truck
Hired Barge
FIGURE 3. Framework of modal split model at second layer
To calculate the probability of particular transportation mode in upper level, so the Nested
Multinomial Logit model adopted from Jiang et al (1999) is utilized:
(10)
Travel cost for hired vehicle is fare or tariff in term of economic cost, which is obtained from the
qualitative survey. This assumption is taken because the transport cost of freight transport of hired
vehicle is generally negotiable thus difficult to define precisely (Ortuzar and Willumsen, 1994;
Friedrich et al, 2003). Further, the cost of freight transport of own-vehicle defined as vehicle
operation cost that is calculated using economic cost. Freight transportation modellings are
conducted using software of ArcGIS version 9.3 and FlowMap Educational Edition 7.3.1.
107
5. INITIAL RESULTS OF ANALYSIS
5.1. CASE STUDY OVERVIEW
Recently the government of Indonesia stated that palm oil industry plays more important roles, at least, in
three areas of the economic development. Firstly, it contributes significantly to the creation of millions
employment opportunities and the alleviation of poverty, especially in rural areas. Secondly, palm oil
industry is also one of the engines of regional development. Many areas developed mainly due to the
existence and expansion of this industry. Thirdly, palm oil industry has become major non-oil export
earnings. In 2008, for example, the export earnings from palm oil valued at around US $ 12.4 billion
dollars, which is the largest contributor of export earnings non-oil exports (The Coordinating Ministry for
Economic Affairs, 2009).
The Province of Central Kalimantan is the fourth-largest-palm plantation in Indonesia, after the Province
of Riau, North Sumatra and South Sumatra respectively. In 2009, the total area of palm plantations in
Central Kalimantan is 4,011,032.364 Ha but only 1,687,969.104 Ha that has been operated and licensed.
Crude Palm Oil (CPO), which is the primary products of palm plantation, grew 13.5% by the period of
2000-2004 and it rose up to 15% in 2007. The current productivity of palm plantation to produce CPO in
Central Kalimantan is around 0.205 ton CPO/ha/year (Provincial Government of Central Kalimantan,
2009). Investment and consumption are the main activities of activator economy in Central Kalimantan.
By considering the foreign and domestic investment in the last five years (2005 to 2009) - as reported by
Central Bank of Indonesia, the growth of investment in plantation, which is mostly to support CPO
industries, is as illustrated in Figure 4.
Percentage of Investment on Plantation to the Total Investment

(Foreign and Domestics)
in Central Kalimantan in the Period of 2005-2009
100%
80%
60%
40%
20%
0%
Quarterly
FIGURE 4. Investment on plantation in Central Kalimantan

Source: Central Bank of Indonesia, 2009
Nowadays, most of national and provincial roads in Central Kalimantan, which are used as the main
access for CPO production and distribution, are in poor condition. According to the Public Works Office
of Central Kalimantan Province (2006), it is estimated more than 50% of the existing roads, which consist
of 1,707.57 km-long national road and 1,050.26 km-long of provincial road, are in bad condition due to
truck-overloaded. Most of the trucks transporting CPO are loaded up to 12 tons although maximum of the
prevailed axle load is only 8 ton. This circumstance is not only affecting the transport infrastructures (like
road deterioration) but also the transportation services as well (like longer travel time and higher
transportation cost due to low speed of vehicle). Therefore, an appropriate method of freight
transportation planning (including modelling), in relation to the spatial location of commodities activities
108
is required to achieve efficient freight flow in the region. The data used in the model analysis and vehicle
characteristics are presented in Table 5 and 6.
TABLE 5. Data used for modelling CPO transportation
No Data Format Year
1. Administrative map Geo-database 2006
2. Road network Geo-database 2006
3. River network Geo-database 2006
4. Port map Geo-database 2006
5. Land use map Geo-database 2006
6. Regional spatial plan (RTRW Province) Geo-database 2006
7. Palm plantation map Geo-database 2006
8. Regional economic growth Pdf file 2005-2009
9. Documentation of CPO processing and economics activities Pdf file 2006
10 Documentation of port facilities and performance Pdf file 20052008
TABLE 6. Vehicle characteristics of CPO transportation in Central Kalimantan Province

Transport Mode Operational speed
Max Des. Speed Travel Cost
(kph)
Capacity (kph) Range Average (Rp) (US$)
Medium Truck
7.5 T 60 35 - 50 40 Rp 3200/Ton/Km US$ 0.34/Ton/Km
Tanky Truck
6.8 T 60 35 - 50 40 Rp 6460/Ton/Km US$ 0.68/Ton/Km
Tanky Barge
400 - 800 DWT 30-60 40 Rp 6000/Ton US$ 0.63/Ton
(1000 - 2000 T)
Source: Preliminary survey to CPO firms, 2009
5.2. MODELLING LOCATION CHOICE AND SPATIAL PATTERN OF CPO PRODUCTION
There are 2 steps considered in modelling location choice of CPO factory: identification of optimum
location of CPO factory; and identification of the shortest route of CPO transportation from factory to the
networks. Identification of the most likely location of CPO factories in this study is started from
calculation of production capacity of palm plantation for every year. Production capacity of plantation
was classified in three classes according to area service. Assumption used in this study is that every
plantation should have minimum area 5400 ha to have a factory. Proximity analysis using nearest distance
to closest factory is used to identify the un-served plantations.
Identification of the shortest route of CPO transport from factory to the road and river network that
consist of: a) identification of intermediate points on the closest road and river networks from the
factories; and b) identification of accessibility level of factories to the networks. Intermediate points are
points in the road or river to transport CPO from factories to the final ports. Intermediate points are
calculated from the shortest distance of factories to the road and river networks. The function of
intermediate points in this process is to know the level of accessibility of factories to networks and to
identify the starting points of transport route to the ports.
Accessibility levels show the ease to transport CPO from factories to the closest intermediate points in the
road and river networks. To identify the level of accessibility, the assumption is based on average velocity
of road transport mode (40 km/hour). The accessibility is classified as good if the distance is less than 10
km (15 minutes trip), 1020 km for moderate and more than 20 30 km for bad accessibility. The results
of this process are the map accessibility of factories to the road and river networks.
109
5.3. MODELLING REGIONAL FREIGHT TRANSPORTATION OF CPO COMMODITY
There are several processes required in modelling regional freight transportation for CPO commodity in
Central Kalimantan Province by using network analysis of ArcGIS software. These processes consist of
modelling distance, time, and cost of specific transportation mode, determining the nearest port,
determining the shortest route of using road and river networks, and modelling of probability of choosing
certain modes of transportation.
Identification of efficient CPO transportation from road and river networks by using various modes
considers some criteria, which are distance, travel time, and travel cost. For these criteria, the calculation
conducted by considering the average speed and travel cost of each modes. The average speed used, based
on assumption that there is no different classes among networks either in road or river network, so that all
segments in networks can be passed in the same speed without impedance. In terms of road network, the
distance was calculated by considering the distance from the factories to road networks (intermediate
points in road), based on a detour distance factor of 1,2 (1.2*Eucledian distance). For river network, the
calculation of these criteria is conducted by combination of both road and river networks. The
combination is used because the factories are not directly attach to the river networks, so that for reaching
the river network, CPO are transporting first to the river by using road networks.
The closest facilities analysis is conducted to identify the destination port where the CPO will be sent.
The closest facility option in ArcGIS is used, based on the assumption that people tend to go to relatively
closest facilities around them rather than to farther away facilities, because of distance, time and cost
reasoning. Utility function measures the satisfaction level perceived by the decision-maker upon which he
or she makes the choice decision. In term of transport, people consider time, cost, and other factors for
satisfying them. This study only considers time and cost for determining utility that is indicated in linear
function and is expressed as follows:
(11)
Generalized cost function is regarded as disutility or deterrence functions (Ortuzar and Willumsen, 1994)
and it is considering the following formula:
(12)
Estimation parameter is minus because it represents the disutility function of time and cost.
Probability of choosing certain modes is calculated by using the logit model from utility function of
transportation by road and river. The logit model used for determining the probability of choosing certain
mode is:
Exp(Vroad )
Pr obability road = (13)
( Exp(Vroad ) + Exp(Vriver ))
Exp(Vriver )
Pr obability river = (14)
( Exp(Vroad ) + Exp(Vriver ))
Where: V road = utility of road network, and V river = utility of river network
5.4. RESULTS AND DISCUSSIONS

According to the analysis, there are 113 of CPO factories identified. The factories are classified into 3
classes based on their production capacity: small (<60 ton FFB/hour); medium (60-90 ton FFB/hour); and
big (>90 ton FFB/hour). From 113 factories, 80 factories are small (71%); 26 factories are medium
(23%); and 7 factories are big (6%). According to the analysis, it is about 86 factories (76%) have good
accessibility to the road networks (<10 km distance and 10 minutes time). The rest is about 20 factories
110
(18%) having moderate accessibility (10-20 km and 10-30 minutes time), and 7 factories (6%) have bad
accessibility level to road (>20km and >30 minutes time). On the other hands, the accessibility level of
factories and plantation served to the river network is mostly bad. It is about 43 factories (38%) are in bad
accessibility. They need more than 30 minutes to reach the closest river. The rest is 34 factories (30%)
have good accessibility, and 36 factories (32%) have moderate accessibility to river. Considering the
accessibility on both road and river networks, the plantation accessibility to network can be determined.
There are 91 factories (81%) are identified accessible to the networks, 19 factories (17%) are moderate,
and only 3 factories (2%) are not accessible in both road and river networks.
The CPO is transported from factory to the 2 main ports, which are Kumai and Sampit. The CPO will be
delivered to the closest ports as the destination. The closest port is identifying through both road and river
networks. Using road network, 84 factories send their product to port 1 (Sampit), and the rest (29) send
their product to port 2 (Kumai). However, by using river network, 91 factories send their product to port 1
(Sampit), and 22 factories to port 2 (Kumai). It is interesting in some cases, because the same factory has
different closest port by using different networks (river and road). For example in the case of factory 1,
the closest port by using road network is Kumai, on the other hands, by using river networks the closest
port is Sampit. This condition is closely related to the factory location from road and river networks
(accessibility level to network), and the travel distance by using road and river from factory to port. These
conditions also considered on the latter step for determining the efficient routes from factories to ports,
especially to which port and network the CPO will be delivered efficiently.
The utility of each mode (river, and road) is defined by the equation considering travel time and cost, and
the weighting factor of cost considered 4 times more important than time. The estimated parameters 1
and 2 are derived from trial and error process, considering former assumption in which people regard
cost factor is 4 times important than time, because CPO is categorized into un-perishable goods. Hence
the equation for utility either for road and river obtained are as follows:
(15)
The decimal 6 used because of the different fraction in term of time and cost. Time uses units and cost
uses hundred thousands, and even millions. In fact, there are other factors should be considered for utility
calculation, such as reliability of river network that is influenced by climate and also error estimation, but
this study only limits the analysis on travel time and cost.
The CPO flow is calculated by considering the probability of choosing modes, meaning that a factory
probably will send CPO only by using river or road network to the nearest port based on this probability.
It then yields the volume of commodity flow to certain port. Table 7 and Figure 5 illustrate the flow of
CPO from factories using road and river networks to the Sampit Port and Kumai Port in Central
Kalimantan Province.
TABLE 7. CPO flows on the road and river networks
No Origin Number of Firms Volume of load Total Load Per Port
(ton/day) (ton/day)
1 Factory to port 1, using 46 15,725.492

road flow (33%) 33,732.208
(70%)
Factory to port 1, using 38 18,006.716
river (37%)
2 Factory to port 2, using 14 8,669.2651
road (18%) 14,561.800
Factory to port 2 using 15 5,892.545 (30%)
river (12%)
Total 113 48,294.018
111
FIGURE 5. CPO flows on the road and river networks
The accessibility level depends primarily on the distance from factory to reach the closest networks. In
term of road networks, the plantations with bad accessibility to road are located on the southern part of
the study area, because of the lack of road infrastructures in this area. By considering the travel time and
travel cost, the road network is more efficient for transporting CPO in the study area. It is also proved by
the higher probability in choosing this transportation mode. However, the river network also has
opportunity to compete with road network in term of travel cost. According to the result, there is little
different between probability to choose road or river networks.
This study only considers travel time and cost for determining utility. In fact, there are other attributes
should be considered for utility calculation such as reliability of river network that is influenced by
climate and also error estimation. Moreover, the port (outlet) choice is only dependent on the shortest
route in terms of distance, travel time and cost, and ignored other attributes such as the shipper or carrier
preference, and other transportation service attributes (e.g. network reliability). Therefore, conducting
quantitative survey is highly recommended to get more realistic and the better model in the future.
6. CONCLUSION
To formulate an effective strategy for regional transportation planning, the planners or policy makers
need to know the location of resources, processing or factory locations, and the final outlet (ports) to send
their goods to the outside world. To do this, various characteristics of the different commodities and the
available networks that can be used, need to be known. An understanding of how the relationships
between commodities activities and transportation are linked by spatial structure is essential.
The existing freight transportation planning models, such as promoted by Southworth (2002), Holguin-
Veras and Thorson (2000) and Holguin-Veras et al (2001), give more attention to the transportation
modelling with several stages to be conducted. The locations of industries or economic activities applied
in the transportation modelling are using a given data. As a new innovation in freight transportation
planning, this new method does not only consider the transportation modelling but also involving the
spatial location modelling/mapping, especially in determination of factories/industries location, and
considers the government regulations as the constraint, in formulating the freight transportation planning
for regional scale. Hence this method can produce the more realistic/practical transportation model
112
because the location of economics activities that generate freight transportation is considered in the model
and can be changed dynamically over time.
This new method also offers a more efficient stage of transportation modelling where the distribution,
modal split, and route assignment models are conducted in the same time. All simulation models will be
built on the basis of up-to-date GIS software that is now widely used in the spatial location modelling
(and mapping) and transportation modelling. However, this method faces with the great challenges in data
collection and database construction because it requires large data to be formatted and converted into GIS
database.
ACKNOWLEDGEMENT
The authors would like to thank the World Class Research University (WCRU) of Civil and
Environmental Engineering, Faculty of Engineering, Universitas Gadjah Mada for providing partial
funding to this research, and Prof. Dr. Ir. Martin van Marseveen and Dr. Ir. Mark Zuidgeest for
conducting research at ITC, University of Twente, The Netherlands.
REFERENCES
Ando, A., Meng, B. (2006). Transport sector and regional price differentials: A SCGE model for Chinese
Provinces. Discussion Paper No. 81, Institute of Developing Economies (IDE), JETRO, Chiba,
Japan.
Bontekoning, Y.M., Macharis, C., Trip, J.J. (2004). Is a new applied transportation research field
emerging?A review of intermodal rail-truck freight transport literature. Transportation Research
Part A 38 (2004), 134.
Bontekoning, Y.M. (2000). A jump forward in intermodal freight transport: are hub-terminals an
alternative for shunting? TRAIL Research School, Delft.
Central Bank of Indonesia. (2009). Kajian Ekonomi Regional Provinsi Kalimantan Tengah. Retrieved
from http://www.bi.go.id/web/id/ (23 December 2009).
De Jong, G., Gun, H. F., Walker, W. (2004). National and international freight transport models: an
overview and ideas for further development. Transport Reviews 24 (1),103 124. Retrieved from
http://eprints.whiterose.ac.uk/2015 (16 June 2009).
Friedrich, F. M., Haupt, T., Noekel, K. (2003). Freight modelling: data issues, survey methods, demand
and network models, CD-ROM in Proceedings of 10th International Conference on Travel
Behaviour Research, Lucerne.
Holgun-Veras, J., Thorson, E. (2003a). Modeling commercial vehicle empty trips with a first order trip
chain model. Transportation Research Part B 37 (2003), Vol. 37 (2), 129-148.
Holgun-Veras, J., Thorson, E. (2003b). Practical implications of modelling commercial vehicle empty
trip., Transportation Research Record 1833, 87-94.
Holguin-Veras, J, et al. (2001). An assessment of methodological alternatives for a regional freight
model, Appendix I: literature review on freight transportation demand modelling. New York
Metropolitan Transportation Council Report, New York, NY. Retrieved from
http://www.utrc2.org/research/assets/6/regionalfreight2.html (1 April 2009).
Holgun-Veras J., Thorson E. (2000). Trip length distributions in commodity-based and trip-based freight
demand modelling: investigation of relationships. Transportation Research Record No.1707, 37
48.
Jiang F., Johnson, P., Calzada, C. (1999). Freight demand characteristics and mode choice: analyses of
the results of modelling with disaggregate Revealed Preferences data. Journal of Transportations
and Statistics 2, 149-158.
113
Kreutzberger, E., Macharis, C., Woxenius, J. (2006a). Intermodal versus unimodal road freight transport.
In Jourquin, B., Rietveld, P., Westin, K. (2006). Towards Better Performing Transport Networks
(17 42), Routledge, Oxon.
Kreutzberger, E., Konings, R., Aronson, L. D. (2006b). Evaluation of the cost performance of pre- and
post-haulage in intermodal freight networks: analysis of the interaction of production models and
demand characteristics. In Jourquin, B., Rietveld, P., Westin, K. (2006). Towards Better Performing
Transport Networks (256 284), Routledge, Oxon.
Kreutzberger, E., Macharis, C,, Vereecken, L., Woxenius, J. (2003). Is intermodal freight transport more
environmentally friendly than all-road freight transport? A Review, NECTAR Conference No 7,
Ume, Sweden, June 13-15, 2003.
Lem, L.L. (2002). Promoting Economic Development by Improving Transportation Infrastructure for
Goods Movement, U.S. Economic Dev. Administration, USA.
Lubis, H.A., Isnaeni, M., Sjafruddin, A., Dharmowijoyo, D. (2005). Multimodal transport in Indonesia:
recent profile and strategy development. Proceedings of the Eastern Asia Society for
Transportation Studies, Vol. 5, 46 64.
Ortuzar, J. D., Willumsen, L.G. (1994). Modelling Transport, John Willey and Sons, U.K.
Office of Public Works of Central Kalimantan Province. (2006). Sinergi antara transportasi jalan dan
sungai di Kalimantan Tengah menuju sistem berkelanjutan, Palangka Raya.
Provincial Development Planning Agency of Central Kalimantan. (2009). Rencana Umum Tata Ruang
Kawasan dan Kota Provinsi Kalimantan Tengah, Palangka Raya.
Rodrigue, J.P., Slack, B., Comtois, C. (2004), Intermodal Transportation. Retrieved from
http://people.hofstra.edu/geotrans/eng/ch3en/conc3en/ch3c5en.html (28 May 2009).
Simon, D. (1996). Transport and Development in the Third World. Routledge, New York.
Southworth, F. (2002). Freight Transportation Planning: Models and Methods, Book Chapter, K.G.
Goulias (Ed) CRC Press, USA.
Stead, D., Banister, D. (2006). In Jourquin, B., Rietveld, P., Westin, K. (2006).Towards Better Performing
Transport Networks (136 156), Routledge, Oxon.
Tavasszy. (2006). Freight modelling: an overview of international experiences. TRB Conference on
Freight Demand Modelling: Tools for Public Sector Decision Making, Washington DC. Retrieved
from http://onlinepubs.trb.org/onlinepubs/archive/Conferences/ FDM/TavasszyPaper.pdf (29
March 2010).
The Coordinating Ministry for Economic Affairs the Republic of Indonesia. (2009). The 5th Indonesia
Palm Oil Conference and Price Outlook 2010. Retrieved from http://www.ekon.go.id/ (23
December 2009).
114
REGIONAL ECONOMIC
ACTIVITIES (VALUE ADDED
OF PRODUCTION)
PRODUCTION OF GOODS
AND SERVICES
MOVEMENT OF GOODS PRIVATE

TRANSPORTATION SERVICES
(COMMODITY FLOW) (DESICION MAKER)
COMMODITY FLOW BY PRIVATE

TRANSPORTATION MODES (DESICION MAKER)
COMMODITY FLOW BY SPECIFIC

PRIVATE
TRANSPORTATION MODES AND
OWNERSHIP STATUS (DESICION MAKER)
IDENTFICATION OF TYPE AND PUBLIC

LOCATION OF TRANSPORT
INFRASTRUCTURES (DESICION MAKER)
PERFORMANCE OF
TRANSPORT
INFRASTRUCTURES
GOVERNMENT REGULATIONS,
FINANCE, HUMAN RESOURCES
FIGURE 1. General
MANAGEMENT, ETC. research framework
APPENDIX 1. Conceptual framework
115
I. Socio- Economic Data in the Policy / Regulations:

Study Area Government Regulations,
Regional Spatial Plan
Regional Income (RI) Model
(Income= Consumption +Investment + Export- Import) Spatial Data of Selected

Commodities: Resources,
Factories, Outlets
Selected Commodities in the Region
Transportation Data:
Characteristics of Vehicle,
II. Qualitative Survey on Production and Goods Infrastructures, and
Transportation of Selected Commodities Services
Production Activities and Flow Movement of

Selected Commodities
III. Mapping of Spatial Location of Selected

Commodities using ArcGIS by Considering the
Shortest Route
Map of Spatial Location of Resource, Factory, and

Outlet of Selected Commodities
IV. Transport Modelling of Selected Commodities

using FlowMap and ArcGIS
Utility function of Transportation Cost -1 (TC1)

V. Quantitative
= 0 + 1 + 2 + 3
Survey on
Utility functions of Transportation Cost -2 (TC2) Shippers and
Carriers
= 0 + 1 + 2 + 3
= 0 + 1 + 2 + 3

Freight
Yes/ Optimum
VII. Model Transportation
Optimum
= , + , . calibration & Planning
Model
validation Procedure and
Not Optimum Methodology
VI. Identification of Type and Location of
Transport Infrastructures
APPENDIX 2. Methodology for regional transportation planning
116
EFFECT OF RAINFALL ON MICROSCOPIC TRAFFIC FLOW PARAMETERS
Hashim Mohammed Alhassan 1*, Johnnie Ben-Edigbe 23
ABSTRACT:
Inclement weather is one of the causes of instabilities on freeways and urban road networks.
Rainfall has been known to reduce speed, travel demand and increase travel time and headways.
To understand how rainfall affects traffic flow requires a microscopic examination of traffic
parameters. This paper examines the effect of rainfall on speed and headways of vehicles in a traffic
stream. Data was collected on a principal road in Johor Bahru, Malaysia for two months. The data was
filtered to obtain traffic flow information for both dry and wet operating conditions and these were
analyzed to see the effect of rainfall on the microscopic parameters. The results show empirical
evidence for decrease in speed and increase of headways under rain conditions. A speed drop by
4.2% from dry weather to wet weather and headways of up to 0.5seconds decreased by 5% but
increased for all others. The effect of rainfall on speed and headways are hard to observe in free flow
conditions as vehicle-vehicle interactions are minimal and bearing any incident emerging during rainfall,
it is unlikely to cause bottlenecks to traffic flow.
Keywords: Rainfall, Traffic flow, Headway, Speed, Wet, Dry, Weather.
1
Post Graduate Student, Universiti Teknologi Malaysia, 81310 Skudai, Johor, MALAYSIA
2
Associate Professor, Universiti Teknologi Malaysia, 813109 Skudai, Johor, MALAYSIA
*
Correspondence Author: Hashim Mohammed Alhassan, Universiti Teknologi Malaysia, 81310 Skudai, Johor,
Malaysia.
Tel: +60177412650
E-mail: mahashim2@live.utm.my
117
INTRODUCTION
Traffic studies in adverse weather conditions began with the pioneering work of Tanner (1952),
whose work focused on the effect of rainfall on traffic flow. The pioneering efforts to elucidate the
weather impact of traffic continued with the work of Codling (1974). The weather elements that are
known to affect traffic flow include snow and ice, rainfall, fog, hail, wind and dust storms Perry and
Symons (1994). These occur at different times and locations across the globe and in varying intensity.
They may also occur singly or in combination to cause dramatic effect on traffic. In tropical regions, the
weather elements exclude snow, ice, fog and hail. In South-east Asia, rainfall occurs throughout the year
and it is the major weather element that affects traffic flow. Various studies report that weather elements
and rainfall in particular, contribute significantly to incidents and accidents on highways. The Federal
Highway Administration (FHWA, 2008), reports that 75% of weather related accidents occur on wet
pavement and 47% of these occur during rainfall.
Interest in weather effects on traffic flow has grown in recent times because of the need to
minimise highway incidences emanating from bottleneck locations inclusive of weather effects. Accurate
prediction of the cause and effects of weather-induced bottlenecks and the mechanics of traffic flow under
rainfall could shape the strategies required to implement intelligent transportation systems on freeways
and urban road networks.
LITERATURE REVIEW
Tanner (1952) generated data by observing traffic flow during dry and wet periods in the second
week of August in 1949 and 1950 respectively and showed that traffic flow decreased under rainfall.
Accident related studies in wet weather have been reported by Codling (1974), who reported increases in
accident rates and attributed 31% of injuries to wet weather. Eisenberg (2004) recognised degraded
pavement condition and poor visibility as the major safety concern in wet weather. Chung et al (2005a)
and Keay and Simmonds (2006), identified rain as causing reduction in travel demand and traffic volume.
Lin and Nixon (2008) ascribed 71% of crash rate and 49% of injury rate on highways to rain. Studies on
the effect of rainfall on travel demand and intensity have been reported by Cools et al (2005), Maze et al
(2006) and Cools et al (2010). All of these showed that travel demand and intensity are negatively related
to rainfall. Reductions in traffic volume and capacity have been reported by Keay and Simmonds (2005)
and Chung et al (2006). In the area of rainfall effect on traffic speed Chung et al (2006), Minh Hai Pham
et al (2007) and Billot Romaine (2009) have reported speed reductions during rainfall.
The researches carried out, the manner of data collection, the coupling of weather and traffic data,
the analysis carried out and the observation sites does not allow a consistent and logical comparison to be
made across the research spectrum. For instance, the technology of data collection in Tanners (1952)
time precludes any use of extensive data, so speed and density data were not captured in his study.
Recently, some researchers Chung et al (2006), Dailey (2006), Pham et al (2007), and Romaine
(2009) have discovered the advantage in coupling weather and traffic data to see the effect of rainfall on
traffic flow and Dailey (2006) has developed a relationship between rainfall rate and speed reduction. All
of these used historical traffic and rainfall data generated from weather radar stations in proximity to
traffic loop detectors in the road network. This prevents the observation of real-time traffic flow under
rainfall conditions.
Two gaps have been identified in the study of rainfall effect on traffic flow; first the need to
isolate the effects of other factors on the flow is critical and secondly, the need to explain the mechanism
that leads to flow degradation during rainfall. This paper reports preliminary findings of a comprehensive
Wet Weather traffic studies project currently going on at Universiti Teknologi Malaysia to develop a
database on adverse weather effects on traffic flow and to design weatherresponsive system to
118
ameliorate the impacts of weather on congestion on highways in Malaysia. This paper therefore examines
the microscopic properties of traffic flow during rainfall.
METHODOLOGY
Data for this study was generated on Skudai-Pontian Highway in Johor Bahru, Malaysia. It is a
two-way two-lane highway. The site of the observation is 23km from Universiti Teknologi Malaysia in
the Pontian direction. The roadway is in good motorable condition and it is well marked, signed and
warning and regulatory devices properly located. That is, the condition of the roadway does not in any
way induce traffic instability, thus drivers are properly guided.
Two parallel pneumatic tubes installed in the middle of a 2km straight section of the highway
captures traffic as they transverse the tubes. The tubes are connected to a MetroCount 5600 vehicle
classification system that detects vehicles that hit the tubes through air pulses that are sent to the sensors.
The roadside unit records and stores particulars of the vehicles which can be retrieved later for analysis.
The data collection and observation site is shown in figure 1. Traffic data for the months of May and June
2010 were retrieved and observations of traffic behaviour made during rainfall events within the period.
In all, 71 rainfall events were observed during the data collection period of 61 days.
The traffic data generated were synchronised with the rainfall data delivered by the department of
Irrigation and drainage over the same period. The nearest rain gauge station to the observation site is
within 0.75km. To minimise spatial variation of rainfall over the catchment of the gauge station and the
detector site, one observer at the detector site recorded rainfall start and end times. These start and end
rainfall times were then used to select traffic flow data captured by the MetroCount Road Side Unit
(RSU) to serve as data for rainfall conditions. Other periods were regarded as dry traffic flow conditions.
Pairs of wet and dry traffic flow data of the same time period were then carefully abstracted and analyzed.
Observation Site
Figure 1: Data Collection and Observation Site
RESULT ANALYSIS
General Road Features
Microscopic traffic flow parameters include speed and time headways. These are under the control of the
driver and thus the driver may select and change his speed according to prevailing road conditions. The
119
chosen speed will then determine the time between vehicles ahead and behind the driver. The Skudai-
Pontian Highway is not congested and operates at free flow regime most of the times. The traffic flow
profiles for the period are shown in figures 2 and 3. The operating condition of the road during dry
weather is free flow with 85% of the traffic being cars. Roads in Malaysia generally have high proportion
of curvatures and grades which discourage overtaking manoeuvres at such sections. This causes a
platooning effect on the flow until sight distances becomes adequate for safety. Truck presence in the
traffic stream causes a similar effect. The dry weather traffic operating conditions are shown in figure 4 to
7. Clearly the road operates in the free flow region and microscopic features are easily discernible. The
posted speed limit on the road is 60km/hr. The speed distribution along the road and its statistical
parameters are shown in figure 7. With a total of 322,329 vehicles observed in June alone, the maximum
speed observed was 159.6km/hr and a minimum of 10.4km/hr. A mean speed of 63.6km/hr was recorded
for the road which has a standard deviation of 10.51km/hr.
Figure 2: Traffic Flow Profile in May
Figure 3: Traffic Flow Profile in June
120
Figure 4: Speed -Density Plot
Figure 5: Speed - Volume Plot
Figure 6: Volume- Density Plot
121
Speed | Bin | Below | Above |

0 - 10 | 0 0.0% | 0 0.0% | 322239 100.0% |
10 - 20 | 99 0.0% | 99 0.0% | 322140 100.0% |
20 - 30 | 355 0.1% | 454 0.1% | 321785 99.9% |
30 - 40 | 2240 0.7% | 2694 0.8% | 319545 99.2% |
40 - 50 | 16095 5.0% | 18789 5.8% | 303450 94.2% |
50 - 60 | 99784 31.0% | 118573 36.8% | 203666 63.2% |
60 - 70 | 138189 42.9% | 256762 79.7% | 65477 20.3% |
70 - 80 | 45885 14.2% | 302647 93.9% | 19592 6.1% |
80 - 90 | 13012 4.0% | 315659 98.0% | 6580 2.0% |
90 - 100 | 3865 1.2% | 319524 99.2% | 2715 0.8% |
100 - 110 | 1437 0.4% | 320961 99.6% | 1278 0.4% |
110 - 120 | 810 0.3% | 321771 99.9% | 468 0.1% |
120 - 130 | 271 0.1% | 322042 99.9% | 197 0.1% |
130 - 140 | 118 0.0% | 322160 100.0% | 79 0.0% |
140 - 150 | 48 0.0% | 322208 100.0% | 31 0.0% |
150 - 160 | 31 0.0% | 322239 100.0% | 0 0.0% |
160 - 170 | 0 0.0% | 322239 100.0% | 0 0.0% |
170 - 180 | 0 0.0% | 322239 100.0% | 0 0.0% |
180 - 190 | 0 0.0% | 322239 100.0% | 0 0.0% |
190 - 200 | 0 0.0% | 322239 100.0% | 0 0.0% |
Figure 7: Speed Distribution in Dry Weather
Also the 85% and 95% percentile speeds are 72.4 and 81.7km/hr respectively. 63.2% of all the vehicles
exceeded the mean speed and drove with a mean speed of 69.04km/hr. The dry weather headway
distribution similarly confirms the operating conditions on the road to be in the free flow region.
Effect of Rainfall on Speed.
Rainfall characteristically causes a reduction in speed of vehicles. The process by which speed
reduction is attained is explained by Alhassan and Edigbe (2010a). To see the effect of rainfall on speed
of traffic, similar periods of dry and wet traffic data were compared. Figures 8, 9 and 10 show the
relationships between speed, volume and density in wet weather. It is easy to see that traffic behaviour is
in the free flow region. However, there is some clustering of traffic as a result of the rainfall indicated by
the density legend on the right of the plot.
Figure 8: Speed - Density Plot in Wet Weather
122
Figure 9: Speed-Volume Plot in Wet Weather
Figure 10: Volume-Density Plot in Wet Weather
The corresponding plots in dry weather for the same period as the wet weather are shown
in figures 11, 12 and 13 and depicts a similar behaviour as in wet weather. Detailed statistical analysis
show a difference in the wet and dry weather speed profiles and is shown in table 1. Of the number of
vehicles observed, 40.4% and 59.6% travelled below and above the posted speed limit (PSL) in dry
weather respectively, whereas in wet weather, 63.4% and 36.6% travelled below and above the PSL.
There is a drop in mean speed from 62.2km/hr in dry weather to 58km/hr in wet weather. This represents
4.2% drop in mean speed. Table 2 shows the summary of the speed profiles in both wet and dry weather.
There is evidence that rainfall diminishes traffic speed. However, in free flow regime, this is hard to see
as interaction between vehicles is minimal.
Figure 11: Speed-Density Plot in Dry Weather
123
Figure12: Speed-Volume Plot in Dry Weather
Figure 13: Volume -Density Plot in Dry Weather
Effect of Rainfall on Headways.
Headway is the positioning of vehicles relative to each other on the highway. Table 3 and 4 show
the results of the headway analysis for both wet and dry weather. Vehicles that travelled with headways of
up 0.5seconds were 8.1% in wet weather. In dry weather, there were 13.1% with up to 0.5seconds
headway. This portrays the drivers perception of the risk associated with driving in adverse weather. Thus
headways of 0.5 seconds and less were reduced by 5%. Headway classes higher than 0.5 seconds
increased in wet weather for all classes than was observed in dry weather. This agrees with Hogema
(1996) except in this case the increment went beyond the 5seconds observed by Hogema.
The implication is that drivers perceive driving during rainfall as a risk situation and drive more
calmly than in dry weather. Generally drivers are not deterred by rainfall until there is significant drop in
sight distance caused by higher intensity rain or longer rainfall duration. The action of the wind shield
wiper is to restore the drivers sight distance. This action is however, overwhelmed by higher intensity
rain or longer duration rainfall and the sight distance drops. Correspondingly, the driver responds to this
situation by reducing speed to cope with the changing situation.
CONCLUSION
Implications for road quality performance and congestion
Traffic flow in wet weather reduces the speed and headways of vehicles. In light traffic conditions,
these are hard to observe. This is because interaction among vehicles is minimal and the traffic stream as
a whole is less affected. It is reasonable to suppose that as the flow rate increases in adverse weather,
greater interaction will occur and the effect on the traffic stream will be more visible and consequential.
The behaviour of traffic under rainfall should be investigated for roads with higher flow rates in the
vicinity of congestion. Interesting features could be revealed that gives insight into traffic behaviour
under rainfall.
124
Table 1: Speed Profiles in Wet and Dry Weather

WET WEATHER DRY WEATHER
Speed Bin Below Above Bin Below Above
0-10 0 0.0% 0 0.0% 2475 100.0% 0 0.0% 0 0.0% 6165 100.0%
10-20 0 0.0% 0 0.0% 2475 100.0% 2 0.0% 2 0.0% 6163 100.0%
20-30 3 0.1% 3 0.1% 2472 99.9% 6 0.1% 8 0.1% 6157 99.9%
30-40 18 0.7% 21 0.8% 2454 99.2% 38 0.6% 46 0.7% 6119 99.3%
40-50 342 13.8% 363 14.7% 2112 85.3% 371 6.0% 417 6.8% 5748 93.2%
50-60 1205 48.7% 1568 63.4% 907 36.6% 2071 33.6% 2488 40.4% 3677 59.6%
60-70 726 29.3% 2294 92.7% 181 7.3% 2681 43.5% 5169 83.8% 996 16.2%
70-80 144 5.8% 2438 98.5% 37 1.5% 779 12.6% 5948 96.5% 217 3.5%
80-90 33 1.3% 2471 99.8% 4 0.2% 173 2.8% 6121 99.3% 44 0.7%
90-100 4 0.2% 2475 100.0% 0 0.0% 39 0.6% 6160 99.9% 5 0.1%
100-110 0 0.0% 2475 100.0% 0 0.0% 4 0.1% 6164 100.0% 1 0.0%
110-120 0 0.0% 2475 100.0% 0 0.0% 1 0.0% 6165 100.0% 0 0.0%
120-130 0 0.0% 2475 100.0% 0 0.0% 0 0.0% 6165 100.0% 0 0.0%
130-140 0 0.0% 2475 100.0% 0 0.0% 0 0.0% 6165 100.0% 0 0.0%
140-150 0 0.0% 2475 100.0% 0 0.0% 0 0.0% 6165 100.0% 0 0.0%
150-160 0 0.0% 2475 100.0% 0 0.0% 0 0.0% 6165 100.0% 0 0.0%
160-170 0 0.0% 2475 100.0% 0 0.0% 0 0.0% 6165 100.0% 0 0.0%
Table 2: Statistical Summary of Speed Profiles in Wet and Dry Weather

PARAMETER MEAN MAXIMUM MINIMUM 85% 95% MEDIAN VARIANCE STD POSTED
SPEED SPEED SPEED SPEED SPEED SPEED DEVIATION SPEED LT
km/hr km/hr km/hr km/hr km/hr km/hr km/hr
WET
WEATHER 58.0 98.4 21.0 65.5 72.0 57.2 67.67 8.23 60
DRY
WEATHER 62.2 114.1 16.4 70.2 77.8 61.6 81.74 9.04 60
Table 3: Wet Weather Headway Analysis

WET WEATHER
Headway Bin Below Above
0.00 0.50 214 8.1% 214 0.1% 2423 91.9%
0.50 1.00 796 30.2% 1010 38.33% 1627 61.7%
1.00 2.00 598 22.7% 1608 61.0% 1029 39.0%
2.00 - 4.00 313 11.9% 1921 72.8% 716 27.2%
4.00 8.00 270 10.2% 2191 83.1% 446 16.9%
8.00 16.00 201 7.6% 2392 90.7% 245 9.3%
16.00 32.00 82 3.1% 2474 93.8% 163 6.2%
32.00 64.00 5 0.2% 2479 94.0% 158 6.0%
64.00 128.00 0 0.0% 2479 94.0% 158 6.0%
128.00 1000.00 0 0.0% 2479 94.0% 158 6.0%
Table 4: Dry Weather Headway Analysis
DRY WEATHER
Headway
Bin Below Above
0.00 0.50 960 13.1% 960 13.1% 6373 86.9%
0.50 1.00 2121 28.2% 3081 42.0% 4252 58.0%
1.00 2.00 1411 19.2% 4492 61.3% 2841 38.7%
2.00 - 4.00 888 12.1% 5380 73.4% 1953 26.6%
4.00 8.00 788 10.7% 6168 84.1% 1165 15.9%
8.00 16.00 489 6.7% 6657 90.8% 676 9.2%
16.00 32.00 116 1.6% 6773 92.4% 560 7.6%
32.00 64.00 6 0.1% 6779 92.4% 554 7.6%
64.00 128.00 0 0.0% 6779 92.4% 554 7.6%
128.00 1000.00 0 0.0% 6779 92.4% 554 7.6%
125
REFERENCES
Alhassan, H. M., and Edigbe, J. B. (2010a). "Impact of Rainfall Intensity on Traffic Flow Rates
Modelling Techniques." Proceedings of the Third International Conference on Engineering,
Science and Humanities (IGCESH), Johor Bahru, Malaysia.
Billot, R. (2009). "Integrating the Effects of Adverse Weather Conditions on Traffic:

Methodology, Empirical Analysis and Bayesian Modelling."
Chung, E., Ohtani, O., and Kuwahara, M. (2005a). "Effect of Rainfall on Travel Time and
Travel Demand.5th ITS European Congress, Hannover, Germany.
Chung, E., Ohtani, O., Warita, H., Kuwahara, M., and Morita, H. (2005b.). " Effect of Rain on
Travel Demand and Traffic Accident. ." 8th IEEE Intelligent Transportation Systems
Conference, Vienna, Austria.
Chung, E., Othani, O., Warita, H., Kuwahara, M., and Morita, H. (2006). "Does Weather Affect
Highway Capacity."
Cools, M., Moons, E., and Wets, G. (2010). "Assessing the Impact of Weather on Traffic
Intensity." Weather,Climate and Society, 2, 60-68.
Dailey, D. J. (2006 ). "The use of Weather Data to Predict Non-recurring traffic Congestion."
Technical Report, Agreement T2695,Task 54,Washington State Transportation Center, USA.
Eisenberg, D. (2004). "The Mixed Effects of Precipitation on Traffic Crashes." Accident Analysis
and Prevention, 36(4), 637-647.
FHWA. (2008). "Road Weather Management Overview." http://ops.fhwa.gov./weather/overview.
assessed 14th September 2009.
Keay, K. and Simmonds, I. (2005). "The Association of Rainfall and Other Weather Variables
with Road Traffic Volume in Melbourne, Australia." Accident Analysis and Prevention, 37, 109
124.
Lin, Q., and Nixon, W. (2008). "Effects of Adverse Weather on Traffic Crashes: Systematic
Review and Meta Analysis." Proceedings of the 87th Annual Meeting of the Transportation
Research Board, National Academies, Washington D.C., 2015.
Maze, T. H., Aggarwal, M., and Burchett, G. (2006.). "Whether Weather Matters to Traffic
Demand, Traffic
Safety, and Traffic Operations and Flow." Journal of the Transportation Research Board, , TRB,
National Research Council, Washington, D.C., pp. 170-176., 1948., 170-176.
Minh, H. P., Chung, E., de Mouzon, O., and Dumont, A. G. (2007). "Season Effect on Traffic: A
case study in Switzerland." Seisan-kenk Yu, 59, 214-216.
Tanner, J. C. (1952). "Effect of Weather on Traffic Flow." Nature, 4290, 107.
126
DESIGN OF BATTERY PACK fOR ELECTRIC VEHICLE BASED ON LITHIUM-ION

TECHNOLOGY
Z.Y Phuan*1, G.D Gan2, C.K Leong3, B.K Cheah4, M.K Yoong5, Y.H Gan6 and K.W Chew7
ABSTRACT:
In these modern days, electric vehicles are well known for its efficiency, it would be a cost saving choice
compared to the ICE car, due to its advantages such as silent engine, zero emission which are totally
green to the environment. However the selection of a suitable battery type such as lead acid, NiMH,
Lithium ion in the long run plays a very important role in the construction or design of the electric vehicle
due to its function and also the physical aspects. In terms of the functionality, the vehicle must be able to
provide sufficient energy to allow the car to move or accelerate and at the same time, having a long
lifespan or constant power supplied to the vehicle. Taking this into consideration, lithium ion battery have
been selected as the main power source when designing an electric car, this can be viewed from the
characteristics of a lithium ion battery such as fast charging rate, light weight, high capacity and long
charge-discharge cycle, etc. Hence the design of battery pack using lithium ion battery inclusive of its
compartment design, ventilation and the monitoring system will be discussed in this paper.
Keywords : Lithium Ion, fast charging rate, charge-discharge cycle
127
INTRODUCTION
In these modern days, the vehicles that can be seen running on the road are mostly powered up by
petrol or diesel as the source of power. As years goes by, these energy source will be depleted where it
originates from fossil fuels that takes years to produce naturally. Besides that, the current source of energy
also causes pollution to the air which is carbon monoxide directly from internal combustion engine (ICE)
vehicle. As the gasoline is not a renewable energy, it can be expected that the price of petro will rise to
sky high in the future. Hence, the electric car using battery as main electric source has to be developed in
order to improve fuel economy and reducing wastage emission to the environment. In line with the latest
battery development, the lithium battery is chosen to be used in electric vehicle. The lithium ion battery
has a long charging-discharging cycle, fast charging system, and no memory effect. The characteristics of
a lithium ion battery must be fully understood before any power management experiment is done on the
battery. There are certain factors that will cause the lithium ion battery to either perform better or worst
such as the overcharge, deep discharge, high current charge-discharge, and temperature. In this paper, the
characteristics of lithium ion battery will be further discussed.
LITERATURE REVIEW
2. Lithium Ion Battery
Many kinds of battery can be used to power up an electric vehicle and until now, many manufacturers
would prefer to option for the lithium ion battery. However, there are advantages and disadvantages of
using lithium ion battery which can be seen from the physical aspect of this battery. This battery is very
compact and light weight such as laptop battery pack, hand phone batteries and other mobile devices.
Hence this solves the problem of having a very heavy weight electric vehicle. This saves up spaces
allocated for the battery allowing a much smaller power needed for acceleration instead of needing a
higher power to be supplied to the motor in order to compliment with the weight of lead acid batteries.
Another advantage of using lithium ion battery is that the charging time is much faster than the lead acid
battery. This is important especially when long distance travelling is required. The estimated and
approximated time for the charging of the lithium ion battery to full charge is around 50min 1hr (fast
charging) and 3hr 4hr (slow charging).
However, there are also disadvantages in regards to the lithium battery where the unit price of a lithium
battery is much higher than lead acid battery. An 11.1V rating battery would cost up to approximately
RM350. Hence the budget allocated for the batteries will increase much more if the manufacturers were
to option for lithium ion batteries as the main source of energy. Besides that, if the batteries are left on
standby mode or not in use, self discharging takes place hence losing the battery power.
3. Charging and Discharging
The lithium ion battery has a high energy density, high cell voltage, a low self discharge rate and also a
very good cycle life. In the lithium ion, there is an anode and cathode that is imbedded in the chemical
compound to make up of two different kind of lithium ion. These two different kinds of lithium ion will
form a positive and a negative pole and a battery reaction will happen between these two poles. There are
many types of lithium ion batteries such as the manganese acid lithium ion, lithium polymer, and cobalt
acid lithium ion battery but the charging and discharging characteristics are the same. The way to charge
up the lithium ion is by using a method called current-constant voltage charging where it is charged using
two continuous steps. In the first step, the current will be kept constant until the voltage reaches the limit
of the top-of-charge condition (approximately 4.2V). Once the voltage reaches the limit, it is kept
constant instead until the current drops to a pre-determined value.
128
Percentage of Discharging Capacity and Cell Voltage based on Various Source of Battery. [5]
As seen from the figure shown above, the most stable, high and consistent voltage output supplied from
the battery is from the lithium ion battery compared to the rest due to the flat discharging curve. We can
approximate that the upper voltage limit (supreme limit) of battery charging for a lithium ion is 4.2V with
a tolerance of 0.05V and the lower voltage limit is approximately 2.9V. When the charging of a battery
reaches approximately 4.2V, the battery pack is at 60% - 70% charged. Then, the battery should undergo
continuous charging until it current drops. If the lithium ion does not reach the upper voltage limit of
4.2V, it will not be fully charged and this happens due to the characteristics of the anode and the cathode
of the battery pack where it needs a higher voltage input to insert the ions into the anode and cathode
crystals resulting to more ions that can be inserted into the pack to a certain extend. This technique of
constant current charging at the first step uses a charger that supplies a uniform current despite of the
battery state of charge or temperature and at the same time, it helps to eliminate imbalances of cells and
batteries connected in series. The advantage of using this constant current is that it is supplying a single
rate current which is appropriate for cyclic operation where it is necessary for an overnight full charge of
a battery. But this causes omission of gases when high rates of charge is supplied leading to positive grid
oxidation as temperature of the battery increases or if overcharging occurs which will be discussed next.
The trickle charging method is not possible for the lithium ion battery as it cannot be overcharged
without causing any damage to the battery pack. This trickle charging method can be defined as a
charging current that can be applied to the battery without damaging it or better known as slow charge.
This charging technique also is always used to recharge a battery for losses from self discharging as well
as to restore the energy discharged during the intermittent use of the battery. Once the lithium battery is
overcharge, lithium metal plating will occur on the positive pole and the negative pole releases oxygen
due to the cathode acting as an oxidizing agent hence losing its stability. This problem of overcharging
causes the battery to heat up. At low temperature, the electrolytes in the battery might freeze up, affecting
the performance of the battery while at high temperature; the degradation of the battery takes place hence
resulting in a capacity loss. These can be further explained in the table below showing the effect of
temperature changes at different discharging current on the capacity of the lithium ion. We can say that at
low discharging current (10A) and at a temperature at 20 degrees, the capacity (Ah) is the maximum
compared to the low discharging current (10A) at -20 degrees [1].
129
Title: Effect of Temperature Changes at Various Discharging Current on the Capacity of Lithium Ion.[1]
Title: The effect of temperature on the lifespan of a lithium ion battery. [5]
In the figure above, it can be seen that the effect of temperature on the lifespan of a lithium ion battery.
These shows that self discharging happens in the battery if it is left on standby mode or not in used for a
period of time causing a cumulative error when we try to estimate the residual capacity (Ah) of a battery
at a given discharging rate. Hence this self discharging issue will be inevitable where the predicted shelf
life for the lithium ion is 10 years approximately and this battery will self discharge at a rate of 2 % 3%
every month. The temperature also plays an important part in the self discharging cycle in the battery
where at higher temperature; the battery tends to lose more current through self discharging.
Title : The capacity variation with temperature of a lithium ion battery [5]
130
In the graph, we can see that the upper limit is set to 4.2 volts and lower limit is set at 2.5 volts with
indicates as the fully charged and empty conditions of the cell using the current-constant voltage charging
method and at various temperature. As seen also, there is two different value of C-rates that is involved at
the point where the battery is discharged from full capacity to empty. In a real world, we have to take into
consideration that the cell may be charged at one temperature and it can be discharged at different
temperatures where it will affect the effective cell capacity. The cell is very inefficient in giving up the
charge at high discharge rates and low temperatures. This results in the available capacity is reduced at
high discharge rate and although there is a reduction in capacity at high temperatures, but even at low
temperature there is also a rather high reduction of capacity at low temperatures.
The lithium ion has no memory effect where the energy stored in the battery pack does not have to be
completely drained out before recharging. However, if the battery pack is completely discharged, the
battery pack cannot be used or in other words, degradation of the battery pack will start to take place
when it is discharged below 2V. This memory effect is also known as the battery effect or lazy battery
effect where it causes them to hold less charge and gradually lose their maximum energy capacity if they
are repeatedly charged after only being partially discharged. The disadvantage of having this memory
effect is that it might cause voltage depression due to long term overcharging where the peak voltage of
the battery drops more quickly than normal even if the total energy remains the same. This voltage
depression is caused by a repeated process of overcharging of a battery causing a formation of small
crystals of electrolyte on the plates, choking the plates from its process hence resulting in higher
resistance and low voltage provided by the battery. This causes the battery as a whole to discharge
rapidly.
In overall, this lithium ion battery is design to operate safely within a normal operating voltage but tends
to be unstable if it is charged to higher voltages and as it is charged above 4.30V the cell causes the
plating of metallic lithium on the anode and the cathode tend to become an oxidizing agent where it loses
stability and produce oxygen. Overcharging of the battery causes the cell to heat up. The charging,
discharging, overcharging and overdischarging characteristics is given much attention where a
commercial lithium ion packs has a protection circuits that limits the charge voltage to 4.3V per cell
which is approximately 0.1 volt higher than the voltage threshold of the charger. There is also a
temperature sensing device in the cell to disconnect the charge if the temperature of the cell increases to
90 degrees or 194 Fahrenheit. A mechanical pressure switch is embedded into the cells to permanently
interrupt the current path if a safe pressure threshold is exceeded. There is also a safety circuit in the cell
to prevent extreme low voltage. If the battery is discharged below 2.5 volts per cell, the current path is cut
off through the safety circuit. Besides that, some of the advanced lithium ion batteries contains a wake-up
feature where the protection circuit will start to draw current after the battery have been activated with a
short period of charging. This will allow the battery to have prolonged storage. It has been explained by
battery experts that charging lithium ion is much simpler and straightforward than the nickel-based.
Besides meeting the tight voltage tolerances, we can see that the charge circuit can be design much
simpler with fewer variables to consider. Full charge detection is done by applying voltage limits and
observing the current saturation on full charge is easier than analyzing many complex signatures. Also,
the charge currents are less critical and are variable where low current still permits proper full charge
detection but takes a longer time to charge up to top of charge condition.
4. State of Charge
To further understand the state of charge or known as SOC, we have to look into other battery parameters
such as discharge rate, depth of discharge and also state of discharge or known as SOD. The discharge
rate can be defined as the current at which a battery is discharged and can be expressed as Q/h rate where
Q is the battery capacity and h is known as the discharge time in terms of hours. On the other hand, the
depth of discharge is known as the percentage of battery capacity (rated capacity) to which a battery is
131
discharged . If the withdrawal of at least 80% of the battery (rated) capacity is known as deep discharge.
State of Charge is closely related to this two terms where SOC is defined as the present capacity of a
battery where it is the amount of capacity that remains after discharge from a top-of-charge condition or
can be defined in laymans term as the percentage estimate of how full the battery is. Every cell have
different range of SOC hence leading to a performance difference among batteries in the pack. This helps
to balance up the charging and discharging in order to maximize the battery pack lifespan and the energy
efficiency. In order to estimate the most accurate measurement for the value of SOC, a method known as
the combination algorithm is shown below to have an efficient estimation using the open circuit voltage
method circuit voltage (OCV) and ampere hour counting method with the particular working conditions
in an electric vehicle.
Title: Combination Algorithm Flowchart. [3]
[3, 6] From the flowchart above, we can briefly describe the process of estimating the SOC value for the
cell. The system is first initialized, calculating the original value of SOC (SOC o ) using the OCV method
before turning on the vehicles engine. Then it will check for every cells SOC o for its individual
minimum value where a cell that has the lowest value of SOCo is assumed and estimated to be the
minimum value of the battery packs SOC o . After undergoing this process, the extended Kalman
Filtering( EKF) algorithm will take sampling data of at least 200 data to correct the value of SOC based
on the feedback system of SOC o providing a more accurate SOC depending on the number of sampling
data taken. The ampere hour counting method is used to calculate the SOC when the battery voltage is
changed slowly. When the SOC drops to 15%, the EKF algorithm will function to estimate the SOC until
the discharging process is done. In this combination algorithm flowchart, the EKF algorithm functions as
the optimum state estimator for non linear system where it corrects the value of SOC o and also gives an
accurate initial value to the ampere hour counting method. We can also monitor the states of the battery to
avoid over discharging when the SOC value is below 15%. However this technique of measuring the SOC
of the battery helps to measure as accurate as it can but not 100% accurate due to some of the factors that
influenced the SOC of the battery. One of it is its usable capacity where the capacity of a cell is not
constant and changes from time to time depending on the temperature, charging and discharging rates
with the age of the cell. This will definitely affect the capacity of the cell significantly if we consider the
effect of other parameters on the cell. Besides that, the charge and discharge rates also plays an important
role in influencing the SOC of lithium ion batteries where the electrochemical action in the cell takes a
period of time to complete and causing inability of the electrochemical action to follow instantaneously or
support the load placed on the cell.
132
Float is the a charging system where it senses when the battery voltage is at a appropriate float level and it
will temporarily ceases charging, maintaining the charge current at zero or a very minimal level until the
battery output voltage drops resulting in the continuation of charging. This appropriate float voltage
varies much with the design of the battery and the surrounding temperature where with the appropriate
voltage for the battery type and proper temperature compensation, a float charger may be kept connected
indefinitely without causing damage to the battery. With a voltage drop of 0.1V for a 5 degree rise in
ambient temperature, if this is not compensated, itll shorten battery life or over/under charge. If the SOC
is above 100%, hence it is due to the battery inefficiency and self discharge losses. But this is unusual and
happens all the time where the SOC reaches 115% when your battery comes into float for example. This
15% is the extra power your charging source has to put in to make up for the battery losses.
It is important to determine the SOC or the life of the battery accurately through the number of cycles it
has to perform and also the depth of discharge. The optimum battery life to utility ratio will occur if the
battery is not discharged lower than 50%. If we are using an expensive battery such as lithium ion, it is
encouraged to have a monitoring system for the battery life or SOC to prevent any excessive discharge or
premature battery failure. There are many other ways to monitor this such as using an amp hour meter.
The advantage of using an amp hour meter is its capability to allow the user to keep check on the charging
equipment to monitor if it is working properly or if there is any failure in the process. It also makes it
possible to be aware of any components that the battery supplies is using excessive amount of power so
that it can be switched off or reduced. Besides that, the battery efficiency can be calculated using an amp
hour meters that provides an early warning in a case of battery failure. This will definitely help in the
troubleshooting process by reviewing the data collected from the amp hour meter. First, an amp hour
meter works in a way that it counts the rate and time of current flow from a battery being discharged and
it works the same way when it is being charged back up. Once it determines the full point, the capacity
and the efficiency of the battery is stated on the amp hour meter to produce an instantaneous readout of
the SOC in amp hours or percent of the total at any time in the cycle. This is probably the most effective
way to get accurate readings of charge from a battery that is constantly in use. But on the other hand, the
calculation of the readings acquired is slightly tedious because the battery capacity is a moveable target
and also the full charge of the battery is not accurately defined where it is always possible to put more
current in but this increases the current pumped into the battery causing heat to be produced and will not
be available on discharge. The amount of current that can be drawn from the battery depends on the
temperature of the battery and rate of discharge set. Fuzzy logic is required much to do the calculations
but the amp hour meters uses periodic and frequent resets that are usually performed every time the
battery reaches its full capacity. So in other words, the longer a battery remains without getting fully
charge, the more inaccurate the amp hour meter reading will produce. The meter have to calculate the
efficiency of the battery and this can be done by comparing the charge to discharge total from once cycle
to the next and hence, this calculation have to be averaged over a number of cycles where the more the
battery get cycled, the more it provides us with a more accurate reading as the number of cycles increase.
Another way to measure the SOC is to use a voltmeter where it can be used to get a quick snapshot of the
state of charge of the lithium ion. Usually a top-of-charge condition battery of 12 volts will have an open
circuit voltage of around 12.6 volts and a battery that is completely discharged is when there is a voltage
drop to 10.5 volts at a 20 hour discharge rate. Hence from here, we can estimate that at 50% state of
charge, it will be at around 12.2 volts. The only disadvantage using this voltmeter is that in order to obtain
a accurate reading, the battery has to be at rest or not in use for 24 hours. If the battery is used without
waiting for 24hours, we will observe a false reading on the voltmeter due to the surface charge on the
plates. But overall, the voltmeter doesnt do it accurately enough for most purposes unless an expanded
range voltmeter is used.
133
CONCLUSION
From this paper, it can be concluded that the design of the lithium ion battery pack have to be
done in the most delicate way possible to ensure achieving a maximum efficiency of the battery. This
requires a full understanding of the characteristics of the lithium ion before proceeding to the fast-
charging of the battery where the construction and design of the battery have to be lightweight, cost
saving, and small for the physical side and it must be placed under a desirable temperature condition to
avoid any battery degradation or battery freezing. Without the memory effect, we can say that this is one
of the best advantage of the lithium ion that allows customer to recharge their vehicle before their battery
source is drained out and also the C-rate of the electric vehicle must be set to a proper low value C-rate to
ensure that it provides enough power to the entire vehicle and at the same time, having a long time period
for the battery to discharge before reaching the deep discharge region.
REFERENCES
1. Zhenpo Wang, Ke Chen, and Fengchu Sun. The Technological Development of Domestic Li-
ion Power Battery and Its Application on the Electric Vehicle. June 2005.
2. Marco S.W Chan, K.T Chau, and C.C. Chan. Design and Implementation of Neural Network
Based Capacity Indicator for Lithium-Ion Battery. December 2004.
3. Jie Xu, Mingyu Gao, Zhiwei He, Jianbin Yao, Hongfeng Xu. Design and Study on the State of
Charge Estimation for Lithium-ion Battery in Electric Vehicle. 2009 International Conference
on Artifical Intelligence and Computational Intelligence.
4. Sheng Shui Zhang. The effect of the charging protocol on the cycle life of a Li-ion Battery.
Received 16 April 2006, Received in revised form 22 May 2006, Accepted 6 June 2006.
5. McGraw Hill TAB Electronics Build Your Own Electric Vehicle Sept 2008.
6. Iqbal Husain. Electric and Hybrid Vehicles Design Fundamentals. 2003, published by CRC
Press LLC.
7. Zhenpo Wang, Ke Chen, and Fengchu Sun. Study of the EV Battery Pack Attended Mode June
2004
8. Chi-Kyun Park, Zhiwei Zhang, Zhiqiang Xu, Archana Kakirde, Kenny Kang, Chul Chai, George
Au, Laura Cristo Variable Study for the Fast Charging Lithium ion Batteries 20 December
2006
134
STUDIES OF ELECTRIC MOTORS FOR LIGHT-WEIGHT ELECTRIC VEHICLE
Gan Guo Dong 1*, Gan Yu Han 2, Leong Chee Ken 3, Phuan Zheng Yi 4, Vincent Cheah Beng Keat 5,
Yoong Mun Kiat 6, Chew Kuew Wai 7
ABSTRACT:
In the development of the electric vehicle (EV), the parameters of electric motor, such as its torque,
maximum speed, rotation per minute (rpm) and etc played an important role in determining the
performance of a good EV. The electric motor is the heart of the EV which offers the driving element that
moves the EV in various conditions. The electric motor comes in different shapes, driving methods, types
and functions. However, the problems faced lies in the selection of the perfect electric motor for a four-
seater, urban-bound vehicle that can muscle out the sufficient torque and deliver wide range of speed.
Also, maintaining high efficiency in the long run to overcome the dominance of the combustion engine.
The solutions are to do engineering research in the field of electric motor. Various experiments are
performed to determine the motor parameters. In this paper, the arguments between the usage of DC
motor and AC motor have been briefly discussed. The suitability of DC motor for light weight EV have
been investigated and discussed.
Keywords: DC/AC motor, torque, maximum speed, rotation per minute (rpm), light weight EV.
1 Student, Universiti Tunku Abdul Rahman, Setapak, 53100 Kuala Lumpur, MALAYSIA
3
Student, Universiti Tunku Abdul Rahman, Setapak, 53100 Kuala Lumpur, MALAYSIA
4
5
6
7
Assistant Professor, Universiti Tunku Abdul Rahman, Setapak, 53100 Kuala Lumpur, MALAYSIA
*Correspondence Author: Gan Guo Dong, Universiti Tunku Abdul Rahman, Malaysia. Tel: +60122068462. Email:
kenny_ggd@hotmail.com
135
INTRODUCTION
According to a new study from the Centre for Entrepreneurship and Technology at University of
California, Berkeley, electric cars could comprise 64-86% of US light vehicle sales by 2030[3]. As seen
in Fig.4, the study shows rapid adoption for electric vehicles with switchable batteries, quantifies how the
electrification of the U.S. transportation system will decrease America's dependence on foreign oil,
increase employment, and reduce the environmental impact of transportation emissions. From this study,
the electric vehicle and the development of high performance electric motor is a force to be reckoned
with.
Thus, development of the electric vehicle industry depends on the heart of every electric vehicle (EV)
which is the drive train or also known as electric propulsion system (EPS) and this EPS greatly affects the
performance of the EV. EPS plays a role in transferring energies in either direction as required, under
control of the driver at all times. Due to diversity of EVs, the design of electric motor and propulsion
system has to be carried out at system level, which should take EV types, performance, configuration and
electric motor parameter into consideration. And, among the EPS-system, it is very important to select the
proper type of motor with suitable rating because it is the crucial step in designing the overall system.
Figure 1. Functional Block Diagram of a Typical Electric Propulsion System. [2]
For the electric motor, many criteria such as efficiency, cost, reliability, power density, maturity of
technology and controllability must be taken into consideration. Usually, electric cars are driven by large
electric motors usually rated between 3.5 and 28 horsepower. For those accustomed to gas engines, this
may not seem like much power, but the rating systems used for gas engines and electric motors are so
different that the numbering system is almost meaningless [6]. Gas engines are rated at their peak hp,
electric motors are rated at their continuous hp. The peak hp of an electric motor is usually 8 to 10 times
its continuous rating. Electric vehicle drive motors can be divided into two basic groups, DC or direct
current motors, and AC or alternating current motors.
136
Figure 2. Primary EV Power Train [2]
Figure 3. Conceptual Illustration of General EV Configuration [2]
137
Figure 4. US Light Vehicle Sales and Fleet Composition under Baseline Scenario
LITERATURE REVIEW
EV Demands and Design Outline of EPS
TABLE I
EVS BASIC REQUIREMENT ON ELECTRIC DRIVE TRAIN AND THEIR IMPLICATIONS ON PERFORMANCE
Requirement Implications
High torque at low Driving performance.
speeds and constant
power at high speed.
Continuous and Comfort.
smooth drive control.
High efficiency, Longer driving range.
regenerative braking.
High specific power Overall efficiency and
and specific torque. consistency.
Maintenance free. Safety and
maintenance cost.
Reasonable Cost. Price and Budget.
With this, a description of motor parameters and control strategy can be discussed and also to identify the
parameters according to the performance requirement. Lastly, optimization can be made to improve the
drive train and efficiency.
138
Characteristics of Electric Motor for EV
In EV, the vehicle performance is completely determined by the torque-speed characteristic of the
motor which is much closer to the ideal as compared to ICE (Internal Combustion Engine). In order to
meet EV requirement, operation entirely in constant power is needed. However, operation fully in
constant power may be impossible for any practical vehicle. For EVs, the desired output characteristics of
electric motor drives are illustrated in Fig.5. It can be observed that the electric motor is expected to be
able to produce a high torque at low speed for starting and acceleration because as it increase to the base
speed, the voltage increases to its rated voltage while the flux remain constant. Also, electric motor
generates a high power at high speed for cruising due to the fact that beyond the base speed, the voltage
remains constant and the flux is weakened. The result is constant output power while the torque declines
hyperbolically with speed. Thus, a single-gear transmission will be enough for a light-weight EV [2]. In
Fig.5, as we know from above, the electric motor can generate constant rated torque up to its base speed.
At this speed, the motor reaches its rated power limit. The operation beyond the base speed up to the
maximum speed is limited to the constant power region. This region of the constant power operation
depends primarily on the particular motor type and its control strategy.
Figure 5. Typical Performance Characteristics of Electric Motor for Traction. [2]
From the output characteristics of electric motor for EVs, the following valuable results can be concluded
as follows: a) the power requirement (rated power) for acceleration performance (acceleration time and
acceleration distance) decreases as constant power region ratio increases. b) Conversely, the torque
requirement (rated torque) for acceleration increases as constant power region ratio increases. This results
in a larger motor size and volume. c) The maximum speed of electric motor has a pronounced effect on
the required torque of the motor. Low speed motors with the extended constant power speed range have a
much higher rated shaft torque. Consequently, they need more iron and copper to support this higher flux
and torque. d) As motor power decreases (due to extending the range of constant power operation), the
required torque is increasing. Therefore, although the converter power requirement (hence the converter
139
cost) will decrease when increasing the constant power range, the motor size, volume, and cost will
increase. e) Increasing the maximum speed of the motor can reduce the motor size by allowing gearing to
increase shaft torque. However, the motor maximum speed cannot be increased indefinitely without
incurring more cost and transmission requirements. Thus, there is multitude of system level conflicts
when extending the constant power range [4].
Next, it is the tractive effort of the electric motor is discussed. The tractive effort which is also known as
tractive force is described as the pulling force exerted by a vehicle. The tractive effort generated by a
traction motor on driven wheels and vehicle velocity are expressed as:
(1)
And
(2)
where and are respectively the motor torque output and speed (rpm), is the gear ratio of
transmission, is the gear ratio of final drive, is the efficiency of the whole driveline from the motor to
the driven wheels and is the radius of the drive wheels [2].
The efficiency of an electric motor varies with the operating conditions on the speed-torque curve as
shown in Fig.6 below where the most efficient operating area exist. Thus, the EPS design should be at
least be as close as possible to the area with the most efficiency [2].
Figure 6. Typical Electric Motor Efficiency Characteristics[2]
140
AC Motor and DC Motor
In this section of the paper, the basic principles of the many types of electric motor are discussed.
Furthermore, the advantages and disadvantages of each type of the AC and DC electric motors are further
discussed.
1) DC Motor
DC motors have a long history in EV use. The most commonly used version is what is
known as a series-wound motor, which means the armature and field windings are wired in series.
Other designs include shunt-wound, compound-wound, and permanent magnet motors. One of
the advantages of using the DC motor is that the design concept of the DC motor is easy to
understand due to the maturity of this technology in electric vehicle industry [7]. Thus, making it
easy for us to troubleshoot the problems related to DC motors. Besides that, DC motors provide
simple and cheaper drive design that is also easier to manufacture. Varying the speed can be
meaning a variation in a large enough potentiometer. Whether it is using Pulse Width Modulation
(PWM) or Silicon Controlled Rectifier (SCR), the speed and torque can be easily controlled. The
other good thing about the DC motor is that it has a direct and easy way in controlling speed and
torque [8]. Speed is proportional to armature voltage of the motor while the torque is proportional
to the supplied current. However, DC motors have their drawbacks which include high
maintenance due to the commutator and brushes which will wear in time and causes dust from
brushes to accumulate and soon deteriorate the motor efficiency.
2) AC Motor
At the present time AC motors are most commonly found in commercially built EVs, as
they require more sophisticated and complex control systems than DC motors. AC control
systems on the horizon and will likely become available soon in the future as it is good prospect
to venture in [8]. One of the advantages of the AC motor is its reliable operation. The design of
an AC motor requires no brushes and commutators which mean that it has low maintenance.
Thus, no brushes to be replaced and no dust accumulation form the brushes. Owing to the fact
that the there are no commutators and brushes, there will be no frictional losses and thus the
overall efficiency is higher than the DC motor. Even though the AC motor technology in electric
vehicle is still growing, the AC motor has a distinctive advantage which is the regenerative effect
which can be used to recharge the battery of EV when the car slows down [8]. However, there are
also disadvantages to the AC motor. One of them is the cost of the motor and its driver circuit.
The speed control and torque control of the AC motor is expensive due to the complexity of the
circuit. Converter circuits are needed to convert the batterys DC-to-AC and also single-phase to
three-phase. The standard AC motors cant be operated in speeds less than about 1/3 of base
speed due to thermal considerations [9].
Comparative Study on the Specific Type of Electric Motor for EV
In this section, there will be further comparisons between electric motors and their advantages and
drawbacks of implementing them in the EV.
1) Brushed DC Motor
141
The brushed series DC motor is the most affordable and readily available current
production motor. Brushed DC motors are good in achieving high torque at low speed and their
torquespeed characteristics suitable for the high starting torque situation and high load
conditions. And, these motors are easy to control their speed through variation of voltage levels.
The characteristic of series, shunt and compound motors are shown in Fig.7. Series motors are
good in applications requiring high starting torque and heavy torque overload. Shunt motors have
better controllability than series motors. Shunt motors also known as separately excited motors
are suited for field weakened operation, due to its decoupled torque and flux control
characteristics. Moreover, a range of extended constant power operation is obtained by separate
field weakening [4].
Figure 7. Characteristics of DC Motor Types
Moreover, the speed characteristics of DC motors can be observed in the Fig. 8 below.
On the other hand, permanent magnet motors are very efficient, but only in a very narrow rpm
band, and quickly lose their efficiency in the varying speeds of normal driving. Permanent
magnet motors resemble the shunt motor in their torque, speed, reversing, and regenerative
braking characteristics. However, the permanent magnet motors have starting torques several
times that of shunt motors, and their speed versus load characteristics are more linear and easier
to predict [4] [8].
142
Figure 8. Speed Characteristics of DC motors[2]
2) Brushless DC Motor (BLDC)
In a BLDC, the brush-system/commutator assembly is replaced by an intelligent

electronic controller and making the BLDC the most efficient of all electric motor. The controller
performs the same power-distribution found in a brushed DC-motor, only without using a
commutator/brush system. The controller may contain a bank of MOSFET devices to drive high-
current DC power, and a microcontroller to precisely orchestrate the rapid-changing current-
timings. Because the controller must follow the rotor, the controller needs some means of
determining the rotor's orientation/position (relative to the stator coils.) Some designs use Hall
Effect sensors to directly measure the rotor's position. Others measure the back EMF in the
undriven coils to infer the rotor position, eliminating the need for separate Hall Effect sensors,
and therefore are often called "sensorless" controllers.
Figure 9. Basic Block Diagram of a Classical Speed and Position Control for BLDC [1]
BLDC motors can be constructed in two different physical configurations: In the 'conventional'
configuration, the permanent magnets are mounted on the spinning armature (rotor) and the stator
coils surround the rotor. In the 'outrunner' configuration, the radial relationship between the coils
and magnets are reversed; the stator coils form the center (core) of the motor, while the
permanent magnets spin on an overhanging rotor which surrounds the core. In all BLDC motors,
143
the stator-coils are stationary and thus making current circulation only in the stator and not the
rotor, making the rotor to not heat up. This will further ease the cooling design for the BLDC as
the stator in conventional configuration is on the periphery of the motor and it will be easier to
cool when the stator heats up. Improvement of efficiency can be done using the inverted rotor
topology/outrunner (where the rotor rotates on the outside of the stator) has a larger magnetic
field. The additional radius provides higher torque and power density and leads to better
efficiency.
Figure 10. (a) outrunner configuration. (b) conventional configuration
3) AC Induction Motor (IM)
Squirrel cage IM have been the most suitable candidate due to their reliability,
robustness, almost maintenance-free characteristics, low manufacturing cost and the ability to
work in various conditions. The IMs are the only AC motor that has the most mature technology
in EV industry and having several features such as lightweight, small volume, low maintenance,
high reliability and efficiency. Besides, the IM offers a higher efficiency in the regenerative
effect.
Figure 11. AC Squirrel Cage IM [8]
The absence of brush friction allows the motors to have higher maximum speed, and the higher
rating of speed enable these motors to develop high output. Thus, IM does not have so much of
the speed limitation as in the dc motors. Frequency of the voltage is varied to offer the speed
control of the IM. Extended speed range operation beyond base speed is accomplished by flux
weakening, once the motor has reached its rated power capability. Field orientation control (FOC)
of IM can decouple its torque control from field control. This allows the motor to behave in the
same manner as a separately excited dc motor [4] [10]. A properly designed IM, with field
oriented control can achieve field weakened range of 3-5 times the base speed [4]. However, the
controllers of IMs are at higher cost than the ones of DC motors. Moreover, the breakdown torque
limits its extended constant-power operation. At the critical speed, the breakdown torque is
reached and any attempt to operate the motor at the maximum current beyond this speed will stall
144
the motor. Although FOC may extend constant power operation, it results in an increased
breakdown torque thereby resulting in an over-sizing of the motor. In addition, efficiency at a
high speed range may suffer in addition to the fact that its efficiency is lower than that of
permanent magnetic (PM) motors and switched reluctance motors (SRMs) due to the absence of
rotor winding and rotor copper losses [4].
4) Switched Reluctance Motor (SRM)
The SRM is an interesting candidate in the EV industry due to the high performance, low
cost, rugged structure, fault tolerant, outstanding torque-speed characteristics and reliable
converter topology. The design of SRM is similar as the stepper motor. Also, SRM drives can run
in extremely long constant power range and the torque-speed characteristic matches with the
needs of the EV [4]. SRM has high starting torque and high torque-inertia ratio. The simplicity of
the design owes to the rotor without any windings, magnets, commutators or brushes. The fault
tolerance of SRM is also extremely good. Owing to its simple construction and low rotor inertia,
SRM has very rapid acceleration and extremely high speed operation. And, SRM is suitable for
single gear transmission in EV propulsion due to its wide speed range operation. One important
feature of the SRM is the different number of poles on the stator and the rotor. The windings of
diametrically opposite stator poles are connected in a series to form the electric phases of the
motor. For each electric phase, a power electronic circuit with one or two electronic switches is
necessary for the control of unidirectional current flowing during appropriate intervals for the
torque production. Exciting the stator coils in sequence produces the torque in the motor. This
requires an inverter (i.e. a set of low cost electronic switches such MOSFETs or IGBTs) that
drives the motor from the batterys DC voltage supply [11].
Figure 12. SRM Drive System [2]
Figure 13. Cross Section of Common SRM [2]
145
5) Comparisons between Motors
DC motor will continue to be used in EVs because DC motor drives are available at the
lowest cost. From the point of view of efficiency, BLDC motors are the best choice as BLDCs
have several advantages over brushed DC-motors, including higher reliability, longer lifetime (no
brush erosion), elimination of ionizing sparks from the commutator, and overall reduction of
electromagnetic interference (EMI.). However, among the four types of motors discussed, the
lightest of all is the SRM which is a bonus point for EVs. If the choice of motor for EVs is
determined by weight, efficiency and cost, it is clear that SRM is the victor. Besides of the
efficiency, weight and cost, SRM also have the upper hand in cooling, maximum speed, fault
tolerance, and reliability. However, of the four major electric EV motors, BLDC motor has the
highest torque density, but if compared to SRM which have better speed-torque characteristic for
traction application. Even though the BLDC has a higher efficiency than SRM, taking into that
the SRMs fault tolerance plays a big role in EV. SRM are naturally fault tolerant. As for IM and
BLDC motors, their electromechanical energy conversion is interdependent upon proper
excitation. While, SRM drives have discrete phase windings and thus phase windings are
independent of each other. Therefore, if one phase in the SRM fails the SRM drive can still
operate at a lower performance until repair work is done on it. Additionally, the converter
topology used for an SRM protects it from serious electrical fault of shoot-through, which is not
eliminated fully in IM and PM BLDC motor drives [4].
METHODOLOGY
Market Research and Case Studies
Investigation and research is done on all of electric vehicles that are available in the current
market and also the up and rising technologies. The main objective is to observe and explore on the
various type of electric motor that are used in the drive train of the electric vehicle. Furthermore,
determine the value of parameters, such as torque, rpm, input power and the nominal requirement for
voltage and current.
Various Types of Motor Testing
Measurement of Winding Resistance

Magnetization Characteristics
No Load Test
Full Load Test
These tests are to determine the profile of the motor and the load capacity of the motor.
MATLAB Simulation
The simulation will determine the characteristic of the motor in the drive train and the compatibility of it
in the EVs system itself.
146
RESULT ANALYSIS
Market Research and Case Studies
Figure 14. Peugeot Partner
MATLAB Simulation
Figure 15. An Example of the Resulted Simulation of the EV Drive Train for MATLAB
CONCLUSION
As the comparisons stated, the DC motors seems to be more mature in the industry of electric
vehicles and less expensive in constructing the prototype itself. AC motor driver circuit is very
complicated and the components are expensive since the technology is not yet mature. DC motor driver
offers a more straightforward approach to speed and torque control while BLDCs are maturing to replace
DC brushed motors to overcome the problems of high maintenance. However, even though AC motor
controller technology is not yet mature in the electric vehicle industry, thus making the cost very high.
There are several car companies are pushing their way and creating their own path in developing this
147
technology as the AC motor are still can be improved further and someday overtake DC motors in the EV
industry. However, the SRM which is a very powerful candidate in the EV motor struggle proves to have
the upper hand in nearly every category of the comparisons:
a) Lower weight than BLDC and IM.

b) Lower cost than BLDC and IM due to the absence of winding and rare-earth permanent magnets.
c) Fault tolerance and safety.
d) Higher reliability and efficiency than brushed-DC motor.
The only losing points of SRM are the cost of manufacturing is lower than brushed DC motors, the
efficiency of BLDC is higher than SRM and the lacking of maturity of SRM in the EV industry as most
car companies opt for BLDC and IM while EV enthusiasts favour the low cost brushed DC motor.
REFERENCES
[1] Handbook of Automotive Power Electronics and Motor Drives, Edited by Ali Emadi, CRC Press,
Taylor & Francis Group, 2005.
[2] Ehsani, Mehrdad. (2005). Modern electric, hybrid electric, and fuel cell vehicles. CRC.
[3] Becker, T. A., & Sidhu, I. (2009). Electric vehicle in the us - a new model with forecast to 2030.
[4] Xue, X. D., Cheng, K. W. E., & Cheung, N. C. (2008). Selection of electric motor drives for electric
vehicles.
[5] Cheng, Y., & Duan, F. The Design principle of electric motors and drive systems for electric vehicles.
802-805.
[6] Halstead, J. (2005, October 25). Electric car motors. Retrieved from
http://www.evconvert.com/eve/electric-car-motors
[7] Rippel, W. (2007, January 9). Induction versus dc brushless motors. Retrieved from
http://www.teslamotors.com/blog4/?p=45
[8] Larminie, J., & Lowry, J. (2003). electric vehicle technology explained. John Wiley & Sons.
[9] Leitman, S., & Brant, B. (2009). Build your own electric vehicle. McGraw Hill.
[10] Boley, B. L. (1996). Overview of motor types tutorial. Retrieved from
http://www.oddparts.com/acsi/motortut.htm
[11] Ehsani, M., Gao, Y., & Gay, S. (2003). Characterization of electric motor drives for traction
applications. 891-896.
[12] Cho, C. Peter, & Johnson, Ralph H. Electric motors in vehicle applications. 193-198.
[13] Hashemnia, N., & Asaei, B. (2008). Comparative study of using different electric motors in the
electric vehicles.
[14] Keoun, Bradley C. Designing an electric vehicle conversion. 303 - 308.
[15] Buja, G. (2008). Light electric vehicles.
[16] Honda, Y., Nakamura, T., Higaki, T., & Takeda, Y. (1997). Motor design considerations and test
results of an interior permanent magnet synchronous motor for electric vehicles. 75-82.
148
QUICKBIRD IMAGE-BASED ESTIMATION ON NUMBER OF BUILDINGS IN EACH ZONE

FOR TRIP GENERATION MODELLING
(A CASE STUDY IN MAKASSAR CITY)
Qadriathi Dg. Bau 1*, Danang Parikesit 2,

Hartono 3 and Totok Gunawan 4
ABSTRACT:
There are two approaches commonly used to estimate the travelling need on trip generation modeling:
aggregate approach and disaggregate approach. Both approaches, however, have disadvantages. The
aggregate approach which uses administration boundary-based zone has disadvantages. Meanwhile, the
disaggregate approach shall be done with census which will inevitably demand long duration, high cost
and numerous personnel for wide areas. Therefore, a remote sensing image which refers to Quickbird
imagery is used to define the zone system. Of each zone within the Quickbird imagery, the land use type
which is divided into two criteria is analyzed. The criteria cover generation for settlement consisting of
15 criteria and attraction for non-settlement consisting of 16 criteria. The number of buildings in the
generation is estimated based on zone width, while the type of buildings in the attraction is already
specific. The research was conducted in Makassar city in which 4 (four) sub-districts were taken as
samples. The interpretation and analysis of the Quickbird image result in 643 zones with their own
characteristics. The number of smallest buildings reaches 9 units in dispersed and planned-patterned
settlement in Wajo Sub-district whereas that of the largest buildings reaches 1,271 in very dense and
unplanned-patterned settlement in Bontoala Sub-district.
Keywords: Zone, Quickbird Imagery, Number of Buildings, Generation Trip, Makassar City
1
Ph. D Candidate, Dept. of Geograhy, UGM, Bulak Sumur, Yogyakarta, 55281, INDONESIA
2
Prof. Dr., Dept. of Civil and Enviromental Engineering, UGM, Jl. Grafika No. 2 Yogyakarta, 55281, INDONESIA
3
Prof. Dr., Dept. of Geograhy, UGM, Bulak Sumur, Yogyakarta, 55281, INDONESIA
4
Prof. Dr., Dept. of Geograhy, UGM, Bulak Sumur, Yogyakarta, 55281, INDONESIA
*Correspondence Author: Qadriathi Dg. Bau, Bulak Sumur,Yogyakarta, 55281, INDONESIA. Tel: +62 274 545965
Fax: +62 274 545965. E-mail: athieq23@yahoo.com
149
INTRODUCTION
Land use and transportation have an interactive correlation in which land use is one of the
determiners of movement and activities commonly called as trip generation. The trip generation
determines type of transportation facilities for the movement, and when the additional facility is available,
the accessibility rate increases. The changes in accessibility will determine changes which will contribute
to the land use, so the level of generation trip will also change if the land use changes (Khisty dan Lall,
2005). Therefore, prior to the transportation planning as the basis for traffic generation analysis, land use
management shall be analyzed first as different land use management results in different traffic generation
characteristics in term of traffic density, and traffic type and duration.
Two approaches are commonly used for estimating trip need on trip generation modeling namely
Aggregate Approach and Disaggregate Approach. The Aggregate Approach is in overall conducted by
understanding zone attributes either origin zone or destination zone such as social-economic condition of
a zone, zone population, area development and land use pattern of a zone. Meanwhile, the Disaggregate
Zone is done per individual by directly understanding factors contributing to trip but embedded to the
individual doing the trip. Both approaches have weaknesses. De la Barra (1989) said that errors in
Aggregate Approach are generated by:
a. Intra-individual variations;
b. Intra-choice variations;
c. Individual within a group posses similar information over a particular choice;
d. Variables in the utility function does not have same weight;
e. Exact location of consumers even when they are in the same zone;
f. Social and economic position of individual within the group;
g. The criteria to define a group is varied;
h. The best option is not always available.
The Disaggregate Approach for wide areas will demand long duration, high cost, and numerous
personnel, therefore, the approach conducted to review and perform modeling towards
generation/attraction trip in the research is based on zone system resulted from Quickbird image
(Qadriathi, 2009).
Thomas (2001) states that ten aspects enable the use of remote sensing data on transportation:
(1) availability of data; (2) cost; (3) licensing; (4) Image- processing software ; (5) user conservatism;
(6) training; (7) database management; (8) real-time imagery; (9) currency, and (10) Communication.
Referring to Quickbird image-based zone system, the number of buildings is then estimated
especially the settlement buildings to be the basis for finding out the size of generation trip resulted from
the settlement in the zone.
LITERATURE REVIEW
Remote Sensing System

The development in remote sensing science particularly in satellite system allows the users to
obtain data in a considerable quantity with short repetition duration and better precision. The system can
be used to get data and information about symptoms and surface of the Earth because the sensing tool
installed can capture energy of electromagnetic wave rebound by objects on the surface of the Earth
(Dulbahri, 2003).
150
Currently, various remote sensing images have different capability generated by their resolution.
There are 4 (four) crucial resolution concepts in remote sensing which consist of spatial resolution,
spectral resolution, radiometric resolution, and temporal resolution (Danoedoro, 1996, Purwadhi, 2001,
Xiong, 2004, Lillesand et al, 2004, Gopalan, 2006).
Spatial resolution is the size of object detectable by imagery system such as SPOT (Systme Pour
LObservation de la Terre) panchromatic image with 10 x 10 m-sized pixel, IKONOS with 1 m
resolution for panchromatic and 4 m for multispectral, and Quickbird image with spatial resolution
0. 61 cm for panchromatic, and 2.40 m for multispectral. The image with 1 m resolution can identify
small roads and rivers, meanwhile the image with 30 m resolution cannot do it. (Xiong, 2004, Lillesand
et al, 2004, Gopalan, 2006).
Spectral resolution is the capability of optical and electronic system to distinguish objects based
on spectral reflection or gleam. The more bands, the higher the spectral resolution is. ASTER (Advanced
Spaceborne Thermal Emission and Reflection Radiometer) image which has 14 bands gives more detailed
spectral information than Landsat ETM image which has 7 bands only.
Radiometric resolution is the capability of censor to convert intensity of spectral reflection or

gleam into digital coding stated in bit. Remote sensing system with 8 bit radiometric resolution will have
grey scale between 0 255 (28 = 256) (Xiong, 2004), for example, Landsat TM. 11 bit system produces
brightness scale between 0 2048 (211 = 2048) as IKONOS and Quickbird have (Lillesand et al, 2004).
Temporal resolution is the capability of remote sensing system to do repeated recording to the
same areas. High temporal system as on NOAA (National Oceanic and Atmospheric Administration)
satellite can do repeated recording to the same area within 12 hours better than Landsat ETM+ image and
SPOT-5 image which do repeated recording to the same areas within 16 days and 26 days respectively,
while Quickbird image needs 1 to 3 days only.
Therefore, each research using remote sensing images should adjust to its objective, theme and
final scale as each remote sensing image has different characteristics based o the four resolutions.
Zone
In transportation modelling, resolution level used in a research site is subject to attention. Hence,
the approach is done by aggregating individuals in an area into area unit commonly called as Zone.
Characteristic of a movement is the mean (or the one representing) of the whole area of a Zone.
Determining research site and zone boundary often needs compromise of contradictory requirements. On
one hand, there is a demand to improve the model accuracy by enlarging size of research site and its
complexity, on the other hand, there is practical consideration to keep minimizing costs and meet the
required time scale and accuracy (Tamin, 2000).
Several key criteria to consider in determining zone system within the research area suggested by
IHT and DTp (1987) in Tamin (2000) and Ortuzar and Willumsen (1990), include:
a. Zone size shall be designed in a way that the possible group error generated by assumed centralised
activities in a zone center shall not be increasingly too large. Zone system with numerous small zones
can be applied;
b. Zone boundary should be adjusted to census boundary, regional administration boundary, natural
boundary and zone boundary used by the previous research considered as the key criteria;
c. Zone size shall be adjusted to network density to model in which the size of zone is increasingly
bigger if the zone is farther from the downtown;
151
d. Zone size shall be bigger to enable the traffic density to move to the road network with the required
accuracy;
e. Zone boundary shall be made in a way to suit the type of development pattern for each zone like
settlement, industry and offices;
f. Zone boundary shall suit the region boundary used for data collection;
Zone size is also determined by congestion rate; the zone size in the congestion-prone areas
should be smaller than the opposite areas.
METHODOLOGY
The research location in Makassar City falls into 4 (four) sub-districts: Makassar Sub-district,
Bonto Ala Sub-district, Wajo Sub-district, and Mariso Sub-district. The research uses remote sensing data
which is Quickbird image as the research material and supporting maps and other secondary data as the
references, meanwhile, the data are processed with Geographic Information System (GIS).
The steps within the research comprise steps for image processing, visual interpretation of
Quickbird image, and analysis. The first step is processing image in which ArcGIS 9.3 software is used.
Geometric correction of Quickbird image is an effort to correct recording errors geometrically so that the
produced image has uniformed coordinate and scale system. Geometric correction is based on process of
spot coordinate in image which still contains geometric errors to be geometrically correct image assisted
by certain data references or rectification process with the following steps:
1. Determining 48 Ground Control Points (GCP)

2. Calculating and testing transformation with a first-order polynomial transformation;
3. Calculating root mean square (RMS) Error; and
4. Resampling process uses bilinier interpolation
The second step is interpreting Quickbird imagery. Basically, interpreting image consists of two
processes: formulating object and element identity detected on image and finding the importance of the
object and element. The image characteristics such as size, shape, shadow, association, color, pattern,
texture, height and site are used as object identification which are further analyzed and deducted to find a
meaningful correlation in the next process. It results in classification leading to theorization. This is the
basis to interpret and delineate road block, land use type, building order, and building density.
Classification of land use and building density refers to the criteria proposed by Sutanto (1981) with a
modification, and that of building order refers to the classification proposed by Suharyadi (2001) with a
modification.
The final step is the analysis step by accuracy test to land use mapping in the zone of Quickbird
image assisted by GIS and accuracy test to building quantity estimation in zone based on width of
samples assisted by SPSS. The research uses three approaches: (1) processing Quickbird imagery to
obtain map of tentative land use, (2) visually interpreting Quickbird imagery assisted by GIS to gather
data of road block, building density and building order precisely, and (3). ) Data processing for estimating
number of buildings in the zone.
The research results in a zone map consisting of land use map in every road block divided into 2
major groups: trip generation map covering all building of settlement types, and attraction map covering
non-settlement buildings completed with their width. On each land use particularly generation trip, the
number of buildings is estimated, whereas no number of buildings is estimated for the land use
categorized into trip attraction because the object is found out through individual building. The research
uses proportionate stratified random sampling for sampling technique. The maps are further made the
152
basis for formulating correlation between generation/attraction based on Quickbird Image on generation
trip modeling. GIS processes two types of data: spatial data or graphic data and attribute data. In this
research, the processing of attribute data only includes calculation, whereas that of graphic data consists
of (1) change of scale; (2) re-classification; and (3) graphic data calculation. Detailed data can be seen in
Table 1 and Table 2.
Figure 1: Corrected Quickbird Imagery Describing Research Area.
Table 1: Building Density Classification

No Building Density Density Class
1. 20% Dispersed
2. 21% - 40,5% Sparse
3. 41% - 60,5% Semi dense
4. 61% - 80% Dense
5. > 80% Very Dense
Table 2: Building Order Classification
No Building Order Building Pattern
1 > 60% buildings facing the neighborhood road Planned
2 40% - 60% buildings facing the neighborhood Semi planned
road
3 < 30% buildings facing the road unplanned
RESULT ANALYSIS
The research results in a zone map consisting of land use map in every road block divided into 2
major groups: trip generation map covering all building of settlement types, and attraction map covering
non-settlement buildings completed with their width. On each land use particularly generation trip, the
number of buildings is estimated, whereas no number of buildings is estimated for the land use
categorized into trip attraction because the object is found out through individual building. The research
uses stratified sampling for sampling technique. The maps are further made the basis for formulating
correlation between generation/attraction based on Quickbird Imagery on generation trip modeling.
153
Land Use Map
Information about land use is deemed necessary as land use is one of the movement determiners.
In USA, the most common system of closing classification and land use is United States Geological
Survey (USGS) System; however, this research uses modified Sutanto Classification (1981). The
approach done to land use mapping is photo key in which field survey is based on samples selected
through introduction to images or specified photomorphic characteristics in Quickbird Imagery. The
result of object introduction in field with the samples becomes the photo key for land use mapping in all
research areas. The criteria for land use are adjusted to trip generation/trip attraction map. Appendix 1
presents the classification of land use.
Trip Generation/Actraction Map
Each type of land use will generate different number of movements. Therefore, they shall be
broken into trip generation and trip attraction for detailed analysis. Trip generation consists of 15
criteria in which each criteria has certain width and 16 criteria for attraction in which each criteria has
specified width in each zone. On the generation trip, the building density and order are further
analyzed based on zone meanwhile on the attraction map, the building individual is so distinct that only
interpretation test will be done.
Figure 2: Trip Generation Map of Makassar City
Figure 3: Trip Attraction Map of Makassar City
154
Table 3: Type of Trip Generation and Width in 4 (four) Sub-districts in Makassar City
No Type of Land Use Sub-district and Width of Land Use (m2)
Wajo Makassar Bontoala Mariso
1. Dispersed and semi - - - 54,374.86
planned settlement
2. Sparse and semi 1,206.64 - - 134,292.17
planned settlement
3. Semi dense and 1,944.55 32,199.68 1,944.55 58,041.80
semi planned
settlement
4. Dense and semi 2,328.62 110,379.08 21,406.33 44,089.35
planned settlement
5. Very dense and semi 218,480.20 448,305.25 233,385.44 99,619.15
planned settlement
6. Dispersed and 1,031.10 125.69 - 24,885.2
planned settlement
7. Sparse and planned - - - 6,163.43
settlement
8. Semi dense and 6,403.48 1,425.45 1,140.37 17,130.66
planned settlement
9. Dense and planned 27,328.11 11,993.45 15,631.36 6,316.57
settlement
10. Very dense and 145,646.80 365,983.74 131,409.22 145,794.35
planned settlement
11. Dispersed and - - - 66,087.66
unplanned
settlement
12. Sparse and - 20,104.35 - 58,644.29
unplanned
settlement
13. Semi dense and 9,150.99 - - 85,668.53
unplanned
settlement
14. Dense and 9,661.02 88,110.06 39,325.02 27,513.04
unplanned
settlement
15. Very dense and 25,463.99 161,137.06 241,719.06 170,672.93
unplanned
settlement
Accuracy Test
The accuracy test of interpretation result towards land use is deemed necessary in processing data of
remote sensing either in a manual or digital way. To find out the data validity of Quickbird imagery,
accuracy test shall be done by comparing map of interpretation result with quality reference data. The
reference data can be obtained from field measurement or large-scale image. The accuracy test to land use
map interpreted from Quickbird imagery results in 98.5% accuracy. The high accuracy is generated by
spatial resolution of 61 cm-sized Quickbird imagery which implies that 61 cm is the size of the smallest
spotted object. In addition to accuracy test for interpretation result, statistical test is also done towards
estimated number of buildings based on width of building unit and that of zone. The reference is
comparing visual calculation of number of buildings clearly visible in Quickbird imagery to calculation of
zone width number divided by roof width of individual building. Assisted by SPSS, the result shows that
sig value (2-tailed) = 0.887 > 0.05 (taken = 0.05) thus, H o is accepted. The result implies that no
155
difference between the mean of estimation data and the factual data in samples; therefore, the method can
be used for the overall data. Appendix 2 presents the detailed result of estimation to sample building
number.
Simulation of Movement Size on Trip Generation Modeling
Traffic generation depends on two aspects of land use management: type of land use management and
number of activities on the land. Further, the land use and width obtained from the zone based on
Quickbird imagery is translated into transportation size. Hence, the generation trip serves as a function of
the width and type of zone land use management in Quickbird imagery. After generation model of each
type of land use is attained, joint to attribute data in each zone is done assisted by GIS so as to complete
each graphic data with generation width and model. Afterwards, 3-dimensional figure like in Figure 4 is
presented with ArcScene assistance. Figure 4 displays 3-dimensional perspective simulation of each land
use in zone in Bontoala Sub-district with color symbolization for each type of land use, meanwhile the
Height represents number of movement resulted from land use of each zone based on Quickbird imagery
Sports Hall
Hotel
Industry
Recreational Area
Cemetery
Campus
Unoccupied Land
School
Shopping Area
Mosque
Market
Dispersed and Semi Planned Settlement
Sparse and Unplanned Settlement
Semi Dense and Unplanned Settlement
Very Dense and Unplanned Settlement
Gas Station
Hospital
Figure 4: 3-Dimensional Perspective Simulation of Movement Number in Zone in Bontoala

Sub-district
CONCLUSIONS
Quickbird image can be used to capture zone-based data on the basis of road block and
width, land use type and width, building density and building order with image processing within
98.5% accuracy level for each land use.
Number of buildings obtained from calculating zone width divided by roof width of
individual building based on Zone of Quickbird image is very important for an analysis of
detailed movement need-based approach. It is considered better than using Aggregate Approach
based on administration area as zone boundary or Disaggregate Approach by census which
demands individual approach towards movement element in trip generation modeling.
The estimation to building number of each zone is then used to calculate the mean of
movements by considering the social economic condition based on land use type. After being
156
identified, the model of each land use is multiplied by the width so that the final result can
indicate the number of movements generated by each zone.
ACKNOWLEDGMENT
The research is a part of a dissertation; therefore, my gratitude shall go to Prof.Dr-Tech. Ir. Danang
Parikesit, M.Sc for his academic assistance for the dissertation and financial assistance for procuring
Quickbird image and field work with my engagement in World Class Research University (WCRU) and
the research conducted for Competitive Grant for International Cooperation Research. I shall also extend
my gratitude to Prof. Dr. Hartono, DEA., DESS and Prof. Dr. Totok Gunawan, MS as Ph.D supervisors.
REFERENCES
Danoedoro, P. (1996). Digital Image Processing: Theory and Application in Remote Sensing.
Yogyakarta : Geography Faculty, Gadjah Mada University.
de La Barra, T. (1989). Integrated Land Use and Transport Modelling: Decisions Chains and
Hierarchies, Great Britain: Cambridge University Press,
Dulbahri, (2003). Digital Analysis of Remote Sensing Data. Yogyakarta: Puspics of Geography Faculty,
Gadjah Mada University.
Gopalan, (2006). High Resolution Imagery for Developmental Planning with Special Reference to
Developing Economies. Retrieved from F://Jurnal%20Transportasi/Transport 3.Htm (23
Oktober 2007)
Khisty, C.J., & Lall, B.K. (2002). Basic Concepts in Transportation Engineering. Jakarta: Erlangga
Publisher.
Lillesand, T.M, Kiefer R.W. & Chipman J.W. (2004). Remote Sensing and Image Interpretation, New
York: John Wiley and Sons.
Ortuzar, J.D. and Willumsen, L.G. (1994). Modelling Transport. England: John Wiley and Sons.
Purwadhi, F. S. H. (2001). Interpreting Digital Image. PT. Gramedia Widiasarana Indonesia,
Jakarta.
Suharyadi. (2004). Mapping of Urban Building Density in Yogyakarta based on Landsat Thematic
Mapper Image. Geographic Information Science. Yogyakarta: Geography Faculty,
Gadjah Mada University.
Sutanto. (1981). Use of Aerial photos for Making Land Use Map in Yogyakarta Municipality Yogyakarta:
Puspics of Geography Faculty, Gadjah Mada University.
Sutanto. (1994). Remote Sensing 2nd Edition. 2nd Edition. Gadjah Mada Yogyakarta: University
Press
Tamin, O. Z. (2000). Planning and Modelling of Transportation. 2nd Edition., Bandung:ITB
Publisher
Thomas, M.R. (2001). Top 10 Issues in Remote Sensing for Transportation (with Predection):
Remote Sensing and Challenges. Remote Sensing for Transportation-Product and
Result:Foundation for Future.TRB Conference Proceeding 29, Washington, DC
Qadriathi D. B., Parikesit, D., Hartono., & Gunawan, T. (2009). Using Quickbird Image for Zone
Analysis on Transportation Modeling. Proceedings of Simposium in
Geoinformation Science I, 23.
Xiong, D., Lee, R., Saulsbury, J. B., Lanzer, E. L., and Perez, A. (2004). Guidance on Using
Remote Sensing Application for Environmental Analysis in Transportation
Planning. Research Report W A-RD 593-2. Tennessee: Oak Ridge National
Laboratory.
157
No Land Use Type

1. Dispersed and semi planned settlement
2. Sparse and semi planned settlement
3. Semi dense and semi planned settlement
4. Dense and semi planned settlement
5. Very dense and semi planned settlement
6. Dispersed and planned settlement
7. Sparse and planned settlement
8. Semi dense and planned settlement
9. Dense and planned settlement
10. Very dense and planned settlement
11. Dispersed and unplanned settlement
12. Sparse and unplanned settlement
13. Semi dense and unplanned settlement
14. Dense and unplanned settlement
15. Very dense and unplanned settlement
16. Sports Hall
17. Hotel
18. Industry
19. Campus
20. Recreational Area
21. Cemetery
22. Unoccupied Land
23. School
24. Shopping Mall
25. Mosque
26. Market
27. Port
28. Business Area
29. Shopping Arcade
30. Hospital
31. Gas Station
Appendix 1: Land Use Classification
158
Zone Average Building Number of

Width Width in Zone Building in
No Sub-district Land Use for Generation (m2) (m2) Zone
1. Bontoala Dense and semi planned settlement 8,069.42 337 52
2. Bontoala Very dense and semi planned settlement 37, 130.60 337 457
3. Bontoala Dense and planned settlement 7,777.23 1,537 36
4. Bontoala Very dense and planned settlement 20,641.09 347.75 233
5. Bontoala Dense and unplanned settlement 26,386.16 365.28 224
6. Bontoala Very dense and unplanned settlement 143,041.51 1,822.28 1,271
7. Makassar Dense and semi planned settlement 59,766.01 1,520.14 502
8. Makassar Very dense and unplanned settlement 105,696.59 1,687.65 906
9. Makassar Semi dense and semi planned settlement 18,843.85 1,090.04 208
10. Makassar Dense and planned settlement 7,948.81 364.27 61
11. Makassar Very dense and planned settlement 24,147.73 528.82 174
12. Makassar Semi dense and planned settlement 1,425.45 135.75 11
13. Makassar Sparse and unplanned settlement 8,226.04 1,778.59 49
14. Makassar Sparse and planned settlement 2,180.81 139.76 21
15. Makassar Sparse and unplanned settlement 9,697.50 1,103.49 63
16. Makassar Dense and unplanned settlement 55,746.86 103.83 419
17. Makassar Very dense and unplanned settlement 117,002.99 1,157.66 1,009
18. Makassar Dense and unplanned settlement 18,715.87 877.24 112
19. Mariso Sparse and semi planned settlement 49,969.41 992.22 403
20. Mariso Dense and semi planned settlement 27,516.87 349.43 248
21. Mariso Dispersed and semi planned settlement 27,429.61 835.31 249
22. Mariso Very dense and semi planned settlement 10,285.75 91.63 112
23. Mariso Semi dense and semi planned settlement 40,112.85 570.52 164
24. Mariso Sparse and planned settlement 5,074.73 349.59 24
25. Mariso Dense and planned settlement 6,316.58 382.82 33
26. Mariso Dispersed and planned settlement 26,625.65 2,884.27 368
27. Mariso Very dense and planned settlement 40,983.77 656.64 247
28. Mariso Semi dense and planned settlement 13,391.02 524.28 113
29. Mariso Sparse and unplanned settlement 29,560.15 410.67 275
30. Mariso Dispersed and unplanned settlement 27,513.04 130.60 211
31. Mariso Dispersed and unplanned settlement 47,041.08 2,970.10 291
32. Mariso Very dense and unplanned settlement 90,498.90 435.29 1,152
33. Mariso Semi dense and unplanned settlement 56,365.95 186.86 615
34. Wajo Dense and semi planned settlement 5,018.11 40.00 126.52
35. Wajo Sparse and semi planned settlement 1,206.64 72.60 17
36. Wajo Very dense and semi planned settlement 105,209.65 1,497.72 980
37. Wajo Dense and planned settlement 2,187.99 151.89 29
38. Wajo Dense and unplanned settlement 21,059.92 617.04 22
39. Wajo Dispersed and planned settlement 1,031.10 108.81 9
40. Wajo Very dense and planned settlement 22,596.96 238.59 177
41. Wajo Semi dense and planned settlement 4,273.31 83.25 51
42. Wajo Very dense and unplanned settlement 14,886.3 113.94 131
43. Wajo Semi dense and unplanned settlement 9,150.99 109.44 84
Appendix 2: Result of Number of Building Estimation on Zone based on Quickbird Image
159
160
TEST PROTOCOL FOR MALAYSIAN ARMED FORCES (MAF) VEHICLES - HEALTH HAZARD
ASSESSMENT (HHA) AND VEHICLE PERFORMANCE
Adam Gani1*, Shamsul Akmar Ab Aziz2 and Abdul Hamid bin Hassan3
ABSTRACT:
In Malaysia, there is no standard for scientific test protocols for Malaysian Armed Forces (MAF)
vehicles. This test protocol is made based on test procedures employed by the Science and
Technology Research Institute for Defence (STRIDE) Vehicle Branch team, which is considered as
the best practise at this moment. The test protocol is divided into two categories, Health Hazard
Assessment (HHA) and vehicle performance. HHA will focus on vibrations, noises and heat stress
evaluation. Whereas, vehicle performance will be assessed by brake efficiency performance,
maximum speed and turning circle capability. All tests must be performed in-field with certain test
ground conditions.
Keywords: MAF vehicles protocol test; Health hazard assessment; Vehicle performance.
1
Research Engineer, Acceptance of Sample & Inspection of Equipment Branch,Menara TH Selborn, Jalan Tun Razak, 50400
Kuala Lumpur,MALAYSIA
2
Research Engineer, Science & Technology Institute for Defense, Taman Bukit Mewah Fasa 9, 43000 Kajang, Selangor,
MALAYSIA.
3
Abdul Hamid bin Hassan, Senior Research Officer, Science & Technology Institute for Defense, Taman Bukit Mewah Fasa 9,
43000 Kajang, Selangor, MALAYSIA.
*Correspondence Author: Adam Hj Gani, Acceptance of Sample & Inspection of Equipment Branch, Malaysia. Tel: +6003
26872806. E-mail: ag_stride@yahoo.com
161
INTRODUCTION
As the government body responsible for the scientific tests of Malaysian Armed Forces
(MAF) vehicles, the Science and Technology Institute for Defence (STRIDE) is committed to
ensuring that all vehicles that are to be used by the MAF are safe and comfortable. To fulfil these
requirements, several tests have to be performed.
In Malaysia, there is no standard for scientific test protocols for MAF vehicles. Hence, this
paper is aimed at proposing a suitable test protocol for MAF vehicles. This test protocol is made
based on test procedures employed by the STRIDE Vehicle Branch team, which is considered as the
best practise at this moment.
HEALTH HAZARD ASSESSMENT (HHA)
Whole Body Vibration (WBV)
WBV is defined as the mechanical vibration that, when transmitted to the whole body, entails
risks to the health and safety of workers, in particular lower-back morbidity and trauma of spine2. The
mentioned ailments will not happen in a short period of time but will happen a few years later.
Frequent exposure to excessive vibrations is the main cause of the problems. In order to minimize the
risks, MAF personnel shall perform their work within a permissible exposure action limit. The daily
exposure action value standardised to an eight-hour reference period shall be 0.5 m/s2 (2). Figure 1
shows the exposure duration and caution zone for the WBV.
In conducting WBV measurements, vibration measurements shall be made as close as

possible to, or through, the point at which the vibration is transmitted5. For a seated person, the
measurement is made using a seat pad containing 3-axis accelerometers (vibration transducers). The
vehicle vibration will be transmitted to the body through 3-axis transducers. The WBV data will be
recorded by the transducers. The measurement will be carried out in four conditions; idle, and at
velocities of 20 km/h, 40 km/h and 60 km/h 1. The vehicle speed shall stable for a few minutes at
every condition. The pavement road condition shall be straight for at least 3 km. For WBV
measurement, three positions will be selected, at the driver, co-driver and passenger positions. The
MAF personal is seated facing sideways in the vehicle. The data obtained will be analyzed and,
comparisons will be made with Figure 1 in order to get the exposure limit. The conclusions shall be
made according to Table 1.
Table 1: Comfort reactions to vibration environments (Source: ISO 2631-1 (1997))
Vibration Value (m/s2) Comfort reactions

< 0.315 not uncomfortable
0.315 - 0.63 a little uncomfortable
0.5 1.0 fairly uncomfortable
0.8 1.6 Uncomfortable
1.25 2.5 very uncomfortable
>2.5 extremely uncomfortable
162
Figure 1: Health guidance zone (Source: ISO 2631/1(1997))
Hand Arm Vibration (HAV)
HAV is defined as the mechanical vibration that, when transmitted to the human hand-arm
system, entails risks to the health and safety of workers, in particular vascular, bone or join,
neurological or muscular disorders2. The exposure of vibration to that particular part in long periods
of time is the main cause of these ailments. The daily exposure action value standardised to an eight-
hour reference period shall be 2.5 m/s2. The assessment of the level of exposure to HAV is based on
the calculation of the daily exposure value normalised to an eight-hour reference period A (8) 2. In
conducting HAV measurements, vibration measurements should be made at or near the surface of the
hand (or hands) where the vibration enters the body6 . The transducer should be located at the middle
of the gripping zone; it is at this location that the most representative evaluation of the vibration
entering that hand is obtained6. Figure 2 shows the transducer location for vehicle steering.The
measurements will be carried out in four conditions; idle, and speed at velocities of 20km/h, 40km/h
and 60km/h. The vehicle speed shall stable for a few minutes at every condition. The pavement road
condition shall be straight for at least 3 km. The measurement shall be taken at least three times for
each condition.The data obtained shall be analyzed, and data comparisons with Figure 3 shall be
made.
Figure 2: Measurement location for vehicle steering. (Source: ISO 5349-2 (2001))
Figure 3: Hand-arm vibration exposure limit.

(Source: (WISHA, 2003))
163
Pass-by Noise
When the vehicle is passing in front of us, or surrounding us, the sound produced by the
vehicle (e.g engine, exhaust and tyres) is measured. The measurement of that sound is known as the
pass-by noise test. The pass by noise test is performed to meet the Environmental Quality Act 1974
requirements. The basic instrument for sound measurement is the sound level meter (SLM). This is a
microphone connected to a voltmeter with some additional features. The pass-by noise test shall be
performed according to Environmental Quality Act 1974-Environmental Quality (Motor Vehicle
Noise) Regulation 1987. The main parameters that should be understood before the test is carried out
are the instrumentation, the test site, background noise and wind interference, and the test procedure.
The detailed instructions described in Environmental Quality Act 1974 shall be followed. Failure to
do so will cause the results to be invalid. The maximum level of sound which may be emitted by
motor vehicles having more than three wheels shall be the sound levels of Standard C as prescribed in
the Second Schedule of Environmental Quality Act 1974 as shown in Table 2.
Figure 4: Microphone positions for measurements.

(Source: Environmental Quality Act 1974 (1987))
In performing the test, the microphone shall be placed in the positions shown in Figure 4. The
vehicle shall approach the line AA with the path of its centre line following as closely as possible the
line CC3. When the front of the vehicles reaches the line AA, the throttle shall be fully opened as
rapidly as practicable, and held fully opened until the rear of the vehicle reaches line BB, where the
throttle shall then be closed as rapidly as possible3. The data obtained shall be interpreted by
comparison with the Standard C shown in Table 2.
Maximum
Table 2: Standard C (Source: Environmental Quality Act 1974 (1987))
Sound
No Level
Category of Vehicle Permitted
(dB (A))
1. Used for the carriage of passengers and comprising not more than 9 seats (including 80
the driver's seat)
2. Used for the carriage of passengers and comprising more than 9 seats. Permitted 81
maximum weight does not exceed 3.5 tonnes
3. Used for the carriage of goods. Permitted maximum weight does not exceed 3.5 81
tonnes. Engine is less than 200 h.p. DIN
maximum weight exceeds 3.5 tonnes. Engine is less than 200 h.p. DIN
maximum weight exceeds 3.5 tonnes. Engine is 200 h .p. DIN or more
6. Used for the carriage of goods. Permitted maximum weight exceeds 3.5 tonnes. 86
Engine is less than 200 h.p. DIN
7. Used for the carriage of goods. Permitted maximum weight exceeds 3.5 tonnes. 88
Engine is 200 h.p. DIN or more
164
The measurement of engine noise shall be performed in the stationary position. During the
measurement, all other equipment inside the vehicle shall be switched off. The sound of air-
conditioning and radio, and other unwanted sounds, may interfere with the reading. The test site shall
be of any open space with a flat area made of concrete, asphalt or hard material having a high
acoustical reflectivity7. The edges of the test site shall be at least 3m from the vehicle and there shall
be no object present on the test site7. The microphone positions shall be as shown Figure 5 and
according to engine location and driver position. The microphone shall be placed at a height of 0.5 m
from the ground7. Nowadays, most of vehicles have their engines in the front; hence, the test only will
be conducted for this type of vehicles. The microphone positions shall be as shown Figure 6.The noise
emitted by the engine will be measured in two conditions: idle and full throttle. Before the
measurement commences, the engine shall be stabilized at the idle position for a few second before
the reading is taken. At least three measurements shall be carried out, with only the highest level
being noted.
Figure 5: Microphone positions for noise measurement. (Source: ISO 5130

(1982))
Figure 6: Microphone positions A, B and C for front engine.
Measurement of Noise near the Exhaust
The measurement of exhaust noise is similar to engine noise measurement, but the difference
is the microphone position. The test site, measurements methods and other conditions are similar to
what had been described in Section 2.4. The test method shall refer to ISO 5130 as a detailed
guideline of the measurement. For this purpose the microphone height shall be not less than 0.2m and
at the angle of 45010 from the gas flow (Figure 7). At least three measurements shall be carried out,
and will be consider valid if the range of three measurements made immediately one after the other is
not greater than 2 dB 7.
165
Figure 7: Microphone positions for exhaust test (Source: ISO 5130 (1982))
Measurement of Noise inside the Vehicle
Monitoring of internal noise is very important in order to check that the noise of vehicles is
still within prescribed limits8. To perform the test, the vehicle shall be in three conditions, and every
condition shall be tested separately. The conditions are steady speeds, full throttle acceleration and
stationary (ISO 5128, 1980). At least five steady speeds shall be evaluated in the range of 60km/h to
120km/h or 80% of maximum vehicle speed, whichever is lower (ISO 5128, 1980). Full throttle
acceleration shall be tested at 120km/h or 90% of the engine speeds, whichever is lower8 and at
stationary condition, two procedures, at idle and full throttle shall be tested. Other parameters that
need to be considered are microphone positions, tyres and the engine itself. The details of the test
procedures are discussed in ISO 5128 (1980).
Noise Ranging
Quieter vehicles provide a significant advantage as the possibility of vehicles sound being
heard by the enemy is less. By knowing the distance from which the engine noise can be heard, the
operation can move as close as they can to an enemy base without being heard by them. To perform
the test, the test site shall be up to 1 km in range. Microphones are positioned every 25m from point 0
up to 1 km to determine if the sound can be measured. Two vehicle conditions shall be tested; the first
condition is the front of the vehicle facing the microphones while moving away in reverse mode, and
the second condition is the rear of the vehicle facing the microphones while moving away in forward
mode. In both conditions, noise ranging for both idle and full throttle conditions is measured. Before
the measurement, the wind speed, background noise and temperature shall be noted.
Measurement of the Heat Stress
In Malaysia, the temperature can reach up to 400C on sunny days. Hench, it is compulsory to
provide each MAF vehicle with an air conditioning system (Adam et al., 2008). The heat stress
equipment consists of a Natural Wet Bulb Thermometer, a Black Globe Bulb Thermometer and a Dry
Bulb Thermometer.To perform the measurement, the equipment shall be placed inside the vehicle half
an hour before the measurement commences. Two locations shall be monitored; at passenger and
driver cabin. The vehicle shall be under direct sunlight for at least 2 hours during the measurement.
The details of the test method are discussed in (Adam et al., 2008).
166
VEHICLE PERFORMANCE
Brake Efficiency
Brake efficiency is defined as the total the braking effort recorded from all the wheels of the
vehicle when the brake is applied divided by the vehicle weight, and then multiplying the result by
1009. The test will be conducted in steady speed at 30 km/h. Speeds exceeding 30 km/h is not
advisable for all types of vehicles for safety reasons. An important condition that will be taken into
account in choosing the right speed is the vehicle weight. If the vehicle weight is more than 7000 kg,
the speed may be revised. The parameters shall be noted in the report are the tyre pressure, type of
brakes, the brake pad and disc conditions. Any other related parameters may also be considered if
necessary. Before the test commences, the tyre pressure shall be as recommended by the manufacturer
and the tyres condition shall be good. The road shall be pavement type, straight and levelled so that
the gradient influence can be reduced. A pedal force transducer shall be placed on the top of the brake
pedal and shall be tighten properly. The vehicle shall be in a constant speed at 30km/h for a few
seconds before the brake pedal is pressed. The drivers foot shall be placed on the top of the
transducer for immediate stopping by pressing the transducer as well as brake pedal. The force
generated by the braking effort that is transmitted to the accelerometer is call pedal force. In theory,
the applied brake force should be equal to the vehicle weight.
Acceleration Performance Test
The acceleration test is carried out to confirm the vehicles engine capabilities and power.
Underpowered vehicle means that the movement is slow because the power to weight ratio is small.
Vehicle pick up performance test can be done through acceleration test on a straight, levelled and
smooth road. Thus should be of tarmac or concrete surfaces for better grips and stranglehold between
the tyres and the surfaces. The vehicle engine horse power and the vehicle weight shall be recorded.
To perform the test the vehicle must be moved from 0 m to 400m. The throttle shall be fully opened
when starting from stationary position and held fully open until the front of the vehicle reaches the
marking line 400 m. The time and the highest speed are recorded once the vehicle reaches the
marking line 400m. The test shall be repeated three times. The vehicle shall be unladen except for the
driver and the tester. Air-conditioning or any other equipment shall be not operating during the test.
Turning circle is the measure of how easily the car will be to park or to perform a U-turn in
confined spaces. In military practice, a vehicle that has the capability to perform a U-turn or make a
turning in a confined space is an advantage as it provides easy manoeuvrability and faster movements
especially during operations. Usually measured kerb to kerb using the front outside tyre is in practice.
The reading shall be taken three times for the right turn as well as the left turn. Before the
measurement, the vehicle shall be in stationary position and the steering shall be turned to the
maximum, either to the right or to the left. After the completion of three readings at the right turn,
three readings to the left turn shall be performed. The average reading for each side shall be the final
result. The results shall be compared to the manufacturer data in the browser. A notable exception in
this description is of vehicles that are capable of spinning around their central axis, such as a tank or
certain lawnmowers as they do not form a circular path as they turn. In this case the vehicle is referred
to as a "zero turning radius" vehicle, although whether or not the turning radius is actually nonexistent
is unclear11. Figure 8 shows how to measure the turning circle.
Figure 8: Turning Radius(Source: Wikipedia (2008))
167
CONCLUSION
All the tests mentioned in this paper are based on the STRIDE Vehicle Branchs current test
procedures, which have proven to be highly effective, and satisfies the MAFs strict requirements.
Most of the test methods and approaches are adapted from ISO standards, and some are adapted from
directives and journals. For the tests which are not provided with specific references, such as brake
efficiency, acceleration performance, turning circle and noise ranging, these tests are adapted from
current methods that has been practiced for many years by the STRIDE Vehicle Branch. It should be
noted that, this test protocol describes the general test procedures required for testing MAF vehicles;
the conditions and procedures can vary for different vehicle requirements. Hence, the STRIDE
Vehicle Branch works closely with the MAF to ensure that the best test procedures are employed for
the testing of different MAF vehicles.
REFERENCES
A.Shamsul, G. Adam & A.Hamid.(2008).InDefenceS&T Technical Bulletin. Whole Body

Vibration(WBV) on Malaysian Armed Forces (MAF) Vehicles., 1: 7-15.
Directive 2002/44/EC. ( 2002). In On the minimumhealth and safety requirements regarding

theexposure of workers to the risks arising from physical agents (vibration). Brussels.
EnvironmentalQualityAct 974.(1987).EnvironmentalQuality (motor vehicle noise).Malaysia.
G. Adam, A. Shamsul & A. Hamid (2008). In Defence S&T Technical Bulletin. Heat Stress on
Various Types of Military Vehicles in Malaysian Climate., 1: 16-29.
ISO 2631/1 (1997). In Evaluation of human exposure to whole-body vibration. International

Organization for Standardization, Geneva, Switzerland.
ISO 5349-2 (2001). In Mechanical vibration-Measurement and evaluation of human

exposure tohand-transmitted vibration. International Organization for Standardization, Geneva,
Switzerland.
ISO 5130 (1982). Acoustics-Measurement of noise emitted by stationary road vehicle. International
ISO 5128 (1980). Acoustics - Measurement of noise inside motor vehicles. International
Ukmot, (2004). Brake System. Available online at: http://www.ukmot.com/bike_3-3.asp (Last access
date: 7th January 2008)
WISHA (2003). Hand Arm Vibration Analysis. Available online

at:http://personal.health.usf.edu/tbernard/HollowHills/WISHA_HAV.pdf (Last access date:
23rd February
2009)
Wikipedia (2008). Turning Radius. Available Online at:http://en.wikipedia.org/wiki/Turning_radius

(Last access date: 17th November 2008)
168
IMPROVISED ENGINEERING SPECIFICATION DESIGN FOR ROAD HUMP:

A CASE STUDY IN RESIDENTIAL STREET SHAH ALAM
Nor Izzah Zainuddin 1*,

Muhammad Akram Adnan , Mohd Yusof Abd. Rahman2 and Jezan Md Diah 2
2
ABSTRACT:
Generally, an excessive speed and high traffic volume are the most common problems reported on
residential streets. Traffic calming is a term most commonly associated with physical features placed on a
roadway to influence vehicle speed and the cut-through traffic volume. Thus the purpose of traffic
calming is to improve safety and comfort levels for pedestrians as well as for motor users into residential
streets. However based on observation and preliminary investigation on the study area there are problems
associated with the installation of traffic calming devices such as road hump. The problem is that there is
different hump profiles currently in use and the effectiveness of such devices is largely dependent on the
size and shape of the devices. As such, different design of road hump will give different impact to the
driver and if not properly design, it can cause unnecessary discomfort to drivers and passengers, frequent
exposure to humps can cause back pain the national osteoporosis society has stated that sufferers are
being put at risk by road humps and serious danger to bus passengers, especially the elderly. Due to
limited and rare guidelines of road hump in Malaysia a research to improvise road hump design is needed
to identify the effectiveness and a suitable design speed hump that is less hazardous at acceptable speed,
in residential street. Therefore this paper gives some overview in enhancing the understanding of the
relationship between different hump design, speeds and vibration (shock) so that measures can
be taken to address problems related safety and health hazard.
Keywords: Traffic Volume, Traffic Calming, Road Hump, Speed, Hump Design.
1
Master Student, Faculty of Civil Engineering, Universiti Teknologi MARA, 40450 Shah Alam, Selangor, MALAYSIA
2
Senior Lecturer, Faculty of Civil Engineering, Universiti Teknologi MARA, 40450 Shah Alam, Selangor, MALAYSIA
*Correspondence Author: Nor Izzah Zainuddin, Universiti Teknologi MARA, Malaysia. E-mail:
izzahzdin@gmail.com
169
INTRODUCTION
The demand by the public to reduce traffic volumes and speeds in their restricted areas has
necessitated traffic engineers and the local authority to come out with effective and safe traffic calming
strategies. Physical traffic calming devices are often installed to overcome traffic flow problems in
restricted areas but the safety and effectiveness of such devices is largely dependent on the size and shape
of such devices. Speed hump is one of the traffic calming devices used to reduce vehicular speed and is
quite common in residential areas. The design concept of hump is to control vehicular speed by introduce
shock while traverse through it. As such, high vibration level is expected when a vehicle passes over it
at higher speed than the allowable limit. Hump geometry is a major factor in altering the level of shock
in-line with the anticipated speed limit. Currently, in Malaysia there are limited studies and guidelines are
rare on the relationship between hump geometric designs, speeds and vibration incur while passing over a
speed hump. The intention of this study is to evaluate the inter-connecting factors related to hump design
parameters with the effectiveness as well as comfort (level of shock). Such an approach (an ergonomic
approach) may explain the needs to harmonize the designer adoption in designing speed hump and the
road user expectation. The terminology of ergonomic means that the designer needs to take into
consideration of road user expectation and ability or consideration of human factor in engineering design
so that the discomfort of passing over a speed hump can be reduced or eliminated.
OBJECTIVE
The goal of this study is to enhance understanding of the relationship between different hump
design, speeds and vibration (shock) so that measures can be taken to address problems related safety
and health hazard. The specific objectives of this study are:
a. To identify and illustrates the technical details of various type of speed hump designs in Shah
Alam residential area.
b. To determine the most effective hump geometric design at reducing vehicular speed from
different hump profile at study area.
c. To analyze, correlate and evaluate the relationship between hump designs, speeds and vibration
(shock).
d. To formulate mathematical model which able to indicate and explain hump design efficiency in
terms of effectiveness (speed reduction), comfort (level of shock) and maximize overall road
safety.
SCOPE OF WORK
This study involves investigation on the existing speed hump designs in residential area. Spot
speed studies using radar gun will be use to record speed as motorist travel over existing speed hump. The
85th percentile hump-crossing speed for automobiles will be determined for each hump design. Horizontal
and vertical accelerations will be measure using the accelerometer place under the drivers seat and travel
at the observed 85th percentile speed for each of the hump design. The approach allowed relative
comparisons to be made between various speed hump designs, speed and vibration. The accelerations
measure will be entered into multiple linear regression models to estimate the test data and predict
additional accelerations. The existing speed humps is hoped to produce acceleration levels to show the
discomfort criteria, where the improvised design of speed hump can made.
170
LITERATURE REVIEW
Definition of Traffic Calming
Speed humps, along with raised intersections and lateral shifts in the roadway such as chicanes,
narrowing and mini-traffic circles belong to a family of design techniques known as traffic calming
measures. Traffic calming measures can quite literally calm the behavior of motorists, eliminate
opportunities to speed, and increase overall road safety.
According to the ITE March, 1997 International Conference held in Tampa, Florida, traffic
calming is defined as the combination of mainly physical measures that reduce the negative effects of
motor vehicle use, alter driver behavior and improve conditions for non-motorized street users. Traffic
calming measures can be divided into three categories: vertical measures, horizontal measures, and road
narrowing. Vertical measures use forces of vertical acceleration to discourage speed. Types of vertical
measures include speed humps, speed tables, raised intersections, and textured pavements. Speed humps
are a very efficient way of slowing down cars, and, especially in residential areas, may help reduce
accidents (ITE, 1993).
History of Traffic Calming
The desire to slow down vehicles speed on certain streets is not new; the idea of calming motor
vehicle traffic is nearly as old as the existing of the automobile itself. In 1928 an English lord, Cecil of
Chelwood, published a Road Vehicle Regulation Bill in which he suggested the use of .speed hump
or similar devices in the road surface in order to force drivers to slow down (Hass-Klau, 1990, as sited in
Weber, 1998).
Modern traffic calming got its start with the Buchanan Report Traffic in Towns, published in
1963 in the United Kingdom. Colin Buchanan expanded on the ideas of Lord Cecil and another
Englishman named Alker Tripp when he warned of the consequences of increasing automobile use. He
recommended measures for managing traffic and developed the concept of an environmental capacity for
streets (Hass-Klau, 1990).
The first example of a traffic calming project was created in the late 1960s. The Dutch
community of Delft, in response to speeding traffic on one of its residential streets, installed a series of
obstacles to physically alter its character and slow traffic. The street was called a woonerf, which
translates into residential yard. The scheme was a success, and several other communities quickly
followed with woonerven of their own (Hass-Klau et al, 1992).
Soon afterwards woonerven were constructed in Germany, and by the 1970s and 1980s traffic
calming was being applied in larger-scale verkehrsberuhigung which translates into traffic
tranquilization projects. The practice has expanded into shopping areas, school zones, village centers and
even main roads, and speed hump have come into increasing use (Pharaoh & Russell, 1991, as sited in
Weber, 1998).
Speed Humps on Residential Street
Of all the traffic calming measures that have been developed, the best at reducing speeds has been
the speed hump. However many concerns have been voiced about speed humps, particularly their
effectiveness and their potential to create unwanted noise and vibration. Figure 1 shows typical speed
hump dimensions. In this figure the speed hump has a parabolic shape. The speed hump, however, can
take several additional shapes including a sinusoidal, circle, or flat-top shape as shown in Figure 2.
171
Figure 1: Typical speed hump dimensions (Ewing, 1999)
Figure 2: Design profile of speed hump (Ewing, 1999)
Theory of speed hump
A road hump works by transferring an upward force to a vehicle, as it transverses the hump. The
force translates into a vertical acceleration and displacement and induces a front to back pitching motion
in vehicles having a wheelbase similar to the length of the hump (Jarvis, 1992). At low speed the
acceleration is of small amplitude. As speed increase the amplitude and pitching also increase and
therefore increase the displacement. If the acceleration is great enough it is interpreted as an
uncomfortable sensation. Acceleration acting on seated individuals is primarily interpreted through the
lower part of the hipbone. The degree of discomfort perceived depends on the frequency of the
acceleration and increase with it duration and amplitude (Griffin 1990).
RESEARCH METHODOLOGY
This empirical research involves case study within the residential street in Shah Alam to evaluate
the effectiveness and shock (vibration) experience or incur while traversing various design of road humps.
The location and the road hump design will be determined for the first phase of the experiment. Spot
speed studies using radar gun will be use to record speed as motorist travel over existing speed hump. The
85th percentile hump-crossing speed for automobiles will be determined for each hump design.
172
The methodologies for collecting speed data at three identified spots are before (30 meter
approaching hump), at and after the hump (0-5 meter) for all the speed hump locations using video
recording techniques and spot speed study using radar speed gun as to compare the data. The data
collection will be recorded for an hour for each selected site. Three sites for every speed hump design
which are Sinusoidal, Circular, Parabolic and Flat-Topped (total sites = 12 nos.). Data collection will be
taken during day time, weekday and off-peak period. Road inventory at study location will be taken
besides, the hump geometric data parameters such as length, width, height of the speed hump at each site.
In the second phase of the experiment, The vibration levels/shocks will be measured by the
accelerometer place under the drivers seat and travel at the observed 85th percentile speed for each of the
hump design. The accelerometer are attached to the VBOX II SX that could be installed in cars, off-road
vehicles, trucks, boats, bikes or earth moving vehicles. The VBOX II SX comes with a 20Hz or 10Hz
update rate, featuring high contrast screen display, USB interface and SD car logging. It is also equipped
with no logging facilities allowing the user to record data using an external data acquisition system
(external memory card). This means that the vibration from the accelerometer is then transmitted into the
data logger for processing and further analysis.
According to Md Diah J. et. al.(2008), when all the data have been collected, the data reduction
process will be carried-out using Semi Automatic Vehicle Analyser (SAVA) software. Further analysis
using statistical software called MINITAB is used for detail data statistical analysis and verification as
well as model formulation. The flow chart of the experimental procedure is shown in figure 3.
Figure 3: Scope of Work and Methodology to be adopted for this study
173
SIGNIFICANCE OF STUDY
This study will enhance the understanding on different types of speed hump and how it can reduce the
speed of vehicle as well as understanding level of discomfort (vibration) for road user while crossing
different speed hump design. This understanding can be use in developing geometric design standards
that consider an ergonomic approach. Furthermore deterministic models to estimate the accelerations
measured that relate and explain ergonomic parameters with safety elements may be crucial to be
considered in speed hump design.
REFERENCES
Ewing, Reid, and Charles Kooshian, (1999) U.S Experience with Traffic Calming. ITE Journal, August
1999, pp.28-33.
Griffin, M.J., (1990). Handbook of Human Vibration. Academic Press, London.
Hass-Klau, Carmen, (1990) The Theory and Practice of Traffic Calming: Can Britain Learn from the
German Experience? Rees Jeffreys Road Fund, Transport and Society Discussion Paper 10, United
Kingdom, January 1990.
Hass-Klau, Carmen, Inge Nold, Geert Bocker and Graham Crampton, (1992) Civilised Streets: A Guide
to Traffic Calming. Environment and Transport Planning, Brighton, United Kingdom.
ITE Technical Council Task Force on Speed Humps, Guidelines for the Design and Application of Speed
Humps, Institute of Transportation Engineers (ITE), Washington, DC, USA, 1993.
Jarvis, J.R., (1992). An Investigation of Road Humps for Use on Bus Routes. Research Report ARR
No.222, Australian Road Research Board (ARRB) Ltd., Victoria, Australia.
Md Diah J, Adul Rahman M.Y. and Adnan M.A. (2008). Traffic data reduction technique using portable
vision based traffic analyser (PVBTA) at Malaysian conventional roundabout. EASTS Int. Symp. on
Sustainable Transportation Incorporating Malaysian Universities Transport Research Forum Conf.
2008 (MUTRFC08),Universiti Teknologi Malaysia, Johor, Malaysia.
Pharaoh, Tim M., and John R. E. Russell, (1991) Traffic Calming Policy and Performance. Town
Planning Review 62, No. 1, pp.79-105.
Weber P, Braaksma J (1998) Towards a North American geometric design standard for speed hump. ITE.
174
CASE STUDY OF EV CONTROLLER AND POWER MANAGEMENT SYSTEM
Vincent Cheah Beng Keat 1*, Gan Yu Han 2, Leong Chee Ken 3, Phuan Zheng Yi 4, Gan Guo Dong 5,
ABSTRACT:
The application for the designed Pulse Width Modulation Motor Control circuit including Battery
Management System is in the field of battery powered cars. This new move towards renewable energy
and environmental friendliness is definitely a goal worth aiming for as the Earths natural resources
continue to diminish. Battery powered vehicles are indeed a very important component in the future of
electrical and electronics engineering in this nation. This is evident considering Protons latest offering,
the Proton Saga EV Green Propulsion, and is Malaysias first very electric vehicle. It is hoped that
through the execution of this assignment, we as engineering students may someday contribute towards the
technological development in the field of electronics of our very own country, Malaysia.
In an Electrical Vehicle (EV) design, the power management system and controller plays an
important role as a brain of the EV, to ensure the car can run smoothly and in order. The main functions
of a controller is basically to control the main contactor and kill switch contactor, incorporating
appropriate safety measures and interlocks, control the reversing contactor, incorporating appropriate
safety measures and interlocks, power the vacuum pump in the EV braking System, power the toaster
heater contactor and power DC-to-DC-contactor, where as the power management system ensured that
the power distribution are in order and at the meantime achieving maximum transfer efficiency. The
power management system includes converter circuit, charging circuit, motor drive and base drive circuit.
In this report, the works are mainly focus on power management system and base drive circuit inclusive
of their controller and converter circuit. The functionality of each of the controller will be investigated
and discussed.
Keywords: BMS, BCU, BMU, SOC, CAN, cell protection, isolation of gate and base drive circuit.
7 Assistant Professor, Universiti Tunku Abdul Rahman, Setapak, 53100 Kuala Lumpur, MALAYSIA
*Correspondence Author: Vincent Cheah Beng Keat, Universiti Tunku Abdul Rahman, Malaysia. Tel:
+60164775558. Email:vincentcheah1@gmail.com
175
INTRODUCTION
Recently, Malaysia is owned their first homemade pure electric car, officially known as Proton Saga
EV Green Propulsion concept. The whole concept behind this is to produce a fully functioning electrical
vehicle with zero emission incorporates with key technology such as battery management system. In
order to achieve to this, proton place more attention on the heart of the vehicle (Lithium Balance). This
Lithium Balance battery is placed together with plug in and play concept, with built in management
electronic.
Figure 1: Proton Saga EV Green Propulsion
The whole idea behind this concept is basically to develop a system that does battery monitoring, in
which it keeps track on the operational parameters charging and discharging. This may also include
voltage, currents, battery internal and ambient temperature (temperature surrounding the vehicle). The
controller in this case, must be smart enough to provide proper protection to the system by providing an
appropriate indication to sound an alarm or electronically disconnect the battery from the load if any of
the parameters exceeds its limit.
There are mainly three objectives of having a BMS (Battery Management System) inside an EV
which are protecting Lithium Battery from any damage typically by protecting the cells, prolong the
lifespan of that battery and to maintain the battery states in which allowed the system to run specific
functions and applications specifically attached to it.
The primary application of this BMS is to provide necessary monitoring and control, avoiding the
cells from damage due to over rating temperature condition. This is essential as the vehicle may be
working under harsh condition or even differences in climate (Tropical Climates, Subtropical Climates,
Mediterranean Climates, Temperature climates, Arctic temperature and Desert Climates) within a
particular continent which affect the overall ambient temperature. Hence, individual cell in the
automotive must be protected by isolating the battery detect the cause of the fault when an external fault
takes place. To illustrate this, disconnect the battery source when the heat generated within a system is
severe. However, for non-severe cases, we may turn on an additional fan attached in front of the battery.
176
Figure 2: Charging profile of Lithium Battery
Secondary major development of BMS refers to state of charge (SOC) determination. This basically
providing the user some indication the amount left over before performing any recharging. This SOC
estimation is also known as Gas Gauge or Fuel Gauge functions. In this design, SOC basically
calculates each of the individual cells within the battery besides ensuring them from overstressed
condition. Apart from that, with SOC indicator, over-charging and over-discharging condition can be
prevented since each of the charging and discharging cycle can easily be monitored.
Apart from that, Electric cars are driven by large electric motors usually rated between 3.5 and 28
horsepower. Compared to gasoline engines, this may not seem like much power, but the rating systems
used for gasoline engines and electric motors are different because gasoline engines are rated at their peak
hp and electric motors are rated at their continuous hp. The peak hp of an electric motor is usually 8 to 10
times its continuous rating.
Electric vehicle drive motors can be divided into two basic groups, DC or direct current motors, and
AC or alternating current motors. DC motors have a long history in EV use. The most commonly used
version is what is known as a series-wound motor, which means the armature and field windings are
wired in series. Other designs include shunt-wound, compound-wound, and permanent magnet motors. At
the present time AC motors are most commonly found in commercially built EVs, as they require more
sophisticated and complex control systems than DC motors.
177
LITERATURE REVIEW
BATTERY MANAGEMENT SYSTEM
Figure 3: Conceptual representation of BMS primary function
According to the figure 3 above, there are basically three main building blocks needed to be considered
throughout this design. They are Battery Monitoring Unit (BMU), Battery Control Unit (BCU) and
Controller Area Network (CAN) bus vehicle communication network including its internal configuration
to the vehicles management system. Note also that, other parts of the vehicle systems may not only be
connected to one another within a system, but more importantly, they are able to communicate with BMS
via CAN bus such as Thermal Management System, including anti theft devices which has the capability
to disable the battery. The ECU (Engine Control Unit) module may not be needed, in the case where the
parameter concerning Internal Combustion Engine (ICE) is not being examined.
BATTERY MONITORING UNIT
The Battery Monitoring Unit is microcontroller based which is mainly used to monitor health of
the battery bank. In other words, BMU calculates battery capacities apart from battery efficiency from
time to time and simultaneously monitoring each of the cells upon cell deterioration prior to failure as
illustrated by three sub-module of BMU. This sub-module is separated for the purpose of clarification.
Battery failure can easily be determined by monitoring the cell voltage during discharging. Note that, in
this case, the charging voltage drop should not be long enough. This type of failure, however, can be
classified as battery abuse in which the main causes are high temperature and excessive charge current.
Other typical factors that may give rise to battery failure are ageing and premature-failure where they are
caused by corrosion and manufacturing defect respectively.
178
Figure 4: Multiplexer to reduce component count
Apart from monitoring all the cells in parallel, a multiplex architecture can easily be interfaced to the
BMU in order to reduce cost. Based on this design, only single analog or digital output can be monitored
from time to time. One of the drawbacks of this system is that only one voltage cell can be determined at
a time. Hence, a very high speed mechanism is required such that each can be monitored sequentially.
BATTERY CONTROL UNIT
Power electronics circuitry is the main component that lies within Battery Control Unit (BCU). From
figure 3, it is observed that BMU obtains its instruction in terms of control signal from the earlier block
that mentioned earlier, BMU to carry out a specific task, such as controlling battery charging profile. In
this case, controlling the voltage and current charging profile, providing a top up charge to each of the
individual cell serving as a purpose to equalize all the charge within the battery, fault isolation,
regenerative braking switching to charge battery whenever it is required, dump out excessive braking
charges generated during breaking and last but not least able to respond corresponding to the vehicles
operation mode. In order to have these functions working properly, each of the battery cells must be equip
with highly expensive current switch, having the capability to do switching around 200 AMPS or more in
order to perform necessary interconnect.
CONTROLLED AREA NETWORK
Controller Area Network (CAN) was developed back in 1985, by Bosch to be used in-vehicle network.
Back in the past, automotive manufacturers start of by using point to point wiring. As time goes by, it was
realized that having a lot of wires not only result in bulky, heavy but also expensive. Today, CAN is used
instead of traditional wiring as CAN not only cheap, less complexity, reduce in weight but more
importantly it has high-integrity serial data communication for real-time control application integrated
into standard in-vehicle network.
179
BATTERY PERFORMANCE MANAGEMENT SYSTEM
In order to achieve a smart monitoring system having the ability to manage individual cells and isolating
weaker one are strongly needed in todays EV. The main components of a BPMS include: -
Precision fuel gauge

Battery charge balance or battery capacity balance for out-of-step batteries and if possible individual
cells
A reference to a standardized data set as the voltage and temperature cut-off control parameter
(particularly for a rapid battery charger)
A data logger for evaluation and processing of battery performance data
A supervisory data acquisition and control system for battery pack thermal management
Apart from that, BPMS block also comes together also has cell balancing concept where it ensures that
each of the cells are evenly distributed as discussed. Furthermore, the declines of weak cells are
monitored from time to time until it has reached to certain point where it cannot be used. This however
reduces the probability of damage due to overcharging process. Apart from that excessive discharges rate
mainly due to during regenerative braking where it is characterized as shorted as short but powerful bursts
of charging current at excessive voltages. In this condition it may exceeds the battery charge capability.
Next, SOC determination and Thermal Management System are also embedded into this block as well.
There are two main controllers used in this design. One of those controller works. One is to active
mode; meanwhile the others work under hot standby mode. Apart from that, each of these batteries is
interface with battery monitor also known as BMONs. Each BMON can monitor up to five 12 V batteries
in the battery pack and up to four auxiliary temperature and vent pressure sensors. This connection is
made based on Daisy Chain methodology, where data communication is made through high speed data
bus (HSDB). Note that, in a typical system, at least 7 to 8 BMONs are used to monitor the battery pack
every 12V increment along with 2 MC.
The MC incorporates dual-processor architecture to perform all its functions: communications,

processing and control, data storage, and data retrieval. Historical data concerning the battery are stored in
the RAM (non-volatile memory). Nowadays, 30 charge and discharge cycle most recent cycles, are
grouped together, meanwhile the other data, older data is compressed every two weeks. Well, in this case,
BMPS can control the amount of fast-mode DC charging including rate of charging from high-voltage
within AC power grid. Hall Effect sensors are used to measure current within each cell since it can
measure both charging and discharging profile precisely. Current limiter circuits are possible to be done
along together with this module in order to prevent excessive current generated from regenerative
breaking as discussed. Once this is detected, warnings, alarms, and diagnostic messages will notify the
user.
BMPS THERMAL MANAGEMENT SYSTEM
Thermal Management System is designed to monitor variation in temperature due to battery

chemistry and performing necessary task such as heating or cooling the battery cells depending on its
state. Tests conducted in the laboratory and with EV urban driving suggest that using a thermal
management system improves the mileage and battery life by at least 20%. Thermal management system
plays a very crucial role when performing rapid charging. Imagine, during the charging process, large
amount of charge is delivered to the EV, hence temperature issued is inevitable.
180
The primary design is to keep the battery insulated. The insulation will help to enhance heat
generation especially during winter or summer. On the other hands, as for second criterion design, air
flow circulation has to be properly distributed to ensure minimum temperature, achieving equilibrium
between battery and the surrounding. Note different design criterion should takes place under various
operating conditions.
ISOLATION OF GATE AND BASE DRIVE
The whole idea behind this design is typically to operate a power transistor as switches, an
appropriate gate voltage or base voltage must be applied to the transistor so that we are able to drive the
circuit into the circuit for low on-state voltage. Figure below shows the topology of single phase bridge-
inverter (bidirectional motor driver).
Figure 5: Single phase bridge inverter and grating signals
The logic generators basically are designed such that it produces four pulses. As shown, these pulses
are shifted in time so they are able to drive the motor circuit (in this case, ) in two directions. Note also
that, at any instance of time, we only operate, only two diagonal transistors, leaving the other two to be in
the ground state. With this, there is no doubt that the motor will turn only at one direction, at any instance
of time. Braking can easily be achieved, by setting the entire gate to either high or setting them to low.
Bear in mind that, the position of the reference point keeps changing from time to time depending on the
logics applied to the transistors. Note also that, according to some of the sources, there is need for
isolation and interfacing circuits, such as pulse transformers or optocouplers, between the logics and
power transistors. However, transistors, and can be gated directly without isolation or interfacing if the
logic signals are compatible with the gate drive requirements of the transistors. The figure below simply
illustrates the overall circuit connection within a car.
181
Figure 6: A simple DC controller connected to batteries and DC motor
The whole concept concerning the implementation of base drive circuit on the electric vehicle is as
illustrated in figure 8 above. As seen, potentiometer (variable resistor) is used to determine the amount of
power to be delivered to the motor. For instance, zero power when the vehicle is in a stop position. PWM
(Pulse Width Modulation) generated does not only have the ability to control direction of the vehicle but
more importantly, it able to produce variation in speed. An experiment has been carried out to verify this
statement. A very simple design of DC controller for base drive circuit can easily be achieved by
integrating 3 basic fundamental blocks which are PWM generator, logic generator and H-bridge circuit.
The use of astable mode of timer IC555 to generate PWM is indeed the easiest yet most cost effective
method. Take, note that the PWM generated needs to be converted to logics in order to operate the
transistor, this where op-amp comes into picture.
Figure 7a : Schematic of basic base drive and motor drive design
182
Figure 7b: basic base drive and motor drive design to model EV
As discussed, reducing the supply say, from 12V to 6V, will eventually lessen the motor to run at half
of its speed. Of course, practically, this approach is not applied directly. Instead of decreasing the voltage
level of supply, the motor speed in general is controlled by turning on and off the supply quickly with the
use of PWM since it is very inefficient for us to do so. An analogous concept lies behind switching also
lies behind florescent light. They are turn on and off with a very rapidly to ensure that the filament does
not overheated. Furthermore, they are also able to generate average amount of glow compared to direct
supply attached to it.
Figure 8: Variation in supply corresponding to the speed of motor
Generally, the time taken for a motor to speed up or even slow down under switching conditions
totally depends on the inertia of the rotor (normally how heavy is it) including the amount of friction and
load torque generated. Apparently, there are still a lot of improvements have to be done in order to
produce a more efficient, effective yet reliable system.
183
DC-DC CONVERTER
Full bridge (H-bridge) transfer circuit is naturally the main converter circuit in any electrical
vehicle design. Typical input voltage and input current in the converter lies within the range of 100-
150VDC and 0-50 A respectively. Output voltage is 0-48VDC and output current is 0-100A. Basic
construction of a DC-DC converter is as shown as below. As seen from the figure below, there are four
basic structure of the converter namely main power transformer circuit module (including transformer),
control module, derived and protection module, and isolation module. Well, as observed the main circuit
in this design is typically none other than full bridge circuit. Having the use of double close loop control
of voltage and control taken into the design, the controller is able to overcome over-current protection and
the temperature protection. Furthermore, DC-DC converter also able to completes the voltage
transformation from unstable DC voltage to stable.
Figure 9: Principle scheme of DC-DC converter
METHODOLOGY
LOW COST BMS FOR MULTICELL
With the use of commercial battery monitoring device integrated with Maxims DS2571, a device
which uses optical isolation methodology, a low cost yet high performance BMS can easily be achieved.
Note that in this design, DC-to-DC converter plays a very important role to provide power to the opposite
isolated side. In other words, since only single battery source is used, extra cost, complexity and size need
not take into account.
184
Figure 10: Schematic diagram for low cost BMS
Based on the circuit diagram above, DS2751 is used to monitor current flow locally. The DO (Data
Out) and DI (data in), on the most left of the schematic diagram are yet to be connected to a
microcontroller. This circuit is basically a simplified version of BMS daisy chain, ring topology. Hence,
two connectors are used in each side.
When DO is low, IC2 will drive the data in to be low as well. As a result, IC1 will be turn on leaving
data-line connected to the DS2751 to be low. It is observed that when once the IC1 is activated, DI will be
turned as high in the microcontroller itself. The digital output of the DS2751 can either be used to as a
base drive circuit to enable the cell balancing circuit.
Figure 11: Connected to the microcontroller based on BMS daisy chain (ring topology)
The limitation of the output port of the microcontroller can easily be solved with the use of daisy chain
methodology where by the cell monitoring boards are connected to one another. In order to increase the
drive capability, FETs or other driver circuits may come into the picture. Last but not least, CAN bus is
used in this case, so that communication to a test PC is possible. Better illustration of a BMS Daisy Chain
is as shown in the figure 12(a) below.
185
(a) BMS Daisy Chain (ring topology)
(b) BMS Master and Slaves (star topology)

Figure 12: Practical BMS implementation
In figure 12, it obvious that each of the cells in BMS Master and Slaves design are attached to at least one
slave before it is connected to the master meanwhile in the BMS Daisy Chain, each of the cells and slave
are connected together to form a ring before they are employed to the master. The terms slaves simply
refers to the temperature and voltage measurement sub-module within each of these cells to monitor and
yet implementing cell balancing. This slave module does not carry out any signal processing; all these
processes are performed within the master sub-module. SOC determination is also included into this
module.
Unlike, BMS Master and Slaves, the BMS Daisy chain is not only easy to construct, and design
but more importantly they have a higher reliability when an automobile is concerned.
Cell balancing is indeed another essential parameter throughout this development. When multi
cells are concerned, there is a probability that the rate where one of the cells breakdown is much higher
186
rather than having a single cell. Likewise, having only a single cell, an EV can is impossible to obtain
100V to 200V or even practically more in order to power up the EV. Hence, series type of battery
connection is more practically seen in today. This connection eventually will leads to premature failure to
the cells as there are interaction among themselves causes by overlapping operating characteristic of an
individual cells such as production of tolerance, temperature distribution and ageing profile which may
leads to the subsequent battery. These problems can be reduced by considering parallel chain.
Another method to avoid Long term unbalance rather than short term is to consider the term
equalization throughout the implementation. First, ensure that batteries used are from the same batch
yet same manufacturer. Next, thermal management system should be brought into the whole picture, to
monitor the ambient distribution temperature surrounding the battery ensuring that it does not have any
significant change. Introducing BMS system to monitor the SOC of each of the chain cell as discussed
above. Once it detected the odds, switching circuit is then activated to control charges across each of
those cells during its charging state. This charging may also comes from regenerative which may occurs
more once in a day. Hence, the system must be able to withstand that as well.
MAIN AND CONTROL CIRCUIT FOR EV
Generally, the main circuit uses H-bridge topology comprising four IGBT transistors. Note that
output of the circuit will eventually undergo bridge rectifier. This is indeed a DC-DC converter by itself.
During operation, the average current that flows through each diode is half compared to the average
current flows through the inductance, L. The structure of the circuit is as shown as below.
Figure 13: main circuit for DC-DC converter
The controller circuit is certainly the most essential piece in the design. They are made of core
converter in which they are able to decide its dynamic stability of the converter. This can be achieved
with the use TL494. Note also that PWM can be generated within TL494. Apart from TL494, it also made
up of simple filter circuit and RC electric circuit discharge loop. The controller circuit is as shown below.
187
Figure 14: Controller circuit for DC-DC converter
RESULT ANALYSIS
The controller circuit DC-DC converter circuit side by side with main DC-DC converter in DC-DC
converter of 5KW fuel cell sightseeing car. The results of the experiments are shown as below under most
common condition.
Figure 15: IGBT driver waves
Figure 16: Output wave of the transformer
188
CONCLUSION
Battery management system comprises of two main block namely Battery Monitoring Unit and
Battery Control Unit via CAN bus. As the term BMU implies, it basically used to monitor the state of the
battery. However, the main module perform the necessary function would not be other than BCU. Today,
these features are enhanced another new module better known as BPMS. This is done in order to achieve
smart monitoring system, which takes into account cell balancing, precision cells gauge, managing
thermal management system and many more. However, in order to ensure that EV having the capability to
withstand any climate or even prevention the occurrence damage cells during rapid charging, the BMPS
thermal management is introduced. As discussed, there are typically two main types of BMS
configuration namely BMS Daisy Chain or BMS Master and Slaves. Of course, they have pros and cons
among one another.
REFERENCES
Battery Management Systems (BMS) [Battery Management System for EV]. (n.d.). Retrieved July 8,
2010,from Woodbank Communications.Batteries and Energy Storage Magazine (BESTMAG) website:
http://www.mpoweruk.com/bms.htm
Dallas Semiconductor. (n.d.). DS2751 Multichemistry Battery Fuel Gauge [Data Sheet Specification on
DS2751]. Retrieved July 20, 2010, from http://www.alldatasheet.net/datasheet-pdf/pdf/114374/
MAXIM/DS2751.html
Guokai Xu, Xiuchun Zhao, Hang Su, & Zhenqiang Yang. (2008, June 1). [Review of the controller main
and control circuit]. A Pulse-width Modulation Based DC-DC Converter for Electric Propulsion, 6(1), 1-
3.
LITHIUM BALANCE A/ S. (n.d.). Case Study High speed Electric Vehicle [Lithium Battery
Management System]. Retrieved July 17, 2010, from
http://www.lithiumbalance.com/attachments/article/88/CaseStudyHighspeedElectricVehicle.pdf
M J Schofield. (n.d.). The Controller Area Network (CAN) [key feature of CAN]. Retrieved July 15, 2010,
from http://www.mjschofield.com/
Microchip Technology Inc., T. C., & Microchip Technology Inc., S. D. (n.d.). "Developing Affordable
Mixed-Signal Power Systems for Battery Charger Applications" [Lithium Ion Charging Profile].
Retrieved July 21, 2010, from
http://www.microchip.com/stellent/groups/designcenter_sg/documents/market_communication/en027883.
pdf
Mr PenyuBiru . (2009, July 20). Proton Saga Ev concept [Online forum message]. Retrieved from
http://carca-marba.blogspot.com/2009/07/proton-saga-ev-concept.html
Sandeep Dhameja. (2002). ELECTRIC VEHICLE BATTERY PERFORMANCE. In ELECTRIC

VEHICLE BATTERY SYSTEMS (p. 133) [Afterword]. (Original work published 2002) Retrieved from
http://www.pdfoo.com/ ebook-16617/electric-vehicle-battery-systems.html
189
Zanthic Technologies Inc., S. D. L. (2008, June 25). A low cost Lithium Ion battery management system
for a multi-cell, high voltage battery pack [Low Cost Lithium Battery Management]. Retrieved July 20,
2010, from http://www.zanthic.com/images/ A%20low%20cost%20Lithium%20Ion%20batter
%20management%20system.pdf
190
A STUDY ON ELECTRIC CAR CHASSIS AND DESIGN PRINCIPLE
Gan Yu Han 1*, Gan Guo Dong 2, Leong Chee Ken 3, Phuan Zheng Yi 4, Vincent Cheah Beng Keat 5,
ABSTRACT:
In this project, the placement of components in an electric car (EV) and the system chassis design will
be studied and explained. The scope covered is divided into 5 main portions:
Part 1: Placement of important components such as Lithium-ion polymer batteries, DC brushless motor,
controller and driver circuits strategically in the chassis of an electric car.
Part 2: Study to maintain balance, aerodynamic and spacing optimization in the design of an electric car.
Part 3: Important criteria to be considered in the positioning of batteries, motor and controller in the
chassis such as the size, weight, type of components used, specification requirements, etc.
Part 4: Balancing the front and back weight ratio of the car to ensure stability and safety of the driver and
passengers.
Part 5: Design of good air ventilation and heat dissipation to prolong the lifetime of expensive equipments
such as batteries and provide safety to the users.
Commercially available electric vehicles and prototypes still do not have a defined standard of component
placements unlike engine vehicles. Therefore, these studies will provide a fundamental platform for
further studies on commercially available electric cars in the market.
Keywords: balance, stability, safety, heat dissipation, spacing optimization
3
4
5
6
7
Assistant Professor, Universiti Tunku Abdul Rahman, Setapak, 53100 Kuala Lumpur, MALAYSIA
*Correspondence Author: Gan Yu Han, Universiti Tunku Abdul Rahman, Malaysia. Tel: +60163453054.
Email:ganyuhan@live.com.my
191
INTRODUCTION
The chassis and system design is the foundation of an electric car conversion. There are
fundamental principles to convert an engine car into an electric car and necessary to take into
consideration important factors such as weight, aerodynamic drag, rolling resistance and drivetrains. It is
very important to optimize, design and select the correct chassis for an EV conversion.
OPTIMIZE Getting familiar with the chassis trade-offs involved.

DESIGN Making sure the components selected accomplish the objectives.
SELECT Buying a suitable chassis according to required specifications.
Having grasped the knowledge involved in all these steps will help in the chassis and system
design of any EV conversion since the principles are universal and can be applied in buying, building or
converting.
Generally, a chassis design should cover all these important factors:

strong;
light;
rigid;
not vibrate, particularly at frequencies and harmonics of rotating parts;
aerodynamic;
resistant to impact;
crumple evenly in an accident, minimising forces on driver/passengers;
strong enough to fix components to easily;
impact and roll resistant;
cheap;
aesthetically pleasing;
corrosion-proof.
Figure 1:
Conceptual
Illustration of a
general EV
Configuration
192
LITERATURE REVIEW
Since the chassis is the foundation for an electric car, it needs to be chosen wisely. It is much
easier to make an EV conversion compared to building the chassis from scratch.
Therefore, there are a lot of factors to be considered in a chassis selection and it is utmost
important to be clear of the exact specifications and requirements needed to be accomplished before
making the ultimate decision of buying.
Examples of important factors needed to be taken into consideration are such as:
- Chassis with an aerodynamic shape (low coefficient of drag) and thin wheels (low resistance to ground
surface from tyres, brakes and steering)
- Big, strong and lightweight chassis for stable support
-Adhere to federal and state laws and certain safety standards
- Minimum drivetrain losses for high speed, long range, higher power transfer efficiency
- Weight management and balance ratio
- Motor and battery placement
- Handling, gearing, and safety
The most important components such as battery, motor and controller must be placed strategically in
the chassis to maintain its balance, aerodynamic, heat dissipation, spacing optimization, etc.
Most of the real electric car designs available in the real world and all of them have their own
specific placement of the components. In another words, there is no defined way to place the motor,
batteries and controller for different electric cars because they are unique, and not standard like engine
vehicles where their engines are situated at the bonnet and petrol tank situated beside the boot.
Therefore, the placement of motor, batteries and controller must take into consideration the size,
weight, type of components being used, specification requirements, air and heat ventilation etc.
A suitable motor that can be used in an electric car is a DC brushless motor. The motor should be
coupled to the back wheels and therefore should be placed at the back of the car to reduce conversion
power losses and maintain convenience for maintenance.
The batteries used are Lithium-ion and can be placed either in front or at the back of the car
depending on its size and weight so that it is easily accessible to ease maintenance and changing after a
few years time.
The controller which consists of converters, inverters, motor drive and driver circuit can be
placed at the bonnet of the car where the engine is supposed to be. By doing this, it will be much easier to
troubleshoot for problems in the controller such as short-circuit etc.
Other accessories such as air-conditioner, sensors and gauge, lighting, car entertainment, center
lock, wiper, alarm system can be placed in the bonnet like in an engine car.
The front and back weight ratio of the car must be balanced to ensure stability and safety of the
passengers. The design and placement of components must also take into consideration of good air
ventilation and heat dissipation to prolong the lifetime of expensive equipments such as batteries and
provide safety to the users.
193
METHODOLOGY
1. Optimizing the EV
The following factors are the key to optimizing an EV:

Weight and climbing and acceleration
Aerodynamic drag and wind
Rolling and cornering resistance
Drivetrain system
Weight, Climbing and Acceleration
In an EV conversion, the single most important thing needed to be considered is the weight.
Doing a conversion means replacing key components from a car such as a petrol engine and petrol tank
with motor, controller and batteries.
If the weight management is not planned properly, the converted car will inherit the end result of
the original cars weight trade-offs. Therefore, it is very important to consistently monitor the weight
management at three different times:
-Before purchasing the conversion vehicle
-During the conversion
-After the conversion
Weight affects nearly every aspect of an EVs performance: acceleration, climbing, speed, and range.
Weight and Acceleration

F = ma where F = force, m = mass, a = acceleration
Weight and Climbing

F = W sin where F = hill climbing force, W = vehicle weight, = angle of incline
Weight and Speed

hp = FV/550 where hp is motor horsepower, F is force in pounds, and V is speed in ft/sec.
Weight and Range

D = Vt where D = distance, v = speed, t = time
There are two other important weight-related factors to keep in mind when doing EV conversions:
- front-to-rear weight distribution
- the 30 percent rule
194
Figure 2: Front-to-rear Weight Distribution Trade-off
The vehicles front-to-rear weight distribution must always be kept intact and must not exceed its
total chassis and front/rear axle weight loading specifications.
The 30 percent rule only applies to EV cars using lead acid batteries where the battery weight
should be at least 30 percent of the gross vehicle weight.
Aerodynamic drag and wind
The aerodynamics design of a vehicle is of utmost important when the battery supply source is
limited and barely enough to support the vehicle movement and acceleration in the case of an electric car.
Engine vehicle does not need to worry about the power of their source since they have plenty of internal
combustion engine horsepower at their disposal. Therefore, they do not focus much on the aerodynamics,
but more on style and sleek.
The optimum aerodynamic shape can be seen from our surroundings such as droplets of rain
where it is rounded in front and pointed at its rear. Bicycle racing helmets also adhere perfectly to this
principle. Airplanes, submarines and bullet trains on the other hand have long adhered to the raindrop
principle into their chassis design.
The aerodynamic drag force, F can be expressed as

F = (CAV)/391
where F is the aerodynamic drag force in lbs., C is the coefficient of drag of the vehicle, A is its frontal
area in square feet, and V is the vehicle speed in mph.
195
The equation defining the relative wind factor, Cw, is

Cw = (0.98(w/V) 1 0.63(w/V))Cr 0.40(w/V)
where w is the average wind speed of the area in mph, V is the vehicle speed, and Cr is a relative
wind coefficient that is approximately 1.4 for typical sedan shapes, 1.2 for more streamlined vehicles, and
1.6 for vehicles displaying more turbulence or sedans driven with their windows open.
Figure 3: Coefficient of Drag

Summary for Different Vehicle
Types
To summarize how we can improve the aerodynamic drag and wind:

- choose the lowest coefficient of drag
- choose smallest frontal area
- shape rear airflow
- block and shape other airflows cover underneath
Rolling and cornering resistance
Normal engine vehicle tires are fat, have wide tread and have higher rolling resistance because
they have been optimized for good adhesion instead. However, electric cars cannot follow these
characteristics and have to go against this current way of thinking to enhance the performance on an EV
car.
The rolling resistance force is defined as

F = C Wcos
where C is the rolling resistance factor, W is your vehicle weight in lbs., and is the angle of incline.
Tires are important to any vehicle especially EVs. They support the vehicle and battery weight
while cushioning against shocks; develop front-to-rear forces for acceleration and braking; and develop
side-to-side forces for cornering.
196
Other ways to reduce the rolling and cornering resistance are:

-use largest and thinnest radial tires
-use high tire inflation pressures
-reduce brake, suspension drag and steering scuff
Drivetrain system
The drivetrain of the vehicle, which is also known as the powertrain of the vehicle is the
transmission elements and the propulsion unit combined. The transmission is the mechanical linkage that
transmits power between the electric motor shaft and the wheels. The drivetrain of an electric vehicle
(EV) consists of the electric motor, gearbox, driveshaft (only in rearwheel drives), differential, half-shafts,
and wheels. The capability of electric motors to start from zero speed and operate efficiently over a wide
speed range makes it possible to not use the clutch that is used in internal combustion engine vehicles
(ICEV). A single gear ratio is sufficient to match the wheel speed with the motor speed. Therefore,
electric cars can be designed without a gear, but the use of a speed reducer allows the electric motor to
operate at much higher speeds for given vehicle speeds. This minimizes the motor size because of the low
torque requirement at higher speeds.
Since the power source for engine vehicle(petrol) and electric vehicle(batteries) are different, the
drivetrains for electric motors opposed to those used for internal combustion engines.
In practical point of view, the power available from the batteries must be equal to the job of
overcoming the tractive resistances for any given speed.
The obvious mission of the drivetrain is to apply the batterys power to driving the wheels and
tires with the least loss and highest efficiency. To sum it up, the drivetrain must perform a number of
tasks:
Convert torque and speed of the engine to vehicle motion-traction
Change directions, enabling forward and backward vehicle motion
Permit different rotational speeds of the drive wheels when cornering
Overcome hills and grades
Utilize optimum battery power
The function of each component related to the drivetrain is as follows:

Electric MotorProvides the raw power to propel the vehicle.
ClutchFor internal combustion engines, separates or interrupts the power flow from the engine so that
transmission gears can be shifted and, once engaged, the vehicle can be driven from standstill to top
speed.
Manual TransmissionProvides a number of alternative gear ratios to the engine so that vehicle
needsmaximum torque for hill-climbing or minimum speed to economical cruising at maximum
speedcan be accommodated.
DriveshaftConnects the drive wheels to the transmission in rear-wheel-drive vehicles; not needed in
front-wheel-drive vehicles.
DifferentialAccommodates the fact that outer wheels must cover a greater distance than inner wheels
when a vehicle is cornering, and translates drive force 90 degrees in rear-wheel-drive vehicles (might or
might not in front wheel- drive vehicles, depending on how engine is mounted). Most differentials also
provide a speed reduction with a corresponding increase in torque.
Drive AxlesTransfer power from the differential to the drive wheels.
197
Figure 4:
Simplified EV
Drivetrain
Layout
In summary, the criteria needed to be taken into consideration when designing the drivetrain are:
- choosing manual or automatic transmission
- choosing front wheel drive for higher efficiency
- minimize drivetrain lubricant viscosity
2. Designing the EV
To design an electric vehicle, it is important to match the motor-drivetrain combination to the

body style selected by going through the following steps:
Learn when to use horsepower, torque, or current units, and the reason of using them
Look at a calculation overview
Determine the required torque needed from loads for the selected vehicles chassis
Determine the available capabilities of selected electric motor and drivetrain torque.
It is also important to consider the function of the electric vehicle whether it is needed for speed,
range or utility and if it meets the needs of the user in the end.
3. Selecting the EV
EV was mainly converted from the existing ICEV by replacing the IC engine and fuel tank with
an electric motor drive and battery pack while retaining all the other components.
Converting an existing internal combustion engine vehicle saves money in terms of large capital
investment, a large amount of labor and time. By using a used or second hand engine vehicle together
with an EV conversion kit, the advantages obtained is a structure that comes completely with a stable
body, chassis, suspension, steering, and braking systems - all designed, developed, tested, and safety-
proven to work together.
If the converted electric vehicle does not exceed the original vehicles overall weight by much, all
systems should continue to deliver their previous performance, stability, and handling characteristics
198
without any problems. EV conversions also inherit another body benefit - its bumpers, lights, safety-
glazed windows, etc. are already approved and tested to meet all safety requirements by the federal laws.
Another advantage is saving more money because automobile junkyards make money by buying the
whole car and selling off its pieces for more than they paid for the car.
Building rather than converting an EV cost a lot more than buying the exact same new car
because manufacturers buy their components in bulk while we buy them separately. Unless an EV
conversion is done using a kit, building from scratch means buying individual components such as chassis
tubing, angle braces, and sheet stock plus axles/suspension, brakes, steering, bearings/wheels/tires,
body/trim/paint, windshield/glass/wipers, lights/electrical, gauges, instruments, dashboard/interior
trim/upholstery, etc.parts that are bound to cost a lot more compared to buying them already
manufactured and installed in a completed vehicle.
RESULT ANALYSIS
1. Wind Tunnel
A wind tunnel is a research tool used in aerodynamic research. It is used to study the effects of air
moving past solid objects.
Wind-tunnel testing was applied to automobiles, mainly to determine ways to reduce the power
required to move the vehicle on roadways at a given speed and also to determine aerodynamic forces per
se. In these studies, the interaction between the road and the vehicle plays a significant role, and this
interaction must be taken into consideration when interpreting the test results. In an actual situation the
roadway is moving relative to the vehicle but the air is stationary relative to the roadway, but in the wind
tunnel the air is moving relative to the roadway, while the roadway is stationary relative to the test
vehicle. Some automotive-test wind tunnels have incorporated moving belts under the test vehicle in an
effort to approximate the actual condition.
Traditionally, wind tunnel testing was a sizeable trial and error process, ongoing throughout the
development of a vehicle. Today, with the high level of CAD prediction and pre-production evaluation,
coupled with a greater human understanding of aerodynamics, wind tunnel testing often comes into the
design process later. The wind tunnel is the proving ground for the vehicle's form and allows engineers to
obtain considerable amounts of advanced information within a controlled environment.
Whilst advanced design processes can anticipate a large proportion of aerodynamic performance,
it is still crucial to assess a vehicle in the wind tunnel. Many elements of a vehicle's form only reveal their
behaviour in air flow when carefully tested and cannot be anticipated on computer. The reality of
production, tolerances in components and accuracy of build can all play a part in affecting the
aerodynamic behaviour of a car.
Aside from engineering concerns, manufacturers are increasingly looking to see how to improve
the customer-side of aerodynamics. For example, wind noise from door mirrors is considered very
undesirable and can only really be evaluated in a wind tunnel. Other, less obvious issues can also be
examined - such as whether air flow forces water through seals or dirt into door apertures.
199
Figure 5: Sophisticated sound equipment is used in the wind tunnel to compile data on wind noise.
Figure 6: A stream of smoke travels over the vehicle in the wind tunnel as air passes from right to left. It
can be clearly seen that laminar air flow remains attached until the very rear of the vehicle, emphasising
just how carefully aerodynamic performance has been considered.
200
CONCLUSION
In conclusion, understanding vehicle chassis and system design is undeniably one of the most
important aspect in carrying out an EV conversion because it is the shell and body that supports all the
most important components in an electric car.
Therefore, special attention needed to be channeled into this topic to ensure the construction of an
electric vehicle achieved its objectives and is a success.
REFERENCES
Leitman, S., & Brant, B. (2009). Build Your Own Electric Vehicle. New York: The McGraw-Hill
Companies, Inc, pp. 102
Larminie, J., & Lowry, J. (2003). Electric Vehicle Technology Explained. John Wiley & Sons., pp. 226
Husain, I. (2003). Electric and Hybrid Vehicles Design Fundamentals. Boca Raton London: CRC Press
LLC, pp. 229
C. C. Chan and K. T. Chau, Modern Electric Vehicle Technology, Oxford University Press, New York,
2001.
Ehsani, M., Gao, Y., Emadi, A. (2010). Modern Electric, Hybrid Electric and Fuel Cell Vehicles. Boca
Raton London: Taylor and Francis Group, LLC, pp. 105
http://en.wikipedia.org/wiki/Wind_tunnel
http://www.cardesignonline.com/design/aerodynamics/car-wind-tunnel-testing.php
201
202
EXPLORATION OF 85th PERCENTILE OPERATING SPEED MODEL ON CREST VERTICAL

CURVE TWO-LANE RURAL HIGHWAYS
Syed Khairi Syed Abbas 1*, Muhammad Akram Adnan 2 and Intan Rohani Endut 3
ABSTRACT:
Road accidents are a major cause of loss of human life, property damages and money throughout the
world. One of the reasons the road accidents related with the problem interaction between drivers and
road alignment. Facts showed that road accidents have become a major source of social and econometric
losses in recent years. In Malaysia, accidents in rural highway linked with the problem of inconsistencies
on the operating speed and geometric design of highways. The concept of design consistency is one of the
several tools used by highway designers to improve roadway safety. Despite that, more intention to road
safety has to be made so that the traffic collisions due to sub-standard and inconsistency of highway
design can be reduced and eliminated. The purpose of this study was to develop a set of geometric design
standard for an 85thpercentile speed on the crest vertical curve that reflective on Malaysian two-lane rural
highways conditions. The model allowed the designer to be more efficient in identifying relationship
between operating speeds with length of curve, gradient, andk-value. Parameters related to highway
characteristics, speed behavior and highway alignment has been conduction by laser gun meter detector.
Multiple linear regression analysis was conducted to develop the 85thpercentile speed (V 85C )
modelbycombination of all data under different types of the crest vertical curves alignment. The
developed V 85C model presented intends to detect geometric design inconsistencies in most of the crest
vertical alignment. And lastly, this research is an attempt to harmonize the designer adoption in selecting
and designing highway and enhanced the current design practice in Malaysia.
Keywords: Crest Vertical Curve, Geometric Design, Ergonomic Design, Operating Speed
1
Student Postgraduate, Faculty of Civil Engineering,Universiti Teknologi MARA, 40450 Shah Alam, Selangor, MALAYSIA
2
Senior Lecturer, PhD, Faculty of Civil Engineering, Universiti Teknologi MARA, 40450 Shah Alam, Selangor, MALAYSIA
3
Lecturer and Head of MTRL,PhD, Malaysia Institute of Transport, Faculty of Civil Engineering, Universiti Teknologi MARA,
40450 Shah Alam, Selangor, MALAYSIA
*Correspondence Author: Syed Khairi Syed Abbas, Faculty of Civil Engineering, Universiti Teknologi MARA,
MALAYSIA. E-mail: s_k_5421@yahoo.com
203
INTRODUCTION
The geometric designs are an important aspect to consider in construction of road or highways. The
designer should take into consideration the aspect of operating speed changes especially when vehicles
moving upgrade or downgrade in vertical curve. Determination the effect of operating speed especially
for vehicles while moving at crest vertical curves is essential to provision of adequate geometric design of
road especially in the scope of ergonomic and consistency of geometric design.
For this research paper, the operating speed model will be developed to evaluate the design speed
consistency and can be used to predict the speed that will match the acceptable design speed on the two-
lane rural highways. The vertical alignment of road consists of tangent grades and vertical curves. The
operating speed of the vehicles at the crest vertical curve is depends on the gradient of the curve. The
upgrade of the gradient for each curve also will affect the operating speed especially for heavy vehicles
because it is mostly affected by long grades.
Based on the current design practice in Malaysia, Jabatan Kerja Raya (The Public Works Department of
Malaysia) JKR Arahan Teknik (Jalan) 8/86 'A Guide on Geometric Design of Roads', (JKR, 1986) and
The Road Engineering Association of Malaysia, REAM Guideline 2/2002 'A Guide on Geometric Design
of Roads', (REAM, 2002) concentrate on the individual elements for which the geometric parameters
should be selected and determined the satisfy the minimum requirement.This approach has shown it
weakness due to the potential geometric inconsistencies because of that design product not guarantee the
users expectation. Location where there are large speed changes are the indicator of the inconsistencies
of the geometric design elements. Speed- profile models based on the operating speeds have been
suggested to replace the design speed concepts to ensure the consistency of the design.
LITERATURE REVIEW
85th Percentile Speed
The 85th percentile speed is the speed at or below which 85 percent of the free-flowing vehicles travel.
Traffic engineers have assumed that high percentage of drivers will select a safe speed on the basis of the
conditions at the site. The 85th percentile speed has traditionally been considered in an engineering study
to establish a speed limit. In most cases, the difference between the 85th percentile speed and the average
speed provides a good approximation of speed samples standard deviation. The 85th percentile speed is
considered the appropriate posted speed limit even for those sections of roadway that have an inferred
design speed less than the 85th percentile speed. Posting a roadways speed limit based on its 85th
percentile speed is considered good and typical engineering practice. This remains valid even where the
inferred design speed is less than the resulting posted speed limit. In such situations, the posted speed
limit would not be considered excessive or unsafe (Fitzpatricket al., 1997).
Operating Speed
Operating speed can be seen as the highest overall speed at which a driver can travel on a given highway
under favorable weather conditions and under prevailing traffic conditions on a section-by-section basis
(Khisty and Lall, 2003). Based on American Association of State Highway and Transportation Officials
(AASTHO, 2004), operating speedis the speed at which drivers are observed operating their vehicles
during free-flow conditions. The 85th percentile of the distribution of observed speeds is the most
frequently used measure of the operating speed associated with a particular location or geometric feature.
204
Crest Vertical Curve
There are two conditions exist for the minimum length of crest vertical curves. There are the sight
distances is greater than the length of the vertical curve and the sight distance is less than the length
of the vertical curve.Based on REAM Guideline, (2002) minimum lengths of crest vertical curves are
determined by the sight distance is the major control for the safe operation at the design speed chosen.
Passing sight distance is not used as it provides for uneconomical design. An exception may be at
decision areas such as sight distance to ramp exist gores where longer lengths are necessary. Crest vertical
curve have been the principal focus of research because accident rates have been found to be higher on
crest curves than on sag curves. Sight distance is a function of K and its importance in geometric design is
confirmed by research. Messer (1980) reports that sight distance is the primary factor affecting driver
expectancy when looking at the vertical alignment of a road.
K-value (Rate of Vertical Curve)
K-value represents the horizontal distance along which a one percent change in grade occurs on the
vertical curve. It expresses the abruptness of the grade change in a single value. Speed tables or other
design tools often provide a target minimum K value. Based on REAM Guideline, (2002) the rate of
change of grade to successive points on the curve is a constant amount for equal increments of horizontal
distance, and equals the algebraic difference between the intersecting tangents grades divided by the
length of curve or A/L in percent per meter. The reciprocal L/A is the horizontal distance meter required
to effect a one percent change in gradient and is a measure of curvature. This quantity (L/A), termed k, is
used in determining in horizontal distance from the beginning of the vertical curve to the apex or low
point of the curve. The k value is also useful in determining the minimum length of vertical curves for the
various design speeds.
METHODOLOGY
Introduction
This study had been performed to develop empirical model on the geometric characteristics at vertical
alignment along upgrade. The data collected was used to develop the model, and finally the statistical
results were obtained. The methodologies are focused on develop model of 85th percentile operating speed
on the crest vertical curves and the data reduction techniques will be done by using the data recorded into
spot speed data form. This study is focused on the 85th percentile operating speed for vertical alignments
along upgrade of two lane highway at rural area at Lenggong to Sauk, Perak based on R5 road. The most
important of this study is the research process includes the data collection from the field work, analysis
the 85th percentile operating speed and developing the new model of speed based on the case
study.Multiple linear regression analysis was conducted to develop the 85thpercentile speed (V 85C ) model
bycombination of all data under different types of the crest vertical curves alignment.
Site Selection
Three crest vertical curves are selected as study areas of two-lane rural highways. The study areas have
geometric information in construction drawing plan from the respective The Public Works
Department of Malaysia (JKR) officials. For the vertical curve, the geometric data include the K-
value, gradient of the curve (G), radius of circular curve (RC), length of the circular curve (LC), the mile
point of the beginning of the curve (VPC), the point of intersection (VPI), and the end of the curve (VPT).
A sample of the drawing is as shown in Appendix 1 to show that some of the data parameter can be
extracted from original construction plan. All these geometrical data are needed for the analysis and the
205
model development. The data is also to make the regression analysis of operating speed with this
geometric data parameter. Flow chart that adopted for this study is shown in Figure 1.
Selection and Identification of Study Area

Three crest vertical curves from Lenggong to Sauk, Perak at
Literature Review two-lane rural highway are selected as study areas according
to their characteristics. The crest vertical curve condition
along upgrade was observed with difference gradient of
curve.
Inventory Study
The inventory information such as location based on the
vehicle classification, naked eyes observation of crest vertical
curves and gradient etc were conducted
Data Collection
-Spot Speed Method using laser gun detector
Data Analysis
Calculate the 85th percentile operating speed
Presented in graph, chart, table
Model Development
-Regress the
relationship of
operating speed with
other parameters using
Statistical Analysis
Software
Model Validation
-conducted using new data set (independent
data)
Research Finding
Figure 1: Flow Chart that adopted for this study
206
Data Collection
Crest vertical curves are the main concern for this study. The crest vertical curves of the road consist of
tangent grades and vertical curves length. The data collections at site involve observation and take the
speeds of the vehicles. The speed data were collected at every 15 minutes in one and half hour for every
curve. The total number of hours speed data that were collected about ten hours. The speed of different
type of vehicles was collected by observation value using laser radar meter detector. The data collected
will be recorded in the data form to facilitate further analysis. The observations of speed are recorded at
three points along the curve, which is in beginning point curve, middle point curve, and end of the curve.
These types of curves are usually the segments of circles which have gradient those provide a smooth
flow of traffic along the curve. The operating speed of the vehicle at the curve is depends on the gradient
and length as shown in Figure 2 sketch. It is important to know the gradient of the curve to distinguish the
difference between the operating speeds of the vehicle at different gradient. A preliminary survey was
conducted along different two-lane rural roads to locate the section in different alignment of selected sites
area. The summary of available geometric design data of the test sections is presented in Table 1.
Figure 2: Sketch of Observation Point for Speed Measurement at Crest Vertical Curve
Table 1: Detail of selected section

Data Vertical
No. of crest vertical curve 3
K-value (%/m) 30-75
Length of curve, LC (m) 100-500
Gradient, G (%) -6 to +6
Data Reduction
The major tasks of these studies are the collection and reduction of field data to be used in the
development of 85th percentile operating speed models. These data include operating speed and geometric
data at the selected area. Systematic filling spreadsheet systems were applied to organize all the data. The
geometric data from the data collection sheets were entered into a spreadsheet too. In order to achieve the
objectives, data from three sites were used for further analysis in multiple linear regression analysis.
These three sites were carried out for the operating speed observation with different crest vertical curves
parameter such as, length of the curve, gradient of curve and the K-value at each curves section were
carefully investigated. Observations of operating speed behavior in horizontal and vertical curves were
carefully investigated. The speeds data were collected in 15-minute period. The 15-minute period was
chosen because speeds data from these different periodic counts make the number of sample for used in
regression analysis to develop the models. The 85th percentile speed were calculated using the analysis of
spot speed data process. The data collected in spot speed studies are usually taken only from sample
207
vehicle at the selected area in which the study was conducted. These data are used to determine the speed
characteristics of the whole population of vehicles traveling on the selected area. So, it necessary to use
statistical methods in analyzing the 85th percentile speeds (Garber and Hoel, 2002). The multiple linear
regression analysis was performed by using Minitab release 14.12.0 (Minitab Inc., 2004) statistical
software to develop a model that will represent the speed characteristics for the selected curves. Several
models were performed and only the significant and logical models were presented. The models were
developed to predict speed on crest vertical curve by taking consideration the data 85th percentile speed at
the mid-curve of vertical curve, (V 85C ).
RESULT AND ANALYSIS
Development of 85th percentile operating speed, V 85C
Descriptive analysis for 85th percentile operating speed on crest vertical curve, V 85C
To develop an 85th percentile operating speed on the crest vertical curve, the 85th percentile speed at the
point of the beginning of the curve was determined by select the 85th percentile speed at the point of the
beginning of the curve (Point 1). Then, the regression analysis was carried out using Minitab release
14.12.0 (Minitab Inc., 2004). Data screening need to be conducted before the reduced data could be used
for analysis to correctly identify data errors (Norusis, 1994). The data screen included removed the
unusual observation. The unusual observation indicate that, some of data collection missing because of
the careless of enumerator to take the speed value at site location and taking speed of police car,
ambulance or fire brigade truck. The speed is higher than the normal operating speed because they travel
for an emergency case. Histogram was generated for V 85 speed as shown in Figure 3. The number of
observation, mean, median, maximum value, minimum value, standard deviation, skewness and kurtosis
for parameter V 85 are as shown in Table 2.
Histogram (with Normal Curve) of V85

Mean 98.01
14
StDev 10.18
N 43
12
10
Frequency
0
70 80 90 100 110 120
V1
Figure 3: Histogram for 85th percentile operating speed at crest vertical curve, V 85C
208
Table 2: Descriptive Analysis for operating speed on crest vertical curve, V 85C
Variable Number of Mean Median Min Max Standard Skewness Kurtosis
observation deviation
V 85C 43 98.01 100.90 71.50 113.50 10.18 -0.90 0.11

(km/hr)
Interpretation of the Multiple Regression Statistics for V 85C model
The results of the final regression model are studied in further detail in this section. The multiple
regression equation for V 85 model is shown in equation 1. Table 3 shows constant value and K-value
(K)are very significant independent variable forpredicting V 85 where the P-value is less than 0.05 means
that the predictor is significant for both independent parameters and these parameters can be included in
the model for estimating V 85 .
V 85C = 114 0.275 K (1)
The equation above shows that the coefficient for the predictor variable K has a negative sign, implying
that an increase in the K-value will lead to a decrease in V 85 operating speed. The standard error contains
the same units as the coefficients. The relative value of standard error to the coefficient is important in
determining the reliability of the test statistics in estimating the population parameter (Faria, 2003). In
general, the smaller the values of the standard error in relation to the test coefficient produce better
results.
Table 3: Regression Analysis for final modelfor estimating V85 C
Predictor Coefficient Standard T-Value P-Value

Error
Constant,V 85C 114.179 5.150 22.17 0.000
K -0.27473 0.08419 -3.26 0.002
Analysis of variance test for V 85C model
Analysis of variance consists of calculations that provide information about levels of variability within a
regression model and form a basis for tests of significant. The analysis of variance portion of the output is
as shown in Table 4. The degrees of freedom are provided in the DF column, the calculated sum of
squares terms are provided in the SS column, and the mean square terms are provided in the MS
column. The P-value for the F test statistics = 10.65 for crest vertical curve which is less than 0.001,
provides strong evidence against the null hypothesis or H 0 is rejected. Hence, the regression model is
significant and can thus be used to explain or predict theV 85C .
Table 4: Analysis of Variance for Final Model V 85C
DF SS MS F P
Regression V 85 1 896.59 896.59 10.65 0.002
Residual error V 85 41 3452.28 84.20 - -
209
DISCUSSION
The 85th percentile operating speed on crest vertical curve at beginning curve(V 85C ) linear regression
models for Malaysian rural highway traffic have been successfully developed in this study. This model
highlights importance in various aspects of operating speed on rural highway and geometric design of
crest vertical curve alignment. The model developed must be evaluated to check its ability to represent
actual condition and to explain the variability present in a sample other than the used for its calibration.
Therefore, the design speed concept need to be understood well especially in rural highway and the
outcomes from this study are as valuable as a tool to assist the practitioners towards best practice in
highway geometric design and planning.
REFERENCES
American Association of State Highway and Transportation Officials (AASTHO). (2004). A Policy on
geometric design of highways and streets. United States of America.
Faria, D. (2003). A Framework to transform real time GPS data derived from transit vehicles to determine
speed-flow characteristics of arterials. The University of Texas Arlington, U.S: Ph.D, Dissertation.
Fitzpatrick, K., Carlson, P., Brewer, M., Wooldridge, M., and Miaou, S. (2003). Design speed, operating
speed, and posted speed practices. College Station, TX: Texas Transportation Institute.
Garber, N., and Hoel, L. (2002). Traffic and highway engineering. United States: Thomson Learning.
Jabatan Kerja Raya - Public Works Department Malaysia (JKR) (1986). Arahan Teknik (Jalan) 8/86 'A
guide on geometric design of roads'. Kuala Lumpur, Malaysia.
Jabatan Kerja Raya - Public Works Department Malaysia (JKR) (2001). Drawing plan road construction
phase 3 Grik to Lenggong. Kuala Lumpur, Malaysia.
Khisty, J., and Lall, B.K. (2003). Transportation engineering-An introduction, Pearson Education Inc.
3rded.
Minitab Inc. (2004). MINITAB Release 14.12.0. Canada.
Messer, C. J. (1980). Methodology for evaluating geometric design consistency. Transportation Research
Record 757, Transportation Research Board, National Research Council, Washington, D.C., 714
Norusis, M.J. (1994). SPSS Professional Statistics 6.1. Chicago,U.S: SPSS Inc.
Road Engineering Association of Malaysia. (2002). REAM Guideline 2/2002 'A guide on geometric
design of roads'. Shah Alam, Malaysia.
210
Appendix 1: Crest Vertical Curve Profile Plan
(JKR drawing plan road construction phase 3- Sauk to Lenggong, 2003)
211
212
BUS TRANSIT SYSTEM MODELING USING PARAMICS: A CASE STUDY OF KLANG

VALLEY NETWORK
Ho Yun Li *, Toh Boon Leong2
ABSTRACT:
This paper describe the way of the microscopic traffic simulation model to build a bus transit system in
certain area. Although the bus transit system is the oldest mode of public transportation used by man-kind
but a lot human will also needed it because it was the most economic transportation in this world.
Currently, there have several types of microscopic traffic simulation models were used to do research and
practical purpose in this world. The PARAMICS will be used to build the bus transit system in this paper.
PARAMICS is a microscopic traffic simulation model, which simulate the movement and behavior of
individual vehicles on road network. In Malaysia, there are many urban cities that are always congested
with people and vehicles. In this paper, the Klang Valley, the most congested city, will be chosen to be a
case study to build a bus transit network.
_______________________
Project Research Assistant, University Tunku Abdul Rahman (Setapak Campus), Jalan Genting Kelang, 53300
Kuala Lumpur, Malaysia.
Undergraduate student, Department of Civil Engineering, University Tunku Abdul Rahman, Jalan Genting Kelang,
53300 Kuala Lumpur, Malaysia.
* Correspondence Author: Ho Yun Li, University Tunku Abdul Rahman (Setapak Campus), Malaysia. HP:
+60162380816, E-mail: winnieho1987@hotmail.com
213
INTRODUCTION
The Klang Valley is an area comprises the entities of Federal Territory of Kuala Lumpur and
seven others municipal councils of the state of Selangor. The Klang Valley was potrayed by Titiwangsa
Mountains to the north and east and the Strait of Malacca to the west. In Malaysia, it is a city central that
are the heartland of industry and commerce. There have four different operators, i.e. the monorail, the
Putraline (light rail), the StarLine (light rail), and the KTM Commuter (heavy rail), about 182 km of rail
network operated within the Klang Valley
A simulation models like "A computer program that uses mathematical models to conduct
experiments with traffic events on a transportation facility or system over extended periods of time.
Simulation models are to simulate the behavior of traffic in a transportation system over time and space to
predict system performance. Taking the real-world system to consider the mathematical and logical
concepts associated with the operation of the system and simulates it with complicated computer software
in the laboratory rather than in the field.
There have three types of the traffic simulation models that are microscopic traffic simulation,
macroscopic traffic simulation and mesoscopic traffic simulation. Microscopic traffic simulation is the
dynamic and stochastic modeling of individual vehicle movements within a system of transportation
facilities. It predicts the state of individual vehicles and primarily focuses on individual vehicle speeds
and locations by continuously or discretely. Macroscopic traffic simulation aggregates the description of
traffic flow or the measures of effectiveness, i.e. speed, flow, and density, and their relationship to each
other. Macroscopic is simulating on a section-by-section basis rather than microscopic tracking by
individual vehicles. Mesoscopic traffic simulation is the simulation that is combines the properties of both
microscopic and macroscopic simulation. Mesoscopic less consistent than microscopic but it is superior
to some other traffic analysis techniques. This simulation emulates in individual vehicles but it can
describes their activities and interactions based on aggregate relationships. Typical applications of
mesoscopic models are evaluated of traveler information systems.
LITERATURE REVIEW
PARAMICS (PARAllel MICroscopic Simulation) is the one of microscopic traffic simulation

that is suite of software tools. PARAMICS is a microscopic urban and freeway traffic simulation that is
use to model the movement and behavior of individual vehicles on road networks. Quadstone along with
SIAS Ltd., a private engineering consultancy also based in Edinburgh, developed a Cray supercomputer
version of PARAMICS at the University of Edinburgh's Parallel computing center in the early 1990's.
Since 1998 Quadstone and SIAS have developed and marketed independent versions of PARAMICS.
Quadstone distributes PARAMICS outside of the United Kingdom and Ireland, whereas SIAS handles
marketing within the United Kingdom and Ireland.
Quadstone PARAMICS providing a powerful, integrated platform for modeling a complete range
of real world traffic and transportation problem. There are total eight modules in PARAMICS to improve
usability, integration and productivity allowing users to get added value from the modeling process. The
eight modules are as following:
1. PARAMICS Modeller: is the modular suite of microscopic simulation tools to operate at the
microscopic level and integrate with the core network building post simulation tool. There have three
fundamental operations are provided in the Paramics Modeller, which is model build, traffic simulation
with 3D visualisation and statistical output accessible through a powerful and intuitive graphical user
214
interface. Load a few of images in the background of the Modeller and draw the network on top, that is a
case "draw what you see".
2. PARAMICS Processor: is a pair batch simulation productivity tool to run a simulations without
ant graphical interface on any number of networked PCs that created in Modeller. Processor reducing user
down time, and speeding up the model development lifecycle because it can be used to automate
simulation and analysis processes. Processor is split in to two parts, which is Processor Manager that is
used in the configuration and control of the simulations, and the Processor Node is to executes the
network simulation.
3. PARAMICS Analyser: supplies the user with a large collection of simulated data in binary
format that is derived from the Modeller tool. It is a post simulation data analysis package that is
specifically designed to operate at the microscopic level, and to integrate with the core Paramics tools.
With the Analyser, simulation networks, simulated results and observed data will be held, which is able to
created a multiple views that display data in simple or comparison format, either on a common network or
on across similar networks.
4. PARAMICS Designer: is a model conversion and editing tool for converting 3D Studio format
model files (3DS) in to Paramics model format files (PMX). PMX files are personal model file format for
use by the Paramics suite of software tools. In order to use a PMX model in a Paramics application a 3DS
file must be imported into Designer and then saved to PMX format. 3DS files can have multiple files for
the model and its textures but PMX files store all data in a single file.
5. PARAMICS Programmer: is an Application Programming Interface (API) to generate external

applications. It is a comprehensive development API for the Paramics suite and it's the most powerful
research and customisation tool available for users interested in microscopic traffic simulation.
Programmer allows users to augment the core Paramics simulation with new functions, driver behaviour
and practical features.
6. PARAMICS Estimator: is an OD matrix estimation package to operate at the microscopic level

and integrate with the core Paramics tools. Estimator utilizes all aspects of the core Paramics simulation,
ensuring maximum compatibility between the OD matrix, model structure, and underlying assignment
techniques.
7. PARAMICS Converter: work to "convert" the existing network data from a range of sources
into a basic Quadstone Paramics network. In the Quadstone Paramics range, Converter is the new entry-
level tool in the Quadstone Paramics range. Converter reads network data and allows the user to "teach"
the application how that data should be interpreted and transformed into a basic Quadstone Paramics
Network. Through an interactive GUI and graphical display the user can then edit the basic network data.
8. PARAMICS Monitor: is a pollution evaluation framework module that integrates directly with
the core Paramics simulation. Monitor can be used to collect traffic emission pollution data at the most
detailed per-vehivle level on a road network. These levels can be written to a statistics life at regular
intervals, and can also be viewed graphically while the simulation is running.
The PARAMICS Modeler is a stochastic, microscopic, time step, and behavior-based simulation
model. PARAMICS is a suite of high-performance software tools that is to model the movement and
behavior of individual vehicles in detail. With stochastic modeling the model can be run many times and
mean values of output can be calculated. By the research that is conducted at the Transportation Research
Laboratory (TRL) which concluded that the driver behavior can be describes with two parameters, which
is aggression and awareness. These two parameters will affect on quantities such as target headway, top
215
speed, propensity to change lane and gap acceptance of the individuals. In PARAMICS, there have three
assignment techniques, which are "all-or-nothing" assignment, stochastic assignment and dynamic
feedback assignment. All-or-nothing assignment is assumes that all drivers are traveling between two
zones with the same knowledge base for route choice and there is no congestion effect and the link costs
do not depend on the flow levels. Stochastic assignment is to emphasize the variability in drivers'
perceptions of costs and the composite measure that they try to minimize the distance, travel time, and
generalized costs. Dynamic feedback assignment is assumes that drivers who are familiar with the road
network will reroute if the current traffic information condition provide to them.
The key feature of PARAMICS which is include the high speed simulation, integrated software,
interface to macroscopic data formats, interface to point-count traffic data, and support large scale models.
PARAMICS has been created to run at high-speed that is no loss of detail with the real-time simulation of
hundreds, thousands or millions of vehicle. In a single package, PARAMICS provides simulation,
visualization, interactive network creation and editing, interactive adaptive signal control, on-line
simulation data and statistics gathering, vehicle following, traffic control strategy evaluation, and
interactive simulation parameter tuning. PARAMICS can load network data direct from standard node
and link data sets. PARAMICS is capable to simulate the individual movements with the large number
vehicles over a road network faster than real time.
METHODOLOGY
PARAMIC modeler will be used in this paper to develop the bus transit network at the area of
Klang Valley as a study case. Here, the way or step will introduce to indicate that how the PARAMICS
modeler simulate from the beginning to present its output as a real time visual display for traffic
management, road network design and analysis.
Before starting create or build the network, that need to open a new network at the New Network
Wizard. Set the properties, units and orientation, configuration, demand specification, vehicle types,
measurements, categories, and others in the New Network Wizard. All the setting can be set again when
building the network or finish build the network in the Core Model Attributes. After all the setting finish
and save the network in the location that are wanted, there have a default network appear in the Paramics
Modeller for every new network created. That are need to open the Editor Palette before build the
network in the Paramics Modeller, it is used to update the fundamental objects in a network that is to
added, updated and removed each object from the network. The Editor Palette include eight different
editors, which is Junctions, Links, Lane Attributes, Control Points, Zone Elements, Infrastructure, Public
Transport and Signal Control.
The users are allowed to place the graphics file over the network display in Modeller, this called
overlay. It is an initial step in Paramics modeller used as a starting point for the design of a network.
Typically, this file is a map image download from the Google Earth, Google Map, or etc. Overlay is give
the user as a basis and easily to put the element when start to develop the network. It also gives the benefit
of clear understanding when going to public presentation. For this study case, the Google Map was taken
as the aerial photo for overlay because the network structure of main road and arterial road were provided
clearly in the Google Map
Nodes represent points in the network where condition change, either physically or behaviorally.
Normally, nodes should be place at junctions and at locations where the road layout changing, which is
change in number of lanes, change in use of lanes, roundabouts, etc. If road layout is the fly over or other
type that are need to creating a three dimensional model, so that the height above sea level should be
specified for each node. Beside that, if the nodes represent as a entry point to the network should be coded
as zone connector types, designating them as route zones.
216
Link were then created by connecting two nodes together and it should be coded initially with the
correct link categories, which is number of lanes, highway and urban link types, major and minor link
types, speed limits, lane width, etc.
Zones form the start and end points for each vehicle movement and hence define the Origin-
Destinations (OD) matrix pair in the trip matrices. There are two types of the zones, which is area zones
and route zones. Area zones represent area in the network where vehicles entry and exit informally and
without specific location. Route zones represent specific points in the network where vehicles are
generated in traffic flow conditions such as network entry points for freeways, highways, arterials, car
parks, and etc.
Figure 1 Nodes, Links and Zones
A junction represents an intersection where more than two links meet in the network and traffic
interacts or makes route choice decisions. Each junction is represented by a basic configuration detailing,
which is turning movement, their lane allocation and the associated priority. Priority at junctions is coded
as major, medium, minor or barred movement to specifying the range of lanes that are available to make
each prescribed turn. The major types of junctions needed to be coded, which is Priority Junction,
Roundabout Junctions or Signalised Junction. At priority junctions the priority to each link was manually
fed to ensure the validity of the model. Roundabout junctions which are additional behavior used to
represent the merging effects of roundabout behavior. The road intersections controlled by traffic signal
that is a signalised junction.
217
Figure 2 Priority Junction Setting
Figure 3 Roundabout Junction Setting
218
Figure 4 Signalized Junction Setting
The public transport editor that is used to create the bus transit system in the network. In the
public transport palette is used to create the routes, stops and demand for all PT traffic flow in a network.
The bus stop must be create first to represent any point on the network where fixed route vehicles stop for
a period of time or where fixed route vehicles enter or exit the model. Bus routes were then added
according to the routes that are running by the specific bus.
Figure 5 Bus Stops and Bus Route.
In this paper, the data are collect by the way of the video recording and making appointment with
the traffic engineering. From the video recording to make sure on the road structure, which is how many
lanes in the certain link or junction, which type of the junction, the location of the bus stop, how many
bay at the bus stop, where have bus only lane and etc. Other that, the appointment that are make to the
data that needed in simulation that is the bus routes, the fleet size for each bus route, bus type, the
capacity of each bus type, the origin-destination (OD) demand for each bus stop, the daily rider ship of
219
buses and the bus terminals or depots. All this data that use to as an input, i.e location bus stop, bus routes,
fleet size, vehicle type, schedule, passenger rates, arrival rate (p/hr), alight rate (%), pay time (secs), dwell
time, when develop the bus transit system in the PARAMICS Modeler.
Figure 6 The network created in the Paramics Modeller of the area Klang Valley.
CONCLUSION
Bus transit system in Klang Valley had been build by microscopic stochastic traffic simulation
tool that is PARAMICS. After simulation the result will record and analyses are investigated. The
PARAMICS simulation model is used as an example of how the latest simulation models can be applied
in real world situations. The PARAMICS model is a very versatile simulation model. In this paper, the
experience with the use of a PARAMICS has been record.
A detailed analysis of the internal, i.e. car-following, gap acceptance, mean headway, mean
reaction etc, can be carried out in the future study. With the extension of the analysis to include freeway
weaving sections can be work in the future. Beside that, extension of the analysis to include ramp
metering that have many restriction on the ramp.
REFERENCES
Christoff Krogscheepers and Kent Kacir (2001). Latest Trends in Micro Simulation: An Application of
the PARAMICS Model.
Jun-Seok Oh, Cristian E. Cortes, R. Jayakrishnan and Der-Horng Lee (1999). Microscopic Simulation
with Large-Network Path Dynamics for Advanded Traffic Management and Information System.
Robert L. Bertini, Roger Lindgren and Sutti Tantiyanugulchai (2002). Draft Application of PARAMICS
Simulation at a Diamond Interchange. Eastern Asia Society for Transportation Studies, Vol. 7, 2009
Tran Vu TU, Kazushi SANO, Nguyen Cao Y, Doan Thanh TAN (2009). Simulation of Bus Operation in Urban
Roads Using PARAMICS. Eastern Asia Society for Transportation Studies, Vol. 7, 2009.
220
KEEPING CITIES MOVING: A LITERATURE REVIEW ON GREEN CAMPUS
Siti Hafnida Binti Razali 1*
ABSTRACT:
The simply way of keeping cities moving is by achieving sustainable transport. In this paper, the city is
represented by campus with focusing on UiTM Shah Alam. The main aim of this paper is to conduct a
literature review on green campus to be implemented in order to improve the campus transport and to
achieve sustainable transport. Apart from the main aim, objectives of this literature is to review deep
definition on sustainable transport, concept of sustainable transport in landuses, issues and problems of
campus transport, effects of the problems on campus environment and campus community and to find
strategic alternatives to improve campus transport. Analysis is performed based on primary and secondary
data. The primary data are collected through observation on campus transport in Uitm Shah Alam. Mean
while, the secondary data used in this paper are collected from previous research and study on the green
campus, sustainable transport, problems of campus transport, and the perceptions of campus community
toward the current transportation in campus. The campus community is consists of three cluster groups of
students, staffs and administrators. The finding of this work paper of literature review on green campus
found that there are several issues and problems of campus transport. Among the problems are
accessibility problems, single occupancy vehicles, increasing numbers of private vehicles, congestions,
pedestrian and non-motorised transports safety and effect on campus environment and community.
Sustainable green campus transport can be achieved if all of the identified problems solved. Car free
planning, green concept and pedestrian friendly campus, park and ride out side campus, car sharing and
etc are among the strategic alternatives identified.
Keywords: Green campus, sustainable transport, campus transportation
1
Student, Universiti Teknologi MARA (UiTM), 40450 Shah Alam, Selangor Darul Ehsan, MALAYSIA
*Correspondence Author: Siti Hafnida Binti Razali, Universiti Teknologi MARA (UiTM) Shah Alam, Malaysia. Tel:
014-2346168 @ 019-7553654. E-mail: shafnida@gmail.com
221
1 INTRODUCTION
Sustainable transport is derived from sustainable development where it is achieved when a

development meets the needs of the present without sacrificing the ability of future generation to do the
same (Brundtland Comission, 1987). Accomplishing sustainable transportation is very momentous
because of its significant roles to provide accessibility for people, services and goods. It is essential to
ensure the transportation planned for an area is able to helps people, services and goods reach their
destination easily by providing viability of diverse and plentiful transport choices, encouraging the use of
public transport and non-motorised transport such as pedestrian and bicycles (Paul Barter, 2000). In order
to achieve sustainable transport, campus transport must implement four vital characteristics of sustainable
transport stated by Centre for Sustainable Transportation (CST) at Canada, 1997. Those strategies are:
i. Able to provide accessibility and development needs of a person, organisations and community.
ii. Able to provide good, convenient and affordable transportation services.
iii. Able to promote comfortable and safety environment.
iv. The planned for sustainable transport must considering the needs of future generation.
2 PROBLEM STATEMENT
2.1 Inaccessibility
In our everyday lives, we encounter lots of issues and problems related to transportation. Days
from days, the problems appears to be more critical and seems to have no solution. Among the major
problem faced by campus transport is lack of accessibility. This problem creates difficulty for people to
reach to their destination especially by using public and non-motorised transport and by walking. A case
example of home base journey, campus community has awareness to use public transport from home to
workplace. However, due to the walking distance and accessible time from home to the nearest bus stop
or commuter station lead the community to ignore their awareness. In addition, public transport user
normally students have to walk more than 10 minute to reach to the nearest bus stop before they can
board the bus to enter campus area. This is because of inaccessibility due to the lack of public transport
route coverage. Therefore, it forces them to use private vehicles. It is supported by Kuala Lumpur statistic
which indicates that there is only 20% of Kuala Lumpur Population use public transport compared to
other 80% is using private vehicles (Kuala Lumpur Structure Plan, 2020).
2.2 Increasing Number of Private Vehicles
The problems of inaccessibility is encouraging people to use private vehicles rather than public
transport because it is more convenience and easy accessing destination. Although nowadays there are lot
of alternatives to encourage campus community to use public transport however, there are some areas are
not in public transport route coverage. A study conducted by University of Victoria, Campus Security
Services found, although there are lots of programs promoted by The Universitys TDM (Transportation
Demand Management) in order to provide various public transportation options for people, however there
are certain area that are not well served. This situation is forcing people to drive and contributes to the
increasing number of vehicles on road. Apart from that, the increasing standard of living is also one of the
reasons. Campus community including students is capable to own private vehicles. Malaysia statistic
shows, 22% of registered vehicles is increase from year 2000 until 2009 (Refer Figure 1.1). It shows that
the number is increasing by years hence it leads to congestion problem.
222
Figure 1.1: The increasing number of car ownerships from 2000 until 2009
Source: Malaysia Automotive Association, 2010
2.3 Congestion
Due to the present increasing number of private vehicles on road it leads to traffic congestion
problem. Malaysia is currently facing congestion crisis. The roads are over capacity because of the
increasing number of vehicles on road especially private vehicles. As result, more serious congestion
problems occur. For example, previous study indicates that the traffic movement in UiTM campus is
congested with 1.1 level of service (LOS). Single car occupancy is the reason for the increasing number
of private vehicles entering UiTM campus. In fact, according to a study conducted by National
Infrastructure Company Ltd (SPNB), there are 2.2 million cars entering city every day especially during
the peak hour from 8.00 a.m to 9.00 a.m and from 6.00p.m to 7.00 p.m.
2.4 Environmental Effect
Apart from congestion, environment effect is one of issues occur due to the unsustainable
transport. The smoke, dust and noises from vehicles contribute to environment pollution and accelerate
the process of global warming. Previous study found that transportation sector produces 24.1% of CO2
emissions worldwide which is a high number (Eva Kassens, 2009). Therefore, planning sustainable
transport is important to improve the transportation as well as to decrease the negative effects of
transportation sectors on environment.
3 LITERATURE REVIEW
3.1 Sustainable Transport and Landuse
From literature review, researcher found that lot of research has been conducted on academic
genres, which mainly focuses on Ph.D. and masters theses as well as research articles on sustainable
transport. However, literature shows that comparatively fewer studies have been conducted on sustainable
transport that mainly concentrates on landuse. Most of the research on academic and professional genres
concentrates on the issues, strategies and policies to create sustainable transport.
223
As we know, transportation and landuse cannot be separated. Based on the rigorous literature on
various definitions of sustainable transportation, it can be deduced that sustainable transportation system
must able to promote accessibility with good and affordable transportation services in convenient,
comfortable and safety environment, as well as without neglecting the needs of future generation. This is
also supported with statement made by European Conference of Ministers of Transport, a sustainable
transportation system must able to provide accessibility and development needs of a person, organisations
and community with promoting safe environment and consistent with the long-term ecological balance of
the natural environment.
Therefore, Sustainable transport in landuses can be described as an efficiency transportation

system that providing suitable, convenience and affordable choices of transportation modes to users,
besides able to support transportation needs and its connection to landuses while considering surrounding
environment and future generation needs. Therefore in order to plan a sustainable transportation system, a
good planning on transportation system and landuses is needed.
People, services and goods are moving from origin to destination which related to landuses, such as
from home to work, institution, recreation area, business area and etc (Paul Barter, 2000). Each types of
landuse have different requirement on concept of sustainable because each types of landuse play different
role base on its determined function. Examples for application of sustainable transportation concept in
landuses are follows (A study on sustainable transport in Landuse Guideline, 2000):
City centre and urban area consists of various types of landuses such as business and service area,
residential, institutions, recreation area and etc. Consequently, it requires various choices of
transportation modes. The transportation modes must suitable, convenient affordable. The
accessibility must able to access people, services and goods from an area to another with easily and
fast. Using public transport, non-motorised vehicles and walking are the best solutions to alleviate
congestion in city centre and urban area besides helps saving valuable spaces and energy.
Residential area frequently is the origin and destination of a journey. This area requires a good
transportation system which helps to encourage social interaction between the communities.
Normally lots of short distances journey occur in residential area, using private vehicles such as car
is inefficient over the short distances. Hence, providing pedestrians walks with considering safety
elements is important.
Recreation area requires good and strategic accessibility and transportation system in order to
promote visitors.
3.2 Campus Transportation
This article is focused on literature review on green campus. The institution or campus area is a
large community with high density area which consists of three clusters transportation users; the student,
staff and administrator. Campus is the major traffic generator in a way that is significantly different from
standard peak hour commuting traffic (Kaplan, 2008). This is because campuses are large and contain
places to work, learn, socialize and live. Therefore, many of trips are generated with its own purpose.
College campuses nowadays are highly depends on automobile either by public and private
vehicles (Balsas, 2002). Normally, the uses of private vehicles especially car base transport creates
masses of problems such as parking problem, congestion, inefficient over short distances, expensive,
contribute to environment pollution and etc. Those problems not only affect the colleges traffic but also
traffic outside college.
224
Another problem is the environmental impact of transportation such as disturbance to teaching,

loss of natural environment and greenery, despoliation of the visual environment by parking provision,
and health effects on staff and students (Trolley, 1996). Therefore, alternatives and solutions are needed
to tackle the problems without destroying campus green environment and campus community value such
as green campus planning, parking management, car sharing, park and ride, mass transit, pedestrian,
bicycles etc.
4 ANALYSIS
4.1 Public Transport
Since this paper work is a literature review, analysis will be done based on secondary data. From
literature review, it is found that one of reason for campus transport problems is lack of public transport
serving the community. In fact, some campuses are only allowing certain public transport for example in
UiTM Shah Alam, only buses from Syarikat Rapid KL and Shah Alam Taxi are allowing entering the
campus (Hazani, 2008). This restriction is encouraging campuses community to use private vehicles
because the limited number of public transport available.
Apart from that, also from literature review researcher found that the types of public transports
normally serve in campuses are bus and taxi. However, it can be analysed, between these two
transportation modes, bus is the priority rather than taxi due to the fare charged. Bus is more affordable
for campus community especially for students. However, students normally complain on the lack of bus
frequency and bus sometimes late from time. Therefore it forces them to find another alternative to move
inside the campus. The most cited alternative is driving a car or motorcycle. However, this alternative will
lead to congestion problem. On top of that, normally there is only one passenger for a car. The single
occupancy vehicle increases the vehicles number in campus. On top of that, the increasing usage of
private vehicles will continuously increase if the public transport provided did not upgrade and improve
their services.
4.2 Congestion
The congestion problems in campus is normally cause by the increasing number of campus
population. When the number of population increases, indirectly the need of mobility is also increases.
This is supported by a study on The Implementation of Total Public Transportation Services in Uitm Shah
Alam by Rahman, 2008 found that the over population in UiTM creates transportation problems.
Therefore, it can be analysed that due to the over increasing number of campus population larger than the
requirement number causes public transport provided did not capable to cover the demand. Hence, some
students, staffs and administrators prefer private vehicles rather than public transport.
4.3 Parking
Mean while, the parking lots provided are not enough due to the increasing number of private
vehicles in campus. Table 1.1 shows only 2173 parking lots provided in serving 1499 number of car.
With low parking turn over (4.3 6.3 car per lot) the parking lots provided in UiTM campus are
insufficient to fulfill the demand. Therefore, drivers tend to park their cars illegally straight disturbed and
slower the traffic flow and cause congestion as well as air pollution in campus.
225
Table 1.1: Registered car in UiTM, 2000
No. Parking lots user No of car No. Of Parking Lots

provided
1. Students 4165 2173
2. Staff/administrator 7723
3. Contractor 3111
Total 14999
Source: Safety Unit Office, Universiti Teknologi MARA, Shah Alam, 2000.
4.4 Pedestrian
Normally, campus community refuses to walk due to several identified problems which are lack
of proper pedestrian linkages between main landuses, not including friendly pedestrian elements, not
functional when rainy day and insufficient space for proposing pedestrian walkway. UiTM campus
community usually prefers to use vehicles rather than walking inside the campus because there is no
proper pedestrian walk provided. Only side walk is provided for users, in fact the side walk is not fully
connected with each other. For example, the location of student hostel, faculty and administration office is
not convenience for walking. Apart from that, campuses are highly depending to private vehicles.
Therefore the physical development is priority to road network (Refer Figure 2.0) rather than focusing in
providing proper pedestrian walkway.
Figure 2.0: Typical Road Cross-section of Universiti Teknologi MARA, Shah Alam
Source: Harris, 1998
226
5 CONCLUSION
The finding from rigorous literature review on sustainable transport and landuse in campus it can
be deduced that there are lots of weakness that need to be improve in order to achieve the transport
sustainability in campus. Among the strategies proposed to improve the transportation in campus are like
below:
Promoting Car Free Planning project to encourage campus community to abandon their
private vehicles when entering campus area.
Park and Ride provide outside campus
Encouraging Car Sharing concept
Providing good public transportation service to helps students, staffs and administrator
access to their destination easily.
Using concept of green and pedestrian friendly campus.
In future, when designing layout for campus must take into account the elements of
friendly pedestrian.
All of above strategies is rigorously believed able to minimize and solving transportation
problems, circuitously help improving transportation in campus or other than campus area. Besides,
straightly accomplish the goal of sustainable transport. Therefore, our cities will move forward and never
stops. Besides, it is recommended that the further study should be done on the effectiveness of strategies
and policies implemented in order to achieve sustainable transport.
6 REFERRENCE
A.Rahman Paul Barter (2000). Takin Steps: A community Action Guide to People-Centred, Equitable and
Sustainable Urban Transport. Sustainable Transport Action Network for Asia and the Pacific
Abd. Haris Bin Shamsuddin (1998). Towards A Pedestrian-Friendly Campus: A Proposed Re-
development of Campus Circulation for ITM. Shah Alam: UiTM
Cheng-Min Feng (2005). The Evaluation Indicators for sustainable Intercity Transportation in Social
Equity Aspect. Taiwan: National Chiao Tung University
Dinesh Mohan (2007). Traffic Safety as a Pre-Requisite for Sustainable Urban Transport: An
International Analysis. New Delhi: Indian Institute of Technology Delhi
Eva Kassens (2009). Sustainable Transportation: An International Perspective. Canton: Sherman Printing
Foon Weng Lian (2010). Kesesakan Jalan Raya Tingkatkan Penggunaan Minyak dan Pelepasan Gas
Rumah Hijau. Persatuan Pengguna-pengguna Standard Malaysia
Kajian Garispanduan Pengangkutan Mampan Dalam Konteks Guna Tanah. Selangor: JPBD
Ken Gwilliam (2002). Cities on the Move: A World Bank Urban Transport Strategy Review.
Washington: The Word Bank
Nor Hazani Mat Daud (2008). The Implementation of Total public Transportation Services in UiTM
Main Campus, Shah Alam. Shah Alam: UiTM
Rodney Tolley (1999). Green Campuses: Cutting the Environmental Cost of Commuting. UK:
Staffordshire University
227
Zaridah Bte Zakaria (2001). Kajian Isu dan Masalah Tempat Letak Kereta di UiTM Shah Alam. Shah
Alam: UiTM
Internet Sources
Car-Free Planning: Reducing Automobile Travel at Particular Times and Places. Retrieve at 9 p.m on
July 26, 2010 from http://www.vtpi.org/tdm/tdm6.htm
Balsas, Carlos J. L. (2002). Sustainable Transportation Planning on College Campuses. Retrieve at 8 a.m
on 5 July, 2010 from http://www.sciencedirect.com/science?_ob= ArticleURL&_udi=B6VGG-
4783KDV1&_user=2512477&_coverDate=01%2F31%
2F2003&_rdoc=1&_fmt=high&_orig=search&_sort=d&_docanchor=&view=c&_acct=C00002747
8&_version=1&_urlVersion=0&_userid=2512477&md5=aa9ad681c7cbd1f84051e12586d5a06f
New Study Examines Traffic Congestion on a University Campus. Retrieve at 9.30 a.m on July 20, 2010
from Physorg.com
University of Victoria (2010). Campus Security Services. Retrieve at 6 a.m on July 20, 2010 from
http://web.uvic.ca/security/parking/FAQ.html
Victoria Transport Policy Institute (2010). Campus Transport Management: Trip Reduction Programs on
College, University and Research Campuses. Retrieve at 3 a.m on July 5, 2010 from www.vtpi.org
Victoria Transport Policy Institute (2010). Ridesharing: Carpooling and Vanpooling. Retrieve at 3 a.m on
July 7, 2010 from www.vtpi.org
228
E-ASSIST: A TRANSPORT DEMAND MANAGEMENT EXPERT ADVISORY SYSTEM
Resdiansyah Mansyur 1*, Riza Atiq O.K. Rahmat 2,

Amiruddin Ismail 3 and Mohamad Raduan Kabit 4
ABSTRACT:
In view of the fact that supply strategies alone could not solve urban congestion, many cities around the
globe have adopted Transport Demand Management (TDM) strategies as part and partial of their
congestion mitigation plan. TDM comprises several strategies and policies that aim to modifying travelers
behaviour. . TDM comprises strategies and policies that are different in nature which can be divided into
several categories according to how they affect travelers behavior. Selecting and determining suitable
TDM strategies for a particular congestion mitigation goal can be a complex task; thus requires expertise.
In this regards, the effectiveness of a TDM strategy is primarily depending on whether its selection was
appropriately examined prior to its field implementation. This paper presents the development of an
Expert Advisory System for Selection of Implementation Strategies for TDM (E-ASSIST). The process
of organizing the available knowledge of TDM strategies, as well as the process leading to the selection
of one or more strategy advice, is encoded in the knowledge based expert system shell developed for the
purpose by using shell expert system Kappa-PC version 2.4 which was adopted object oriented and high
resolution graphical user interface. The advice given from the working system was evaluated and
validated by comparing the output of the system against the recommendations made by transportation
professionals. The evaluations indicate favourable results for the system. The E-ASSIST system can be
used as a decision support system as well as a teaching tool for junior transportation engineers, planners,
private developers, and government officials.
Keywords: Transport Demand Management (TDM), Knowledge-Based Expert System
1
Senior Lecturer, Linton University College, BUTL, Batu 12, Mantin, 71700, Negeri Sembilan, MALAYSIA
1
Sustainable Urban Transport Research Center (SUTRA), UKM , Bangi, Selangor, MALAYSIA
2
Professor, Sustainable Urban Transport Research Center (SUTRA), UKM , Bangi, Selangor, MALAYSIA
3
Professor, Sustainable Urban Transport Research Center (SUTRA), UKM , Bangi, Selangor, MALAYSIA
4
Lecturer, Universiti Malaysia Sarawak, Kota Samarahan, MALAYSIA
*Correspondence Author: Ir.Resdiansyah Mansyur, Linton University College, Malaysia. Tel: +6006 758-7888
ext. 116, Fax: +606758-7599. E-mail: dianlya@legendagroup.edu.my/dianlya@yahoo.com
229
INTRODUCTION
As traffic grows around the world, congestion becomes more widespread and occurs significantly
longer during weekdays. As such congestion and traffic-related pollution are increasingly becoming
major issues in towns and cities world-wide. According to Kuala Lumpur Structure Plan 2020 released by
Kuala Lumpur City Council, private cars are rising dramatically. The increasing reliance on private
transportation, in particular private cars, has created considerable pressure on the road network which
consequently has contributed to the problems of traffic congestion. Providing more road spaces to keep
pace with traffic demand is not the answer. It would be far too expensive and socially disruptive, and
would exacerbate the long term problem which was initially trying to be tackled. Some people argued that
we should let traffic find its own level, thus forcing drivers to find alternative ways of travel, or not to
travel at all. However, we believe this would mean abdicating our economic, social and environmental
responsibilities. In the meantime, the countryside and urban areas would both suffer significant
deterioration. The only realistic option is to seek ways of first controlling, then reversing, the growth in
traffic. This needs to be done in the most equitable way, acknowledging that there are sections of society
for who travel choices are more limited.
LITERATURE REVIEW
TDM as Solution
In the last decades, traffic and transport engineers and practitioners had come to the conclusion
that traffic problems can be mastered only through the introduction of appropriate Transport Demand
Management (TDM) measures rather than through the provision of new highway infrastructure
(Papaioannou.P & Georgiou.G, 2003). Demands on travelling increasing with the assumptions that
private car is the best or the only solution for mobility. The supply for the infrastructure increases
proportional to traffic volume. At one point, supply cannot meet demands due to certain factors such as
budget, land or space issue (Litman, 2009). Allowing traffic to grow to levels at which there is extensive
and regular congestion is economically inefficient (Frame.G, 2010). Although the construction of
additional road capacity can alleviate some of the effects of congestion, the benefits may be
counterbalanced unless growth in traffic volumes can be restrained. Therefore, another alternative is by
implementing TDM, which is to make sure that people still travel but at the same time reducing the
private car usage (Litman, 2009).
How Expert Advisory System Can Help Us
The implementation of TDM requires experts knowledge, skills and experiences; as they are able
to undertake a complex scenario analysis and subsequently can provide reliable TDM solutions based on
the specific context of the problem (e.g. purpose and location details). As argued, expert are not always
available, nor do they always have the time to consult all possible references, review available data, etc.
(Nazrul Islam 2004). This will lead towards time consuming on scheduling appointments and interviews,
thus delaying one project. There is also a possibility that there might be no more adviser persons left upon
death since most of them are at elderly age.
TDM Measurement (Supply and Demand)
Traditionally, transportation strategies adopt supply-side tactic, which addresses the increased
demand for transportation facilities by supplying more, in other ways, increasing the capacity of the
facilities (Wilson and Shirazi, 1991). Now, it is being realized that this strategy will not be able to meet
the long-term goals and objectives of the transport planners and policy makers. The experiences of many
230
large cities of the world have revealed that as capacity is increased, demand also needs to be increased at
a similar rate, otherwise, in the long-term, transport users experience no net travel time advantages and
society suffers the impacts of costly road bills and environmental degradation (Mahmood et.al, 2009)
Nowadays, it reiterated in the literature to put more emphasis on transportation demand

management strategies rather than the traditional transportation strategies which are based on supply-
side tactics to solve the problems of transportation systems (Behnke, 1996; Beroldo, 1990; VTPI, 2005).
In fact, TDM measures were introduced because of the failure of traditional solutions in addressing the
huge magnitude of congestion in major metropolitan areas (Giuliano and Wachs, 1992 in Mahmood,
2008) where most of the traditional solutions were either financially or practically infeasible.
Furthermore, Broaddus et.al (2009) also emphasize that we cannot build our way out of congestion; we
need paradigm shift from supply side measures to demand management as shown in Figure.1 below
Figure 1: Paradigm shift from supply side measures to demand management
METHODOLOGY
Development of Expert Advisory System
Advisory system is an expert system that provides advice to a user (El Alfi & El Alami 2009).
Advisory systems provide the advices and assist for solving problems that are normally solved by human
experts. They can be classified as a type of expert systems (Forslund, 1995). According to Aronson &
Turban (2001), both advisory systems and expert systems are problem-solving packages that mimic a
human expert in a special area. These systems are constructed by eliciting knowledge from human experts
and coding it into a form that can be used by a computer in the evaluation of alternative solutions to
problems within that domain of expertise. On the other hand, Gregg and Walczak (2006) were stated that
advisory systems are designed to support decision making in more unstructured situations which have no
single correct answer. Various methods have been applied in knowledge acquisition process in order to
collect data from books, encyclopedia, and expert people in transportation, also known as expert domain.
There are lots of TDM strategies listed and understanding each of it is very important. Based on the
summary provided, the most suitable and appropriate TDM strategies are then identified based on specific
objectives and geographical area. The final advices are chosen according to the most appropriate three
strategies among very appropriate list determined by expert domain. Suppose there are less than three
very appropriate strategies, and then the less appropriate TDM strategies are chosen in order to fill in the
advising spaces.
A computer tool used to develop the adviser system is by using open source software called
Kappa-PC version 2.4. The first step is to create hierarchy from TDM strategies towards specific
231
objectives, main purpose, choice of area and the final three advices respectively. Kappa-PC provides the
developer with the flexibility and power required in supporting complex applications (Intellicorp. Inc,
1997). It allows the knowledge engineer to develop prototype using hybrid knowledge representation
technique.
RESULT ANALYSIS
System Architecture and Application of E-ASSIST
The expert advisory system E-ASSIST is developed for young and inexperienced transport
engineers and planners, who are involved in transportation demand management (TDM) strategy planning
and implementation. Such system is able to provide them with reliable advices in order to effectively
decide TDM strategies during the planning and implementation phases. The object browser window in
Figure.3 represents the top-level hierarchical in the object model developed for the E-ASSITS domain.
The class E-ASSIST represents the overall prototype that is comprised of nine major subclasses.
Figure 2: Top-level object hierarchy of the E-ASSIST
The module names in Figure. 2 above are boxed to indicate that their subclasses are not shown.
Example of object hierarchy of subclasses, which are components of different types of major categories of
TDM; improved transport options, incentives to use alternative modes and reduce driving, and land use
management as shown in Fig.4 for improved transport options. Total are 27 hierarchies in subclasses
according the main objective of TDM Strategies.
232
Figure 3: Example of Hierarchical architecture of improved transport options for congestion reduction
Illustration Example of E-ASSIST Domain
As an illustration of typical object representation in the E-ASSIST domain, the hierarchy of

CRimprovedtransportoption class which is a descendant of the CongestionReduction class is shown in
Fig.4.
Figure 4: Object hierarchy of CRimprovedtransportoption class
233
CRITOLargeUrban class is geographical area entity that represents appropriate condition of area
need to be implementing TDM strategies. It contains the slots: (i) PublicTransportImprovement1,which
indicates the strategy on Public Transit (also called Mass-Transit) includes various services using shared
vehicles to provide mobility to the public; (ii) NonMotorTransportPlan1, which indicates Non-motorized
Transportation (also known as Active Transportation and Human Powered Transportation) includes
walking and bicycling, and variants such as Small-Wheeled Transport (skates, skateboards, push scooters
and hand carts) and Wheelchair travel; (iii) Ridesharing1, which refers to carpooling and vanpooling (the
term is sometimes also applied to public transit, particularly commuter express bus), in which vehicles
carry additional passengers. Carpooling uses participants own automobiles. Vanpooling usually uses
rented vans (often supplied by employers, non-profit organizations or government agencies). Most
vanpools are self-supporting operating costs are divided among members. Vanpooling is particularly
suitable for longer commutes. These slots are inherited from the parent class CRimprovedtransportoption
which is a descendant of the CongestionReduction class.
Knowledge Based Modules of E-ASSIST
The E-ASSIST knowledge base is made up of nine class module as mentioned in previous
section, each deal with specific strategy of TDM domain. First, launch the E-ASSIST by opening the
Kappa-PC software and open the E-ASSIST file (.kal) continued by open the session to start the expert
advisory system. After the program is loaded, the general information interface appears (Fig. 5) in the
center of the displaying area if the computer has resolution of 1366x768 pixel (32 bits).
Figure 5: General Information interface
By clicking on the OK button closes the general information interface and invokes the main E-
ASSIST window, as shown in Figure 6. Two way communications between user and E-ASSIST has been
started in this interface.
234
Figure 6: Main interface
Fig.6 shows the E-ASSIST main menu from which the advisory system for TDM can be accessed
with the help of appropriate button. The main menu screen also has e-library and ABOUT button to give
information about all TDM strategies used in this system and the system information in E-ASSIST.
Advisory System Module
Since the E-ASSIST system used in this project based on the objective to implement of TDM as
the input, clicking the on the Advisory System button leads the user to nine main objective of
implementing TDM, as shown in Fig.7 below. There are nine main objectives associated in this interface
as inputs define. Select the one that matches the problem the user encountered in current situation in order
to see the full description about the objective, in this case Energy Conservation and Emission Reduction.
Figure 7: TDM main objective interface
235
Example of this situation is when the new transport planner facing with the pollution problem
in one area triggered by heavy vehicle such as trucks. To solve this issue, the user needs to select Energy
Conservation and Emission Reduction as the input in line with the objective of this strategy to reducing
vehicle energy consumption and pollution emission. After the selection, the description of this TDM
strategy is appeared on the specific box allocated. If user sure and confirmed with the input (correct main
objective) then clicking on the NEXT button calls up the viewing window that displays the Specific Aim to
Achieve the Main Objective interface, as show in Fig.8. If the user cannot confirm the description is
his/her objective to implement TDM then it is necessary to click BACK button, click on the other main
objective as the input, read the TDM main objective description, and do the same step as before, until the
actual main objective is found and confirmed. Once interface of Specific Aim appeared, clicking on the
bullet button to see the specific aim description to implement the TDM on how to achieve the main
objective as shown in previous step. This dialog window is designed to ensure the user choose appropriate
input to avoid wrong advise in appropriate strategy.
In this case, user choose Incentives to use Alternative Modes and Reducing Driving as the
Specific Aim to achieve main objective which is describe on the box it can give commuters resources and
incentives to reduce their automobile trips by providing more efficient alternative modes of travel. Again,
if user sure and confirmed with the selection, then clicking on the NEXT button calls up the viewing
window that displays the GEOGRAPHICAL AREA interface as shown in Fig.9.
Figure 8: TDM specific aim interface
236
Figure 9: Geographical Area interface
After selection of specific objective, the size of study area or area that want to be implement are
listed on the left of interface window. The geographical area is become important in this step since all
TDM strategies in particularly only can be effective in certain area in term of size and population. If the
user cannot expect the size of their study area or not sure about the geographical condition in particular
area, then they can refer to the description box provided. If the size is identified and selected; for example
TOWN, then by click NEXT button cause calls up the viewing window that displays the ADVISE ON
TDM STRATEGIES as shown in Fig.10 below.
Figure 10: Advised TDM strategies result interface
237
Fig.10 shows the advice strategy window. It is display three appropriate strategies which is the
most suitable TDM strategies to be implemented based on the input selection. The advice system
considers three strategies for user to judging and makes decision of suitable TDM strategy. Description
for all three strategies also provided in box dialog to help and guide the user. The user also can click the
VIEW IMPACT button and ADVISE STRATEGY IMPACT SUMMARY will display on the interface as
shown on Fig.11 below.
Figure 11: Impact summary interfaces
Figure 11 above shows the advice strategy for fuel tax as an example. Result shows that travel
impact in term of reducing total traffic significantly as rating 2 was shown (beneficial). On benefit
summary, environmental protecting given rating3 which is significant reduce in fuel use and related
pollutants. Advice strategy summary shows the travel impact, benefit and equity if the strategy chosen
will be implemented. Based on the travel impact summary, user can easily to identified how effective
reduction of trip or total traffic which is stated in the summary using the rating scale; rating from 3 (very
beneficial) to -3 (very harmful); 0 indicates no impact or mixed impact on the objective if user implement
this strategy. In this case, from the beginning user has been identified the objective was Energy Efficiency
with specific aim is Incentives to use Alternative Modes and Reducing Driving; implementing in TOWN.
As result mentioned, three advises were given according to the input selected such as DISTANCE
BASED-PRICING, FUEL TAX and FINANCIAL COMMUTER INCENTIVE.
If the user can confirm this strategy is the most suitable strategy then it is necessary to clicking
PRINT button to viewing window spawns a Notepad text editor with analysis result (rating scale). Once
finishing the use of E-ASSIST, user can click on the CLOSE WINDOW button on the right bottom
interface to end the system.
238
Validation and Evaluation
The prototype testing or validation is the process of evaluating the performance and utility of the
program and revising it as necessary. Testing the prototype also involves looking at the ease of handling
the user interface and understanding the explanatory text (Basri, 2000). Generally, the performances of
the E-ASSIST were comprehensive and satisfactory to the authors judgment and evaluator who is expert
in transportation field. The evaluators are different people (not domain expert) who have same
qualification with the domain expert. The evaluation results demonstrated that the expert decision
compared well with the E-ASSIST output. This is a positive indication of the accuracy of the knowledge
base of the E-ASSIST. The user friendliness of E-ASSIST was the assessed to be satisfactory as a
working prototype.
CONCLUSION AND RECOMMENDATION
The expert system developed is user friendly, and it does not require the user to have any
programming knowledge to use it. The user can smoothly navigate through nine different input screens
before the system outputs its conclusion. Without referring to any TDM expert, the system can help
young and inexperienced engineers and planners in their decision making. In addition, such systems are
less costly to use and saves time.
Moreover, the successful development of an expert advisory system for implementations of TDM
(E-ASSIST) demonstrates that the application of knowledge-based expert system in this domain is
promising. Frame-based knowledge representation method in conjunction with the power of object-
oriented programming enables the knowledge to be abstracted and represented for the computer to
manipulate efficiently and for the knowledge engineer to update the knowledge base with great flexibility.
Since the knowledge in this domain is dynamically growing, the E-ASSIST is by no means complete at
this stage.
REFERENCES
Broaddus. A, Litman. T & Menon. G (2009), Training Document On "Transportation Demand

Management, Sustainable Urban Transport Project (www.sutp.org) and GTZ (www.gtz.de).
Basri, H. (2000). An expert system for landfill leachate management. Environmental Technology, Vol.21,
pp 157-166, Elsevier Science Ltd, London.
ElAlfi. A.E.E & El Alami. M.E (2009). Intelligent Advisory System for Supporting University Managers
in Law. (IJCSIS) International Journal of Computer Science and Information Security, Vol. 3, No. 1,
2009.
Frame. G (2010). Traffic Management and Transport Demand Management. Distance Learning Course
in Urban Planning, Chapter 4, pp 1-19.
Forslund. G (1995). Toward Cooperative Advice-Giving Systems. IEEE Expert: Intelligent Systems and
Their Applications. Volume 10, Issue 4. Pp 56-62. IEEE Educational Activities Department
Piscataway, NJ, USA
Gregg. G.D & Walczak. S (2006). Auction Advisor: an Agent-Based Online-Auction Decision Support
System. Decision Support Systems 41 (2006) 449 471. Elsevier
IntelliCorp, Inc. 1997. Kappa-PC Ver. 2.4 User Manual. CA: IntelliCorp Inc.
239
Litman. T (2009). Are Vehicle Travel Reduction Targets Justified? Evaluating Mobility Management
Policy Objectives Such As Targets to Reduce VMT And Increase Use Of Alternative Modes .
Victoria Transport Policy Institute. (www.vtpi.org); at www.vtpi.org/vmt_red.pdf.
Mahmood. M, Bashar M.A, Akhter. S (2009). Traffic Management System and Travel Demand
Management (TDM) Strategies: Suggestions for Urban Cities in Bangladesh. Asian Journal of
Management and Humanity Science, Vol. 4, No. 2-3, pp. 161-178, 2009. Pp 161-179
Mahmood. M, Bashar M.A, Akhter. S (2009). Travel Demand Management (TDM) For Improved
Transportation System: Lessons for Bangladesh. Journal of Quality and Technology Management
Volume IV, Issue I1, Dec, 2008, pg. 26 - 43
Nazrul Islam (2005). An Expert System for Mix Design of High Performance Concrete. Journal of
Advances in Engineering Software. Volume 36 , Issue 5. Pp. 325 337. ISSN: 0965-9978. Elsevier
Science Ltd. Oxford, UK
Papaioannou.P & Georgiou.G (2003). Are Traffic Demand Management (TDM) measures a solution for
traffic problems in urban areas? The Greek experience. Transport and Management Section,
Department of Civil Engineering, Aristotle University of Thessaloniki.
Turban. E & Aronson. J (2001). Decision Support Systems and Intelligent Systems. 5th Edition.
ISBN:0137409370. Prentice Hall
VTPI (2005). Why Manage Transportation Demand? Retrieved from http://www.vtpi.org.
Wilson. C & Shirazi. B (1991). Transportation Demand Management: Implications of Recent Behavioral
Research. UCTC No. 29. The University of California Transportation Center, University of
California at Berkeley, USA.
240
TIMES SPENT IN DRIVING TAXI

A CASE STUDY IN PUTRAJAYA SENTRAL
Chuah Pooi Yee 1, Sulistyo Arintono 2* and Hussain Hamid2
ABSTRACT:
Taxi drivers normally spend times for driving while searching or waiting at taxi stands to get passengers,
driving with passengers onboard, and driving empty returning to the base. So far, there is no detailed
study in Malaysia regarding the amount of times spent in driving taxi. This research aims to fill the gap.
Data for analysis is collected in Putrajaya Sentral, an Express Rail Link station where any taxi can bring
passenger(s) in or out of the station and passenger needs to register at the taxi counter to get the service.
Vehicle registration number, times of arrival and departure of all taxis were recorded the whole day
during weekday and weekend. Results of the analysis suggest that the average taxi drivers waiting time is
just under 30 minutes in both weekday and weekend, and the average cycle times range from 51 minutes
during weekday to 85 minutes during weekend. This operating condition is better than at the Low Cost
Carrier Terminal where an exclusive airport taxi operation is applied. In Putrajaya Sentral, it is worth for
taxi drivers to wait if the existing number of vehicles in the queue does not exceed 10 or 8 during
weekday and weekend, respectively.
Keywords: Taxi Drivers Waiting Time, Taxi Operation, Non-Exclusive Taxi Service
1
Student, Department of Civil Engineering, Faculty of Engineering, Universiti Putra Malaysia, 43400 UPM
Serdang, Selangor, Malaysia.
2
Lecturer, Department of Civil Engineering, Faculty of Engineering, Universiti Putra Malaysia, 43400 UPM
Serdang, Selangor, Malaysia.
*
Correspondence Author: Sulistyo Arintono, Universiti Putra Malaysia. Tel: +603 8946 6362, Fax: +603 8656
7129, E-mail: sulistyo@eng.upm.edu.my
241
INTRODUCTION
Taxi drivers normally spend much time waiting at taxi stands in various parts of the city or
cruising the streets to get customers to convey to their destinations. How much waiting or cruising times
can be considered as being normal or excessive depend on individual drivers that they finally make a
decision either to leave or stay in that particular mode. One way to judge whether taxi drivers waiting
time is already excessive or not is by comparing to the total drivers working time or total driving time
carrying passengers onboard.
In Malaysia, there is limited data regarding taxi operating characteristics. Therefore it is not
possible to assess how good or bad driving taxi is in terms of the waiting time relative to the other times
spent in doing the job. The following research with data collection conducted in Putrajaya Sentral, one of
the Park and Ride facilities for users of the ERL (Express Rail Link) services, is to fill this gap in the
transportation system database. It is expected that taxi drivers and the government authorities in charge in
taxi management can make use of the recommendations to improve the taxi operating environment.
METHODOLOGY
Most taxi drivers rely on waiting at stands in various locations in the city (shopping malls, hotels,
markets, bus or train stations) to get customers, rather than cruising the streets for the same purpose. It is
more beneficial in terms that, at least, it saves gasoline and human energy, reducing the road traffic,
minimizing pollution, etc. Similar situation prevails in Malaysia.
It is noticed that some taxi drivers keep coming to the same stand almost after every trip carrying
customers to their destinations. The favourite taxi stands vary among drivers ranging from university
students college, commuter station, shopping centres, etc. If sufficient data can be collected from these
drivers regarding their arrival and departure times at the particular stands, then it is possible to do the
analysis on taxi drivers waiting time, driving time and total working time. This method can be
considered more preferable than data collection through interview of taxi drivers where the sample size
needs to be large enough while the honesty of the answers may still be questionable.
Data collection regarding taxi operating characteristics was conducted in Putrajaya Sentral.
Surveyors were located in front of the taxi counter where customers purchase the tickets and get the
intended services from the taxis queuing at the dispatching bay. The data consisting of vehicle registration
number, arrival time (with or without passengers) and departure time (with passengers) were manually
recorded for three days comprising two weekdays and one weekend, from 06:00 until 23:00 during which
time most of the trips occurred. The data analysis consists of drivers waiting time (time difference
between arrival and departure with passengers), cycle time (time difference between two consecutive
arrivals or departures), driving time per cycle (time difference between departure with passengers and the
next arrival time after conveying the passengers), number of trips per vehicle per day, and queue length.
Comparing various drivers times in taxi operation will enable us to judge whether drivers waiting time
is considered normal or excessive.
The operation of taxi in Putrajaya Sentral is a free for all system. It means any taxi can bring
passengers in or out from the station. This is different from the system applied in LCCT (Low Cost
Carrier Terminal) within KLIA (Kuala Lumpur International Airport) where only those registered as
airport taxi are allowed to carry passengers out of the airport. The outsiders (non-airport taxi) may bring
passengers into but not out from the airport. The different regulation in taxi operation governs the
behaviour of the taxi drivers in maximising their potential profit.
242
Taxi Operating Characteristics
To find potential passengers taxi drivers can either cruising the streets or come to places where
there are demands for the service, e.g. shopping malls, rail or bus stations, hotels, etc. When cruising the
streets is selected they face an uncertainty of whether they will or will not find customer(s) within a
reasonable time or travelled distance. Meanwhile, the other option, waiting for customer(s) at designated
locations, spends little or no gasoline in addition to the waiting time which could be comparable to or
longer than the cruising time. Some taxi drivers have their own favourite locations to wait for their
customers and keep coming to the same places after every trip conveying passengers to their destinations.
In previous studies involving taxi operation in Selangor (Chuah, 2009 and Tan, 2008), it is
evidenced that some taxi drivers keep coming to their favourite waiting locations (train station or
residential college for example). It seems that they rely on getting passengers from those points as they
may feel convenient. Similar behaviour is also practiced in the case motorcycle taxi (Utama, 2003 and
Dewi, 2005) and human-powered tricycle in Indonesia (Kusumaningsih, 2003).
In certain places like shopping malls, hotels, etc. there are parking spaces specifically designed
for taxi. There could be four or five spaces provided and taxi drivers normally arrange themselves to serve
customers in a first come first served basis. There is no fee charged for carrying passengers out of the
premises. The fare to be paid by customers can be either based on the meter (official fare level) or
negotiable depending on the agreement between the two parties (customer and taxi driver). In most cases,
especially in developing countries, taxi drivers prefer to negotiate (and refuse to use the meter instead)
where they can charge higher than the meter-based fare level, blaming the long waiting time to get
customers from the premises. The customers on the other hand, have the options of either accepting the
offer after several attempts to bargain or they have to walk down the street nearby and try to flag down a
passing empty taxi (if any), hoping to get a slightly cheaper deal. The most common outcome is that
customers agree to take the taxi with the meter off, given that they carry heavy or large luggage and doubt
that no taxi can be found nearby if the location is somewhat remote.
Some taxi drivers may have waited long enough and finally decide to leave the queue for some
reasons. First, they conclude that the waiting time is too long for which they prefer to take an alternative
action, e.g. move to other waiting place, hoping that on the way there he will meet a customer by the road
side. Second, they receive phone call from a known customer who needs a taxi service. And third, if it is
close to their off-time, they will decide to stop operation for that shift. Whatever the underlying reason is,
they must have their own judgement on how to adjust the waiting time such that it is still within
reasonable limits. This paper will try to discuss on how the adjustment could be made and what action a
driver can make when confronted to a given situation.
Demand for Taxi Services
The ERL service between KL and KLIA ceases operation around midnight daily. So the taxi
counter at Putrajaya Central opens until a few minutes after the last train departs from the station.
Accordingly, it opens early in the morning before the first train arrives, to give service to the needy
customers. The demand for taxi service fluctuates during the day. Table 1 shows average demand as
derived from the observation survey.
243
Table 1: Demand for Taxi Service in Putrajaya Sentral
Time Number of Departure Headway (min)

Interval Weekday Weekend Weekday Weekend
06-07 11 2 5.45 30.00
07-08 36 3 1.67 20.00
08-09 80 4 0.75 15.00
09-10 55 7 1.09 8.57
10-11 36 5 1.67 12.00
11-12 26 10 2.31 6.00
12-13 24 13 2.50 4.62
13-14 21 12 2.86 5.00
14-15 22 9 2.73 6.67
15-16 22 13 2.73 4.62
16-17 15 15 4.00 4.00
17-18 15 20 4.00 3.00
18-19 22 28 2.73 2.14
19-20 17 27 3.53 2.22
20-21 11 11 5.45 5.45
21-22 18 19 3.33 3.16
22-23 7 8 8.57 7.50
Total 438 206
Average 25.76 12.12 2.33 4.95
As seen from Table 1, during normal working day, the peak demand starts from 07:00 and lasts
until 11:00 (with a clear peak at 08:00-09:00) after which the demand becomes relatively constant until
22:00 when it starts to fall closer to midnight, and a total demand of 438 departures. During weekend on
the other hand, as expected, the higher demand comes later which starts from 11:00 and lasts until 22:00,
having its peak around 17:00-20:00. For practical purpose, calculated based on Table 1, it can be assumed
that, the average headway at peak time weekday (08:00-11:00) is 1.16 minutes, and that for the off-peak
time is 3.10 minutes. Similar calculation for the average headways during weekend are 3.73 and and
12.41 minutes for the higher and lower demand period, respectively.
Drivers Waiting Time
In the free for all system, taxi drivers need to check the length of the queue at the time they
enter the facility, recall the average waiting time they ever experienced previously, calculate the
anticipated waiting time at the current condition, and then decide whether they will leave (if the waiting
time is too long, to get passengers from somewhere else) or join the queue (if the waiting time is still
acceptable). In the exclusive airport taxi system on the other hand, drivers do not have any other option,
since the only way for them to get customers is from the airport. So, no matter how long the current queue
is, they will certainly join the queue and wait for their turn to convey passengers out of the airport. As a
result, the average drivers waiting time for the exclusive airport taxi system will be much longer than in
the free for all system. Table 2 shows the average waiting time for taxis operating in Putrajaya Sentral.
As seen from Table 2, in all cases (weekday and weekend), drivers waiting time varies between
10 minutes or less to slightly over one hour with an average of just under 30 minutes, well below the
average drivers waiting time at LCCT (123 to 185 minutes following Soo, 2009). The statistics is quite
convincing due to the high sample size. In order to assess whether this amount of waiting time is
reasonable or not it needs to compare with the other times involved in driving taxi as will be discussed in
the next section.
244
Table 2 Taxi Drivers Waiting Time
Statistics Weekday Weekend

Maximum (min) 71 74
Minimum (min) 10 6
Average (min) 26 29
Standard deviation (min) 15 12
Sample size 896 189
Cycle Time and Driving Time
The cycle time for taxi operation can be defined as time difference between 2 consecutive
arrivals to or departures from a particular taxi stand of the same vehicle. Another important item is
driving time defined as time difference between departure from and the next arrival to the same taxi
stand of the same vehicle. Given the cycle-time and driving-time, assuming that the taxi drives back to
the stand right after delivering passengers to their destinations, the utilized time (defined as the time
spent by taxi drivers with passengers onboard) can also be calculated. These cycle-time and utilized time
are useful for the assessment of the times involved in taxi driving.
Based on the information collected during the survey, i.e. vehicle registration number, arrival
time and departure time carrying passengers, the cycle-times (combined arrival-based and departure-
based) and driving-time were calculated and presented in Table 3.
Table 3: Cycle Time and Driving Time (hh:mm:ss)
Statistics Cycle Time Driving Time

Weekday Weekend Weekday Weekend
Maximum 2:07:00 3:35:00 1:11:00 3:30:00
Minimum 0:10:00 0:17:00 0:10:00 0:10:00
Average 0:50:58 1:24:52 0:26:18 1:06:14
Standard deviation 0:25:55 0:45:46 0:15:17 0:50:40
Sample size 972 234 448 111
From Table 3 it is evidenced that the average driving times are 26 and 66 minutes for the
weekday and weekend, respectively, against the average cycle times of 51 and 85 minutes for the
weekday and weekend, respectively. In other words, on average, drivers spent 25 minutes waiting (equal
to 49.01% of the cycle time) during weekdays and 28 minutes (22.35%) during weekend. It shows that the
operating condition is better during weekend than weekday. This is due to the less number of taxis
operating during weekend (47 compared to 122 during weekday), though the demand is also lower. The
longer average driving time during weekend compared to weekday also suggest that taxi passengers travel
farther distance during weekend, beyond the locations of the government office complex, which is also
reasonable.
Driving taxi involves three different modes within a cycle, e.g. driving with passengers onboard,
driving without passengers to the next waiting point, and waiting. Ideally, the distance travelled or the
time spent driving with passengers should take the highest portion compared to the other two modes
(driving empty and waiting). However, in any case, it may be hard for most drivers to achieve at least
50% driving with passengers. What could be reasonably expected are probably equal shares among the
245
three modes, one-third each for the driving with passengers, driving without passengers, and waiting.
Some drivers may sacrifice more time for waiting than the others. But, logically, the waiting time should
be no more than 50% of the cycle time, which could be considered as the maximum waiting time. As
such, taxi drivers operating in Putrajaya Sentral (49% of the cycle time spent as waiting during weekday)
are actually working between favourable and unfavourable conditions. Therefore, there is a need to
properly decide whether to join or not to join the queue.
The utilized time (time spent with passengers onboard) is roughly half of the driving time (in
practice drivers spent slightly less time to get back to the taxi stand than on the outbound travel carrying
passengers). From Table 3 it can be concluded that the average utilized time is 13 and 33 minutes for the
weekdays and weekend, respectively. Again, when proportioned to the cycle time, they come to merely
25.49% and 38.82% during weekday and weekend, respectively, clearly a low rate of utilization. This
means, most of the time (during weekday), drivers fail to achieve equal shares among the three modes,
and only during weekend is this possible.
Queuing Decision
Upon arrival at Putrajaya Sentral, with or without passengers, a taxi driver needs to decide
whether he likes to join or not to join the queue. The decision is made based on the length of the existing
queue (how many taxis already queuing) and the average headway between departures. Using Tables 1, 2
and 3 it is possible to estimate the expected waiting time for a particular taxi arriving at the stand at
certain time of a day, given the existing queue length. If the expected waiting time will be more than 25
minutes during weekday (50% of the cycle time), then it is not worth waiting. Similarly, during weekend
the cut off point is around 42 minutes. Table 4 shows the recommended guideline for taxi drivers, which
shows that if the existing queue exceeds 10 vehicles during weekday or 8 vehicles during weekend, it will
be better not to join the queue, otherwise the waiting time to get passengers will be excessive.
Table 4 Maximum Queue to Join
Weekday Weekend
Cycle time (min) 51.00 85.00
Max waiting time (min) 25.50 42.50
Average headway (min) 2.33 4.95
Maximum queue 10.94 8.59
CONCLUSIONS
The times spent in driving taxi consist of driving with passenger(s) onboard, driving without
passengers while cruising for passenger searching, and waiting at taxi stands. Mostly, the waiting time is
substantial and, consequently, the driving time with passenger(s) onboard is minimum. Therefore, taxi
drivers need to make decision whether they like to wait or not based on the existing length of queue in a
particular taxi stand.
In Putrajaya Sentral, the average waiting time is just under 30 minutes during weekday or
weekend for the average cycle time of 51 and 85 minutes for weekday and weekend, respectively. Given
the average demand of taxi service at the station and the expected waiting time of less than 50% of the
cycle time, taxi drivers may find it worth to join the queue if there are no more than 10 or 8 vehicles
already at the queue during weekday or weekend, respectively.
246
REFERENCES
Chuah Pooi Yee (2009), Analysis of taxi drivers waiting time case study at Putrajaya Sentral,
Unpublished Bachelor of Engineering (Civil) Project Report, Department of Civil Engineering,
Faculty of Engineering, Universiti Putra Malaysia, Malaysia.
Dewi, E. Y. (2005), Performance analysis of motorcycle taxi: Case study at Pangkalan Tanjung Senang
(Analisis kinerja ojek: Studi kasus di Pangkalan Tanjung Senang), Unpublished Bachelor of
Engineering (Civil) Project Report, Department of Civil Engineering, Faculty of Engineering,
University of Lampung, Indonesia.
Kusumaningsih, E. (2003), The operating characteristics of human-powered tricycle in Bandar Lampung,

Case study in Jalan Pagar Alam (Karakteristik operasi becak di Bandar Lampung: Studi kasus di
Jalan Pagar Alam), Unpublished Bachelor of Engineering (Civil) Project Report, Department of Civil
Engineering, Faculty of Engineering, University of Lampung, Indonesia.
Soo Wan Yin (2009), Analysis of the taxi drivers waiting time case study at LCCT (Low Cost Carrier
Terminal), Unpublished Bachelor of Engineering (Civil) Project Report, Department of Civil
Engineering, Faculty of Engineering, Universiti Putra Malaysia, Malaysia.
Tan Lei Chai (2008), Cost revenue analysis of taxi operation: Case studies in Serdang and Seri
Kembangan, Unpublished Bachelor of Engineering (Civil) Project Report, Department of Civil
Engineering, Faculty of Engineering, Universiti Putra Malaysia, Malaysia.
Utama, E. (2003), Performance Analysis of motorcycle taxi: Case studies in Unila and Jalan
Purnawirawan (Analisis kinerja ojek: Studi kasus di Unila dan Jalan Purnawirawan), Unpublished
Bachelor of Engineering (Civil) Project Report, Department of Civil Engineering, Faculty of
Engineering, University of Lampung, Indonesia.
247
248
ESTIMATION OF DISPERSION OF CARBON MONOXIDE (CO), NITROGEN DIOXIDES

(NO 2 ), AND CARBON DIOXIDE (CO 2 ) FROM PORT KLANG - KLIA ROAD: PRELIMINARY
FINDINGS
Nur Bazla Mohmed Yusof 1, Mohamad Firdaus Ibrahim 2,

Ruslan Hassan 3
ABSTRACT:
Criteria pollutants and greenhouse gases such as carbon monoxide (CO), nitrogen dioxides (NO 2 ), and
carbon dioxide (CO 2 ) are usually the main gases monitored in air quality measurement. The increase of
their concentrations year by year has become a source of concern. The main sources of these gases
identified by Department of Environment (DOE) are due to mobile sources or the transportation sector
contributing more than half of the total percentage of emission load. The purpose of this study is to
estimate the concentration of CO, NOx and CO 2 from the mobile sources by using fuel-based methods
estimation and Gaussian line source dispersion equation. The 64-km stretch of road from Port Klang to
KLIA, Selangor was chosen as a case study. Three-day traffic volume composition study was also
conducted. Finally, the estimated results were then compared with ambient concentration data from two
nearest monitoring stations at Shah Alam and Petaling Jaya, Selangor, Malaysia.
Keywords: Carbon Monoxide (CO), Nitrogen Dioxides (NOx), Carbon Dioxide (CO 2 ), Gaussian Line-
Source Dispersion Model
1
Graduate Student, Malaysia Institute of Transport, Universiti Teknologi MARA, 40450 Shah Alam, Selangor, MALAYSIA
2
Graduate Student, Faculty of Civil Engineering, Universiti Teknologi MARA, 40450 Shah Alam, Selangor, MALAYSIA
3
Professor, Faculty of Civil Engineering Universiti Teknologi MARA, 40450 Shah Alam, Selangor, MALAYSIA
249
1.0 INTRODUCTION
In Malaysia, rapid increase in population in urbanized areas has triggered the development
phase as well, in order to meet the current demand, thus enhancing the economic growth together
with the living standard of the urban community. Modern society is highly dependent on motorized
transportation such as cars, trucks, and railways. Movement of people and goods require energy
which relies mostly on burning of fossil fuels, thus causing emissions with adverse local and global
effects. The emission emitted from various kinds of vehicles comprises many major predominant
gases which will contribute to the air pollution either locally or globally. Carbon dioxide is the most
significant gas produced from the transportation activities and the effect brought by it is attributed to
the global warming phenomenon if adequate controls or mitigation measures are not taken.
The significant increase of vehicles movement in Malaysia has contributed to an increase of

Green-House Gases. There has been an increase in the level of environmental impact which has
contributed to the unstable pattern of weather around the globe and variations in the natural
phenomena such as floods, droughts and the disappearance of glaciers. Malaysia has itself released
of GHG 37.2 tonnes of CO 2 in 2000 and there has been noticeable change in weather pattern, for
example, the flood event in Johore in December-January 2007.
2.0 GREEN-HOUSE GASES
Green house gases are gases in the atmosphere that absorb radiation within the thermal infrared
range causing the greenhouse effect. Carbon dioxides and nitrous oxides are known as principal
greenhouse gases. Even though carbon monoxide is not classified under green house gases, it is one
of major criteria pollutants that is generated from transportation sector .Moreover it is formed when
carbonaceous fuels are burned under less than ideal condition resulting in CO instead of CO 2 (Gilbert
M.M, 2008). Complete fuel combustion produces CO 2 while incomplete fuel combustion will
produce CO. The use of catalytic converters in motor vehicles however generates CO 2 .
Nitrogen oxides form when fuel is burned at high temperatures, as in the combustion process
of the internal combustion engine. One member of the NOx group, nitrous oxide, is a greenhouse gas.
It accumulates in the atmosphere with other greenhouse gasses causing a gradual rise in the earth's
temperature, a phenomenon known as global warming (Jamal, 2004).
Malaysias emission of greenhouse gases (GHG) totalled 144 million tonnes of CO2 in 1994.
Net emissions, after accounting for sinks, totalled 76 million tonnes. On a per capita basis, the net
emissions were equivalent to 3.7 tonnes. The CO 2 emissions from final energy use (excluding
electricity) by various activities of the economy indicate that transportation contributed 49 per cent,
industries, 41 per cent, residential and commercial activities, 7 per cent, and agriculture, 3 per cent of
the overall emissions.
250
3.0 FIELD STUDY
The KLIA-Port Klang road was selected for the field study, being one of the busiest roads and
considered to be generating GHG from a heavy traffic. The availability of the DOE monitoring
stations within the catchment can help in providing essential information on the ambient air
conditions. This 64-km stretch of road will be therefore be used as a representative of a typical road
connecting a marine port and air port, heavily dominated by freight transports over a 24-hour period 7
days a week.
3.1 ROAD DESCRIPTION
The route, Jalan Klang-Banting which connects KLIA to Port Klang is chosen as a case
study. The total distance of this road is 64 km. This Port Klang KLIA route is a double
carriageway with two lanes at each roadway with exception at some intersections where the
numbers of lanes for the roadway are between 3 to 6 lanes. From the observation that had been
made, towards the freight movement, most logistics companies used this route to travel from
Port Klang to KLIA for delivery/supplying goods to KLIA. In addition, other motor vehicles are
plying this road as well because of its preferred short distance causing this route to be
congested at most times.
This route is the most preferable for the traffic as it had the shortest distance to KLIA
from North Port or South Port. The high traffic movement is directly proportional to the
tremendous release of pollutants to the surrounding environment. Moreover, this road has been
upgraded to cater the growth and expansion of urbanisation and the logistics sector. The map
below shows the route from Port Klang to KLIA.
Figure 3.0: Port Klang- KLIA Map
251
3.2 VEHICLE VOLUMETRIC STUDIES
For the purpose of emission prediction from vehicles, a 3- days traffic volume study
was carried out at the convergence junction between Northport and Westport. Traffic
counting included mainly Tally Sheet model to ensure every surveyor recorded 3 types of
vehicles at the same time.
Field motor vehicles counts were continuously monitored for different categories of
vehicles (i.e. passenger cars, vans, lorries and heavy trucks) for each direction. Every surveyor
was provided with a manual counter (mechanical hand tally counter), field data forms and a
stop watch. Traffic counting data are used for prediction of emissions released from the motor
vehicle which was carried out by fuel-based methods. Table 3.1 below is the composition of
vehicles by days.
Figure 3.1: Graph of number of vehicles composition by day
3.3 SOURCES OF GHG
Greenhouse gas emissions from industry, transportation and agriculture are very likely
the main cause of recently observed global warming. In 2008, it was estimated that the
combined pollutant emission load for carbon monoxide (CO) and nitrogen oxides (NOx) were
1,451,746 metric tonnes and 409,972 metric tonnes, respectively. From this overall emission
load, the highest contributor of CO and NOx was motor vehicles 97.1% and 49%,
respectively (DOE, 2008).The estimation of the emissions loading are made by fuel-based
method whereby the consumption of fuel by different types of vehicles for every kilometre is
used for estimation. The emissions predictions are carried out by multiplying the emissions
252
rates per unit vehicles with the volume of traffic in accordance to types of vehicles and with
the stretch of 64 km of the distance covered for the study which is from Port Klang to KLIA.
The emission rate for each pollutant is calculated based on US EPA emission rate
for each category of vehicle. The amount of CO, NOx and CO 2 emitted along the 64 km
stretch by different types of vehicles are shown in Table 3.0(a), Table 3.0(b), and Table 3.0(c).
253
Table 3.0(a): Average emissions and Fuel Consumption for Cars and Vans
Emission Rate and Fuel

Emissions Emissions (lb)
Consumption Kilometres
Components Calculation Per unit
Travel (km) Traffic Volume
Per mile* Per Km vehicles (lb)
Day 1 Day 2 Day3
Carbon Monoxide 20.90g 12.98g 64 12.98g/km X 64km X 1lb/454g 1.829 23,755.05 20,943.88 20,012.92
Oxides of Nitrogen 1.39g 0.86g 64 0.86g/km X 64km X 1 lb/454g 0.121 1,571.55 1,385.57 1,323.98
Carbon Dioxide 0.916lb 0.569lb 64 0.569Ib X 64km 36.416 472,971.00 416,999.62 398,463.87
Source : USEPA, National Vehicle and Fuel Emissions Laboratory
Table 3.0 (b): Average emissions and Fuel Consumption for Light Trucks

Day 1 Day 2 Day3
Carbon Monoxide 27.70g 17.21g 64 17.21g/km X 64km X 1 lb/454g 2.426 5,368.74 3,243.56 6,836.47
Oxides of Nitrogen 1.81g 1.12g 64 1.12g/km X 64km X 1 lb/454g 0.158 349.65 211.25 445.24
Source : USEPA, National Vehicle and Fuel Emissions Laboratory
Table 3.0(c): Average emissions and Fuel Consumption for Heavy Trucks

Day 1 Day 2 Day3
Carbon Monoxide 19.78g 12.29g 64 12.29g/km X 64km X 1 lb/454g 1.733 5,861.00 4,460.74 5,472.81
Oxides of Nitrogen 11.22g 6.972g 64 6.972g/km X 64km X 1 lb/454g 0.983 3,324.51 2,530.24 3,104.31
Source: International Fall Fuel and Lubricants

3.4 MONITORING STATIONS
Ambient air pollutants concentration recorded at two nearest stations which is Shah
Alam and Petaling Jaya stations were used as a comparison with the emission prediction along
the road.
Table 3.1: Average CO and NOx concentrations at Shah Alam and Petaling Jaya stations
Parameters Station Average 1-hr

concentration (ppm)
Carbon Monoxide (CO) Shah Alam 1.002
Petaling Jaya 1.853
Nitrogen Oxides (NO 2 ) Shah Alam 0.019
Petaling Jaya 0.033
Source: S.Z. Azmi et. al (2010)
Higher concentration of CO was recorded at Petaling Jaya with a mean value of 1.853
ppm compared with Shah Alam (1.002 ppm).The other gas which was also recorded at a very
high concentration at Petaling Jaya is NO 2 , which had an average concentration of 0.033 ppm
compared with Shah Alam (0.019 ppm).
4.0 GAUSSIAN DISPERSION MODEL
Gaussian dispersion model defines the amount of pollutants that might reach a receptor/ person
if the person is standing x distance from the line source. The prediction of emissions is carried in this
manner are for control. A line-source dispersion model was used for calculate the emission
concentration of every pollutants. The ground level concentration of pollutant at perpendicular
distance x from the line source can be described by the following:
Where,
C(x) = 2q q =emissions rate per unit of distance along the line (g/ms)
(2) zu z = vertical dispersion coefficient (m)
u = wind perpendicular to line source (m/s)
The calculation is based on stability class D (overcast conditions, regardless of wind speed) and
vertical dispersion co-efficient, z is determined from Table I and II (Appendix A).
255
Table 4.0: Concentration of pollutants (ppm)
C(x) = 2q / (2) z u (ppm)

Distance,
x (km) Carbon Monoxide(CO) Nitrogen Oxides (NOx) Carbon Dioxides (CO 2 )
Cars & Light Heavy Cars & Light Heavy Cars & Light Heavy
Van Trucks Trucks Van Trucks Trucks Van Trucks Trucks
0.2 0.813 1.077 0.770 0.033 0.043 0.266 10.27 12.900 36.584
0.4 0.488 0.646 0.462 0.020 0.026 0.161 6.162 7.739 21.951
0.6 0.348 0.461 0.330 0.014 0.018 0.114 4.402 5.528 15.679
0.8 0.271 0.359 0.257 0.011 0.014 0.089 3.423 4.299 12.195
1.0 0.235 0.312 0.223 0.010 0.012 0.077 2.982 3.744 10.621
Figure 4.0 represent contour iso concentration line of CO from cars and van. From the line source
(vehicle on the roadway), at distances 200m, 400m, 600m, 800m and 1000m, the estimated
concentration of CO by using Gaussian dispersion model, are 0.813 ppm, 0.488ppm, 0.348 ppm,
0.271ppm and 0.235 ppm, respectively.
Vehicle
Figure 4.0: Representative Contour for Iso Concentration line of CO from cars and van
256
5.0 DISCUSSION
From the sources (Table 3.0), emission is calculated based on emission rate and fuel base per
mile which resulted in high amount of emission because they are directly from the vehicles exhausts.
The calculated values for cars and vans, for example ranged from 20,000 23,337 lbs. The amount of
emission is spread over the entire region at some distances from the road under study. The calculated
values from the Gaussian model stipulated a range of 0.813 ppm to 0.235 ppm over distances of 200
to 1000 m.
The concentration of CO calculated by using Gaussian dispersion model shows a lower

amount compared with monitoring stations. From figure 4.0, it can be seen that the concentration of
CO is getting lower as the distance increase. The monitoring stations are located quite a distance from
the source and it did not directly measure the emission from the vehicle itself. The ambient
concentration of CO and NOx recorded at the stations are comparatively in the same scale and
concentrations. The sources can be a combination of other non-point as well as point sources in the
region.
6.0 CONCLUSION
From the sources types, it can be concluded that cars and vans emitted more emission compared
with trucks. This is because the number of cars and vans is higher than trucks on the under study
road for the three days. But if looking at single vehicle emission, trucks produced more emission by
fuel consumption.
As proven in Table 3.0, the most polluting source is carbon dioxide (CO 2 ). The CO 2 emitted by
cars and vans along the 64 km is the highest (398,463.87 lb - 472,971.00 lb). However, CO 2 emitted
by heavy trucks (333,925.02 lb - 438,746.86 lb) is nearly similar to that from cars and vans. This
shows that heavy trucks produced high emission of CO 2 since the number of them on road is
approximately 70% lesser than cars and vans.
There are several uncertainties in this study for instance the exact volume of vehicles on the road
since the traffic volume was carried out at one sampling point only. There might be a possibility that
not all vehicles went along the 64 km stretch of the road. In view of the fact that this is a preliminary
finding, further study will be carried out with the assistance of satellite use, more air monitoring
devices and instruments, together with air modeling software (CALINE-4) to provide more accurate
and better representation of concentration of pollutants derived from mobile sources.
REFERENCES
DOE (2008) Environment quality report, 2009. Department of Environment, Ministry of Science,
Technology and Environment, Malaysia, Kuala Lumpur
Gilbert M. Masters and Wendell P.ELA (2008), Introduction to Environmental Engineering and
Science, Pearson Education,Inc
Jamal H.H., Pillay M.S., Zailina H., Shamsul B.S., Sinha K., Zaman Huri Z., Khew S.L., Mazrura S.,
Ambu S., Rahimah A. and Ruzita M.S. 2004. A Study of Health Impact & Risk Assessment of
Urban Air Pollution in Klang Valley, Malaysia. Kuala Lumpur : UKM Pakarunding Sdn Bhd.
257
S .Z. Azmi, M. T. Latif , A. S. Ismail , L. Juneng, and A.A Jemain (2010) Trend and status of air
quality at three different monitoring stations in the Klang Valley, Malaysia Air Quality Atmosp.
Health 3:5364.
APPENDIX A
Table I: Atmospheric Stability Classifications
Surface Day solar insolation Night Cloudinesse

Windspeeda(m/s) Strongb Moderatec Slightd Cloudy Clear
(>4/8) (<3/8)
<2 A A-B B E F
2-3 A-B B C E F
3-5 B B-C C D E
5-6 C C-D D D D
>6 C D D D D
a
Surface windspeed is measured at 10 m above the ground.
b
Corresponds to clear summer day with the sun higher than 60 above the horizon.
c
Corresponds to a summer day with a few broken clouds, or a clear day with sun 35-60 above the horizon.
d
Corresponds to a fall afternoon, or a cloudy summer day, or clear summer day with the sun 15-35 above
the horizon.
e
Cloudiness is defined as the fraction of sky covered by clouds.
f
For A-B, B-C, or C-D conditions, average the values obtained for each.
Note: A, Very unstable; B, moderately unstable; C,slightly unstable; D, neutral; E,slightly stable; F,
stable. Regardless of windspeed, class D should be assumed for overcast conditions, day or night.
Source: Turner,1970
Table II: Dispersion coefficients (m) for selected distances downwind (km)
258
INSTITUTIONAL RELATIONSHIP IN PORT CITY, PENANG MALAYSIA
Mazuin Dollah 1, Zakaria Ahmad 2
ABSTRACT
This study discussed the role of institutions due to its cooperation they involve in development of port
city. This regard to identify the barrier exist result from lacking in networking between institutions related
to land use planning. It provides the means for land matter institutions and port related institution as well
which is reflect to current practice. With a view of current framework between the local authority and port
authority and also to ensuring that all the institution meet the real need of port cities for what they are
responsible. By providing common basic description of aim and suitable methods, the framework will
improve the network between institutions to be a guide for better land use planning on port cities in the
future. In Malaysia, local authority is agencies are responsible for land use planning and carrying out
development in a city to reach a planning target. When the use of land begun, thus many agencies that
will be involved in order to achieve maximum development. These agencies include the local authority
itself, the agency managing the facilities and stakeholders who manage the sites that just include the port
area, the areas of industry and others.
Keywords: Institutional relationship , Framework, Port City
1
Student, Malaysia Institute of Transport, Universiti Teknologi MARA, 40450 Shah Alam, Selangor, MALAYSIA
2
Lecturer, Faculty of Architectural Planning and Surveying, Universiti Teknologi MARA, 40450 Shah Alam, Selangor,
MALAYSIA
259
INTRODUCTION
Previous study shown the city will success if all the institution has the characteristics of cities consistency
otherwise it might faces difficulties to develop if there are different decisions making from different
institution. This happened when different institution have own interest and aims to achieve. The complex
part will discover when integrity between institutions will be question. Cited by Hoyle (1980) in his book
indicate the growth of port city is fundamental affected by four set of separate factors. They are land
situation, water situation, land site and water side. Port cities is development not only focus on investor to
the port but surrounding areas that support economy and care of port cities strategic & physical planning
(World Bank Project Yemen)
According to Bailey (2005), there are many players in ports like government body and private agencies
that are call institutions dealing with ports every day. Different level of government is involved in daily
operation. The operations have both government as port authority and private agencies as port operator as
well as political initiative in relation to ports. Difference institutions role and limitation can cause
inefficient institution set-up and consequences on support to physical development. Then the integrity
takes place. The successes of European ports not only handle the goods but also in regard to their
organization and administration. They also maintain relations with central and local government and with
private firm operating within the port area (Baily: 2005).
STUDY BACKGROUND
The study was held in Penang which known port city as a case study. The port of Penang was managed
and operated by Penang Port Sdn. Bhd. (PPSB). Corporate entity set up under the Malaysia Government's
privatization policy. Beginning on 1st January 1994, PPSB a government fully owned by the Ministry of
Finance. It took over all facilities and services from the Penang Port Commission, which is now a
regulatory body. The port plays a significant role in the state economy but also contribute to the economic
growth after two decades of its existence. Not surprising if the port given priority by the central
government to depth the channel projects in the RMK-10 (Utusan Malaysia: 2010). Port city basically
combination of port and the city behind the port formed port city. Many cities in the world are growing
due to the influence of the maritime industries and activities generated in and outside the port and its city
around the port. This was agreed by Tai Yong Tan (2007) the modern state of Singapore was embraces
the world by its hinterland remain in essence of port city.
In Malaysia, ministry is the highest as major government agencies and top of the administrative
machinery at the federal level. Establishes ministry has a responsibility and a role to develop policy,
planning, controlling, and coordinating the implementation of duties under the ministry portfolio. This
study will be focus on the various government institutions that govern the Penang port city. They are
Ministry Housing and Local Government (MHLG), Ministry of Transport (MoT), Ministry of Finance
(MoF) and Ministry of Works (MoW). Those ministries play respective roles and have the differences
aspect of development.
Relationships limitations between institutions in Penang port city are also due to the port cities of policy-
related occasions between port and city. In addition, the weaknesses show of cooperation between these
ministries as legal administration. Weaknesses institution networking causes difficulties making planning
and implementation in general. Planning here means that involve land use planning as a component of the
port city. Failure to plan well also affect the components implementation process such as urban road
network facilities used for the transportation, freight logistics, housing, industry, commercial areas etc.
260
MINISTRY
MINISTRY OF HOUSING MINISTRY OF MINISTRY OF MINISTRY OF

& LOCAL TRANSPORT (MoT) FINANCE (MoT) WORKS (MoW)
GOVERNMENT (MHLG)
JPBD MKJR Land LLM

PENINSULAR
MALAYSIA
JAB. KERETAPI Maritime CIDB
JPBD STATE KTMB Aviation JKR
Local Authority PAK PSDC
PAK LJM
LJUM
Figure: Institutional structure in port city LAM

Source: Malaysia Kita (2004)
Based on the framework of government as shown, there are unclear relationships regarding physical
development of port cities in Penang. Every ministries have own respective function and aim.
PROBLEM STATEMENT
There are several problems identified due to current institutional practice.
i. Relevant and Related Authority

There should be a specific authority in a field for smooth administration of the relevant institutions. Some
have referred to appropriate authority member focus on the role and responsibility which has been
provided under the respective departments. The relevant authority related but seems not to carry out
responsibilities that should be. The weakness is caused the failure to achieve goals as a whole
ii. Line Of Administration

Clearly in the current administrative structure there is no vertical or horizontal. Each Ministry is carrying
out responsibilities or their roles without this relationship is important in the administration
iii. Lack of Integration

Conflict, although the ministry has the power to formulate policy but to date there is no longer the policy
of integration between MHLG and MOT, the MoF to develop specific policy areas. The happened when
which authorities are actually responsible for bad access or traffic congestion. The answer may be a local
authority. At the same time which authorities supposedly care for logistics planning from the port or vice
versa? Maybe the answer is local authority. But how cans local authority would know about the capacity
261
to build up the road for suitable access if those institutions do not try to discuss the requirements which
are considered less important.
Abdullah Sanusi (1976) explained federal and state relations are closely linked. Besides economic relation
in terms of planning, ministries of the central administration also deliver services through the state
administration. There is some confusion with the state post office called and paid by the state. However
there are certain positions filled by officers of the Federation in which the salaries and remuneration paid
by the federal Government.
LITERATURE REVIEW
Devolution
The order and pattern of administration now is based on the remains of British colonialism since 1890th.
Ahmad Hussien Meyer (1993) states that perhaps the administration is still remain in a substantive
management system.
Delegation
Representatives must be able to administer, but also in terms of policy making and implementing the
policy. Delegation should also have ability to evaluate policy and project. Governments still prefer to take
a skilled worker from outside the country in economic field. Sometime the expert is not much advanced
from that we have the local administrator
Transfer of Services
Efficient land use planning requires the involvement of all agencies involved. In land use planning, not
only the government directly involved but also involving the private sector. A Decision Maker Guidebook
explain the private sector and agencies are increasingly responsible transport road construction and land
use decision beside those affected as users, business and residence. It should be remember that anything
yang involving land is complex because system involving many parties. National port planning has cited
this leads to several policy decisions which define the role of each ports and ensure that national resources
are used in the most economical manner (UNCTAD:1985)
Related Literature
The way to produce efficient planning and implementation is difficult. A further challenge is the range of
obstacle which limits a citys ability to implement policy instrument. Failure to adopt a logical process for
strategy development can also impose a barrier to effective planning (EUROPIAN COMMISION).
Most port cities nowadays have particular institution administered of responsibility for some role and
policy areas. While others agencies have exclusive responsibility for land use and management of port,
the financial and economic matter or social part. However, those agencies do not have cooperation to
create strong responsibility on network for land use planning together with port when port cities exist
between land use and ports.
Generally, some limit face by cities ability to make own decisions based on three types of constrains
which are typically medium sized cities which suffer most. However smaller cities often have more
freedom while largest one often have more power. The limitation sometimes is distraction on land uses
planning process occurs on port cities situation.
Three types integration are needed to deal with this situation. They are horizontal integration which
means between sectors within authority. Spatial integration which means adjacent authorities and vertical
262
means between tiers of government (DGENV:2005) According to Stephan Schmidt, recent social,
economic & institutional developments have affected the land use planning tools & instruments.
In many cases conflict between institutions can be seen in federal, states and local level because of power
limitations. Abdullah Sanusi (1978) in his book communication and coordination among government
institutions- federal and state involve is power limitation. Most of coordination problems in various
ministries were settling by letter, coordination meeting also arrangement of special meeting among all
related institutions. Some ports in the world need cooperation between local authorities and port authority
to enhance economic aspect or maritime industries as noted by Parson and Douglas (2006). The purpose
of this cooperation exist is not only for port revenues and development but because the expansion of the
land is limited. Expansion requires outside the port area which is under the local authority for example in
New Haven port, Australia.
Peter (2008) was argued that port authorities can actively contribute to better hinterland access. This is
agreed because the ports authority does not have the ability to plan the access but the local authority have.
Different type of involvement of the port authority and the reasons such of involvement discussed. He
added landlord authorities have become more autonomous and take initiative for expansion and
redevelopment of port infrastructure. Port authorities traditionally act as landlord but increasingly operate
beyond the landlord model.
These situations show the power limitation between federal, states and port authority which act as
landlord of the port.
A report review challenge, improvement, innovation and good governance by port authority and local
governance. A key outcome of this restructure was to have better group and integrate a range of function
areas. The current structure aims to over time foster integrated planning, multidisciplinary approach,
improved communication, a breakdown of silos and greater sustainability. For example, a Sustainable
Planning Group was established which combined social, community, strategic land use and transport
planning functions. It is worth noting that various planning processes such as land use planning, social
planning and transport planning have been co-located in the community planning section which is under
local governance. (Review Report Local Government Reform Program: 2005)
PORT CITY INSTITUTIONAL IN PRACTISE
Describing the Institutional Relationship

Institution means more or less stable pattern of social relations between mutually dependent actors which
form around policy problem and or cluster or means & which are form maintained and change through a
series games. It characterized by rule that actor have formed in earlier interactions. Each network has its
own characteristic rules. This rule regulate the behaviour of actor in the network in the sent that they
prescribe what action are permitted, which actor can participate in which games but they also regulate the
networks opinion about the quality of product or services that are produced in the network (Johannes
Franciscus:2004). Whereby institutional network means combination of bodies relations which created
from form around policy problem through a series. At the same time the institutions regulate the
behaviour in the network determine the action they play. Abdullah Sanusi (1978) there is several
administrative levels. The higher is the federal government; state, local, statutory bodies and the
administrative organization's volunteers get help from the government
263
Land and Administration

However, current scenario shows there are less adjustment occurs when it came on network between
institution controllers on land uses matters. Lacking of institutional network can cause an efficient
planning to guide development of port cities.
Ministry of Housing & Local Government (MHLG)

Land is administered under Ministry of Housing and Local Government (MHLG) especially in term of
involving land use planning. It is stipulated Land and Country Planning Act (Act 172) and Local
Government Act (Act 171). The land in Malaysia is under two form of control- federal and state. All state
land is formally administered by National Land Code (1965) and related to law. As we always heard land
is a state matter. It show the power of state to plan for the whole land. State is fully have power and
responsible for approve any development on land use planning and its under local authority. But, as
constitute in National land Code any land reclamation on the land is under federal jurisdiction. It means
reclamation land belong to Penang Port is under federal territory.
Ministry of Transport (MoT)

Ministry of Transport (MoT) is responsible on planning, formulation and implementing rail, maritime,
ports and civil aviation transport policies.
Ministry of Works (MoW)

Responsible for development of infrastructure In this case the JKR is responsible for the construction of
the access road linking the port and its trough various road access to it hinterland. Roads do not have
specific act. According to (Mohd Azam :2003) Development activities related to town and country
planning in this country are not only bounded by the Town and Country Planning Act 1976(Act 172) but
also other existing legislations such as the National Land Code 1965 (Act 56), Street, Drainage and
Building Act 1974 (Act 133), Uniform Building By-Laws 1984 and a few others. The Local Authority in
enforcing planning control is also guided by other mechanisms such as rules and regulations, standards
and guidelines as Guidelines & Geometric On Road Network System by JPBD. While the process of
planning, construction, management and use of express highways in Malaysia are subject to the Federal
Roads (Pengurusan Persendirian) Act 1984.
Port and Administration
Ministry of Finance (MoF)

Officially MoF owned Penang Port since it was corporatized. The ministry responsible to allocate
financial for the port and the aim stress on port growth and profit. In Penang, the operator of the port is
under private sector which manages all services and activity in the port. However the port authority as
observer is only governing the port under MoT and MoF. The conflict will occur outside the port when
the other authority needs to serve the need of freight movement from the port- local authority. How local
authority will provide real needs if they itself does not know the capacity and types of freight mode will
use it? In conjunction of port operator, what will they do if the local authority provides does not meet
their needs? Port authority has no ability on that or vice versa.
Ministry of Transport (MoT)

Clearly state that MoT is responsible for ports and also infrastructure project for ports. It cause the
ministry ignore the important relationship that occurs between the port activity and the networking
activities outside the port that remain important. Examples of the road access need to generated freight
logistics in the city. The relationship actually influenced the development of port cities simply as main
activity of ports is the enabling the transfer of goods from ships to other transport modes and vice versa.
264
Port related an activity includes wholesale, warehousing, logistics and transport are attracted to the direct
proximity of ports. According to Douglas M.Lambert, traffic and transportation, a key logistics activity is
to actually provide for the movement of materials and goods from point of origin to point of consumption,
and perhaps to its ultimate point of disposal as well. This shows the important of integration and
coordination between MHLG and MoT.
ANALYSIS & FINDINGS
i. Vertical and Horizontal Intersection

If we look closer, the institutions involved in the ministry is still looking to establish a coordination
formula between them. Inherited system should be reviewed to meet the development needs or meet the
demands of future development. As an example no actual relationship MHLG ministry and also MoT
cause Penang port city did not become a successful port city yet. How can those institutions integrate
horizontally if there is no communications between them well. Vertical intersection occurs only between
the MoF. This is because the MoF will provide allocation for each ministry. While a horizontal
relationship does not happen to all institutions only perform their respective roles.
ii. Overlapping powers

This is because the responsibility is under the jurisdiction of the local authority as plan in Structure Plan
or Local Plan to provide the access road. It stipulated under Town Country Planning Act 1972. Therefore
the Department of Works (JKR) and the Malaysian Highway Authority (LLM) access likely to handle
road or toll highway maintenance alone. MoW involvement only at the beginning of the process gazette
structure plan or local plan by Chemistry or local authority.
iii. Disjunctions between Institutional

Further analysis of the existence of the gap between the ministry and the role undertaken. What institution
will responsible the development of the Town. The answer are local authorities. What institutions are
responsible for the port? Certainly the MOT. However, when will there is a port city where Successful
institutions will only carry out their duties while the city port development and other infrastructure
development requires a combination of all institutions.
CONCLUSION
Government institutions directly involved in the administration of the port city has been identified. The
potential combination of physical development in terms of road access, road network and other
development components and activities in the maritime industry has seen huge potential in terms of
economy. This was proved again by the researchers before taking the example of Singapore and the other
western countries.
In an effort to create a port city in Penang requires a change from better communication between the
institutions involved. In addition, the actual Respective roles of the institutions involved should be made
in black and white. Respective ministries so clearly the tasks and roles required. Duplication can be
avoided if the duties and roles of institutions involved are the same.
The role of each government institutions involved in land management and administration of ports is
shown. Through it to know the problems are still hovering in the port city government to develop Penang.
Among the problems highlighted is the problem of relevant and related institutions should be directly
265
involved in physical development. In addition to the problem line of administration and transfer of service
affect the efficient administration of these institutions.
Malaysia government needs to have efficient institutional set-up on port cities in order to have sustainable
physical development in a future growth. Related and relevant institutions must have the same goal and
the same time responsible to play the role effectively. It should head from up to bottom level. It also
needs to have specific policy which is important to be a main reference for step of integration between
institutions. In Malaysia, ministry is the highest as major government agencies and top of the
administrative machinery at the federal level. Each establishes ministries have a responsibility and a role
to develop policy, planning, controlling, and coordinating the implementation of duties under the ministry
portfolio.
REFERENCES
1. Abdullah Sanusi Ahmad (1978). Kerajaan Pentadbiran Malaysia, Dewan Bahasa dan Pustaka,
Kuala Lumpur
2. Abdullah Sanusi Ahmad,(1978), Kerajaan Pentadbirab dan Rakyat. Dewan Bahasa dan Pustaka,
Kuala Lumpur
3. Abdul Aziz Zakaria (1974) An Introduction to The Machinery of Government in Malaysia,

Dewan bahasa dan Pustaka, Kuala Lumpur
4. Ahmad Atory Hussien (1993) Pentadbiran dan Politik Pembangunan, Dewan Bahasa dan Pustaka,
Kuala Lumpur
5. British Information Service (1973), Ports in Britain, Central Office Information. London.
6. Barry UBBELS (2005) Institutional Barrier To Efficient Policy Intervention In The European
Port Sector. Amsterdam
7. Department of Local Government (2005) Review Report Local Government Reform Program -
Promoting Better Practice Report. Port Stephens Council. Australia
8. Douglas M.Lambert, J. R. (1997). Fundamentals of Logistics Management. Irwin McGraw-

Hill.
9. Hoyle B. ed. (1996), Cityports, Coastal Zones and Regional Change, England. John Wiley &
Sons Ltd.
10. Joop Koppenjan, Erik-Hans Klijn (2004) Managing uncertainty in Network. New York. Rutledge
11. Kerajaan Malaysia (2004), Malaysia Kita, Petaling Jaya, Selangor: ILBS.
12. Mohd Azam Mohd Abid (2003) Sistem Kelulusan Perancangan Di Bawah Akta Perancangan
Bandar & Desa 1976 (AKTA 172) dan Kanun Tanah Negara 1965 (AKTA 56), Institut Tanah & Ukur
Negara, Kuala Lumpur
266
13. Peter W. DE LANGEN (2008) Ensuring Hinterland Access: The role of port Authorities.
Erasmus University Rotterdam Authority. Netherlands. International transport Forum.
14. Parson& Douglas (2006) Competition & Hinterland Connection. Erasmus University Rotterdam
Authority. Netherlands. International transport Forum.
15. Peter White (1998). Public Transport : Its Planning, Management and Operation (4th Edition).
London: Spon Press.
16. Sady, E. J. (1962). Improvement of Local Government and Administration For 1962,
Development Purpose. Journal of Local Administration Overseas 135-148. London.
17. Stephan Schmidt (1990) Land Use Planning tools & Institutional Change in Germany: Recent
Development In Local & Regional Planning.
18. Tai Yong Tan (2007) Port City and Hinterlands: A Comparative Study of Singapore and Calcutta,
National University of Singapore
19. http://ms.wikipedia.org/wiki/Sistem_Lebuh_Raya_Malaysia
20. Utusan Malaysia (2010) Majlis Pelabuhan Kebangsaan Ditubuh, Kuala Lumpur
21. Utusan Malaysia (2010) Kekal Pelabuhan ke-64 Terbaik, Kuala Lumpur
22. Utusan Malaysia (2010) Lokasi Pelabuhan Pulau Pinang Kekal, Kuala Lumpur
23. (http://ms.wikipedia.org/wiki/Sistem_Lebuh_Raya_Malaysia)
267
268
THE PHYSICAL DEVELOPMENT OF PORT CITY
Raja Noor Hafizah Raja Amir Shah 1*,

Jamalunlaili Abdullah2 and Zakaria Ahmad 3
ABSTRACT:
The initial geographical advantages have given rise to many port cities, and few have continued to remain
alive without expanding productive region, as noted by Huff (1994). Although, the process of
urbanization and industrialization within a port city indirectly give an impact on the development
expansion of the port city, this usually leads to uncontrolled development. The development expansion
occurs in order to meet the current demand of a port city from time to time, and because of that, the port
city faces sensitive and controversial issues that need careful planning solutions. As such, this research is
undertaken in order to identify the physical development of port city. The focus of this research is to
examine the urban growth within the port city in Port Klang. This research takes into account the change
in land use form within the urban area, distribution of physical activities and the direction of land use
expansion of the port city area. The data has been collected from the 1980s until the most recent available
data. The findings of this research will be used to adapt the element of green logistics in Port Klang in
order to achieve the sustainable development within the port city area. In order to control the development
and to achieve sustainable development, suitable recommendations include planning control on the port
city growth and implementing the Integrated Supply Chains as the efficient physical distribution systems.
Keywords: Urban Growth, Port City
1
Raja Noor Hafizah Raja Amir Shah, MITRANS, Universiti Teknologi Mara, 40450 Shah Alam, Selangor, MALAYSIA
2
Jamalunlaili Abdullah, MITRANS, Universiti Teknologi Mara, 40450 Shah Alam, Selangor, MALAYSIA
3
Zakaria Ahmad Principal Lecturer, MITRANS, Universiti Teknologi Mara, 40450 Shah Alam, Selangor, MALAYSIA
269
INTRODUCTION
In the last 30 years port city were re-shaped in their interface being the port area contributes to the
urban activities and access to the water. Port cities are nodes of international trade, channeling
movements of cargo and flows of migration. Port activities often give rise to distinctive regions within
their town or city and might be expected that they would have an effect on the spatial differential and
spatial structure of urban area.
The urban condition and its relation with the coast is demonstrated by the urban shape, which
shows the way the city relates, get and corporate it into its own structure by the existence uses in the
transition between the continuous urban element and the lineal aspect of the coast, with some singular
elements (Grau, 2004 cited by Paula Pacheco, 2007).
Other than that, Jean Paul Rodrigue, Claude Comtois and Brian Slack (2009), explain that there is
a relationship between the size of a port and the size of the metropolitan area it is located in, particularly
for coastal cities having good port sites. Maritime activities were a direct driver of urban growth. The
containerization was one of the factors that support the urban growth by providing much larger port
terminals and also expanding the hinterland access. Moreover, maritime activities is also referring to
planning perspectives, which have incorporate concerns like the degree for which port and city effect one
another in land-use terms that giving rise to a good deal of controversial debate (Hoyle and Pinder, 1992).
It can be said that port activities have a role to play in shaping the spatial structure of the urban area (M.B.
Gleave, 1997).
The maritime industry in Malaysia has grown significantly since the 1970s after the government
announcement in the Third Malaysian Plan to make Malaysia as Maritime Nation. Malaysia has taken an
approach of emphasizing the expansion of capacity to provide a supply-driven environment and
upgrading the facilities of its ports to ensure the efficiency. In addition, ports have been established in the
form of Port Klang as the national load centre which has aggressively spread their wings to enhance their
global connectivity and also contribute to the physical development of Port Klang surrounding area.
Together with the government's intention to make the port of Klang as a national load center for
Malaysian, a balanced development plan is required between the port and the city. Cooperation between
the port authority, port operators, local Municipal council and other related agencies would be able to
develop Port Klang as an effective port city planning beside achieve the government intention.
PROBLEM STATEMENT
Although, the process of urbanization and industrialization within a port city indirectly give an
impact on the development expansion of the Klang port city, this usually leads to physical development
issues. The development expansion occurs in order to meet the current demand of a port city
continuously, and because of that, the port city faces sensitive and controversial issues that demand for
highly careful planning solutions. As such, this research is undertaken in order to identify the physical
development of port city towards green logistics. The focus of this research is to examine the urban
growth, over congestion, problem that related to port expansion issues in Port Klang.
270
Figure 1: Port Klang Map

Source: Port Klang Authority
1 Over congestion
The development of City in Port Klang was influenced by the rapidly developing area surrounded
by the port activities. Persiaran Raja Muda Musa and Jalan Pelabuhan Utara is the main road that
functions as the separator between the residential and ports areas. Although the Port Klang is linked
through an accessible land bridge transport, there is still lacking in term of systematic system of internal
traffic and transportation facilities. This has caused congestion, road safety and transport efficiency.
Therefore, traffic circulation planning, hierarchy of roads and junctions needed to improve the
effectiveness of traffic management system of Port Klang.
271
According to Port Klang Specific Area Plan, traffic congestion occurs at peak hours and the situation
become worse when the heavy vehicles passing through the city to go to the industry area. Through the
improvement of traffic facilities systems, especially those linking with the port and its hinterland would
help balanced development for a port city.
2 Urban Growth
However, with the process of economic growth, increase demand for the port facilities and urban
activities, there are several problems that arose which causes a land use conflict with the surrounding area
and unstructured urban area. One of the issues that arise is many Brownfield sites and land abandoned
impact from the previous activities and the port industry. The issues that occur not only affected the
image of the city but also the visual quality of the port area. Therefore, the systematic and appropriate
land use planning is required to support growth within the Port Klang areas.
Port Klang Specific Area Plan has mention, ad-hoc development resulting the occurrence of
conflict and incompatibility of land use development, especially those involving heavy industry in the
urban area. Increase in the number of industry generates traffic, especially heavy vehicles such as high-
duty trucks.
Growth of population has resulted in considerable expansion of the built-up area of the city and
the emergence of the economic development. From time to time, the built-up area has expended and
become more differentiated from the basis of land-use. The land use zoning is not clear and it caused land
use conflict within the urban area. Conflicts that arise are due to improper land use planning between
industry, housing and commercial area. This situation would lead to traffic congestion, environmental
pollution and decline in the human quality of life.
3 Port Expansion
Government goal is to making Port Klang as the transshipment hub of the region. Expanding the
port facilities will contribute for the creation of business and new job opportunities and generating income
for government as well. Preparation for the new developments were necessary to cater the continued
growth even through there may not be immediately needed for it.
According to Bernama Maritimes New Update (2009), land reclamation was done to a large
swathe of swampland near to fishing village in Pulau Indah which is located on the west coast of
Selangor. In order to make Westport as the nations largest multipurpose port, the development was
followed by a construction work to build an extensive quay complemented by several high-rise
complexes. During the same year, North Port prepare the expansion plan includes the development of
350m container berth and bringing the container quay line at the port to a total of 3.4km. These situations
show that there is a competition between the two ports in developing the port area.
272
Figure 2:Potential and Constrain in Port Klang Development (Urban Area)

Source: Port Klang Specific Area Plan
273
It would be appropriate to prepare a structure plan for port development that would identify the
areas such as the expansion of trade, characteristics of trade flow and the expansion of the port area.
Therefore, the Port Klang authority plays an importance role in preparing the master plan for the Port
Klang area to ensure that existing competition is healthy and balanced with the land use development.
LITERATURE REVIEW
McCalla (1983: 48) define Port as the term cityport, which can be applied to all major ports
within an urban setting, is used to acknowledge the important of the port function in the cities. The
particular conflict that arise in the waterfront as the port attempts to expand in an area where other land
uses, independent of the port is also have land requirement.
Relation between port and city has gained new meanings with the influence of the globalization.
Establishment of strong connections between urban development and port functions shall drastically
improve urban quality of life. Port recognized as a major component of urban setting in the way of
contributes to the form and identity of the city. Refer to Hoyle B.S. in 1998, port city interface sensitive
and controversial and needs careful appropriate planning solutions.
Gleave (1997) explain that urban spatial theory appears to ignore this possibility, urban structure
models having little or nothing to say about the role of port activities in shaping the spatial structure of
towns and cities. In planning for the physical development of port city it is different if compare in
planning for the conventional city because it has a very special correlation between the impacts of the
development in port area that influence the city growth.
Changing pattern in development of the port area will influence the growth in the city area. Bader
noted on his research relationship between a citys location on a body of water and its growth. With the
potential of growth in the port area that give excellent impact on the growth of city area, a concern in term
of planning strategies and special development should be provide.
1 The Scenarios
In order to understand more about the port city growth and relationship within the context of Port
Klang case study, the related literature on the port city relationship need to be examined. There are many
scholar has describe about the port city relationship. Birds (1963) explain about the changing relation
between ports and their host cities, Hoyle (1988) describe about the port/city interface, Hoyle (1989)
port and city linkages model and Murphy (1989) describing socio-economic transformation frameworks
of port cities.
Port infrastructure expanded in time and space was explained on port evolution model in 1963 by
Bird. He identified three major phases on port development process, his model describe the initial
development at the port site and the subsequent expansion from simple terminal facilities for warehousing
and wholesaling to more advance facilities. The changes are caused by changing in maritime technologies
and improvement in cargo handling, spatial relationship between the port and urban center and
specialization in infrastructure and activities. Bird model is suitable and useful as a framework to work
out for the port morphological development.
There are five stages in the in the evolution of the built-up area between the port and the city refer
to Hoyle (1988). Referring on his theoretical and empirical description, the first stage was the primitive
city port which explains about the close association, port and city existed and the city growth that
depending on trade activities at port. While the second stage was explained about the city port expanded
274
rapidly in nineteenth century that based on the port activity that encourages the commercial and industrial
development in the city. Modern industrial city ports illustrate the evolution in technological advancement
in transportation system and services (the third stages). It is involving a larger and deeper harbor besides
the movement of good to new areas that away from the city port location.
Changes to the fourth stage occur when the land available in urban settings was not enough for
modern industrial development and then the development would be distant away from the port area. As
the time goes by, advancement in technologies, economic change and older facilities closer to the heart of
the city this is the stage where the redevelopment of the water front is need to ensure the city port area not
abandoned and still attractive (the fifth stage).
Hoyle (1996), Port function has normally give contribution to some degree of urban growth
because of that port-city relationships become more complex and well establish. Port and local authorities
function as the party that responsible to control redundant port space (Pinder 1988) for the purpose of
reusing abandoned area.
2 Development of the Urban Area
According to Port Klang Specific Area Plan development history, Port Klang known as the Kuala
Klang in the past. It is a small village inhabited by several ethnic groups which are fishermen and farmers.
Before the existence of Port Klang, there is a port located in the Klang town fuction as the place of
unloading of tin ore in the early 1880s.
However, the functions of the port started to decline when the completion of the Belfield bridge that
enhances the relationship between North Klang and South Klang. New port area in Klang Kuala founded
by British colonial rule, Sir Frank Swettenham in 1900 completed all construction of the railway from the
Klang town to Kuala Klang. Port with a jetty and three berths were opened on 15 September 1901 and
called as Port Swettenham, in conjunction with the name of the British Resident of Selangor at that time,
the Sir Frank Swettenham. Port Swettenham has changed its name to Port Klang after Malaysia's
Independence Day. Since it is opening it was located under the administration of the Railway Department.
In order to improve the administration and management of Port Klang, Port Klang Authority has
developed on 1 July 1963 as a statutory body under the parliament, the Port Authorities Act 1963 has
created to ensure the administration and management of the port to be smooth, efficient and systematic. In
the same time, the status of this port has been upgraded from the coastal and ocean port to federal port.
3 Port Development
The post-war years and particularly the period after Malaysia independence saw the port has
gaining increasing importance as gateway. The increasing commercial and industrial development in the
Klang Valley has forms the port direct hinterland demand for further development and improvement of
existing facilities and services. Port Klang has grown in just about every aspect of it operations to emerge
as the leading port of Malaysia.
The 1960s was the decade of great significance to the port. It was a decade of transition of hard
working and success. This was the time that left a memory impression on the ports history that changed
the expression of the port. 1960s was decade that saw the beginning of the transformation of the Port
Klang from a coastal port to its current status as Malaysias premier port.
275
The 1970s saw the setting up of the Klang Port Authoritys own security force, the takeover of
cargo-handling and pilot age services from the private sector, the renaming of Port Swettenham to Port
Klang and the event of the decade. The arrival of the first third generation container ship, the Tokyo Bay,
on August 5, 1973 has contributed to the beginning of containerization to Malaysia. This decade was the
extensive development which changes the face of the port beyond recognition. It was also the time which
the port received a new name.
The expansion and rationalization was in 1980s. Further expansion of North Port has developed
into the construction of the dry bulk terminal, additional general cargo berths and roll on roll off ramp.
Better productivity was the battle cry and the port saw the introduction of administrative and work
procedures to facilitate trade. In 1986 the era of privatization of its container terminal is began.
RESULT ANALYSIS
There are several criteria to verify the physical development of Port Klang. The criteria has been
choose to analyses this research are land use and urban form, port infrastructure, transport and land use
connection and sustainable development. Analysis carried out is to see the preparation of Port Klang as a
planned and sustainable port city.
1 Land Use and Urban Form
Port Klang land use areas should be upgrade and improve in line with the transformation of the
current development and as preparation to meet the demand in the future. Therefore, there are several
activity that has conflicts with land use development should be removed so that the quality of the
environment can be preserved and enhanced. There are vacant land that has dilapidated, unmanaged and
unplanned area that need priority in planning the redevelopment of Port Klang area.
Looking at the history and activities of the port that contribute for the growth of urban area, the
Town Center area has the potential to be developed based on the concepts associated with the port as the
main attraction. Meanwhile, the port and industrial activities are concentrated as purpose to avoid land
use conflict and easy to control the development. To control and protect the environment, the industrial
activity which is harm the environment is not recommended inside the Port Klang area. Given the existing
business has been solid with 80 percent occupancy rate, the less stable areas between the coastal and
highway will be developed as a mixed development area to meet demand.
2 Port infrastructure
Malaysian government has provided the investment since the 1960s for port to cope with cargo
handling changes bough about by containerization. In addition, the following have been implemented.
Klang Port Authority (KPA) has carried out a Master Plan for development of the West Port area of Port
Klang. Twelve berths (3 containers) were commissioned in 1994, followed by 3 more container berths in
1996. It is expected that by the year 2010, a total of 30 berths would be developed.
Trade facilitation was used to reduce documentation and procedures for cargo consolidation and
to attract value-added services, the North Port and West Port areas were converted into a Free
Commercial Zone (FCZ) in 1993 and 1996 respectively. Distribution ports have been established by a
major shipping line in a joint venture with a private operator. A port Klang community based EDI system
introduced in 1994. Electronic lodgment of Customs Import Declaration and manifest was introduced in
1995. Further expansion and improvement of electronic trading is in hand.
276
Moreover, port development explain in Ninth Malaysian Plan (RMK-9) continue to focus on
increase capacity, upgrade equipment and facilities beside improve the efficiency of port services and
related services. Major projects will be undertaken to improve the capacity of the ports including
construction of additional berths in Port Klang and several federal ports.
3 Traffic and land-use connection
Road play an important role as the main communication channel as connector between Port
Klang and its surrounding area. Providing modern, economical and better access between port areas,
urban centres and states will assist in facilitate the development of Port Klang. Together with the aim of
reducing traffic congestion and eliminating regional economic disparities, investment in the road
infrastructure has always been on the top of the Governments infrastructure development agenda.
There has been an increasing emphasis on inter-urban highways, Kuala Lumpur Klang Federal
Highway, New Klang Valley Expressway (NKVE), Shah Alam Expressway (KESAS) and North South
Expressway (ELITE) was the four existing highways have been built to facilitate the flow of traffic to
Port Klang. However, traffic flows into Port Klang will be more smoothly with the construction of the
South Klang Valley Expressway (SKVE) is expected to be completed in 2012. This special provision for
highway planning study was carried out before it construction is to ensure that the construction of
highways SKVE is concern about road safety aspect and bring minimum impact on the environment.
4 Sustainable Development
Partnerships between local authorities, public and private agencies are importance to achieve a
consensus and then act together on the issues of sustainable development. Therefore, all new development
will be reviewed before approval as steps to ensure a well planned development and will not pose a
problem in the future. The complexity of urban development will present main challenges in future
development.
Therefore, Port Klang area has taken an approach in several importance sectors such as land use,
transportation and also the infrastructure. Improvements and construction of roads and junction are the
steps taken to reduce traffic congestion besides providing the heavy vehicle terminal. Meanwhile, Urban
waterfront and pocket park was the step taking in order to provide the green spaces within the urban area.
CONCLUSION
It seem sufficiency clear that the port area was one of the spatial structure that contribute to the
urban area emerge. In order to control the development and to achieve sustainable development, suitable
recommendations include planning control on the port city growth.
REFERENCES
British Information Services (1973), Ports in Britain, London, Central Office of Information.
Robson B. T. (1973), Urban Growth: An Approach, London, Methuen & Co LTD.
Daamen T. (2007), Sustainable Development of the European Port-City Interface, ENHR Conference,
W19, pp 1-20.
277
Ling O. G. and Shaw B. J. (2004), Beyond The Port City: Development and Identity in 21st Century
Singapore, Singapore, Pearson Prentice Hall.
Hoyle B. ed. (1996), Cityports, Coastal Zones and Regional Change, England, John Wiley & Sons Ltd.
Mckinnon A. (2009), Logistics and Land: The Changing Land Use Requirements of Logististical
Activity, Annual Logistics Research Network Conferences, 14, pp 769-775.
Malaysia, Kementerian Perdagangan Antarabangsa & Industri (2006), Industrial Master Plan 3: Logistics
Chapter 2006 2020, Malaysia, Ministry of International Trade and Industry
M.B. Gleave (1997), Port Activities and the Spatial Structure of Cities: The Case of Freetown, Sierra
Leone, Pergamon, Journal of Transport Geography
Nas P. J. M. (2005), Port Cities, Asia Europe Foundation, European Alliance for Asian Studies, 37.
Shamsudin K. (2003), Sustainable Land use Development In The Klang Valley: An Elusive Dream,
Federal Department of Town and Country Planning Peninsular Malaysia, Malaysian Town Plan, 1(1),
5-9.
PAG Consult Sdn Bhd (2010), Port Klang Specific Area Plan, Kuala Lumpur, Department of Town and
Country Planning.
http://www.bernama.com/maritime/news.php?id=441863&lang=en&cat=ex (Tuesday, 13/07/2010)
http://www.btimes.com.my/Current_News/BTIMES/articles/pkand/Article/(Tuesday, 13/07/2010)
http://www.bharian.com.my/bharian/articles/SKVEpercepatakseskeselatanLembahKlang/Article
278
THE HEADWAY PATTERNS AND POTENTIAL PARAMETERS OF BUS

TRANSPORTATION IN PENANG
Suria Haron 1*, Sarina Muhamad Noor 2,

Ahmad Farhan Mohd Sadullah 3 and Leong Lee Vien 4
ABSTRACT:
Bus transportation is one of the various means of public transportation. Its efficiency in
following its route is vital to ensure customers are satisfied with the service provided. This study
is conducted in Penang among Rapid bus passengers. The objectives are twofold, first it attempts
to investigate the headway pattern of bus services and secondly, to identify the potential
parameters of urban bus performance. In conducting this study, initial observation was conducted
to identify the headway pattern. Then, a set of questionnaires was distributed to bus passengers
in order to tap into the satisfaction parameters. The data was compared with Highway capacity
manual and was analyzed for construct validity using Factor analysis. The findings suggest that
most of the services are rendered according to the schedule with only one route found
unattractive. Factor analysis discovered five construct for measuring bus performance. The result
is beneficial for relevant authorities in developing effective measure for better bus performance
Keywords: Bus transportation, headway, service quality, bus parameters.
1
Student, Universiti Sains Malaysia, Seri Ampangan, 14300 Nibong Tebal, Seberang Perai Selatan, Pulau Pinang, MALAYSIA
2
Principle Lecturer, Universiti Teknologi MARA Perlis, Kampus Arau, 02600 Arau, Perlis, MALAYSIA
3
Professor, Universiti Sains Malaysia, Seri Ampangan, 14300 Nibong Tebal, Seberang Perai Selatan, Pulau Pinang,
MALAYSIA
4
Principle Lecturer, Universiti Sains Malaysia, Seri Ampangan, 14300 Nibong Tebal, Seberang Perai Selatan, Pulau Pinang,
MALAYSIA
*Correspondence Author: Suria Haron, Ph.D Candidate, School Of Civil Engineering, Universiti Sains Malaysia,
Malaysia. Tel: +60124740921, Fax: + 045991009. E-mail: sue_riea@yahoo.com/suria.wd07@student.usm.my
279
INTRODUCTION
Transportation and its environmental impacts are major component of urban

environmental management. Many efforts (Canada, the United Stated, European countries,
World Bank and etc) were devoted to develop sets of transportation indicators which can be used
to label their performance. Developing comprehensive evaluation system is still much needed
especially in Malaysia as few attempted effort has been slow (Lee et.al, 2003). Moreover,
through promoting public transportation, it offers a method to reduce the use of private cars in
urban and eventually the adverse impacts of transportation.
The rationale of this study relies on two main issues. First is on the traffic congestion in
the city. This is because the planning and provision of the transportation for urban travel in
Penang was targeted for the need of private car users resulting in a fairly serious level of
congestion (JICA, 1986 and Rasagam ,1999). This situation leads into urban traffic congestion,
increasing rates of injury and death from traffic accidents as well as environmental issues.
Secondly, the facilities provided through public transportation are poor (Selvanathan,
2007). The woes and complaints from publics and tourists need further investigation. As Penang
received thousands of tourists each year, the bus transportation as one of the main public
transportation needs some up-grading. The bad condition of bus network system in Penang due
to no legible schedules at the main bus interchange, adoption of laissez-faire systems by raising
the rates of fare, doing away with the ticketing system and do not follow timetable and planned
routes make the public transportation exacerbate problem and make solutions very difficult.
Hence, the decision to set up Rapid Penang in mid 2007 is a right step to reduce the states traffic
woes. Accordingly, there is a need to bring about significant improvements on the public
transportation in Penang in order to encourage people to use public transport as the main mode of
transportation. Hence, the objectives of this study are twofold:
i. To determine the headway pattern of bus services in Penang.

ii. To determine the potential parameters of urban bus performance indicators that
explains passenger satisfaction and related to sustainable transportation.
This study is important in order to encourage more people to use the bus services and to
obtain insights from actual users of the system about the changes they would like to see to better
meet their needs. Moreover, the results would alleviate the present congestion and related
problems due to public transportation issues. It is hoped that this study will become the
benchmark by public transport operators and planning authorities, for academics and other
researchers in order to improve the service quality of the public transportation system in Penang.
LITERATURE REVIEW
Transportation is one of the most important components of urban infrastructure necessary

to ensure the vitality of an urban area. An efficient network of transportation services are
required to support the complex activity patterns within cities. Furthermore, there is a strong
connection between transportation and city growth. The successive improvements in the
transportation system allowed the city to grow farther and farther (Hamed and Olaywah, 2000).
280
Moreover, transportation can also promote or hinder development and vice versa (Papacostas
and Prevendouros, 2001).
Improving public transportation is seen by many as critical (Ison and Wall, 2002). With a
rise in demand for transportation come congestion and other related problems such as delays,
uncertainty, and stress on both individuals and society (Morris et al., 2005). Hence, the
transportation needs of major cities present significant challenges for policy makers especially
with the unpredictable shifts in population dynamics (Andaleeb et al., 2007).
Reliability Performance
Among the critical issue in public transportation is its reliability performance. Reliability
performance of bus operation refers to consistency of performance, frequency and dependability
(Robert-James, C. (2002). Many of the larger cities have their bus operated on a frequency or
headway rather than time table basis (Randall, et al, 2007). According to United State Highway
Capacity Manual 2000 (HCM, 2000), there are several measurers of reliability used by transit
systems. The most common are on-time performance, headway adherence (the consistency or
evenness of the interval between transit vehicles), missed trips and distance travel between the
mechanical breakdowns. On-time performance is the most widely used measure in the transit
industry. It measures the percentage of trips departing from all schedule time points (Nakanishi,
1997). Meanwhile, headway adherence refers to the degree of actual headway as compared to
scheduled headway. The headway schedule adheres if it met at least 50% of the actual headway
(Hammerle, Haynes & Mc Nell, 2006). Headway adherence is important for frequent service,
since bunched vehicles indicate inefficiency in operation. The bunched buses require the buses to
be in waiting line longer than expected. This will also lead to overcrowding as all the buses
arrive simultaneously at a specific station ,(HCM, 2000). In addition, measuring the reliability
of transit service is also important because both the transit customer and the transit provider
value reliable service (Nakanishi, 1997). With an accurate picture of service performance, it
gives customers valuable information to become active participants in the transit policy and
decision making process (Nakanishi, 1997).
Service Quality
Measuring service quality is important to ensure customers patronizing the business for
repeat purchases. The same principle applies to transportation businesses. Customers who are
satisfied with the bus performance will return and use the services. Studies on service quality can
be traced back to Parasuraman, Zeithmal and Bery (1985) who translate the abstract nature of
service delivery into an instrument capable of measuring the perception and performance of
service quality. The said instrument, SERVQUAL, has gone through various developments.
Most of scholars still agree that the dimensions used in Parasuramans instrument are still much
relevant in measuring service quality. There are five dimensions namely tangibility, reliability,
receptivity, assurance and empathy. These dimensions were adapted to transportation study and
are labeled differently following Andaleeb et al. (2007), Nicholas and Poimboeuf (2003),
Stradling et al. (2007) and (Wang et al., 2006). These are labeled as safety and security for
assurance, facility for tangibility, and economic for receptivity. The dimension of reliability and
accessibility remains as it is.
281
METHODOLOGY
The study was conducted in Penang focusing on the Rapid Penang buses and commuters.
This service provider was used as it provides a higher quality bus that minimize uncomfortable
gap for boarding passengers. The bus was air conditioned with age not greater than 5 years, was
operated by a uniformed staff and have low-floored vehicles. This study used seven routes
patronized by Rapid Penang buses in Seberang Prai, Penang. The routes selected which were
U601, U602, U701, U702, T703, U801 and U802.
The data was collected manually using two stages. First, following Nakanishi (1997),
several point check observation was conducted at transit stop to determine the reliability
performance. Few research assistants were positioned at various bus stops at the respective
routes. These researchers kept a record on the bus arrival time and the subsequent service. The
headway time for the seven routes was measured in order to determine the headway pattern for
each route. The data was collected from 7.00 a.m. until 9.00 a.m. for peak hour and 9.00 a.m.
until 2.00 p.m. for off peak hour. The schedule used was based on Nakanishi (1999), whereby
peak hour is between 7.00 to 8.59 a.m. an midday, 9.00 to 4.00 p.m. is considered as off peak
hour. In addition, most of the respondents are working people and the student who were in the
sample comprised of university students. The survey was conducted on the weekday for three
weeks. The data was then compared to HCM 2000 for both peak and off peak hours. A
comparison with HCM 2000 was used as it provides a method that guides the users in the
analysis of automobiles and transit in a corridor analysis (Dowling, McLeod, Guttenplan and
Zegeer, 2007). On top of that, Malaysian Public Work Department only used HCM 2000 to
measure speed-flow relationship as HCM 2000 application is under research area, this study will
use LOS to assess to highway pattern. Moreover, previous research that used LOS was
conducted in rail transit (Cao, Yuan, Zhang and Zhao, 2009), airport terminals (Correia,
Wirasinghe, Barros, 2008) and among pedestrian (Tan, Wang, Lu, 2007).
According to HCM 2000, there are five categories of level of service (LOS) that
correspond with each bus frequency or headway and number of vehicles per hour. Table 1 show
that LOS A is the most satisfactory as the bus frequency is less than 10 min. Hence, bus
schedules are not required. However, for LOS E and F, the bus headway is between half an hour
to an hour and more than one hour respectively. Service is available yet customers might not
keen in waiting if they have other options. Competition could beat bus operators out if they fall
in these categories.
Table 1: Service Frequency of LOS for Urban Scheduled Transit (HCM 2000)
LOS Headway (min) Veh/h Comments

A <10 >6 Passengers dont need schedules
B > 10-14 5-6 Frequent service; passengers consult
schedules
C >14-20 3-4 Maximum desirable time to wait if bus/train
missed
D >20-30 2 Service unattractive to choice riders
E >30-60 1 Service available during hour
F >60 <1 Service unattractive to all riders
282
During the second stage of data collection, another group of research assistants were
employed to distribute the questionnaires for the bus passengers who were waiting for the Rapid
Penang buses at Rapid Penang bus terminals and bus stops. The purpose is to get the customers
opinions and experiences while using Rapid Penang as their feedback is very important to
improve the bus services. There were five dimensions identified most prevalent by the
literatures. The dimensions covered were safety & security, environment, economic, accessibility
and reliability. These 45 self-completed questions were divided into 2 sections namely Section A
for demographic information and Section B for user perception on level of quality service
offered by the operator.
The data was analyzed using descriptive statistics. Factor analysis was employed during
the second stage in order to group together those variables which are highly correlated with each
other. Then the data was interpreted based on each dimension according to the grouped variables
(Fielding et al., 1985).
RESULT ANALYSIS
The findings are divided into two main subsections. First is on the reliability performance
and secondly, on the service reliability.
Headway pattern
Under reliability performance, a graph was plotted to determine the headway pattern of
bus operations for each of the routes. Figure 1.0 shows the result. The result was interpreted
based on the graph and compared to HCM (2000) table. The figure 1.0 is based on the bus
schedule. The bus arrivals during peak and off peak hour are shown by the bar chart. Then a
comparison was made with LOS based on the schedule and actual and the bus arrivals.
Figure 1.0 Headway pattern during peak-hour and off-peak hour
283
Headway pattern during peak hour
The economic factor also contributes significantly to reduce car dependency by

promoting buses to workplace. Hence, it is a great concern to minimize congestion and
consequently the accident risks. This is relevant to factor 3 on safety and security. Another
attribute, safety while travelling by Rapid Penang bus, is another attractive attribute to attract
people to use public transport.
Figure 1.0 indicates that five routes which are U602, U701, U702, T703 and U801 share
the same LOS E during the peak hour condition. The result indicates that the service was
available during the hour with headway between 30-60 minutes. Hence, reduction in the number
of buses operated during peak hour could lead to headway delay. LOS for U601 and U802 is F
indicating that the service is unattractive to all riders (Table 2)
Table 2: Service frequency LOS for Rapid Penang bus services according to HCM 2000
Route LOS Comments
U601 F Service unattractive to all riders
U602 E Service available during hour
T703 E Service available during hour
Headway pattern during off-peak hour
During off-peak hour, routes U601, U602 and U802 performed poorly as compared to the
other routes (Table 3). It was clear that reducing the number of bus operating during off-peak
hour significantly increased the headway delay and it will not improve the situation.
Table 3: Service frequency LOS for Rapid Penang bus services according to HCM 2000
Route LOS Comments
T703 E Service available during hour
284
(MUTRF C2010), 21 December 2010, Universiti Tenaga Nasional. ISBN 978-967-5770-08-1
Overall, LOS for route T703 performed better as compared to other routes. Customers are
willing to wait if the bus is missed. However, customers who use route U802 are those who have
no other options but to wait for the bus to come. Scoring LOS E indicates poor performance for
bus patronizing routes U602, U701, U702, T703 and U801; though bus is still available during
the hour but the waiting time is within 45 to 50 minutes. According to Nakanishi (1997), the
actual headways must be within 5 min to 30 minutes so that, the customers are served at the
scheduled time.
The results from observation indicate that headway delays caused road congestion during
peak hour. Further observation revealed that headway delay during off-peak hour was strongly
related to reduction in bus frequency. Hence, the bus operator should be targeting operations control
efforts on bus schedule problem.
Service reliability of bus operators
During the second stage of data collection, the Rapid Penang passengers profiles were
tabulated. This was done in accordance to suggestion by Stradling et al., (2007) who mention
that transport patterns and choices vary with person characteristics such as age, gender and
disability, and with journey purposes. The information under Table 4 reveals that majority of the
passengers are young at age 15-24 (56%), female (56%), single (69%) and earning a monthly
income of less than RM500. 46% of the respondents is working at private firms and 44.4% is
students. The data indicates that the majority who patronize the bus services do not have many
options for their travelling purposes due to their personal constraints. Majority of those who use
the buses are workers (32%) and students (30%). As 50% of the respondents do not own any car
with low income, the relationships between income, car ownership and the demand for public
transport are as expected. Income is expected to increase the number of trips and length.
According to Paulley, et.al (2006) rising cost to own a car, driving license ownership and income
growth are the key factors that have shaped personal travel patterns in the last 20 years.
Table 4: Demographics of Rapid Penang passengers
Demographic Item Frequency Percent

below 15 8 4.0
15-24 112 56.0
Age 25-39 41 20.5
40-59 38 19.0
above 6 1 0.5
male 88 44.0
Gender
female 112 56.0
69.0
single 138
30.0
Marital status married 60
1.0
divorced 2
285
government agencies
1 6.1
private agencies
Occupation 90 45.9
pensioner
7 3.6
student
87 44.4
below RM500 94 48.0
RM500-RM1500 67 34.2
Monthly income
RM1501-RM2000 25 12.8
above RM2000 10 5.1
Car/mpv 53 26.5
Van/lorry 3 1.5
Car Ownership
Motorcycle 43 21.5
None 101 50.5
To work place
64 32.0
To
Purpose of 59 29.5
school/college/university
travel 18 9.0
To market
56 28.0
Others
Factor analysis was done on the items in the questionnaires. The Kaiser-Meyer-Olkin
(KMO) measure of sampling adequacy and the Bartletts test of sphericity of sampling indicate
the appropriateness of the Service Quality Survey (SQS) data for factor analysis. The KMO
values obtained signifies the sampling adequacy of this SQS study (above 0.6) while the value of
Bartletts test shows that all the factors are significant (p<0.05). SQS reliability is satisfactory as
Cronbachs alpha coefficient is between 0.75-0.84 (Field, 2002). Table 5 shows the result of
factor analysis.
Table 5: Result of factor analysis for all items.
Factor
Factor Attributes
loadings
The bus station is free from noise pollution. 0.894

The bus station is free from air pollution. 0.856
Do you feel comfortable with the air conditioning on the bus? 0.850
Factor 1
The seats on the bus are always full. 0.843
Environment
The buses are clean. 0.759
There are enough parking spaces at the bus station. 0.618
There are sufficient seats at the bus stop & bus station. 0.607
The use of private vehicles may reduce by using the bus.

Road accident rates may be reduced when more people use the 0.870
Factor 2 bus. 0.834
Economic Financial burden can be reduced when the bus is used. (petrol & 0.800
parking) 0.773
Traffic congestion may be reduced when more people use the bus.
286
Do you feel safe while waiting for the bus? 0.850

Bus drivers are very helpful. 0.832
Factor 3
Do you feel safe while traveling on the bus at night? 0.825
Safety and
Bus drivers are polite to passengers. 0.809
Security
Do you feel safe while crossing the road to the bus stop? 0.802
Bus drivers drive carefully. 0.795
The bus service is convenient to the disable. 0.776

It is easy to reach the bus stop and bus station. 0.768
Factor 4
It is easy to find information on bus route and schedule. 0.746
Accessibility
The bus service is convenient to everybody. 0.715
The schedule of the bus is convenient. 0.632
Table 5 shows that all items load in the dimensions of environment, economic, safety and
security and accessibility. From 34 items, only 22 have the acceptable loadings. The factor
loadings is between 0.618 0.894. As these items represent the construct of service quality, they
are useful in future evaluation on the service provided by the operator.
A high degree of construct validity exhibits the Rapid Penangs passengers significantly. The
environmental factor at the bus station is considered strongest; the noise and air pollution was
low at the time this survey was done. Furthermore, passengers are satisfied over the new bus and
the bus design that is equipped with ultra-low flooring and bus cleanliness.
The economic factor also contributes significantly to reduce car dependency by promoting buses
to workplace. Hence, it is a great concern to minimize congestion and consequently the accident
risks. This is relevant to factor 3 on safety and security. Another attribute, safety while travelling
by Rapid Penang bus, is another attractive attribute to attract people to use public transport.
CONCLUSION
The headway reliability is to maintain the scheduled headways. The intention is to avoid
passengers frustration and discomfort while waiting for the services by minimizing the average
passenger waiting time. The delay in bus arrivals leads into bunching of buses hence, service
cannot be performed towards the users during scheduled hours. Eventually, this will result in
peaks in passenger demand for subsequent bus. Such peak in passenger demand may result in
buses getting full to its capacity and having to bypass subsequent stops due to inadequate
capacity. The result shows that in peak hour, unreliability increases with increasing congestion
and passenger demand.
As suggested by Liu and Sinha (2007), the passenger demand has more serious impact on
headway variability and passenger wait time than on total journey time reliability. Moreover,
according to Wang and Barnum (2006), the importance of reliable bus service to customers is
well known, with arriving when planned being the most important desire of transit riders. Not
surprisingly, therefore, a reliable service delivery indicator is one of the five sets of bus
performance standards that are monitored by most transit systems. Public transit agencies have
287
developed multiple indicators to measure bus service reliability, with indicators of on-time
performance and headway adherence being almost universal, and a third common measure being
running time adherence.
The findings reveal several important attributes. Some of the identified attributes are not
directly under the control of the bus operator. Hence, it offers the challenge to others including
local government (Henser, 2003) in an effort to up-grade the situation. The present study is
parallel with Friman (2004) as it supports an emphasis on delivering high-quality services to
people who use public transport. Indeed, the comprehensive study should be continued to
investigate the present situation of bus operation by pursuing more potential attribute, efficient
policies and effective measure of urban bus public transport performance.
It is suggested that managers and decision makers include provisions from the factors
above in their approach to improve service quality toward sustainability.
REFERENCES
Andaleeb, S.S., Haq, M. and Ahmaed, R. I. (2007), Reforming Innercity Bus Transportation in
Developing Country, Journal of Public Transportation, Vol. 10, No. 1, pp. 1-25.
Cao, S., Yuan, Z., Zhang, C., Zhao L., (2009). LOS Classification for Urban Rail Transit
Passages Based on Passenger Perceptions, Journal of Transportation Systems Engineering
and Information Technology, Science Direct, Vol. 9, No. 2, pp. 99-104.
Correia, A. R., Wirasinghe, S. C., Barros, A. G. (2008). A Global Index for Level of Service
Evaluation at Airport Passenger Terminals, Transportation Research Part E, Vol. 44, No.
4, pp. 607-620.
Dowling, R. G., McLeod, D., Guttenplan, M., Zegeer, J. D. (2007), Multimodal Corridor Level-
of-Service Analysis, Transportation Research Record 1802, Journal of Transportation
Research Board, National Research Council, Washington D.C., pp. 1-6.
Field, A. (2000). Discovering Statistics Using SPSS for Windows, Sage Publication Ltd, 6
Bonhill Street, London.
Fielding, G., Babitsky, T. & Brenner, M. (1985), Performance Evaluation for Bus Transit,
Transportation Research Part A: General, 19, pp. 73-82.
Friman, M. (2004), Implementing Quality Improvements in Public Transport, Journal of Public

Transportation, Vol. 7. No. 4.
Hamed, M.M., Olaywah, H.H. (2000), Travel Related Decisions by Bus, Servis Taxi, and Private
Car Commuters in The City of Amman , Jordan. Cities, Volume 17, Issue 1, pp. 63-71.
288
Hammerle, M., Haynes, M. & McNell, S., (2006). Use of Automatic Vehicle Location and
Passenger Count Data to Evaluate Bus Operation. Transportation Research Record 1903,
Journal of Transportation Research Board, National Research Council, Washington
D.C., pp. 27-34.
Hensher, D., Stopher, P. and Bullock, P. (2003), Service QualityDeveloping a Service Quality
Index in The Provision of Commercial Bus Contracts. Transportation Research Part A, 37,
pp. 499-517.
Highway Capacity Manual (2000). Transportation Research Board, National Research Council,
Washington D.C.
Ison, S. and Wall, S. (2002), Attitudes to Traffic Related Issues in Urban Areas of the UK and
the Role of Workplace Parking Changes, Journal of Transportation, London.
Japan International Cooperation Agency (JICA), 1986. The A Traffic Computerised Feasibility
Study in Penang. Report for Engineering Dept., Municipal Council of Penang Island.
Lee, W.R., Wack, P., Jud, E., Munroe, T., Anguiano, J. and Keith, T. (2003), Towards
Sustainable Transportation Indicators for California, Mineta Transportation Institute. US.,
pp. 188.
Liu, R. & Sinha, S. (2007) , Modelling Urban Bus Service and Passenger Reliability. The Third
International Symposium on Transportation Network Reliability, 19-20 July 2007, The
Hague, Netherlands.
Morris, M., Ison, S. and Enoch, M. (2005), The role of UK Local Authorities in Promoting the
Bus, Journal of Public Transportation, 8(5), pp. 25-40.
Nakanishi, Y. J. (1997), Bus Performance Indicators: On-Time Performance and Service

Regularity, Transportation Research Record 1571, Journal of Transportation Research
Board, National Research Council, Washington D.C., pp. 3-13.
.
Nicolas, J., Pochet, P. and Poimboeuf, H. (2003), Towards Sustainable Mobility Indicators:
Application to the Lyons Conurbation, Transport Policy, 10, pp. 197-208.
Papacostas, C.S. and Prevendoros, P.D. (2001), Transportation Engineering and Planning,
Prentice Hall, New Jersey.
Parasuraman, Zeithmal and Berry . (1985), A Conceptual Model of Service Quality and its
Implication of Future Research, Journal of Marketing, pp. 35-48.
Paulley, N., Balcombe, R.,Mackett, R., Titheridge, H.; Preston, J.M., Wardman, M.R., Shires,
J.D., White, P. (2006), The Demand for Public Transport: The Effects of Fare, Quality of
Service, Income and Car Ownership. Transport Policy, 13 (4), pp. 295-306.
289
Randall, E. R., Condry, B. J., Trompet, M. (2007), International Bus System Benchmarking:
Performance Measurement Development, Challenges, and Lessons Learned. Submitted for
Transportation Research Board 86th Annual Meeting.
Rapid Penang Route Map. Retrieved from http://www.rapidpg.com.my.
Rasagam, G. A Step Towards Sustainable Transport: A Case Study of Penang, Malaysia.

Transport and Communications Bulletin for Asia and the Pacific, 33.
Robert-James, C. (2002), Transport, Local Government and the Regions Committee Inquiry into
the Bus Industry, The Institution of Highways and Transportation. 6 Endsleigh Street,
London.
Selvanathan, S. (2007), Sustainable Transport Index for Penang Island, Report of Final Year
Project. School of Civil Engineering, Universiti Sains Malaysia, Engineering Campus, Pulau
Pinang.
Stradling, S., Carreno, M., Rye, T., Noble, A. (2007), Passenger Perceptions and the Ideal Urban
Bus Journey Experience, Transport Policy, 14, pp. 283-292.
Tan, D. D., Wang, W., Lu, J. (2007), Research on Methods of Assessing Pedestrian Level-of-
Service for Sidewalk. Journal of Transportation Systems Engineering and Information
Technology, Vol. 7, No. 5, pp. 74-79.
Wang, P., Lin, J. & Barnum, D. (2006), Data Envelopment Analysis of Bus Service Reliability
Using Automatic Vehicle Location Data. Submitted to the 86th Annual Meeting of
Transportation Research Board, Washington D.C.
290
COMPARISON OF SPACE MEAN SPEED USING CAR CHASING METHOD AND MOVING
VEHICLE METHOD AT TWO-LANE HIGHWAYS IN MALAYSIA
Mohd Azahar bin Awang 1* and Leong Lee Vien 2
ABSTRACT:
Engineers and planners have used travel time and delay studies since the late 1920s to evaluate
transportation facilities and plan improvement. Travel time and delay are two very important performance
measures in a highway system. A travel time study will provide data on the amount of time taken to
travel through a section of a road or highway and with the knowledge on that particular length of road
section under studied, space mean speed can be obtained. In this paper, estimation of space mean speeds
at various two-lane highways using car chasing method and moving vehicle method is discussed.
Statistical analyses were conducted to compare the means of space mean speeds obtained using both
methods. The results showed that means of speed obtained using the car chasing method and moving
vehicle method based on volume in vph were statistically significantly different but the means of speed
obtained using the car chasing method and moving vehicle method based on volume in pcu/h were not
significantly different at 95% confidence interval.
Keywords: Space Mean Speed, Car Chasing Method, Moving Vehicle Method and Two-lane Highway
1
Postgraduate student, Universiti Sains Malaysia, Engineering Campus, 14300 Nibong Tebal, Penang, MALAYSIA
2
Senior lecturer, Universiti Sains Malaysia, Engineering Campus, 14300 Nibong Tebal, Penang, MALAYSIA
*Correspondence Author: Mohd Azahar bin Awang, Universiti Sains Malaysia, Malaysia. Tel: +6004 5996266,
Fax: +604 5941009. E-mail: azahar47@yahoo.com
291
INTRODUCTION
A two-lane highway is an undivided roadway with two lanes, one for each direction of the traffic
flow. Overtaking a slower vehicle requires the use of the opposing lane as sight distance and gaps in the
opposing stream permit. Travel time and delay are two major measures of highway system performance.
Vehicle speed is directly related to travel time in which travel time varies inversely with travel speed. A
travel time study provides data on the amount of time taken to traverse a section of road or highway.
These data, combined with the length of section under study, produce mean travel speed. Speed is a rate
of motion expressed as distance per unit of time (Transportation Research Board, 2000). Speed is indeed
an important parameter in transportation studies because it influences the time of traveling, safety, cost,
convenience and economics (Cuthbert, 2006). According to the U.S. HCM 2000 (Transportation Research
Board, 2000), speed is important to measure the quality of the traffic service provided to the road users.
There are several methods to conduct travel-time and delay studies. For example the average
vehicle method, car chasing method, moving vehicle method, license plate method, direct observation
method and the interview method. Methods such as average vehicle method, car chasing method and
moving vehicle method require test cars to conduct the study while methods such as license plate method,
direct observation method and the interview method does not require test cars (Robertson 1994).
Nevertheless, this study focuses only on the comparisons between two methods of measurement, namely
the car chasing method and moving vehicle method.
Car chasing method measures the travel time, running time, distance traveled and variables such
as type, location, duration and the cause of traffic delays along the study route (Robertson, 1994). From
the data obtained, the space-mean speed, travel speed and running speed can be calculated. According to
Robertson (1994), the recommended minimum total length of the route is 1 mile (1.61 km).
Moving vehicle method provides estimates of hourly volume, average travel time and space mean
speed as a vehicle makes round trips through the sections being investigated. These estimates are
obtained by measuring travel time, opposing vehicles, overtaking vehicles and passed vehicles. This
method is only applicable on two-way segment length where the opposing traffic is visible at all times.
The test vehicle must also be able to turn around at each end of the test sections (Robertson, 1994).
STUDY METHODOLOGY
There In this study, three criteria of two-lane highways were used to select study site location.
The first and second criteria are that the minimum segment length for each site is 3.5 km and should be at
least has 1 km outside the influence of traffic signal (Refer figure 1). Traffic signals were avoided because
they cause congestion especially during peak hour. High volume intersections and driveways were
avoided because the speed reduction required to make a turning maneuver would skew the average travel
speed (Karjala, 2008). The third criterion for site selection was that the study sites should not be affected
by major geometric features such as grades, rolling ground or hilly site. The study site location should be
on flat and straight segment.
Figure 1: The minimum segment length allowed for two-lane highways
292
Data collections were conducted simultaneously using the car chasing method and moving
vehicle method. Two test cars were used as each method required one test car. The test cars used in this
study are the typical passenger cars found in Malaysia. Each test car will have two observers and a stop
watch to take time travel and stop or delay. For moving vehicle method, a video camera is required to
capture the opposing traffic volume, number of vehicles passed by the test car and the number of vehicles
overtaken by the test car.
Table 1 shows the suggested number of test runs to perform in a travel-time survey. The values
in Table 1 have a confidence level of 95% and are based on desired permitted error and the average range
in running speed. Running speed is used because it is a more stable measure than travel speed. The
required sample size can be determined after a minimum of two test runs but four initial test runs are
recommended. The average range in running speed was then calculated using equation (1).
R=
A (1)
N 1
Where:
R = Average range in running speed (km/hr)
A = Sum of absolute differences between each pair of sequential runs (km/hr)
N = Number of completed test runs
According to Robertson (1994), the suggested ranges of permitted errors in the estimation of mean
travel speed related to study purposes are as follows:
Transportation planning and highway needs studies: 3.0 mph (4.83 km/h) to 5.0 mph (8.05
km/h)
Traffic operations, trend analysis and economic evaluations: 2.0 mph (3.22 km/h) to 4.0 mph
(6.44 km/h)
Before-and-after studies: 1.0 mph (1.61 km/h) to 3.0 mph (4.83 km/h)
Table 1: Minimum sample-size requirements (Robertson, 1994)
Average range in Minimum number of runs for a permitted error of:

running speed 5.0 8.0
(km/hr) 2.0 km/hr 3.5 km/hr 6.5 km/hr
km/hr km/hr
5.0 4 3 2 2 2
10.0 8 4 3 3 2
15.0 14 7 5 3 3
20.0 21 9 6 5 4
25.0 28 13 8 6 5
30.0 38 16 10 7 6
Car Chasing Method
Car chasing method only requires a test car with a driver and an observer to observe travel time
for the study location. Firstly, the beginning and ending points on that road segment will be identified and
the distance between the two points will be measured. For this method, the test car will be driven at
another driver-desired speed in which the driver will have to drive at the speed of a randomly chosen car
in front and will follow the chosen car until that car leaves the area, parks or the chase requires unsafe
293
and/or illegal maneuvers, in which case the next nearest convenient vehicle is followed. During the shift
from one car to another, the test car is driven normally with respect to present traffic condition. The
amount of time to travel between the two points and stopping time will be measured and a stop is defined
as the absolute termination of motion of the test vehicle. Car chasing method is applicable and valid on
two-way directions or one-way direction roads.
The travel speed for each test vehicle run is then calculated using equation (2) while equation (3) is used
to calculate space-mean speed.
D
S = 60 (2)
T
Where:
S = Travel speed (km/hr)
D = Segment length of the test car travelled (km)
T = Travel time taken by the test car (minutes)
ND
S = 60 (3)
T

Where:
S = Space mean speed (km/hr)
D = Segment length of the test car travelled (km)
T = Sum of travel times for all test runs (minutes)
N = Number of test runs
Moving vehicle method
In this method, minimum of 2 observers will be needed to conduct the study. The beginning and
ending points are the same as those identified from car chasing method. The driver will drive at an
average speed and obeyed speed limit of the particular road while another observer will hold a video
camera and at the same time records the starting time and ending time for every segment. The video
camera will capture the opposing traffic volume, number of vehicles passed by the test car and the
number of vehicles overtaken by the test car. This method is applicable and valid on two-way directions
where opposing vehicle is visible at all times. The driver must be able to turn around instantaneously at
end of the pair run. Both volume and speed measurements are obtained simultaneously through this
method. This method is based on a survey vehicle that travels in both directions on the road.
Figure 2: Moving vehicle method (Robertson, 1994)
294
Referring to Figure 2, considering a test car which is travelling N-bound (from Point B to Point A),
volume, travel time and speed can be calculated using equation (4), (5) and equation (6) respectively:
M + On Pn
Vn = 60 s (4)
Tn + Ts
Where:
V n = Volume per hour, northbound
M s = Opposing number of vehicles met when the test car was travelling south
O n = Number of vehicles overtaking test car as it traveled north
Pn = Number of vehicles overtaken by the test car as it traveled north
Tn = Travel time when traveling north (minutes)
Ts = Travel time when traveling south (minutes)
The average travel time for one-directional flow is calculated using the following equation:
O Pn
Tn ( average ) = Tn 60 n (5)
Vn
Where T n(average) is the average travel time of all traffic northbound and the other variables are as defined
for equation (4).
Therefore, the space-mean speed for one directional flow is determined using equation (6).
d
S n = 60 (6)
T
n ( average )
Where S n is the space-mean speed in the northbound direction in km/h and d is the segment length of
test section in km.
However, it should be noted that, the value of average travel speed will be different when the unit of
traffic flow is different. Therefore, the effects of different traffic flow units which are in either vph or
pcu/h will also be examined in this study. The conversion factors used to convert the various type of
vehicle into passenger car unit (pcu) are based on the manual from Road Engineering Association of
Malaysia (REAM).
Table 2: Conversion factors (REAM, 2000)
Equivalence values
Type of Vehicles
Rural Standard Urban Standard
Passenger Cars 1.00 1.00
Motorcycles 1.00 0.75
Light Vans 2.00 2.00
Medium Lorries 2.50 2.50
Heavy Lorries 3.00 3.00
Buses 3.00 3.00
DATA COLLECTION
This study was conducted at 27 different locations of two-lane highways in Perak, Penang and
Kedah were chosen. Data collections were conducted from 9.00 am to 5.00 pm on weekdays during fine
weather. On the average, four test runs were performed at each site. Table 3 shows the details of study
locations.
295
Table 3: List of sites

Site No. Northbound Southbound State
1 Permatang Kerat Telunjuk Lubuk Buntar Kedah
2 Lubuk Buntar Serdang Kedah
3 Sungai Bongak Simpang Lima Perak
4 Taiping Kamunting Perak
5 Bukit Gantang Taiping Perak
6 Bukit Berapit Bukit Gantang Perak
7 Sungai Gadong Semanggol Perak
8 Taman Valdor Tasek Cempedak Penang
9 Tasek Cempedak Junjung Penang/ Kedah
10 Kulim Junjung Kedah
11 Padang Serai (Segment 1) Lunas Kedah
12 Padang Serai (Segment 2) Lunas Kedah
13 Tasik Gelugor Padang Serai Penang/ Kedah
14 Tapah Pekan Temoh Perak
15 Temoh Pekan Temoh Perak
16 Kampar Temoh Perak
17 Jeram (Segment 1) Gopeng Perak
18 Jeram (Segment 2) Gopeng Perak
19 Tikam Batu Bumbung 5 Kedah
20 Padang Tembusu Tikam Batu Kedah
21 Lubuk Kawah Sungai Petani Kedah
22 Gurun Perwaja Kedah
23 Guar Cempedak Gurun Kedah
24 Sedaka Guar Cempedak Kedah
25 Padang Lumat Sedaka Kedah
26 Sungai Kering Sungai Limau Dalam Kedah
27 Yan Besar Sungai Udang Kedah
RESULT AND DISCUSSION
The results for the space mean speed obtained from car chasing method are as shown in Table 4.
S N-S and S S-N represents the space mean speed of northbound and southbound directions of the test
vehicle respectively. The range of space mean speed obtained is between 57.32 km/hr to 83.17 km/hr.
296
Table 4: Space mean speed obtained using car chasing method

Space mean speed (km/h)
Site No.
S N-S (Northbound) S S-N (Southbound)
1 72.97 63.99
2 71.80 70.83
3 75.21 74.24
4 78.14 76.71
5 73.12 70.38
6 69.89 72.65
7 68.70 64.30
8 71.76 62.08
9 64.89 61.71
10 64.78 63.17
11 57.32 78.59
12 65.37 65.34
13 70.40 72.22
14 83.17 75.66
15 73.75 71.99
16 75.73 62.42
17 75.48 76.21
18 80.54 69.42
19 69.48 71.83
20 71.01 79.92
21 61.90 71.13
22 66.51 75.27
23 61.81 66.51
24 68.63 64.47
25 71.86 75.27
26 67.85 68.52
27 59.03 59.79
Average 70.04 69.80
Space mean speed study using the moving vehicle method was also conducted simultaneously at the same
study locations. The space mean speeds were calculated based on volume in vph and pcu/h as shown in
Table 5. From table 5, the range of space mean speed obtained based on volume in vph is 54.61 km/hr to
83.70 km/hr and for pcu/h is from 54.64 km/hr to 84.22 km/hr.
297
Table 5: Space mean speed obtained using moving vehicle method for volume per hour in vph and pcu/h
Space mean speed (km/h)
Based on volume per hour, vph Based on volume per hour, pcu/h
Site No.
S N-S S S-N S N-S S S-N
(Northbound) (Southbound) (Northbound) (Southbound)
1 65.24 74.60 66.71 74.60
2 60.56 61.85 59.79 61.04
3 81.08 70.38 83.27 68.25
4 66.43 70.82 66.00 71.02
5 81.94 81.12 84.22 83.60
6 60.84 65.68 59.94 66.71
7 83.70 74.80 81.02 74.37
8 61.31 66.82 62.54 67.39
9 59.05 57.01 56.75 59.65
10 62.78 60.05 64.34 59.70
11 59.91 62.56 59.10 62.47
12 62.17 62.51 62.24 62.60
13 66.08 64.88 66.39 65.48
14 75.72 69.79 76.57 72.03
15 63.60 62.36 64.22 61.68
16 72.41 72.66 71.05 72.66
17 73.64 81.33 73.72 82.10
18 73.81 75.35 74.02 75.51
19 63.51 63.81 64.72 64.17
20 65.66 76.53 65.97 77.43
21 67.63 67.86 67.77 67.98
22 60.49 66.93 62.00 68.02
23 58.03 57.16 58.83 57.55
24 66.98 63.39 68.09 65.03
25 67.16 71.25 68.61 71.05
26 68.43 69.83 69.05 69.83
27 56.96 54.61 57.00 54.64
Average 66.86 67.63 67.18 68.02
Subsequently, values of space mean speeds obtained using car chasing method and moving vehicle
method were compared by means of statistical analyses. Table 6 shows the values of space mean speeds
calculated from both methods which were used in the statistical analyses.
298
Table 6: Comparison of space mean speed

Space mean speed (km/hr)
Site Moving vehicle method
Direction Car chasing
No. Based on volume Based on volume
method
per hour, vph per hour, pcu/h
Northbound 72.97 65.24 66.71
1
Southbound 63.99 74.60 74.60
Northbound 71.80 60.56 59.79
2
Southbound 70.83 61.85 61.04
Northbound 75.21 60.98 83.27
3
Southbound 74.24 57.78 68.25
Northbound 78.14 81.08 66.00
4
Southbound 76.71 70.38 71.02
Northbound 73.12 66.43 84.22
5
Southbound 70.38 70.82 83.60
Northbound 69.89 81.94 59.90
6
Southbound 72.65 81.12 66.71
Northbound 68.70 65.68 81.02
7
Southbound 64.30 60.84 74.37
8 Northbound 71.76 74.80 62.54
Southbound 62.08 83.70 67.39
Northbound 64.89 66.82 56.75
9
Southbound 61.71 61.31 59.65
Northbound 64.78 59.05 64.34
10
Southbound 63.17 57.01 59.70
Northbound 57.32 60.05 59.10
11
Southbound 78.59 62.78 62.47
Northbound 65.37 62.56 62.24
12
Southbound 65.34 59.91 62.60
Northbound 70.40 62.17 66.39
13
Southbound 72.22 62.51 65.48
Northbound 83.17 64.88 76.57
14
Southbound 75.66 66.08 72.03
Northbound 73.75 75.72 64.22
15
Southbound 71.99 69.79 61.68
Northbound 75.73 62.36 71.05
16
Southbound 62.42 63.60 72.66
Northbound 75.48 72.41 73.72
17
Southbound 76.21 72.66 82.10
299
Table 6: Comparison of space mean speed (cont.)

Space mean speed (km/hr)
Site Moving vehicle method
Direction Car chasing
No. Based on volume Based on volume
method
per hour, vph per hour, pcu/h
Northbound 80.54 81.33 74.02
18
Southbound 69.42 73.64 75.51
Northbound 69.48 73.81 64.72
19
Southbound 71.83 75.35 64.17
Northbound 71.01 63.81 65.97
20
Southbound 79.92 63.51 77.43
Northbound 61.90 63.16 67.77
21
Southbound 71.13 68.84 67.98
Northbound 66.51 69.11 62.00
22
Southbound 75.27 63.11 68.02
Northbound 61.81 60.49 58.83
23
Southbound 66.51 66.93 57.55
Northbound 68.63 58.03 68.09
24
Southbound 64.47 57.16 65.03
Northbound 71.86 63.39 68.61
25
Southbound 75.27 66.98 71.05
Northbound 67.85 67.16 69.05
26
Southbound 68.52 71.25 69.83
Northbound 59.03 68.43 57.00
27
Southbound 59.79 69.83 54.64
Average 69.92 67.24 67.60
In this study, the t-test was conducted to compare the means for two groups of cases. In order to compare
the means of space mean speeds obtained using car chasing method and moving vehicle method for both
volumes in vph and pcu/h, t-test was first conducted between the values of speed obtained using car
chasing method and moving vehicle method for volume in vph. In this test, 95% confidence interval was
used. The group statistics for speed obtained using the car chasing method and moving vehicle method
for volume in vph is as shown in Table 7.
Table 7: Group statistics for speed obtained using the car chasing method and moving vehicle method
(volume in vph)
Std. Std. Error
Group N Mean
Deviation Mean
Car chasing method 54 69.9207 5.91041 0.80431
Speed Moving vehicle method
54 67.2419 7.14755 0.97266
(volume in vph)
300
The results obtained from Levenes Test are as shown in Table 8. Referring to Table 8, based on the
observed significant level of 0.202 which was greater than 0.05 (hence not significant at 0.05 level), null
hypothesis of equal variances was not rejected. Thus, the pooled-variance t-test was used. The results of
the pooled variance t-test are as shown in Table 9. Based on the observed significant level of 0.036 in
Table 9, null hypothesis was rejected and can be concluded that the means of speed obtained using the car
chasing method and moving vehicle method (volume in vph) were significantly different at 95%
confidence interval.
Table 8: Levenes test for speed obtained using the car chasing method and moving vehicle method
(volume in vph)
Levenes Test of Equality of Variances
Speed F = 1.647 Sig. = 0.202
Table 9: T-test results for speed obtained using the car chasing method and moving vehicle method
(volume in vph)
95% C.I. of the
Sig. Mean Std. Error
t d.f. Difference
(2-tailed) Difference Difference
Lower Upper
Equal variances
2.123 106 0.036 2.67889 1.26213 0.17659 5.18118
assumed
Subsequently, t-test was first conducted between the values of speed obtained using car chasing method
and moving vehicle method for volume in pcu/h. The group statistics for speed obtained using the car
chasing method and moving vehicle method for volume in pcu/h is as shown in Table 10.
Table 10: Group statistics for speed obtained using the car chasing method and moving vehicle method
(volume in pcu/h)
Std.
Std.
Group N Mean Error
Deviation
Mean
Car chasing method 54 69.9207 5.91041 0.80431
Speed
Moving vehicle method (volume in pcu/h) 54 67.6017 7.32175 0.99636
The results obtained from Levenes Test are as shown in Table 11. Based on the results in Table 11, the
observed significant level of 0.212 which was greater than 0.05 concluded that null hypothesis of equal
variances was not rejected. Thus, again, the pooled-variance t-test was used and the results are as shown
in Table 12. From Table 12, based on the observed significant level of 0.078, null hypothesis was not
rejected and can be concluded that the means of speed obtained using the car chasing method and moving
vehicle method (volume in pcu/h) were not significantly different at 95% confidence interval.
Table 11: Levenes test for speed obtained using the car chasing method and moving vehicle method
(volume in pcu/h)
Levenes Test of Equality of Variances
Speed F = 1.576 Sig. = 0.212
301
Table 12: T-test results for speed obtained using the car chasing method and moving vehicle method
(volume in pcu/h)
95% C.I. of the
Sig. Mean Std. Error
t d.f. Difference
(2-tailed) Difference Difference
Lower Upper
Equal variances
1.811 106 0.073 2.31907 1.28049 -0.21962 4.85776
assumed
CONCLUSION
In this paper, space mean speeds are estimated using car chasing method and moving vehicle
method at various two-lane highways in Malaysia. Statistical analyses were conducted to compare the
means of space mean speeds obtained using both methods. The results showed that means of speed
obtained using the car chasing method and moving vehicle method based on volume in vph were
statistically significantly different but the means of speed obtained using the car chasing method and
moving vehicle method based on volume in pcu/h were not significantly different at 95% confidence
interval.
ACKNOWLEDGEMENT
The authors wish to express their sincere gratitude to Highway Planning Unit, Ministry of Works,
Malaysia for funding this study under the project entitled Malaysian Highway Capacity Study Stage 3
(Inter-Urban).
REFERENCES
Ahmad Raqib, Wan hashim Wan Ibrahim & Ahmad Farhan Mohd Sadullah (2005). Estimating travel
time of arterial road using car chasing method and moving observer method. Journal of
transportation science society of Malaysia 1, 77-87
Al-Kaisy, A., & Karjala, S. (2008). Indicators of performance on two-lane rural highway. Journal of the
Transportation Research Board, 2088, 87-97.
Cuthbert, T. (2006). Flow, Speed and Capacity. PTRC London Lecture Series. Retrieved 20 may 2010,
from http://www.mvaconsultancy.com/publications/Flow_speed_capacity.pdf
Jerome, L. Catbagan, & Nakamura, H. (2008). Desired Speed Distributions on Two-Lane Highways
Under Various Conditions. Journal of the Transportation Research Board, 2088, 218226
Karjala, S. R. ( 2008). Estimating quality of traffic flow on two-lane highways. Masters thesis.
Montana State University.
Ministry of Works Malaysia. (1986). Arahan Teknik (Jalan) 8/86: A guide on geometric design of roads,
Robertson, H. D. (1994). Manual of Transportation Engineering Studies. (Robertson, H. D., Hummer, J.,
& Nelson, D. C., Eds.). Institute of Transportation Engineers.
Transportation Research Board. (2000). Highway Capacity Manual (2000). National Research Council.
Washington, DC.
302
EXPERT SYSTEM FOR PAVEMENT CONDITION RATING
Norlela Ismail 1*, Amiruddin Ismail 2, and Riza Atiq O.K. Rahmat 3
ABSTRACT:
Airport pavements are one of the important assets in a country. They should be well managed to ensure
the safety of the passengers and aircrafts are kept to their high standard. As the pavement structures
continuously deterioration over time due to traffic load and environmental factor, sustain and maintain
them in a safe and an adequate condition required an enormous investment in time and money. Right
maintenance at the right time can extend the life of pavement and substantial overall cost saving.
Pavement distress condition rating procedure presented by pavement condition index (PCI) was
developed by the U.S. Army Corps of Engineers in 1970s to help manage airport pavement. Since then,
the use of PCI for airport pavement has received wide acceptance and formally adopted as standard
procedure worldwide. Computing the PCI manually is not a tedious operation but the calculations
involved are time consuming. This paper presents the application of expert system for analysis and rating
the existing surface airport pavement condition. The expert system will automatically determine the PCI
value based on the procedure method in ASTM D5340 (Standard design method for airport pavement
condition index surveys). Using several sample unit of pavement, the system was tested and the results
show that this system has revealed satisfactorily findings in a faster PCI determination.
Keywords: Expert System, Pavement Distress, Pavement Condition Rating
1
Lecturer, Universiti Tenaga Nasional, KM 7, Jalan Kajang-Puchong, 43009 Kajang, Selangor, MALAYSIA
Postgraduate Student, Universiti Kebangsaan Malaysia, 43600 UKM Bangi, Selangor, MALAYSIA
2
Professor, Universiti Kebangsaan Malaysia, 43600 UKM Bangi, Selangor, MALAYSIA
3
Professor, Universiti Kebangsaan Malaysia, 43600 UKM Bangi, Selangor, MALAYSIA
*Correspondence Author: Norlela Ismail, Universiti Tenaga Nasional, Malaysia. Tel: +6003 8921 2020
ext. 2286, Fax: +60389212116. E-mail: norlelaismail@uniten.edu.my
303
INTRODUCTION
Airport networks are one of the largest community assets in most country that constitute an
enormous investment of public funds. Preserving this investment and maintaining the airport networks
operation in a safe and smooth condition is clearly a worthy objective. Airport pavements need to be
provided in an adequate load-carrying capacity and good ride quality for safe operation of aircraft under
whatever weather condition. However, the pavement structures begin to experience continuously
deterioration due to aircraft traffic load and environmental factor upon completion. If being unattended,
this deterioration, can eventually pose safety risks to airplanes taking off or landing which may increase
the possibility of aircraft accident and aircraft operation delayed. Serious pavement distress may also
require costly repairs in term of money and closure time. Adequate and timely maintenance can extend
the life of the pavement and substantial overall cost saving.
The pavement surface condition plays a pivotal role in the analysis and design of pavement
rehabilitation strategies (Ritchie 1987). Estimation of pavement surface condition includes the
determination of surface distress, ride quality or roughness, structural capacity, and surface friction
(Michael et al. 1998; Broten et al. 1998; Sanford et al. 2003). As distresses visible on the pavement
surface provide valuable input in the rehabilitation design process, many airport agencies conduct visual
survey and distress rating on pavement condition to support their decision in maintenance and
rehabilitation of a runway. The pavement condition index (PCI) developed by the US Army Corps of
Engineers in mid 1970s seems the most common airfield pavement distress condition rating used by
airport agencies to help manage the airport pavement (Josuelito, W , 2004; Shahin 2002, 1992; Hall et al.
1992; Michael et al. 1998). It has gained widespread acceptance and has been formally adopted as
standard procedure worldwide.
The PCI is a numerical index that rates pavements on a scale of 0 for a failed pavement to 100 for
a pavement in perfect condition. This index is developed to provide a measure of the pavement structural
adequacy and integrity as well as an insight cause of the experienced distresses in the pavement. As per
Shahin et al. (1979), PCI is the first tool used in the pavement rehabilitation process because it provides a
standard measure of the pavement condition in term of structural integrity and operational condition, an
objective and rational method for identifying maintenance and repair needs, and an early warning system
for identifying expensive repair project.
Computing the PCI manually for a single sample unit is not a tedious operation but it is time
consuming especially when involving large volume of data (Shahin 2002). Most airport engineers use
computer program such as PAVER software to speed up the distress evaluation process (Shahin et al.
1982). This paper presents the determination of PCI using an object oriented expert system based on
KAPPA-PC to rate the pavement condition. It is part of the functional modules developed for airport
pavement maintenance and rehabilitation expert system (ES-APAM).
METHODOLOGY
Expert System Design
In the development of expert system for airport pavement and rehabilitation, the knowledge-
based in airport pavement distress rating condition contain information relating to the type, severity, and
density of airport pavement distresses which acquired from various sources of expertise in the field of
airport pavement. The knowledge acquisition then being transfer and transform into an expert system
computer programmed using the rule based method that consist of an IF-THEN part where IF clause
presents premises or condition and THEN clause present conclusions. Rule-base method is commonly
being used in expert system as the rules can be modified easily to meet changing needs.
304
An object oriented expert system based on KAPPA-PC shell is used in this study as it offered the
advantages of ease of manipulation and increase productivity. Besides, it is also suitable to be used in
limited time and has substantial object capabilities (Basri 1999; NurIIyana et al. 2007). This software
package is also very attractive with highly graphic display (Mohamed and Mohamed Hussaini, 1999).
KAPPA-PC can integrate well the object-oriented programming and rules, which allows knowledge
representation using hybrid technique. Moreover, KAPPA-PC is chosen because of its proven reliability
and can be perform by both forward and backward chaining reasoning.
Airport Pavement Condition Rating
In this study, the airport pavement condition rating which expressed in PCI is made through the
visual condition survey that based on the procedures in ASTM D 5340 (The Standard Test Method for
Airport Pavement Condition Index Surveys). The type, severity, and quantity of distress are identified
upon field measurement. Weighted deduct value is then used to combine data on each distress type,
severity level, and distress density into a single condition value. For section samples that have distresses
with deduct value greater than five points, modified deduct value is used. This value is subtracted from
100 to give the PCI value. The steps used to determine the PCI value is illustrated in Figure 1. The
maximum allowable number of deduct, m and the PCI value are determined using the following
equations:
(1)
where is the highest deduct value for sample unit i.
(2)
Input: Distress type, severity level, and density
Step 1: Determine the deduct value (DV) for each distress type and severity
level combination from individual DV curve
Step 2: Determine the number of individual DV that is more than five points (q)
i) If q 1 Total DV = maximum CDV
ii) If q > 1 (Go to Step 5)
Step 3: Determine the max allowable number of deducts (m) (Equation 1)
Step 4: Determine total DV and Corrected Deduct Value (CDV) from CDV
curve
Step 5: Reduced the smallest q to five points
Step 6: Repeat Step 4 through Step 6 until q=1
Output: PCI value (Equation 2)
Figure 1: PCI Determination
305
The PCI is rated as in Table 1. The survey data is also used to compute the structural condition
index (SCI) and foreign object damage (FOD) condition index (FCI). These indices are tools that could
help users to quickly reduce the number of feasible rehabilitation option.
Table 1: Pavement Condition Index Ratings (Sources: Shahin M.Y. 2002)
PCI 100-86 85-71 70-56 55-41 40-26 25-11 10-0

RATING Good Satisfactory Fair Poor Very Poor Serious Failed
RESULTS AND DISCUSSION
The purpose of rating the airport pavement distress condition in this study is to collect and
provide input data of pavement condition in order to further process in analysis and design of pavement
maintenance and rehabilitation (M&R) strategies. The output of the system will include the cause of the
pavement problem, M&R treatment recommendation, cost, and method of performing the M&R
strategies.
The variables involved in the airport pavement distress condition rating are the distress type,
severity level, and quantity of distress. Sixteen types of distresses are considered in the development of
the expert system in this study, which is shown in Figure 2. User are required to specify the type, severity
level, and density of each particular distress that present in the pavement section through the visual
condition survey as in Figure 3. The system also guides the user in identifying the type of distress and
selects the correct severity level by describing each distress by severity. Figure 4 shows an example of
description for alligator crack distress type. After inserting the distress parameter, the user has to press the
forward button (in Figure 2) for making the system calculate the PCI, FCI, SCI and rate the pavement
condition that as shown in Figure 5.
Figure 2: Window for Distress Type Selection.
306
Figure 3: Distress Input Data for Alligator Cracking
Figure 4: Description Window of Alligator Cracking Type Distress
Figure 5: Window of Pavement Condition and Rating
307
In order to verify the result of the expert system in this study, the PCI determination was
analyzed by several samples unit of airport pavement as in Table 2. The result was then
compared to manual calculation and PAVER software. Table 3 shows that the result of PCI
calculated from the expert system are almost similar with the manual calculation and PAVER
program. The comparative result is illustrated clearly on graph in Figure 6.
Table 2: Input data of local sample
Sample 1 2 3 4 5 6 7
Facility Type Runway Runway Runway Runway Runway Runway Runway
Pavement Type AC AC AC AC AC AC AC
Sample Area 450 m2 450 m2 450 m2 450 m2 450 m2 450 m2 450 m2
Distress Type
Alligator Crack (L) - - 3.12 m - - - 2.20 m2
Long/Transverse Crack (L) 29.83 m 6.96 m 356.8 m 33.59 m 21.31 m 21.3 m 68.7 m
Long/Transverse Crack (M) - - 30 m - 7.5 m - -
Patching (L) - 2.03 m2 0.25 m2 2.25 m2 0.25 m 0.25 m 0.25 m
Rutting (L) 13.33m2 - 0.70 m2 - - - -
Bleeding - - - 1.30 m2 - - -
Table 3: Result of PCI Determination
Sample PCI
1 2 3 4 5 6 7 mean
Manual 74.5 93.5 48 76 62 92 78 74.85
Micro PAVER 74 94 48 77 62 92 79 75
ES-APAM 74.54 93.7 47.64 76.16 61.57 91.64 78.22 74.78
100
90
Pavement Condition Index
80
70
60
50
40
30
20
10
0
1 2 3 4 5 6 7
sample unit
Manual PAVER ES-APAM
Figure 6: PCI of Manual, PAVER and ES-APAM
308
CONCLUSION
After completion, airport pavements begin a gradually deterioration due to larger weight
and volumes of aircraft traffics and climatic factors. Failure to perform suitable treatment in the
early stage of deterioration may results in serious pavement distress that may require costly
repairs in term of money and closure time. Since pavement distress plays an important role in the
analysis and design of pavement rehabilitation strategies, many airport pavement conduct survey and
distress rating procedure based on pavement condition index to support their decision in
maintenance and rehabilitation of a runway. The used of PCI which is developed by the US Army Corps
of Engineers in 1970s has received wide acceptance and formally adopted as standard procedure
worldwide. This standard employs the visual distress in which distress type, severity, and quantity
are identified and rating system that rates the pavement condition from 0 for a failed pavement to
100 for a perfect pavement. PCI is also used to measure the structural integrity and surface
operational condition of a pavement.
The method of determining the numerical value of PCI is simple operation but the calculations
involved are time consuming. This paper presents the determination of PCI using an object-oriented
expert system based on KAPPA-PC to speed up the process of rating the pavement condition. The method
used in calculating the PCI is based on procedures in ASTM D5340. The expert system will
automatically calculate the PCI of each sample unit survey once the user enter the distress
information (i.e. type, severity, and quantity). Using several deteriorated sample unit of airport
pavement, this system was analyzed and the results were compared to manual calculation and
PAVER program. The similarity of PCI results indicate that the expert system has revealed
satisfactorily findings in a faster PCI determination.
ACKNOWLEDGMENTS
We would like to thank the Malaysia Airport Holding Berhad (MAHB) for providing
information and data.
REFERENCES
Basri, NEA (1999) An Expert System for the Design of Composting Facilities in Developing Countries.
PhD Dissertation, University of Leeds.
Broten, M. and Zimmerman, K. (1998) Basics of Airport Pavement Management. Proceeding 4th
International Conference on Managing Pavements, Durban South Africa
Josuelito, W (2004) engineering Technical Letter (ETL) 04-9: Pavement Engineering Assessment (EA)
Standards, Department of The Air Force, Tyndall AFB, FL
Hall, J.W., Grau, R. W., Grogan, W. P. and Hachiya, Y. (1992) Performance indicators from army airfield
pavement management program. Pavement Implementation, eds F.B. Holt and W.L. Gramling, STP
1121, American Society for Testing and material, Philadelphia, PA, 297-317.
Michael, G., and Patrick, S. (1998) Airport pavement management systems: an appraisal of existing
methodologies. Transportation Research Part A: Policy and Practice. Volume 32 (3), pp 197-214.
DOI 10.1016/S0965-8564(97)00008-6.
309
Mohamed, A., and H. Mohamad Hussaini (1999) Sistem Pakar Untuk pembelajaran Kualiti Kuasa.. J.
Kejuruteraan Universiti Kebangsaan Malaysia, 11: 2-2.
NurIlyana M.Z and Latifah A.M (2007), The 2nd National Intelligent Systems And Information
Technology Smmposium, ITMA-UPM, Malaysia.
Sanford-Bernhahardt, K.L., Loehr, J.E., and Huaco, D. (2003) Asset Management Framework for
Geotechnical Infrastructure. Journal Infrastructure Systems, ASCE 9(3), 107-116
Shahin, M.Y., Darter, M.I, and Kohn, S.D. (1979) Evaluation of airfield Pavement Condition and
Determination of Rehabilitation Needs. Transfort Research Record 700, pp 1-10
Shahin M.Y (1992) 20 years experience in the PAVER Pavement Management System: Development and
Implementation . In Pavement Management Implementation, eds F.B. Holt and W.L. Gramling STP
1121. American Society for Testing and Material. Philadelphia, PA 256-271
Shahin, M.Y (2002) Pavement Management For Airport, Roads, and Parking Lots. Chapman & Hall,
New York. ISBN-13:978-0387-23464-9.
Ritchie, S.G. (1987) Expert System in Pavement Management. Journal of Transportation Research. 21A
(2). pp 145-152. DOI 10.1016/0191-2607(87)90007-0
310
FREE-FLOW SPEED AT TWO-LANE HIGHWAYS IN MALAYSIA
Leong Lee Vien 1, Mustaqin Alpi 2*,

Hazwaruaida Muhammad
ABSTRACT:
Data were collected at 28 sites (56 lanes) of two-lane highways in the state of Kedah, Perak, Pahang,
Selangor, Negeri Sembilan, Melaka, Terengganu and Johor. Based on the results obtained, on the
average, the mean free-flow speed is 62.66 km/h, minimum free-flow speed is 32.49 km/h and the
maximum free-flow speed is 89.83 km/h. Most of the lower speeds were contributed by motorcycles and
in some cases trailers. As for the higher speeds, major contributors are cars, followed by lorries or large
vans and in some cases, trailers and motorcycles. On the average, traffic composition for Class 1 (cars/
small vans/ utilities) is 54.16% and the mean free-flow speed for Class 1 is 64.49 km/h. The average
traffic composition for Class 2 (lorries/ large vans) is 13.98%, but mean free-flow speed computed for
Class 2 which is 65.51 km/h is higher than that in Class 1. The average traffic composition for Class 3
(trailers) is 5.61% and the average free-flow speed computed is 64.43 km/h. The average traffic
composition for Class 4 (buses) is only 0.16% with mean free-flow speed of 68.51 km/h. The lowest
mean free-flow speed of 53.84 km/h was observed for vehicles in Class 5 (motorcycles), with average
traffic composition of 29.04%.
Keywords: free flow speed, mean free flow speed, traffic composition
1
Lecturer, Universiti Sains Malaysia, Nibong Tebal, 14300, Penang, MALAYSIA
2
Student, Universiti Sains Malaysia, Nibong Tebal, 14300, Penang, MALAYSIA
*Correspondence Author: Mustaqin b Alpi, Universiti Sains, Malaysia. Tel: +6004 5996266
Fax: +6045941009. E-mail: mus_wm@yahoo.com
311
INTRODUCTION
A two-lane highway is an undivided roadway with two lanes, one for each direction of the traffic flow.
Usually this kind of roadway does not have any barriers between the two lanes. Passing a slower vehicle
requires the use of the opposing lane as sight distance and gaps in the opposing stream permit. Two-lane
highways are key elements in the highway system of most countries. They perform a variety of functions.
They are located in all geographic areas and serve a wide range of traffic. In Malaysia, all the two-lane
highways are connected with each other through a network. Usually two-lane highways are located in
rural areas connecting with other rural or urban areas.
Speed is a fundamental measurement of the traffic performance on the road system. Speed is defined as
the rate of movement of a vehicle in distance per unit of time. The more common unit of speed is
kilometer per hour (km/h).
LITERATURE REVIEW
Free-flow speed is a speed that the drivers can drive their vehicle without obstruction and can speed with
their own desired speed and not be influenced by other road users but influenced by characteristics of the
vehicle , the driver, the road and (road) condition such as weather and traffic rules (speed limits)
(Hoogendoorn, 2005). According to U.S. Highway Capacity Manual 2000 (Transportation Research
Board, 2000), free-flow speed for two-lane highway is the speed of traffic during low traffic flow (up to a
two-way flow of 200 pcu/hr) and low density. It is the speed at which drivers feel comfortable travelling
under the physical, environment, and traffic control conditions on an uncongested section of two lane
highways. The important of free-flow speed is that it is a starting point for analyzing capacity and level of
service (LOS) for uninterrupted flow conditions. In a study conducted by Botma study (as cited in
Hoogendoorn, 2005) free-flow speed is presented in terms of how individual drivers choose their free
speed, discussing a behavioural model relating the free speed of a driver to a number of counteracting
mental stresses a driver is subjected to. According to a study conducted by Brilon and Weiser (2006),
travel speed that the passenger cars can maintain under expected peak hour condition is the primary
interest where realistic representations of speed-flow relationship for two lane rural highways are of
considerable importance for the layout of regional road network. The free-flow speed and its variability
across drivers are important safety factors in setting speed limits and designing roadways (Medina and
Tarko, 2005).
According to a study conducted by Medina and Tarko (2005) to investigate speed factors under free-flow
conditions for two-lane rural highways in Indiana, U.S., free-flow vehicles were identified based on time
headways of 5 s or more and free-flow speed were measured on weekdays during daylight hours and
favourable weather condition.
However, Bang et al. (1996) in their study to develop speed-flow relationships for rural roads in Indonesia
stated that free-flow speed was determined for unobstructed vehicles based on the definition of vehicles
with headway to the nearest vehicle in front of more than 8 s and no recent or immediate meeting with a
vehicle in the opposing direction. The speed-flow relationships is determine from empirical speed-flow
data for even a limited number of standard road classes, traffic and environment (roadside) condition and
data collecting is conducted at a very large number of sites. Another way is to use a simulation model
calibrated for actual driver behaviour and vehicle characteristics. In a study conducted by Al-Kaisy and
Karjala (2008), free-flow speeds were also calculated by averaging the speed of all vehicles travelling
with headways greater than 8 s. The average travel speed as a percentage of free-flow speed is an
indicator of the amount of speed reduction due to traffic. If average travel speed is close to free-flow
speed, the interaction among successive vehicle in the traffic stream is small and a high level of service
(performance) is expected.
312
However, in the study conducted by Yagar and Aerde (1983) on the effects of the geometric and
environmental conditions on speeds for two-lane rural highways in Ontario, Canada to control the varying
traffic volumes, the speeds used for estimating the geometric effects were those corresponding to main
directions volumes of 900 pcu/hr and opposing volumes of about 300 pcu/hr.
According to U.S. Highway Capacity Manual 2000 (Transportation Research Board, 2000), the field
study should be conducted in periods of low traffic flow (up to a two way flow of 200 pc/h) and should
measure the speeds of all vehicles or of a systematic sampling (e.g., of every 10th vehicle).
The study conducted by Bang et al. (1996) in Indonesia to develop speed-flow relationships for rural
roads has concluded that factors such as lane width and side friction (public transit stops, pedestrians,
non-motorised vehicles, and entries and exits from roadside properties and minor roads) and road
functional class (arterial, collector or local) reduced free-flow speed significantly.
Apart from that, horizontal curvature is widely documented as a primary speed factor. Other elements
such as spirals, speed limit, enforcement level, cross section width, longitudinal grade, side slope, general
terrain, drivers attitude and vehicle acceleration and deceleration capabilities were identified as less
important factors. The restriction imposed in spot location was the speeds must not be affected by any
traffic signal or stop sign (Medina and Tarko, 2005).
According to U.S. Highway Capacity Manual 2000 (Transportation Research Board, 2000), free-flow
speed will be lower on sections of highway with restricted vertical or horizontal alignment and it is also
tend to be lower when posted speed limits are lower. In a study conducted by McLean (as cited in Staplin
et al., 1997) to investigate driver speed behaviour and rural road alignment design found that there is a
strong effect of curvature on speed. This is true because the drivers tent to slower their vehicle before
moving into the curvature or bend. Curves with radii less than 400 m (1,300 ft) have a particularly high
accident rate and relative to other alignment properties, road curvature has the greatest influence on driver
speed behaviour.
However, Yagar and Aerde (1983) who studied the effects of the geometric and environmental conditions
on speeds concluded that land use and speed limit have the most significant impact on speeds for two-lane
rural highways in which 70% of any variation in speed limit was reflected directly in operating speeds,
while adjacent land used (the presence of land use along the 1500 m of road leading up to the location)
decreased speeds by about 8 km/h. Apart from that, they also found that lane width, gradient and
accessibility were also statistically significant. The sensitivity of speed to lane width was 6 km/h per m.
However, road curvature, presence of extra lane, sight distance, centre lane markings and lateral
obstructions were not found to have statistically significant effects on speed.
Apart from the geometric design of the road, environment effects also were taken into account in
determining the vehicle speed. Ibrahim and Hall (as cited in Kyte et al., 2000) studied the effect of
adverse weather on freeway operations in Canada. They conducted tests on the effects of rain and snow
on speed-flow occupancy relationships, summarizing their findings into three categories, namely clear
and rainy weather, clear and snowy weather, and rainy and snowy weather.
A study conducted by Brilon and Ponzlet (as cited in Kyte et al., 2000) in Germany to assess the effects of
weather condition, daylight or darkness and other factor on speed-flow relationships, concluded that
darkness reduced driver speeds by 5 km/h and wet roadway condition reduce driver speeds by 9.5km/hr.
313
METHODOLOGY
The basic survey equipment at each site consists of a CCTV setup, which includes a laptop, a cell battery
and a set of power inverter and a 10-m pole to place the CCTV. It is important to make sure that the
equipment is properly calibrated before data collection.
Lane widths, shoulder widths (paved and unpaved), length of passing zone and number of access points
were recorded manually and vehicles in the opposing directions were also recorded using a separate video
recorder. This is to enable the segregation process during data reduction in which speeds for unobstructed
vehicles with headway to the nearest vehicle in front of more than 8 s and no recent or immediate meeting
with a vehicle in the opposing direction can be obtained.
Subsequently, by using the TRAIS Advance SRM 2.8 software (which uses the state of the art
intelligent image processing technology), traffic flow and composition, speed, headways and gaps were
obtained automatically. Upon which speed data will be segregated for vehicles more than 8 s and without
meeting any vehicles in the opposing direction, and then averaged and convert to space-mean speed to
obtain the space-mean free-flow speed.
RESULT ANALYSIS
Space-mean free-flow speed were calculated based on the averaged time-mean free-flow speed using the
space-mean speed time-mean speed equation developed earlier in this study. The equation is as shown
below.
vs = 1.030vt 2.667 (1)
Where
vt = time mean speed
vs = space-mean speed
In this study, there are five classes of vehicles as shown below:

Class 1 - Cars/Small Vans/Utilities
Class 2 - Lorries/Large Vans
Class 3 - Trailers
Class 4 - Buses
Class 5 - Motorcycles
Table 1 shows the results of free-flow speed calculated for the 28 sites (56 lanes) of two-lane highways in
Malaysia. Based on the results obtained, the ranges of mean, minimum and maximum free-flow speeds
are between 40 to 75 km/h, 20 to 53 km/h and 75 to 105 km/h respectively. Most of the lower speeds
were contributed by motorcycles and in some cases trailers. As for the higher speeds, major contributors
are cars, followed by lorries or large vans and in some cases, trailers and motorcycles. On the average,
the mean free-flow speed is 62.66 km/h, minimum free-flow speed is 32.49 km/h and the maximum free-
flow speed is 89.83 km/h.
314
Table 1: Mean, minimum and maximum free-flow speeds
FFS(km/h)
Site
Site location Mean Minimum Maximum
No.
FFS FFS FFS
(km/h) Vehicle class Vehicle class
(km/h) (km/h)
1 Changkat Kinding to Chemor 55.43 24.90 Class 4 - Motorcycles 84.74 Class 1 - Cars/Small Vans/Utilities
2 Chemor to Changkat Kinding 52.20 18.90 Class 4 - Motorcycles 76.53 Class 1 - Cars/Small Vans/Utilities
3 Batu Gajah to Kampung Gajah 54.88 25.81 Class 4 - Motorcycles 80.20 Class 1 - Cars/Small Vans/Utilities
4 Kampung Gajah to Batu Gajah 64.96 33.62 Class 4 - Motorcycles 88.96 Class 1 - Cars/Small Vans/Utilities
5 Bota Kanan to Padang Changkat 60.08 24.84 Class 4 - Motorcycles 103.67 Class 3 - Trailers
6 Padang Changkat to Bota Kanan 70.81 30.39 Class 4 - Motorcycles 99.50 Class 2 - Lorries/Large Vans
7 Pangkor to Segari 67.07 35.81 Class 4 - Motorcycles 99.08 Class 1 - Cars/Small Vans/Utilities
8 Segari to Pangkor 58.64 39.45 Class 1 - Cars/Small Vans/Utilities 86.87 Class 2 - Lorries/Large Vans
9 Beruas to Pantai Remis 56.18 26.12 Class 4 - Motorcycles 89.38 Class 1 - Cars/Small Vans/Utilities
10 Pantai Remis to Beruas 62.95 19.17 Class 4 - Motorcycles 86.27 Class 1 - Cars/Small Vans/Utilities
11 Parit Buntar toTiti Serong (Parit Tok Ngah-Tjg Piandang) 58.86 24.62 Class 4 - Motorcycles 88.37 Class 1 - Cars/Small Vans/Utilities
12 Titi Serong to Parit Buntar (Tjg Piandang-Parit Tok Ngah) 56.70 22.64 Class 4 - Motorcycles 87.75 Class 1 - Cars/Small Vans/Utilities
13 Titi Serong to Tanjung Piandang 67.40 41.05 Class 3 - Trailers 92.23 Class 1 - Cars/Small Vans/Utilities
14 Tanjung Piandang to Titi Serong 65.64 33.03 Class 4 - Motorcycles 92.23 Class 1 - Cars/Small Vans/Utilities
15 Kg Hj Sharif (Temerloh) to Kg Jenderak (Jerantut) 71.16 34.80 Class 4 - Motorcycles 103.77 Class 4 - Motorcycles
16 Kg Jenderak (Jerantut) to Kg Hj Sharif (Temerloh) 74.72 40.35 Class 4 - Motorcycles 99.22 Class 2 - Lorries/Large Vans
17 Rumpun Makmur (Jerantut) to Gunung Senyum (Jengka) 67.64 31.61 Class 4 - Motorcycles 88.52 Class 2 - Lorries/Large Vans
18 Gunung Senyum (Jengka) to Rumpun Makmur (Jerantut) 67.75 33.43 Class 4 - Motorcycles 95.15 Class 1 - Cars/Small Vans/Utilities
19 Kg Baru (Benta) to Kg Latar Kasah (Jerantut) 67.81 32.35 Class 4 - Motorcycles 104.53 Class 3 - Trailers
20 Kg Latar Kasah (Jerantut) to Kg Baru (Benta) 60.02 30.47 Class 4 - Motorcycles 88.69 Class 1 - Cars/Small Vans/Utilities
21 Jengka 8 to Bandar Pusat Jengka 64.83 20.12 Class 4 - Motorcycles 88.88 Class 1 - Cars/Small Vans/Utilities
22 Bandar Pusat Jengka to Jengka 8 71.99 41.55 Class 4 - Motorcycles 95.67 Class 2 - Lorries/Large Vans
23 Bandar Pusat Jengka to Jengka 15 61.26 32.40 Class 4 - Motorcycles 89.75 Class 1 - Cars/Small Vans/Utilities
24 Jengka 15 to Bandar Pusat Jengka 59.01 30.56 Class 4 - Motorcycles 86.80 Class 1 - Cars/Small Vans/Utilities
25 Hulu Langat to Kajang 43.02 24.15 Class 4 - Motorcycles 76.83 Class 1 - Cars/Small Vans/Utilities
26 Kajang to Hulu Langat 38.76 19.29 Class 4 - Motorcycles 75.47 Class 1 - Cars/Small Vans/Utilities
27 Semenyih to Hulu Langat 58.50 36.00 Class 4 - Motorcycles 82.18 Class 1 - Cars/Small Vans/Utilities
28 Hulu Langat to Semenyih 51.55 24.69 Class 4 - Motorcycles 89.15 Class 4 - Motorcycles
29 Pedas to Sepri 69.21 35.30 Class 4 - Motorcycles 89.29 Class 3 - Trailers
315
Table 1: Mean, minimum and maximum free-flow speeds (cont..)
FFS(km/h)
Site
Site location Mean Minimum Maximum
No.
FFS FFS FFS
(km/h) Vehicle class Vehicle class
(km/h) (km/h)
30 Sepri to Pedas 57.22 19.78 Class 4 - Motorcycles 86.92 Class 1 - Cars/Small Vans/Utilities
31 Rembau to Pulau Mampat 59.68 31.75 Class 4 - Motorcycles 84.33 Class 1 - Cars/Small Vans/Utilities
32 Pulau Mampat to Rembau 61.78 31.99 Class 4 - Motorcycles 89.14 Class 1 - Cars/Small Vans/Utilities
33 Bongek to Gadong 57.52 31.14 Class 4 - Motorcycles 88.81 Class 1 - Cars/Small Vans/Utilities
34 Gadong to Bongek 58.32 33.98 Class 4 - Motorcycles 86.89 Class 1 - Cars/Small Vans/Utilities
35 Durian Tunggal to Alor Gajah 64.57 29.61 Class 4 - Motorcycles 100.86 Class 2 - Lorries/Large Vans
36 Alor Gajah to Durian Tunggal 55.51 24.56 Class 4 - Motorcycles 88.56 Class 1 - Cars/Small Vans/Utilities
37 Serdang to Lubuk Buntar 68.04 38.66 Class 4 - Motorcycles 101.58 Class 3 - Trailers
38 Lubuk Buntar to Serdang 56.83 22.14 Class 4 - Motorcycles 82.85 Class 1 - Cars/Small Vans/Utilities
39 Lubuk Buntar to Permatang Kerat Telunjuk 61.07 22.57 Class 4 - Motorcycles 87.37 Class 4 - Motorcycles
40 Permatang Kerat Telunjuk to Lubuk Buntar 61.09 30.66 Class 3 - Trailers 84.80 Class 3 - Trailers
41 Merbau Pulas to Kuala Ketil 59.67 31.58 Class 4 - Motorcycles 85.71 Class 1 - Cars/Small Vans/Utilities
42 Kuala Ketil to Merbau Pulas 59.32 36.68 Class 4 - Motorcycles 83.08 Class 1 - Cars/Small Vans/Utilities
43 Bukit Besi to Jerangau 67.40 41.05 Class 3 - Trailers 92.23 Class 1 - Cars/Small Vans/Utilities
44 Jerangau to Bukit Besi 63.26 39.31 Class 3 - Trailers 89.85 Class 1 - Cars/Small Vans/Utilities
45 Simpang Kuala Terengganu to Bukit Besi 72.58 38.71 Class 4 - Motorcycles 90.88 Class 2 - Lorries/Large Vans
46 Bukit Besi to Simpang Kuala Terengganu 73.59 52.37 Class 1 - Cars/Small Vans/Utilities 96.66 Class 2 - Lorries/Large Vans
47 Kluang (Agromas) to Bandar Tenggara (Petronas) 64.79 35.82 Class 4 - Motorcycles 83.21 Class 1 - Cars/Small Vans/Utilities
48 Bandar Tenggara (Petronas) to Kluang (Agromas) 71.07 23.19 Class 4 - Motorcycles 101.57 Class 3 - Trailers
49 Kg Jenglau to Kg Parit Raja 59.69 40.09 Class 1 - Cars/Small Vans/Utilities 88.60 Class 1 - Cars/Small Vans/Utilities
50 Kg Parit Raja to Kg Jenglau 61.50 27.36 Class 4 - Motorcycles 88.41 Class 1 - Cars/Small Vans/Utilities
51 Kg Baru to Pantai Sg Miang 61.80 18.36 Class 4 - Motorcycles 89.75 Class 1 - Cars/Small Vans/Utilities
52 Pantai Sg Miang to Kg Baru 72.96 37.19 Class 4 - Motorcycles 101.60 Class 1 - Cars/Small Vans/Utilities
53 Pekan to Permatang Keledang 60.39 30.28 Class 4 - Motorcycles 88.90 Class 2 - Lorries/Large Vans
54 Permatang Keledang to Pekan 51.44 24.75 Class 4 - Motorcycles 82.96 Class 1 - Cars/Small Vans/Utilities
55 Kg Gadong to Merchong 59.12 31.08 Class 4 - Motorcycles 87.61 Class 1 - Cars/Small Vans/Utilities
56 Merchong to Kg Gadong 65.57 32.51 Class 4 - Motorcycles 89.49 Class 3 - Trailers
316
Table 2 shows the results of mean free-flow speed and traffic composition. Based on the results in traffic
composition, on the average, traffic composition for Class 1 is 54.16% and the mean, minimum and
maximum free-flow speed for Class 1 is 50.23 km/h, 64.49 km/h and 73.17 km/h respectively. The
average traffic composition for Class 2 is 13.98%, but the range of free-flow speed for Class 2 is larger
than Class 1 in which the minimum and maximum values of free-flow speed for Class 2 are 44.11 km/h
and 88.33 km/h respectively. Therefore, the mean free-flow speed computed for Class 2 which is 65.51
km/h is higher than that in Class 1. However, the accuracy of the values obtained is somehow low as the
number of data sampled is small.
The average traffic composition for Class 3 is 5.67% and the average free-flow speed computed is 64.43
km/h while the minimum and maximum values of free-flow speed are 48.67 km/h and 91.93 km/h
respectively. The average traffic composition for Class 4 is only 0.16% with the mean free-flow speed of
68.51 km/h and the minimum and maximum free-flow speeds are 57.05% and 78.49% respectively. The
lowest mean, minimum and maximum free-flow speeds of 53.84 km/h, 34.30 km/h and 70.81 km/h were
observed for vehicles in Class 5 with average traffic composition of 26.04%.
Subsequently, the one-way ANOVA test was conducted. The one-way ANOVA procedure produces a
one-way analysis of variance for a quantitative dependent variable by a single factor or independent
variable. It is used to test the hypothesis that several means are equal where:
H 0 : 1 = 2 = .. = k (2)
and the alternative hypothesis says that there is a difference among the means:
H 1 : 1 2 .. k (3)
The F statistics constructed for testing the hypothesis is as shown below:

Between - groups sum of squares/(k - 1) between - groups mean square
F= = (4)
Within - groups sum of squares/(n - k ) within - groups mean square
where k is the number of groups and n is the total sample size. Therefore, if the means are far apart,
especially relative to the variation within each group, the size of the F statistic is large and the null
hypothesis is rejected. The rejection of H 0 implies that the regression equation differs from a constant
and shows that at least one regressor variable is important. The assumptions required by the analysis of
variance are as follows:
Independent random samples from each population
The populations are normal
The population variances are equal (Levenes test was used to test on the homogeneity of variance)
The alternative hypothesis however, doesnt say which groups differ significantly from each other.
Therefore, post hoc range tests and pairwise multiple comparisons can determine which means differ.
Range tests identify homogeneous subsets of means that are not different from each other while pairwise
multiple comparisons test the difference between each pair of means. Table 3, 4 and 5 show the results
obtained for Levenes test of homogeneity of variance, one-way ANOVA and post hoc test respectively.
317
Table 2: Mean free-flow speed and traffic composition

Vehicle class
Class 1 - Class 2 -
Class 3 - Class 4 - Class 5 -
Cars/Small Lorries/Large All vehicles
Site Trailers Buses Motorcycles
Site location Vans/Utilities Vans
No.
Mean Mean Mean Mean Mean Mean
Total
FFS (%) FFS (%) FFS (%) FFS (%) FFS (%) FFS
(%)
(km/h) (km/h) (km/h) (km/h) (km/h) (km/h)
1 Changkat Kinding to Chemor 58.28 51.52 61.63 12.12 64.48 4.55 - 0.00 47.16 31.82 55.43 100
2 Chemor to Changkat Kinding 56.37 46.55 62.34 6.90 48.67 1.72 - 0.00 46.43 44.83 52.20 100
3 Batu Gajah to Kampung Gajah 58.88 56.92 44.11 3.08 - 0.00 - 0.00 50.03 40.00 54.88 100
4 Kampung Gajah to Batu Gajah 67.05 46.94 68.54 30.61 - 0.00 - 0.00 55.69 22.45 64.96 100
5 Bota Kanan to Padang Changkat 63.35 60.92 68.86 6.90 71.32 4.60 - 0.00 48.80 27.59 60.08 100
6 Padang Changkat to Bota Kanan 70.40 53.75 77.23 23.75 91.93 6.25 - 0.00 54.64 16.25 70.81 100
7 Pangkor to Segari 71.55 45.33 65.74 18.67 61.70 8.00 - 0.00 62.23 28.00 67.07 100
8 Segari to Pangkor 59.12 61.90 62.72 7.14 50.31 4.76 - 0.00 57.92 26.19 58.64 100
9 Beruas to Pantai Remis 63.86 50.00 55.60 7.58 52.17 4.55 - 0.00 46.65 37.88 56.18 100
10 Pantai Remis to Beruas 64.36 57.14 67.45 12.70 49.77 3.17 - 0.00 59.42 26.98 62.95 100
11 Parit Buntar toTiti Serong (Parit Tok Ngah-Tjg Piandang) 64.63 52.94 62.73 6.86 69.49 1.96 - 0.00 49.63 38.24 58.86 100
12 Titi Serong to Parit Buntar (Tjg Piandang-Parit Tok Ngah) 61.44 58.67 52.83 8.00 55.21 2.67 - 0.00 48.78 30.67 56.70 100
13 Titi Serong to Tanjung Piandang 69.69 78.31 64.53 9.64 58.71 7.23 - 0.00 48.97 4.82 67.40 100
14 Tanjung Piandang to Titi Serong 69.71 64.00 62.86 14.00 62.18 7.00 - 0.00 52.50 15.00 65.64 100
15 Kg Hj Sharif (Temerloh) to Kg Jenderak (Jerantut) 70.71 70.73 73.03 10.98 74.98 4.27 - 0.00 70.81 14.02 71.16 100
16 Kg Jenderak (Jerantut) to Kg Hj Sharif (Temerloh) 72.30 69.08 88.33 19.08 74.68 4.61 - 0.00 61.93 7.24 74.72 100
17 Rumpun Makmur (Jerantut) to Gunung Senyum (Jengka) 72.25 59.32 78.03 13.56 63.42 3.39 - 0.00 50.78 23.73 67.64 100
18 Gunung Senyum (Jengka) to Rumpun Makmur (Jerantut) 70.28 67.14 71.79 12.86 73.88 1.43 - 0.00 55.32 18.57 67.75 100
19 Kg Baru (Benta) to Kg Latar Kasah (Jerantut) 66.23 32.26 72.90 28.23 68.06 30.65 - 0.00 56.47 8.87 67.81 100
20 Kg Latar Kasah (Jerantut) to Kg Baru (Benta) 60.70 45.21 61.34 20.55 62.00 13.70 - 0.00 55.90 20.55 60.02 100
21 Jengka 8 to Bandar Pusat Jengka 68.86 52.94 65.56 12.61 54.45 4.20 65.10 0.84 58.73 29.41 64.83 100
22 Bandar Pusat Jengka to Jengka 8 72.22 59.69 77.37 22.48 79.61 3.10 - 0.00 61.24 14.73 71.99 100
23 Bandar Pusat Jengka to Jengka 15 64.04 63.53 60.24 3.53 61.22 7.06 - 0.00 54.58 25.88 61.26 100
24 Jengka 15 to Bandar Pusat Jengka 62.97 55.71 66.96 7.14 85.93 1.43 - 0.00 50.17 35.71 59.01 100
25 Hulu Langat to Kajang 50.23 29.63 53.94 1.85 - 0.00 - 0.00 39.60 68.52 43.02 100
26 Kajang to Hulu Langat 50.45 25.49 52.03 1.96 - 0.00 - 0.00 34.30 72.55 38.76 100
27 Semenyih to Hulu Langat 61.22 42.53 54.94 16.09 54.92 11.49 - 0.00 57.92 29.89 58.50 100
28 Hulu Langat to Semenyih 51.95 52.59 53.15 9.48 51.28 8.62 - 0.00 50.39 29.31 51.55 100
318
Table 2: Mean free-flow speed and traffic composition (cont..)

Vehicle class
Class 1 Class 2
Class 3 Class 4 Class 5
Cars/Small Lorries/Large All vehicles
Site Trailers Buses Motorcycles
Site location Vans/Utilities Vans
No.
Mean Mean Mean Mean Mean Mean
Total
FFS (%) FFS (%) FFS (%) FFS (%) FFS (%) FFS
(%)
(km/h) (km/h) (km/h) (km/h) (km/h) (km/h)
29 Pedas to Sepri 68.25 41.54 71.20 35.38 71.13 3.08 - 0.00 67.38 20.00 69.21 100
30 Sepri to Pedas 59.00 51.11 61.98 33.33 - 0.00 - 0.00 41.18 15.56 57.22 100
31 Rembau to Pulau Mampat 63.61 59.38 - 0.00 - 0.00 - 0.00 53.91 40.63 59.68 100
32 Pulau Mampat to Rembau 65.61 58.33 72.21 8.33 - 0.00 - 0.00 52.46 33.33 61.78 100
33 Bongek to Gadong 60.55 49.09 65.88 10.91 - 0.00 - 0.00 51.52 40.00 57.52 100
34 Gadong to Bongek 60.68 40.63 62.94 25.00 66.60 1.56 - 0.00 51.50 32.81 58.32 100
35 Durian Tunggal to Alor Gajah 66.66 44.87 85.23 7.69 65.11 1.28 76.95 3.85 57.45 42.31 64.57 100
36 Alor Gajah to Durian Tunggal 61.92 46.03 65.98 4.76 - 0.00 - 0.00 48.49 49.21 55.51 100
37 Serdang to Lubuk Buntar 67.54 46.07 72.64 24.72 76.85 5.62 78.49 1.12 61.28 22.47 68.04 100
38 Lubuk Buntar to Serdang 59.52 45.54 58.72 14.29 59.19 8.04 - 0.00 51.57 32.14 56.83 100
39 Lubuk Buntar to Permatang Kerat Telunjuk 64.38 37.14 57.92 19.05 61.94 16.19 - 0.00 58.28 27.62 61.07 100
40 Permatang Kerat Telunjuk to Lubuk Buntar 63.76 37.60 60.47 25.60 63.52 14.40 - 0.00 56.40 21.60 61.09 100
41 Merbau Pulas to Kuala Ketil 62.70 42.17 62.82 13.25 60.12 12.05 - 0.00 54.30 32.53 59.67 100
42 Kuala Ketil to Merbau Pulas 61.83 43.27 62.15 11.54 62.81 12.50 - 0.00 53.67 32.69 59.32 100
43 Bukit Besi to Jerangau 69.69 78.31 64.53 9.64 58.71 7.23 - 0.00 48.97 4.82 67.40 100
44 Jerangau to Bukit Besi 64.70 72.41 63.18 8.62 60.03 8.62 - 0.00 55.95 10.34 63.26 100
45 Simpang Kuala Terengganu to Bukit Besi 73.17 61.54 82.10 17.95 59.38 3.85 - 0.00 63.19 16.67 72.58 100
46 Bukit Besi to Simpang Kuala Terengganu 73.17 82.72 78.42 12.35 72.64 2.47 - 0.00 64.75 2.47 73.59 100
47 Kluang (Agromas) to Bandar Tenggara (Petronas) 66.45 65.79 66.94 15.79 58.83 10.53 - 0.00 54.67 7.89 64.79 100
48 Bandar Tenggara (Petronas) to Kluang (Agromas) 71.58 61.76 74.03 17.65 80.64 5.88 - 0.00 61.53 14.71 71.07 100
49 Kg Jenglau to Kg Parit Raja 60.88 73.53 58.06 11.76 58.41 4.41 57.05 1.47 52.97 8.82 59.69 100
50 Kg Parit Raja to Kg Jenglau 61.91 59.38 70.23 9.38 63.76 4.17 - 0.00 57.22 27.08 61.50 100
51 Kg Baru to Pantai Sg Miang 66.47 64.56 52.73 5.06 55.56 8.86 - 0.00 52.54 21.52 61.80 100
52 Pantai Sg Miang to Kg Baru 70.62 47.83 82.80 26.09 88.09 7.61 - 0.00 63.41 18.48 72.96 100
53 Pekan to Permatang Keledang 62.22 38.64 63.13 38.64 70.52 6.82 - 0.00 49.22 15.91 60.39 100
54 Permatang Keledang to Pekan 55.74 66.67 51.68 7.53 53.94 3.23 - 0.00 39.56 22.58 51.44 100
55 Kg Gadong to Merchong 67.02 53.13 62.03 7.81 54.63 7.81 - 0.00 46.09 31.25 59.12 100
56 Merchong to Kg Gadong 68.45 53.03 66.49 12.12 65.37 10.61 64.97 1.52 58.45 22.73 65.57 100
319
Based on the results obtained from Levenes test of homogeneity of variance, the Levene statistic
obtained is 4.066 with a p-value of 0.003. Thus, at the 0.05 level, the null hypothesis of equal variances
across the five classes of vehicles was rejected. Hence, equal variances were not assumed in the post hoc
test.
Table 3: Levenes test of homogeneity of variance for one-way ANOVA based on 5 classes of vehicles
Levene Statistic d.f.1 d.f. 2 Sig.
4.066 4 214 0.003
The results of the one-way ANOVA are as shown in Table 4. Based on the observed significant level
which is less than 0.05 in Table 4, null hypothesis was rejected and therefore, can be concluded that the
means of free-flow speeds for different classes of vehicles were significantly different.
Table 4: One-way ANOVA based on 5 classes of vehicles
Sum of Squares d.f. Mean Square F Sig.
Between Groups 5243.189 2 1310.797 20.099 4.56 E-14

Within Groups 13956.279 214 65.216
Total 19199.468 218
However, upon knowing that the means of free-flow speed were significantly different for different
classes of vehicles, post hoc multiple comparison procedures (equal variances not assumed) were used to
know exactly which group is different from each other. This is due to the fact that the means of free-flow
speed might be different for all five classes of vehicles, or it may be that only one class that differ from
the rest. However, due to the high standard error obtained for mean free-flow speed in Class 4 as a
consequence of very small number of data sampled, this group was excluded in the post hoc test. The
difference in the mean free-flow speed observed between two groups is shown in the column labelled
Mean Difference. Pairs of means that were significantly different at the 0.05 level from each other are
marked with an asterisk (*). Therefore, based on the results obtained in Table 5, mean free-flow speed
for vehicles in Class 1, 2 and 3 are not significantly different from each other but they are significantly
different from mean free-flow speed for vehicles in Class 5.
Table 5: Post hoc multiple comparisons based on 5 classes of vehicles

Mean Difference Std. 95% Confidence Interval
(I) Class (J) Class Sig.
(I J) Error Lower Bound Upper Bound
2 -1.0286 1.4367 0.9793 -4.8944 2.8371
1 3 0.0566 1.6637 1.0000 -4.4508 4.5640
5 10.6479* 1.1908 8.593E-14 7.4548 13.8409
1 1.0286 1.4367 0.9793 -2.8371 4.8944
2 3 1.0853 1.9289 0.9941 -4.0997 6.2703
5 11.6765* 1.5397 2.870E-7 7.5445 15.8085
1 -0.0566 1.6637 1.0000 -4.5640 4.4508
3 2 -1.0853 1.9289 0.9941 -6.2703 4.0997
5 10.5912* 1.7534 2.870E-7 5.8593 15.3231
1 -10.6479* 1.1908 8.593E-14 -13.8409 -7.4548
5 2 -11.6765* 1.5397 1.013E-10 -15.8085 -7.5445
3 -10.5912* 1.7534 2.870E-7 -15.3231 -5.8593
320
CONCLUSIONS
Preliminary results of free-flow speed were obtained and discussed in this paper. The minimum, mean
and maximum free-flow speed values for each site as well as mean free-flow speed based on traffic
composition for each site were also presented. In this study, five classes of vehicles were considered.
One-way ANOVA test was conducted to check whether the mean free-flow speeds of the five classes are
equal or otherwise. However, due to the high standard error obtained for mean free-flow speed in Class 4
as a consequence of very small number of data sampled, this group was excluded in the test. The results
obtained indicate that the mean free-flow speed for vehicles in Class 1, 2 and 3 are not significantly
different from each other but they are significantly different from mean free-flow speed for vehicles in
Class 5. Nevertheless, further data collection and analyses will be conducted in order to achieve better
results and to produce a regression model to estimate free-flow speed for two lane highways in Malaysia.
ACKNOWLEDGEMENT
(Inter-Urban).
REFERENCES
Al-Kaisy, A., & Karjala, S. (2008). Indicators of performance on two-lane rural highways: empirical
investigation. Transportation Research Record 2071, 87-97.
Brilon, W., & Weiser, F. (2006). Two-lane rural highways: the German experience. Transpoortation
Research Record 1988, 38-47.
Bang, K. L., Carlsson, A., & Palgunadi. (1996). Development of speed-flow relationships for Indonesian
rural roads using empirical data and simulation. Transportation Research Record 1484, 24-32.
Hoogendoorn, S. P. (2005). Unified approch to estimating free speed distributions. Transpoortation

Research Part B, 39, 709-727.
Kyte, M., Khatib, Z., Shannon, P., & Kitchener, F. (2000). Effect of environmental factors on free-flow
speed. Transportation Research Circular E-C018: 4th International Symposium on Highway
Capacity, 108-119.
Medina, A. M. F., & Tarko, A. P. (2005). Speed factors on two-lane rural highways in free flow
conditions. Transportation Research Record 1912, 39-46.
Staplin, L., Gish, K. W., Decina, L. E., Lococo, K. H., Harkey, D. L. & Tarawneh, M. S. (1997).
Synthesis of human factors research on older drivers and highway safety. Human factors and
highway safety research synthesis, 2, retrieved 2 July 2010, from
http://ntl.bts.gov/DOCS/97095/ch02/ch02_01.html
Transportation Research Board. Highway Capacity Manual (2000). National Research Council,
Washington D.C.
Yagar, S., & Aerde, M. V. (1983). Geometry and environmental effects on speeds of 2-lane highways.
Transportation Research Part A, 17(4), 315-325.
321
322
TRAFFIC ACCIDENTS ANALYSIS ALONG EAST COAST EXPRESSWAY
Intan Suhana, M.R.1; Shahril Faiz, M.N2, Azlina I.3
ABSTRACT:
Road traffic accidents have been one of the main causes of death in almost all country in the world after
wars and diseases. Malaysia also have not been excluded from being part of the countries that faced the
same problems and as a preventive measures, Malaysian government have taken a step forward by
initiating a National Road Safety Plan to look into this problems closely and finally suggested the best
possible plan for Malaysia. This study aims to provide an introductory platform for the government by
studying the trends of road traffic accidents that occurred in a newly built expressway which have known
for its highest standard of design and construction procedures. This study will based on road traffic
accidents data collected from Royal Police Force, Traffic Units from year 2004 until year 2007 and will
be focused on East-Coast Expressway Phase 1 only. From the raw data given, researchers have worked
together to sort the data out and finally produced tables contained numbers of road accident data recorded
from year 2004 until year 2007 classified into each district along the expressway and types of injuries.
The results clearly shows that from August 2004 until December 2007, for a 169 km stretch long, East-
Coast Expressway have recorded 17 traffic accidents / kilometers or 4 traffic accidents for each kilometer
per year. This shocking yet disappointing fact really shows that National Road Safety Plan inspired by the
governments should not only focusing on federal roads but also roads with high design standards such as
expressways and highways.
Keywords: Traffic accidents, East-Coast Expressway, Accident Analysis
__________________________
1
Lecturer, Universiti Malaysia Pahang, Lebuhraya Tun Razak, 26300, Kuantan, Pahang, MALAYSIA
2
Student, Universiti Malaysia Pahang, Lebuhraya Tun Razak, 26300, Kuantan, Pahang, MALAYSIA
3
Lecturer, Universiti Malaysia Pahang, Lebuhraya Tun Razak, 26300, Kuantan, Pahang, MALAYSIA
*Correspondence Author: Intan Suhana bt. Mohd Razelan, Universiti Malaysia Pahang, Malaysia. Tel: +6009 549
2944, Fax: +6095492998. E-mail: intan@ump.edu.my
323
INTRODUCTION
Road traffic accidents now have become worldwide public health problems. According to United
Nation report in year 2005, traffic accidents had taken 6,000 lives in Malaysia and cost us RM1.5 billion
of loss in terms of productivity, medical cost, property damage and other expenses.
In Pahang state alone, a total of 445 road fatalities been recorded in year 2004. In overcoming this
situation from being serious and uncontrollable, most of the road transportation agencies and local
authorities have done a lot of awareness programs that also involved an analysis of road accidents trends
as one of the action in preventing accidents. According to PDRM, total road accidents and fatalities in
Malaysia from year 1990-2004 showed a relieving result with fatalities percentage has a major decreased
from 4.6% in year 1990 to only 1.8% in year 2004.
The intent of this study are to provide an analysis of road traffic accidents trends in Malaysians
expressway by taken a newly built East-Coast Expressway Phase 1 which connecting Kuala Lumpur and
Kuantan as a study area.
LITERATURE REVIEW
This study aims to provide an analysis on road traffic accidents along major expressway in
Malaysia. Although this type of road are known for its highest design standards, the occurrences of road
traffic accident could not be avoided.
According to BAEK, Seungkirl et al (2005), the main causes of accidents is a progressive

withdrawal of attention to traffic and the roadway environment, which leads to impaired performance
behind the wheel. Eckhardt, N., Thomas, I. (2004) also concluded that the characteristic of the
environment and the roads significantly influence the occurrence of road accidents, and the changes on
these characteristic are quite important elements in the explanation leading to the suggestion that road
users do not adapt their behavior sufficiently to changes in road conditions.
Besides that, Abdul Kareem (2003) had suggested that lack of attention, reckless driving, lack of
proper protection, speeding, bad personal habits, social and behavioral misconduct and inconsiderate
drivers of larger vehicle are some of the problems that caused accidents in Malaysia.
A study by Azmani, W., et al (2005) had suggested that road quality was found to be significant
associated factors for severe injuries. Uneven surface of road seemed to expose the road users to the risk
of accidents although the road was just opened to the public. Even a small errors of design such as in
curve super elevations may lead the vehicle to lose its control especially during high speed.
OBJECTIVES OF THE STUDY
The main objectives of this study is to analyze and produce a traffic accidents trends for East-Coast
Expressway for year 2004 until year 2007 and consequently identify the most critical district and black
spot area within the district that recorded the highest number of road traffic accidents along East Coast
Expressway.
324
METHODOLOGY
Traffic Unit, Malaysian Royal Police Departments has a statewide database on the accidents occurred
at Pahang State. In this research, the data on the Numbers of Accidents occurred on East-Coast
Expressway (ECE) from year 2004 to year 2007 was used.
Raw data gathered from the Traffic Unit contains information regarding the location of road traffic
crashes together with date and time on which the accidents occurred. As this is the primary data for this
study, the data will be carefully checked and any unsuitable data will be removed from datasets. The
study area which is ECE Phase 1 which consists of 169km stretch dual-carriageway divided road
connecting Kuala Lumpur to Kuantan are spread across 4 district in Pahang which is Bentong, Temerloh,
Maran and Kuantan district.
For the purpose of this study, data were segregated between districts in order to identify which
district experienced the highest number of road traffic accident along ECE. After the above-said district
was located, the data once again were classified into smaller stretch in order to determine the black spot
area within the district. Finally, the accidents prone area along ECE can be identified.
RESEARCH AREA
The study area, East-Coast Expressway is a newly built highway which was officially opened to
public on 1st August 2004. This highway is connecting West-Coast and East- Coast of peninsular
Malaysia beginning from Karak Highway and ended at Kuantan Interchange for Phase 1.
Source: Current Research

FIGURE 1.0: Research Area. ( East-Coast Expressway)
325
ACCIDENT ANALYSIS AND DISCUSSION
A 169km length East-Coast Expressway Phase 1 is crossing 4 major districts along the way from
Kuala Lumpur to Kuantan. From Kuala Lumpur, the expressway will starts its routes at Bentong district
followed by Temerloh, Maran and finally reached Kuantan either through Gambang Interchange or
Kuantan Interchange.
As this research mainly will be focused on developing patterns of road traffic accidents along
ECE from year 2004 until year 2007, the data were classified according to the 4 major districts as in
below Table 1.0.
Table 1: Numbers of road traffic accidents along East-Coast Expressway from 2004-2007
Number of Accidents in Years Total number of

Districts 2004 2005 2006 2007 accidents
Kuantan 93 237 281 350 961
Maran 77 233 276 283 869
Temerloh 95 196 215 207 713
Bentong 16 50 66 111 243
Total 281 716 838 951 2786
From Table 1.0, it is clearly shows that Kuantan districts is always on the top ranking on number
of road accidents cases starting from year 2004 until year 2007 followed by Maran, Temerloh and lastly
Bentong. Not only that, we can also saw an ascending trends on numbers of road traffic accident where
the numbers of accidents always kept on increasing from year to year with the highest number of
accidents is in year 2007 with 961 number of accident cases.
Also, if we can go further, for a total of 2786 numbers of traffic accidents within 41 months of its
operation period (the expressway opened for public in August 2004), this expressway has experienced
about 68 numbers of accidents per months and this something that really needs an expert attention not
only to make sure the safety of the road users is in good hands but also to make sure that the reliability of
the expressway in unquestionable.
In the next figures, the researchers will be showing the numbers of road traffic accidents in
relation to the type of accidents from year 2004 until year 2007 for each district along East-Coast
Expressway. The types of accidents were classified into 4 classes which are fatal accidents, accidents with
deaths, accidents with severe injuries, accidents with slight injuries.
326
14
12
10 Kuantan
Numbers
8 Maran
6 Temerloh
4 Bentong
2
0
2004 2005 2006 2007
Year
Figure 1.0 : Number of Fatal Accidents
With references to Figure 1.0, we can also concluded that the numbers of fatal accidents along
East Coast Expressway shows an ascending trends starts from year 2004 until year 2007 for almost all
district except for Bentong district who shows an downward trends between year 2005 and year 2006.
Apart from that, Temerloh district has been in the top rank by recording 13 numbers of fatal
accidents in year 2007 alone which is extremely high compared to only 1 accident for Bentong district in
year 2006. However, the facts that the length of ECE in Temerloh district are longer than in Bentong
district cannot be put aside. The differences of about 42 kilometers should also be taken into
considerations as the longer the road length the higher the risk of accidents occurrences.
30
25
20 Kuantan
Numbers
Maran
15
Temerloh
10 Bentong
5
0
2004 2005 2006 2007
Year
Figure 2.0 : Number of Accidents with Deaths
327
As for Figure 2.0, the numbers of accidents with deaths recorded a slightly higher number of
accidents compared to number of accidents in fatal accidents with Temerloh district still is in the top list.
This may caused by the total of deaths immediately after the accidents ( died on scene ) is lower
compared to deaths occurred within 24 hours after the accidents took place as most of the victims will be
given treatments as soon as the medical assistants arrived.
14
12
10 Kuantan
Numbers
8 Maran
6 Temerloh
4 Bentong
2
0
2004 2005 2006 2007
Year
Figure 3.0 : Numbers of Accidents with Severe Injuries
In Figure 3.0, the numbers of accidents with severe injuries shows dramatic changes with the
highest number of victim were generated from Maran district. This is quite surprising as most of the fatal
accidents and accidents with deaths happened in Temerloh district. This data widely revealed that most of
the accidents happened within Maran district rarely led to deaths but only caused severe injuries to the
victims compared to accidents happened within Temerloh district that most of the accidents will bring
deaths to the victims.
There must be some logic explanation behind this situation as the length of roads between this
two district is almost the same as suggested by Mekitew (2000) that besides human factor, road accidents
can also caused by weather, sight distance conditions, faulty design and/or inadequate maintenance of the
road infrastructure developments. However, Bentong district once again had recorded the lowest number
of accidents with severe injuries by only having 3 accidents with severe injuries for its 8.9km length
expressway.
328
90
80
70
60 Kuantan
Numbers
50 Maran
40 Temerloh
30 Bentong
20
10
0
2004 2005 2006 2007
Year
Figure 4.0 : Number of Accidents with Slight Injuries
In Figure 4.0, it also clearly shows that Maran Districts also had recorded the highest number of
accident with slight injuries as in previous figure ( Figure 3.0 ) but if we looked at the numbers of
accidents, we can see a huge differences between number of accidents with severe injuries and number of
accidents with slight injuries.
The difference of about 68 numbers of accidents (80 numbers of accidents with slight injuries in
year 2006 compare to 12 number of accident with severe injuries in year 2007) is solely coming from
Maran district alone. Or else, if we looked at different angle, for every kilometers traveled there must be
at least 2 accidents within Maran district in year 2006.
ACCIDENTS PRONE AREA
Determination of accident prone area is made to identify stretches that have a high ranking of
accident cases. Below figure and table shows the number of accidents for every section along East-Coast
Expressway and from that, black spot areas within Kuantan District along East-Coast Expressway were
identified.
The identification of black spot areas will only limits to a stretch starting from Gambang
Expressway until Kuantan Expressway since this areas had been identified as one of the most critical
stretch along the East-Coast Expressway within Kuantan District. This first figure below shows the
number of accidents cases for all 4 major district which is located along the East-Coast Expressway.
329
400
350
300
250 Kuantan
Bilangan
Maran
200
Temerloh
150 Bentong
100
50
0
2004 2005 2006 2007
Tahun
Figure 5.0 : Summary of Accident Cases by District (2004 - 2007)
From Figure 5.0, it is clearly shows that the number of accident cases occurred within Kuantan
district is the highest compared to another district. Although in year 2004 until year 2006, the number of
accident cases in Kuantan, Maran and Temerloh District were quite similar but in year 2007, the number
of accident cases occurred within Kuantan District were having a huge jump by leaving all the other 3
districts far behind.
Numbers of Accidents between Gambang and Kuantan

Interchange
60
50
40
Numbers of Accidents
30
20
10
Figure 6.0 : Number of Accident Cases between Gambang and Kuantan Interchange( within Kuantan
District).
330
As shows in Figure 6.0, the stretch of road between Gambang and Kuantan Interchanges were
segregated into 11 sections with 2 km stretch in each section. The number of accident cases occurred in
each section were identified and the results shows that the sections which are closely located to Gambang
Interchange had recorded the highest number of accidents compared to other section.
49 number of accident cases has been recorded for a section starting from KM 219 until KM
220.9 which makes the differences of about 40 number of accident cases when compared to the safest
section along this stretch which is between 237 until 238.9 (there are 9 accident were recorded along KM
237 until KM 238.9). This figure was quite extreme as for this section alone, the ratio of accidents per
kilometers has becomes 20 accidents in each kilometer.
CONCLUSION
An ascending trend in number of accidents cases along East-Coast Expressway which

connecting Kuala Lumpur in the west and Kuantan in east were caused by many factors. Unsuitable
geometric designs that caused an uneven road surface were suspected to be one of the major causes
contributing to this problem. Bad weather conditions such as raining worsen the condition and the
situation is uncontrollable.
Although a lot of mitigation programs have been conducted especially during festive seasons, but
most of them are only focusing on the drivers or road users with only a little attention were given to the
road condition.
This study were conducted mainly to make all road users realizes that the age of the roads
doesnt reflects its reliability in terms of safety, since a newly built expressway like East-Coast
Expressway can also creates a dangerous situation for the road users especially during the peak hours and
festive seasons.
REFERENCES
BAEK. Seungkirl, CHANG. Hyunho, KANG. Jeong-Gyu, Byoungjo Yoon (2005). An Effect of Trip
Length on Freeway Accidents Rates. Journal of Eastern Asia Society for Transportation Studies,
Volume 6, pp. 3467-3481, 2005.
Eckhardt, N., Thomas, I. (2004). Spatial Nested Scales for Road Accidents in the
Periphery of Brussels.IATSS Research Vol.29 No 1 2005.
Abdul Kareem (2003). Review of Global Menace Accidents with Special Reference to Malaysia
A Social Perspective. Malaysian Journal of Medical Sciences. 10(2). pp. 31 39.
Azmani W., Mohamed Rusli A., Aziz Al-Safi Ismail and Hashim, M. (2005). Factors and Road Traffic
Accidents during Festive Seasons in Kelantan. NCD Malaysia 4 (3). pp. 24 27.
Said Easa and Chandi Ganguly (2005). Modelling Drivers Visual Demand on Complex Horizontal
Alignments. Journal of Transportation Engineering, Volume 131, No.8, August 1, 2005.ASCE
Pg.583-590.
Amy E.Donaldson, Lawrence J.Cook, Caroline B.Hutchings, J.Michael Dean. (2006). Crossing County
Lines: The Impact of Crash Location and Drivers Residence on Motor Vehicle Crash
Fatality.Accidents Analysis and Prevention Article.
331
332
PROPERTIES OF BITUMINOUS BINDER MODIFIED WITH WASTE

POLYETHYLENE TEREPHTHALATE
Zahra Niloofar Kalantar 1*,

Abdelaziz Mahrez 2 , Mohamed Rehan Karim 3
ABSTRACT:
The purpose of this research is to investigate the possibility of using Polyethylene Terephthalate
as polymer additives in Bituminous Mix. The characteristics of PET-modified binder obtained by
fix mixing temperature, was investigated. The binders were prepared by mixing the PET in 2%,
4%, 6%, 8% and 10% (by the weight of optimum bitumen) with 80/100 penetration grade
bitumen at temperature of 150 C. It may be inferred that PET-modified bituminous binders
provide better resistance against permanent deformations and rutting due to their higher softening
point when compared to conventional binders. Using PET-modified binders also contribute to the
recirculation of plastic waste, as well as to the protection of the environment.
Keywords: Polyethylene Terephthalate, modified bitumen, binder test.
1
Master student, Department of Civil engineering, Faculty of engineering, University of Malaya, Jalan university,
Kuala Lumpur, 50603 Malaysia.
2
Senior Lecturer, Department of Civil engineering, Faculty of engineering, University of Malaya, Jalan university,
Kuala Lumpur, 50603 Malaysia.
3
Professor, Department of Civil engineering, Faculty of engineering, University of Malaya, Jalan university, Kuala
Lumpur, 50603 Malaysia.
*Correspondence Author:Zahra Niloofar Kalantar, Tel.: +60 17 280 6072 ; Fax : +60 3 79552182
Email Address : niloofar.kl@gmail.com
333
1. INTRODUCTION
The performance of the road pavement is determined by the properties of the bitumen, as
bitumen is the continuous phase and the only deformable component (Yuonne et al., 2001).
Bitumen is also a viscoelastic material with suitable mechanical and rheological properties for
waterproofing and protective coverings for roofs and roads, because of its good adhesion
properties to aggregates (Gonzlez Uranga, 2008), (Garcia-Morales et al., 2004, (Polacco et al.,
2006), (Navarro et al., 2009).
One of the most important properties of the bitumen mixture is its ability to resist shoving
and rutting under traffic. Therefore, stability should be high enough to handle traffic adequately,
but not higher than the traffic conditions require. The low stability causes unraveling and flow of
the road surface. Flow is the ability of a HMA (Hot Mixture Asphalt) pavement to adjust to
gradual settlements and movements in the sub-grade without cracking. The flow may be regarded
as an opposite property to the stability, determining the reversible behavior of the wearing course
under traffic loads and affecting the plastic and elastic properties of the asphalt (Kuloglu,
1999),(Hinisliglu & Agar, 2004).
Some improvements in asphalt properties have been achieved by selecting the proper starting
crude, to make asphalt. Unfortunately, there are only a few crudes that can produce very good
asphalts and due to the characteristics of the current traffic, neat bitumen do not results suitable
for paving applications. (Yuonne et al., 2001), (Garcia-Morales et al., 2005).
From practical experience of last 4 decades is proven that the modification of the asphalt
binder with polymer additives, offers several benefits. These include enhanced fatigue resistance,
improved thermal stress cracking, decrease in temperature susceptibility and reduction of rutting
(Abtahi et al., 2008), (Gonzalez et al., 2006), (Perez-Lepe et al., 2003). To enhance various
engineering properties of asphalt many modifiers such as styrene based polymers, polyethylene
based polymers, polychloroprene, gilsonite, various oils have been added to asphalt (Al-Hadidy
& Yi-qiu, 2009).
Polymer modified binders show improved adhesion and cohesion properties, rutting
resistance, thermal cracking, fatigue damage, stripping, and temperature susceptibility. These
improvements have led polymer modified binders to be substituted for asphalt in many paving
and maintenance applications.In many cases, they are selected to reduce life cycle costs (Yuonne
et al., 2001), (Awwad & SHbeeb, 2007), (Garcia-Morales et al., 2004*).
Polymer modification of asphalt binders have become increasingly standardized in designing
optimally performing pavements (Yetkin, 2005).
The uses of virgin polymers in bitumen to improve the properties of bitumen, has been performed
for many years. In spite of that, recently there is interest in the replace of commercial virgin
material by recycled polymers. (Gonzlez Uranga, 2008).
In view of plastic materials versatility, relatively low cost and their small duration of life
(roughly 40% have duration of life smaller than 1 month) (Achilias et al., 2007), the consumption
of plastic materials has been growing steadily, (approximately 21.2 million tonnes for 2003)
although the disposal of waste plastics constitutes a severe environmental problem, due mainly to
their chemical inertness (De la Puente et al., 2002).
Over 78 wt. % of this total corresponds to thermoplastics (mainly polyolefins, low density
polyethylene, LDPE-17%, high density olyethylene, HDPE-11%, polypropylene, PP-16%) and
the remaining to thermosets (mainly epoxy resins and polyurethans) (Achilias et al., 2007). Some
aromatic polyesters are biodegradable but PET waste is not biodegradable, thus, it will create
334
environmental hazards if disposed in landfills. Therefore, the only way of addressing the problem
of disposal of post-industrial and post-consumer PET waste is through recycling (Mittal et al.,
2010). Methods for recycling polymer waste have been developed and new recycling approaches
are being investigated (Wei et al., 2010), (Yuonne et al., 2001). Recycling waste polymer in road
construction as a bitumen modifier is a new method.
Polyethylene has been found to be one of the most effective polymer additives (Hinisliglu &
Agar, 2004). Polyethylene is the most popular plastic in the world. Polyethylene is semi-
crystalline material with excellent chemical resistance, good fatigue and wear resistance and a
wide range of properties. It has a very simple structure. A molecule of polyethylene is a long
chain of carbon atoms, with two hydrogen atoms attached to each carbon atom. They are light in
weight and provide good resistance to organic solvents with low moisture absorption rates
(Awwad & SHbeeb, 2007).
Polyethylene terephthalate (PET) is a linear, aromatic polyester (Bing et al., 2006). PET is
commonly recycled, and has the number "1" as its recycling symbol.
2. MATERIALS AND METHODS
2.1 Polymers
There are several kinds of polymers that can be recycled in bitumen (Murphyl et al., 2001).
These include low and high density polyethylene (LDPE, HDPE), widely used in packaging and
plastic bottles; polypropylene (PP) often used in straws, sweet wrappings, textile, automobile
industries and furniture industries (Zheng et al., 2009), polyvinyl chloride (PVC), used in
plumbing pipes and fittings; polyethylene terephthalate (PET), widely used in water and soft
drink bottles and acrylonitrile butadiene styrene (ABS), used in electronic devices such as laptops
and mobile phones. Some of these materials would be unsuitable for use in manufacturing
recycled polymer-modified bitumen. Although the above recycled polymers are available in
sufficient quantities so as to be commercially feasible, not all are suitable for the proposed use
(Casey et al., 2008).
In this investigation, waste PET in the powdered form in 2%, 4%, 6%, 8% and 10% (by the
weight of bitumen) was used as a modifier. About 50 waste mineral water bottles were crushed
and then were sieved. The properties of PET are as follows: density 1370 kg/m3 and melting point
260C. The gradation of PET is presented in Table 1.
2.2 Bitumen
The selected bitumen was 100/80-pen grade Malaysian bitumen with a softening point of
46C. The properties of asphalt binder, which are presented in Table 2, are within the
specification of penetrated bitumen grade 80/100.
Table 1 Gradation of PET
Sieve Percent passing

size
701 m 100
450 m 0
335
Table 2 the results of tests performed on bitumen
Test Test result

Ductility (cm) < 100
Penetration (0.1 mm) 84-100
Softening point (C) 46
Specific Gravity 1.03
(g/cm3)
2.3 Mixing Method
Dry mix is used in this study. The bitumen, contained in a 1-liter cylindrical container, was
put into the oven and heated until it became liquid. The molten bitumen was poured into six small
empty cylindrical containers with a volume of 250 ml. The net weight of bitumen in each small
tin was 120 gr. Before mixing the waste PET with bitumen in small tins, bitumen would have to
be heated up to about 160C first. To achieve this, small tins of bitumen were heated into the
oven for about 1 hour. After that, waste PET in percentage of 2%, 4%, 6%, 8% and 10% was
added into the tins and were mixed manually for about 2 minutes.
2.4 Sample preparing
Immediately after mixing, the mixture was placed in viscosity sample chambers. Following
this, softening point rings, penetration cups and then DSR test specimens were prepared. After
having cooled in room temperature for 1 day, samples were tested.
2.5 Standard bitumen tests
The modified binders were subjected to a series of standard bitumen tests for dynamic shear
rheometer, viscosity, penetration and softening point. The dynamic shear rheometer is used to
characterize the viscous and elastic behavior of asphalt binders at medium to high temperatures.
A Dynamic Shear Rheometer test was conducted by using a temperature sweep starting from
30C to 80C, while the frequency was fixed at 1.159Hz (10 rad/sec). Parallel plates with
diameter of 25 mm and 1 mm gap were used. The viscosity test can be conducted at various
temperatures, but since manufacturing and construction temperatures are fairly similar regardless
of the environment, the test for asphalt binder specification is always conducted at 135C
(Roberts et al., 1996). The test was done under ASTM 2000. This viscosity is reported in Poise,
where 1 Poise = 0.1 Pa s. (http://www.hydramotion.com/pdf/Website_Viscosity_Units_V2.pdf).
The penetration value is reported in units of decimillimetres (dmm). This test is done three times
and the average is taken as the penetration value of the bitumen (Casey et al., 2008). The test was
done under ASTM 2000. The softening point is also done under ASTM 2000.
3. RESULTS AND DISCUSSION
This chapter discusses the results of laboratory tests. Binders are characterised by using a
number of standard physical tests such as penetration test (temperature, load and time are 25C,
100g and 5sec respectively), softening point test, viscosity test using Brookfield viscometer
(temperature range from 90 to 170 C, spindle No.27, and a rotating speed of 20rpm), and also
rheological measurements by using a Dynamic Shear Rheometer (tests conducted by using a
temperature range of 30 C to 80 C, and a frequency of 1.159Hz).
336
3.1 Penetration test
Penetration is a measure of the consistency or hardness of bitumen and is the most common
control test for penetration grade bitumens. Figure 1 shows the variation of penetration value with
the various percentages of bitumen modified PET and it shows that consistency decreases with
the addition of PET. The penetration values for the modified binders decrease as the PET content
in the mix increases. The decrease is 2.4%, 4.9%, 8.5%, 11% and 14.6% with the addition of 2%,
4 %, 6%, 8% and 10% of PET, respectively, as compared to the original bitumen. This means that
the addition of PET makes the modified bitumen harder and more consistent. This is good in one
sense since it may improve the rutting resistance of the mix, but on the other hand this may affect
the flexibility of the bitumen by making the asphalt much stiffer, thus the resistance to fatigue
cracking can be affected.
3.2 Softening point
The softening point is a measure of the temperature at which bitumen begins to show
fluidity. It is also defined as the temperature at which a bitumen sample can no longer support the
weight of a 3.5-g steel ball. Figure 2 shows that softening point increases with increasing PET
content. The results clearly show that the addition of PET to bitumen increases the softening point
value, and as the PET content increases the softening point also increases. This phenomenon
indicates that the resistance of the binder to the effect of heat is increased and it will reduce its
tendency to soften in hot weather. Thus, with the addition of PET the modified binder will be less
susceptible to temperature changes.
The effect of softening point of a binder on resistance to permanent deformation, of bituminous
pavement mixes, has been studied by various researchers. An example is hot rolled asphalt where
it was found that the rate of rutting in the wheel tracking test at 45C, was halved by increasing
softening point by approximately 5C (Fernando & Guirguisl, 1984). Therefore it is expected that
by using the PET in the bituminous mix the rate of rutting will decrease due to the increase in
softening point.
85
pentration (0.1 mm)
80
75
70
65
0% 2% 4% 6% 8% 10%
PET percentages
Figure 1 Penetration grade vs. different percentage of PE
337
51
Softening point (C)

50
49
48
47
46
45
0% 0% 0% 0% 0% 0%
PET Percentage
Figure 2 Softening point vs. different percentage of PET
3.3 Rotational Viscosity
The viscosity-temperature relationship of a binder helps to identify the temperature range for
successful mixing and compaction of bituminous mixtures. If the viscosity of bitumen is too high,
the binder may not completely coat the aggregate in the bituminous mixture; if the viscosity is too
low, binder drainage is likely to occur during the storage and transportation of the mix. Identify-
ing the most suitable temperature range for mixing allows decisions on the suitability of the
binder to be made.
Modified asphalt binders are usually more viscous than unmodified ones (Aflaki &
Tabatabaee, 2008). Figure 3 shows the viscosity curves of Modified bitumen and Figure 4 shows
the viscosity of a modified binder at135C. Table 3 shows the ratio of viscosity of the modified
asphalt binders to the viscosity of the base binder at 135C. This ratio varies between 1 and 2 for
the modifiers used in this study. In the extreme case of adding 10 % PET to the base binder, the
viscosity increased from 275 mPas to 538 mPas. This value is less than 3000 mPas and therefore
satisfies the ASTM D6373 criterion for asphalt binder workability (Aflaki & Tabatabaee, 2008).
10000
0%
8000
2%
6000
Viscosity (mPas)
4%
4000
2000
0
80 90 100 110 120 130 140 150 160 170 180
Temprature (C)
Figure 3 Viscosity vs. Temperature (90-135 C)
338
600
500
Viscosity (mPas)
400
300
200
100
0
0% 2% 4% 6% 8% 10%
PET Percentage
Figure 4 Viscosity vs. PET% @ 135C Temperature
Table 3 Viscosity ratio of modified asphalt binders
Sample Viscosity (mPas) Viscosity Ratio
Original Binder 275.0 --
Modified with 2% PET 290 1
Modified with 4% PET 305 1.1
Modified with 10% PET 538 2
3.4 Dynamic shear modulus
The DSR measures a specimen's complex shear modulus (G*) and phase angle (). The
complex shear modulus (G*) can be considered the sample's total resistance to deformation when
repeatedly sheared, while the phase angle (), is the lag between the applied shear stress and the
resulting shear strain. This means the larger the phase angle (), the more viscous the material
will be. Phase angle () is limited between 0 and 90 degrees. If the phase angle is 0 degree the
material is purely elastic and if the phase angle is 90 degrees, the material is purely viscous. The
specified DSR oscillation rate of 10 radians/ second (1.59 Hz) is meant to simulate the shearing
action corresponding to a traffic speed of about 55 mph (90 km/hr)
339
(http://pavementinteractive.org). Early in pavement life, rutting is the main concern, while later in
pavement life fatigue cracking becomes the major concern (http://pavementinteractive.org).
3.4.1 Complex shear modulus
The temperature dependence of G* is lowest for the unmodified bitumen compared to

modified bitumen. Figure 5 shows that in all samples (unmodified and modified bitumen) the
complex shear modulus G* decreases as the temperature increase. The graphs we can also show
that as PET content increases the complex shear modulus G* increases. This can be clearly seen
from the figure 5 since the graphs of G* follow a certain order.
It can be seen that the lowest graph is for the original sample (without PET) and the highest one is
for the sample with 10% of PET.
The lower slope of the complex shear modulus G* means that the asphalt is softer, and can
deform without developing large stresses. Also, a higher complex shear modulus G* have are
beneficial since it reduces rutting problems (deformations) in the asphalt.
3.4.2 Phase angle
Figure 6 shows that the phase angle for the unmodified bitumen is higher than modified
bitumen. Generally as PET content increases the phase angle decreases and we can verify this
by doing the same comparison as before concerning the order of the graphs for the samples.
We can also see from the unmodified and modified samples that as the temperature increases the
phase angle also increases.
The lower slope of phase angle means that the asphalt is more elastic than viscous, and will
recover to its original condition without dissipating energy. Also, at high temperatures, a low
phase angle is desirable since this reduces permanent deformation.
400
0%
300 2%
4%
6%
G* (kPa)
200
8%
10%
100
0
30 40 50 60 70 80
Temprature C
Figure 5 G* vs. Temperature
340
94
0%
90
2%
Phase angle ( ) 86
4%
82
6%
78
8%
74
10%
70
30 40 50 60 70 80
Temprature C
Figure 6 Phase angle vs. Temperature
4. ANALYSIS OF VARIANCE (ANOVA)
ANOVAs form a set of analytical tests which can be used to identify possible differences in
the mean value of more than two data samples.
To assess the effect of PET on the penetration, softening point, viscosity, complex shear modulus
and phase angle, the ANOVA test at 5% significance level was conducted.
The analysis on the penetration and softening point tests results proved that the percent of PET is
responsible for the differences in result. According to the test result F is bigger than F critical
which means that groups means are not equal. (Table 4)
Table 4 ANOVA results for Penetration and Softening point test
Source of Variation Penetration Softening point
F 26.7 19.2
P-value 8E-13 1.2 E-6
F critical 2.4 2.8
ANOVA on viscosity, complex shear modulus and phase angle tests result is reported in
table 5 which tells us the following:
1. There is a significant difference between temperatures overall because the calculated F

value is more than the critical F value.
2. There is very highly significant effect of PET overall because P value is less than =0.05.
341
Table 5 ANOVA results for Viscosity test, G* and
Source of Viscosity G*
Variation Tem PET% Tem PET% Tem PET%
F 198.2 2.72 54 27 89.4 8.1
P-value 2.9E-19 0.042 9.6E-15 0.04 8.9E-18 6.4E-05
F critical 2.6 2.6 2.42 2.53 2.42 2.53
5. CONCLUSIONS
Based on the laboratory test results, the mixture containing PET resulted in higher resistance
to permanent deformation and higher resistance to rutting. It appears that polyethylene
Terephthalate (PET) decreases the consistency and increases the resistance of the material to
temperature changes, while the resistance to flow also increases. It may be inferred that PET-
modified bituminous binders provide better resistance against permanent deformations and rutting
due to their higher softening point when compared to conventional binders. It also contributes to
the recirculation of plastic waste, as well as to the protection of the environment.
REFERENCES
Abtahi, S.M., Sheikhzadeh, M. and Hejazi, S.M. (2010) Fiber-reinforced asphalt-concrete A

review. Construction and Building Materials. Vol. 24, No.16, 871-877
Achilias, D.S., Roupakias, C., Megalokonomos, P., Lappas, A.A. and Antonakou, .V. (2007)
Chemical recycling of plastic wastes made from polyethylene (LDPE and HDPE) and
polypropylene (PP). Journal of Hazardous Materials Vol. 149, 536-542
Aflaki, S. & Tabatabaee, N. (2008) Proposals for modification of Iranian bitumen to meet
the climatic requirements of Iran. Journal of Construction and Building Materials, vol.
2, No. 1. 1-10
Al-Hadidy, A.I. & Yi-qiu, T., (2009) Effect of polyethylene on life of flexible pavements.
Journal of Construction and Building Materials, vol. 23, No. 1. 1456-1464
ASTM 2000, Standard Test Methode for Penetration of Bitumminous Materials, D5-97, Annual
Book of Standards, Vol.04.03,Eston,MD,USA,2000.
ASTM 2000, Test Method for Softening Point of Bitumen (Ring -and- Ball Apparatus), D36-
06,Annual Book of ASTM Standards,Vol.04.04,Easton, MD,USA,2000.
ASTM 2000, Test Method for Viscosity Determinations of Unfilled Asphalts Using the
Brookfield Thermosel
Awwad, M. T. & SHbeeb, L. (2007) The Use of Polyethylene in Hot Asphalt Mixtures. American
Journal of Applied Sciences,vol. 4, No.6, 390-396.
342
Bing Sh, Hong L, Thomas B. k. and Lawrence K. D. (2006) Surface Properties of Cell-treated
Polyethylene Terephthalate. American Journal of Biochemistry and Biotechnology,vol 4,No.2,
170-174.
Casey, D., McNally, C., Gibney, A., D. Gilchrist, M. (2008) Development of a recycled polymer
modified binder for use in stone mastic asphalt. Journal of Resources, Conservation and
Recycling Vol 52. 11671174
De la Puente, G., Klocker, C. and Sedran, U., (2002) Conversion of waste plastics into fuels:
Recycling polyethylene in FCC. Applied Catalysis B: Environmental Vol. 36, 279-285
Fernando, M.J. and Guirguis, H.R. (1984) Natural rubber for improved surfacing,
Proceedings 12th ARRB conference, Part 2, 121-130.
Garcia-Morales, M., Partal, P., Navarro, F.J., Gallegos, C. (2005) Effect of waste
polymer addition on the rheology of modified bitumen. Fuel, Vol.85, No.11. 936-943
Garcia-Morales, M., Partal, P., Navarro, F.J., Martinez-Boza, F., Mackley, M.R., Gallegos, C.
(2004) The rheology of recycled EVA/LDPE modified bitumen. rheologica Acta Journal
Vol.43 No.5. 482-490
Garcia-Morales, M., Partal, P., Navarro, F.J., Martinez-Boza, F., Gallegos, C., Gonzlez, N.,
Gonzlez, O. and Muoz M. E. (2004*) Viscous properties and microstructure of recycled eva
modified bitumen. Fuel, Vol.83, No.1, 31-38
Gonzalez, O. Munoz, M. E. & Santamaria, A.(2006) Bitumen/polyethylene blends: using m-

LLDPE to improve stability and viscoelastic properties, rheologica Acta Journal Vol.45 No.5.
603-610
Gonzlez Uranga, O. (2008) Rheological property of bitumen modified with polyethylene and
polyethylene based blends.
Hinisliglu, S. & Agar, E. (2004) Use of waste high density polyethylene as bitumen modifier in
asphalt concrete mix. Journal of Materials letters. vol. 58, 267-271.
Kuloglu. N,(1999) Effect of astragalus on characteristics of asphalt concrete,Journal of Materials

in Civil Engineering vol.11 No.4, 283-286
Mittal, A., Soni, R.K., Dutt, K., Singh, S., (2010) Scanning electron microscopic study of
hazardous waste flakes of polyethylene terephthalate (PET) by aminolysis and ammonolysis.
Journal of Hazardous Materials Vol. 178, 390-396
Murphy, M., O'Mahony, M., Lycett, C. and Jamieson, I. (2001) Recycled Polymers for Use as
Bitumen Modifiers. Journal of Materials in Civil Engineering, Vol. 13, No. 4, 306-314
Navarro, F.J., M., Partal, P., Garcia-Morales, M., Martn-Alfonso M.J., Martinez-Boza, F.,
Gallegos, J.C., Gonzlez, N., Bordado, M. and Diogo, A.C. (2009) Bitumen modification with
reactive and non-reactive (virgin and recycled) polymers: A comparative analysis. Journal of
Industrial and Engineering Chemistry, Vol.15, No.4, 458-464
343
Perez-Lepe,A., Martnez-Boza, F.J., Gallegos, C., Gonzalez, O., Muno, M.E., Santamar, A.
(2003) Influence of the processing conditions on the rheological behavior of polymer-modified
bitumen. Fuel, Vol.82, No.11. 1339-1348
Polacco, G., Stastna, J., Biondi, D. and Zanzotto, L. (2006) Relation between polymer
architecture and nonlinear viscoelastic behavior of modified asphalts. Current Opinion in
Colloid & Interface Science Vol 11, 230-245
Roberts, F.L., Kandhal, P.S., Brown, E. Ray, Lee, D., Kennedy, T.W., (1996) Hot Mix Asphalt
Materials, Mixture Design and Construction, NAPA Research and Education Foundation,
Lanham, MD
Wei, T., Wu, K., Lee, Sh., Lin, Y., (2010) Chemical recycling of post-consumer polymer waste
over fluidizing cracking catalysts for producing chemicals and hydrocarbon fuels, Journal of
Resources, Conservation and Recycling Vol 52. 1-10
Yuonne B., M. P. M., Yajaira R. (2001) Polymer Modified Asphalt. vision technologica, vol 9,
No,1, 39-48.
Yetkin Y. (2005) Polymer Modified Asphalt binders. Journal of Construction and

Building Materials, vol. 21, No. 1. 66-72
Zheng, Y., Shen, Zh., Cai, Ch., Ma, Sh., and Xing, Y. (2009) The reuse of nonmetals recycled
from waste printed circuit boards as reinforcing fillers in the polypropylene composites.
Journal of Hazardous Materials. Vol. 163 No. 2-3 600-606
http://pavementinteractive.org
http://www.hydramotion.com/pdf/Website_Viscosity_Units_V2.pdf
344
FINITE ELEMENT ANALYSIS OF FLEXIBLE PAVEMENT SUBJECTED TO HEAVY AXLE
LOAD AT MALAYSIAN FEDERAL ROAD ROUTE ONE
Osama Mahmoud Yassenn 1*, Bahardin Bin Baharom 2,
Intan Rohani Endut 3, Mohamed Ahmed Hafez 4
Md Yunus Ab Wahab5
ABSTRACT:
In the last few years excessive damages were observed on Malaysian federal road route one, including
rutting, cracking and fatigue. The main cause of these distresses was probably the heavy axle load. Given
that finite element analysis has been proven to be useful in pavement analysis, to determine which
damages caused by heavy axle load and to predict possible future damages. The objectives of this paper
are to choose the proper analysis models for each pavement layer to be used in finite element analysis and
to suggest the proper method to determine the parameters needed for finite element analysis. Moreover it
suggests a method for using finite element analysis outputs to determine the relationship between axle
load, and equivalent axle load factor based on fatigue and rutting criteria based on federal road route one
subgrade layer. This paper suggests using DruckerPrager model for base and subbase layer modeling.
The use of dynamic modulus accompanied with viscoelastic model for surfacing layer. For needed
parameters determination, it suggests the use of falling weight deflectometer to determine the elastic
moduli of the base, subbase and subgrade layers, and estimation of frictional angle based on preceding
local geotechnical experience. In addition, this paper suggests using two dimensional axisymmetric
model, and circular tire-pavement contact area for the finite element analysis. Furthermore this paper
describes the way to use the maximum horizontal tensile strain at the bottom of the top layer and the
maximum vertical compression strain at the base of the subgrade layer to determine the relationship
between axle load, and equivalent axle load factor based on fatigue and rutting criteria based on federal
road route one subgrade layer.
Keywords: equivalent axle load factor, finite element, pavement analysis.
Associate Professor, Universiti Teknologi MARA, Faculty of civil engineering, 40450 Shah Alam, Selangor, MALAYSIA
1
Student , Universiti Teknologi MARA, Malaysia Institute of Transport, 40450 Shah Alam, Selangor, MALAYSIA
2
Deputy Director of Malaysia Institute of Transport , Universiti Teknologi MARA, Malaysia Institute of Transport, 40450
Shah Alam, Selangor, MALAYSIA
3
Head of Malaysia Transport research laboratory , Universiti Teknologi MARA, Malaysia Institute of Transport, 40450 Shah
Alam, Selangor, MALAYSIA
4
Lecturer, Universiti Teknologi MARA, Faculty of civil engineering, 40450 Shah Alam, Selangor, MALAYSIA
Honorary consultant, Universiti Teknologi MARA, Malaysia Institute of Transport, 40450 Shah Alam, Selangor, MALAYSIA
5
*Correspondence Author: Malaysia Institute of Transport, Universiti Teknologi MARA, Shah Alam, Malaysia. Tel:
+60193692471 , E-mail: eng.osama.alhusyni@gmail.com
345
1.0 INTRODUCTION
There is a growing tendency towards higher traffic volume and traffic load demands on existing
road and pavement network system. Road network system is considered as important arteries for
sustainable and economic development in any country. For that reason, it is not only vital to plan ahead
for constructing new roads, but also preservation and upgrading of existing road network. The importance
of road transportation in Malaysia is growing rapidly due to the economic growth of the country, resulting
not only in development of infrastructure, but also in terms of transportation trucks technical
developments, such as, heavier trucks, and larger carrying capacity. Therefore, the heavier trucks and the
increase in traffic volume are likely to cause a lot more distresses to roads than ever in the past.
This paper is only concern with flexible pavement. It is so called flexible pavement because the
whole structure flexes under loading. It is a layered structure that consists of a surface layer, an aggregate
base layer, and an aggregate subbase layer, in that order from the peak to the base.
Pavement damage is affected by many factors such as traffic, environment, material and design
considerations; however, the traffic loads play the key role in deterioration. Trucks are the main
consumers of the road network. Their loads transferred to pavement through various types of axle
configurations. Axle load is the total weight imposed on pavement by all wheels joint in one axle. For
current design purposes, different axle load and axle configuration are converted to standard axle load
using load equivalency factors. These load equivalency factors are based on loss of present serviceability
index and were developed for a limited number of load magnitudes, pavement types, ages and
environments [1]. The present of serviceability index is based on serviceability, and only slightly accounts
for other key performance measures such as rutting, fatigue and cracking of flexible pavement.
In the previous few years excessive damages were observed on Malaysian federal road one,
including rutting, cracking and fatigue. It is crucial to identify the defects on this road, and whether it is
caused by heavy axle load or other factors. Given that finite element analysis has been proven to be useful
in studying various pavement failure problems [3, 4, 5, 15], any finite element software can be used to
determine which defects caused by heavy axle load. Thus the first objective of this conceptual paper is to
introduce the finite element method for pavement analysis, and to choose the proper analysis model for
each pavement layer in the finite element model.
Furthermore, using axle load factor defines how much damage caused by certain axle load and
pressure with comparison to the damage caused by standard axle load and pressure, therefore the second
objective of this paper is to determine the relationship between axle load, tire contact pressure and
equivalent axle load factor based on fatigue and rutting criteria based on federal road route one subgrade
layer.
This paper starts with suggesting Drucker-Prager model for the analysis of the granular material,
and viscoelastic model for surfacing layer, and then it introduces using of heavy weight deflectometer for
elasticity modulus determination. It also describes the method of internal friction angle determination.
Moreover, this paper suggests two dimensional axisymmetry for finite element analysis. After that it
identifies the primary response parameters which can be gotten from the finite element analysis and can
help in damage identification and prediction. Finally, it suggests the formulas that can be used for
identifying the relationship between axle load, tire contact pressure and equivalent axle load factor based
on fatigue and rutting criteria. Figure 1 shows the topics discussed in this conceptual paper.
346
FINITE ELEMENT ANALYSIS OF FLEXIBLE PAVEMENT SUBJECTED TO HEAVY
AXLE LOAD AT MALAYSIAN FEDERAL ROAD ROUTE ONE
Finite element procedures Axle load factor

for pavement analysis determination
Choosing the Determination Finite element Axle load factor due Axle load factor due
proper analysis of needed model to fatigue criteria to rutting criteria
model parameters description
Drucker Elasticity Shape

Prager moduli geometry
model for
Base and
subbase
Friction Tire pavement
angle contact area
Viscoelastic
model for
asphalt Response
concrete layer parameters
Figure 1: Topics discussed in this conceptual paper.
2.0 OBJECTIVES
The objectives of this paper are
1- To choose the proper analysis models for each pavement layer to be used in finite element
analysis and to suggest the proper method to determine the parameters needed for finite
element analysis.
2- To suggest a method for using finite element analysis outputs to determine the relationship
between axle load, and equivalent axle load factor based on fatigue and rutting criteria based
on federal road route one subgrade layer.
347
3.0 PAVEMENT ANALYSIS MODELS

3.1 INTRODUCTION FOR PAVEMENT ANALYSIS MODELS (BASE AND
SUBBASE)
Flexible pavement is a complicated structure comprising of different materials with various

properties, therefore perfect mathematical description of the structure could be very complex.
Pavement structural analysis has been significantly developed during the previous years, it started
with modeling of soil as linear elastic material which has been introduced by Boussinesq [2], it is a quite
simple analytical model where vertical and horizontal normal and shear stress can be computed, in the
linear elastic theory, the pavement is assumed to be isotropic and homogenous. Such material can be
completely described by only two elasticity parameters such as Poissons ratio and modulus of elasticity.
Afterward, Due to the work of Burmister, Boussinesqs theory was extended to multilayer elastic
model [3]. Burmister introduced the multilayer elastic model in pavement design problems. The pavement
was represented as a multilayer system. Each layer distinguished by different modulus of elasticity to
represent different material. Burmister developed his solution first for a two-layer system and then
extended them to a three layers system [4].
The weakness of linear elasticity is that it is a poor model to represent the real stressstrain
behavior of soils, except at low stress levels and small strains. since the most suitable values of Youngs
modulus, E, and Poissons ratio, , rely on confining pressure and deviator stress, there is not any fully
acceptable way o f ch oo sin g simp le valu es of E and for a lin ear-elastic analysis. This leads to the
formulation of many nonlinear models over the past forty years, some based on experimental evidence,
and others based on theoretical principles, some are more accurate than others [5].
A nonlinear elastic model was proposed by Duncan and Chang [6] based on hooks law to relate
strain to stress increments, and on the Coulomb failure criterion. Afterwards this model was enhanced by
Rodrguez-Roa[7] who showed that, on condition that the soil mass is not close to failure, the model can
predict granular soil behavior with sufficient accuracy.
More advanced models called elastoplastic models were developed to represent more precisely
the relation between shear and volumetric strains, and the behavior of soils at high stress levels. It is
suggested that the elastoplastic model is used for Malaysian federal road rout one evaluation. This model
is more complicated than linear and nonlinear elastic models. The biggest differences between nonlinear-
elastic and elastoplastic behavior appear near to failure, at failure, and subsequent to failure. This is
because the strains produced from an increment of stress at high stress levels are strongly affected by the
stresses existing in the soil before the new stress increment is applied [7].
For the case of Malaysian federal road one, a typical pavement cross section can be analyzed
using a linear-elastic perfectlyplastic DruckerPrager model [8] which is one of the elastoplastic models.
It can be used for simulation of the stress strain behavior of granular soil layers such as base and subbase
layer. This model was chosen because of its simplicity, since it only requires elasticity modulus and the
internal frictional angle to be determined. It can represent the behavior of soils at high stress levels more
accurately.
3.2 DRUCKERPRAGER MODEL FOR BASE AND SUBBASE LAYERS
There are various plasticity models that can be picked to stimulate the stress strain behavior of
frictional material such as Mohr-Coulomb model, Drucker-Prager model, Lade-Duncan and Matsuoka-
Nakai models, and Hoek-Brown model [9] It is suggested that Drucker-Prager model is used to simulate
348
the stress and strain behavior in base and subbase layer because of its simplicity since it only depends on
elasticity modulus and the internal friction angle.
In favor of modeling the yielding frictional material Drucker and Prager introduced the following
function.
(1)
Where and are the material constants
= First invariant of stress tensor
= Second invariant of deviatoric stress tensor
On a deviatoric plane, the previous equation can be drawn as a circle as shown in Figure 2.
However in principle stress space it can be plotted as a cone as shown in Figure 3.
Figure 2: Drucker-pragers and Mohr-coulomb yield surfaces on a deviatoric plane [9].
Figure 3: Drucker-Pragers yield surface in principle stress space [5].
349
The Drucker-Prager yield surface is matched with Mohr-coulmb yield surface to decide the
material constants and for analysis [9]. Figure 2 shows such a match. Mathematically, this condition
demands the following equations:
, (2)
Where = frictional angel and = cohesion
3.3 VISCOELASTIC MODEL FOR HOT MIX ASPHALT
The viscoelasticity theory first establishment was in the studying of polymers, and a detailed
introduction could be found in the book by Ferry (1980)[10]. However, the application of the
viscoelasticity theory was generally stimulated by the wide use of viscoelastic materials in solid rocket
propellant in space technologies [11]. The viscoelastic model takes in consideration the time rate of stress
and strains. However, the assumption of isotropy and homogeneity still valid [12].
Asphalt binder is a viscous liquid, after mixing with the elastic aggregate, in the process of
making asphalt concrete, viscoelastic behavior is expected. Moreover, polymer, which is viscoelastic
material, is often added to the asphalt concrete to make modified asphalt concrete. Therefore,
viscoelasticity is an unavoidable phenomenon for asphalt concrete. The viscoelastic model can be used
for asphalt layer analysis in our case because it is more realistic for modeling of the asphalt layer than the
elastic model.
One of the moduli that represent the viscoelastic behavior is the complex modulus. It is a
complex quantity, in which the elastic stiffness represented by the real part, and the internal damping of
the material represented by the imaginary part. The absolute value of the complex modulus is usually
called the dynamic modulus [13]. It is basically an elastic modulus taken from a viscoelastic model which
integrates many factors such as loading rate, temperature, grain size characteristics and bitumen viscosity
[5]. The National Cooperative Highway Research Program (NCHRP) suggests the using of the complex
modulus as a design parameter in mechanistic design [14]. In the federal road one case, the dynamic
modulus can be used with viscoelastic theory to represent the viscoelastic behavior of the surface layer,
and the value of the moduli can be obtained from the national Cooperative Highway Research Program
(NCHRP) guide.
4.0 ELASTIC MODULUS DETERMINATION (BASE AND SUBBASE LAYERS)
The traditional methods to determine the elastic modulus of granular materials are in-situ plate
loading test, triaxial compression test, and (California bearing ratio) test. The determination of
elasticity modulus by triaxial compression tests or in-situ plate loading tests is time consuming and might
involve significant expenses. In addition, there is a limitation on the number of samples or measurements
that can be obtained [5].
There are many experimental equations founded based on CBR values have been used for the
determination of elasticity modulus of granular soil [15]. Some of these equations are:
(Shell relationship)
(Poulsen & Stubstad , 1980)[15]
(Powel et al., 1984)[16]
stands for the modulus in MPa.
350
Regardless of these studies, it was proven that there is no obvious relationship between and
, moreover, the test was not suitable for granular materials with large particle sizes. Therefore,
the use of correlation for this type of materials can lead to large errors [5].
Due to the above mentioned reasons, it is suggested that the falling weight deflectometer is used
to determine the moduli of pavement layers. Falling weight deflectometer is a non-destructive testing
device used for pavement properties evaluation including elastic moduli for the different layers. The
falling weight deflectometer is designed to transmit on a load pulse to the pavement layers by dropping a
large weight on pavement surface. Deflection sensors will measure the deformation of the pavement in
response to the loading. By a method called "backcalculation ,the elasticity modulus can be determined.
Figure 4: Falling weight deflectometer
5.0 INTERNAL FRICTIONAL ANGLE DETERMINATION
Another important parameter that should be determined for the analysis is the internal friction
angle, which is a required parameter for Drucker-Prager model. Sensitivity analysis done by Lacey 2008
proved that small variations in frictional angle value do not influence the pavements response
considerably. Two extreme situations were chosen. The first one assumed that frictional angle is
equivalent to its highest value of the expected range for granular layer; the other considered the frictional
angle equivalent to the lowest value of its expected range [5].
Table 1: sensitivity analysis for internal friction angle [5]
351
Table 1, shows that changing the angle of internal friction for granular soil layers result in small
changes in tensile strains and deflections induced on the asphalt layer. This means that for a larger elastic
modulus and a larger thickness of the asphalt layer, the effects of small variation in the internal friction
angles would be even less significant. Thus, it is concluded that frictional angle can be estimated from
preceding local geotechnical experience gained from similar materials. In that case, triaxial tests may not
be required, after the estimation of internal friction angle base on local geotechnical experience [5].
6.0 FINITE ELEMENT METHOD (TYPE ,GEOMETRIC SHAPE, BOUNDARY CONDITIONS)
The pavement layers in the finite element method are considered to be a solid continuum. The
solid continuum domain is divided into sub domains. Each sub domain is divided into a number of finite
elements. These finite elements assemble to the whole problem during the analysis. These elements are
connected by common nodes at their common ends. The analysis gives an approximate solution of the
problem for different boundary conditions and under various loading types using energy or stiffness
formulation [18]. Finite element models have been used intensively pavement analysis and design. There
are three types of models which have been used for multilayered pavement structure analysis and design:
three-dimensional (3-D), plane strain, and axisymmetric [18]. For the Federal Road One case,
axisymmetric model can be used, since it is adequate and to reduce analysis time.
Finite element analysis has been confirmed to be suitable for applications to complex pavement
problems. Even though using the finite elements three-dimensional models can solve all the problems that
can be solved using two-dimensional models, three dimensional models are very expensive to be used in
terms of computational time and data preparation [19]. Thus two dimensional models will be used for our
case.
For solving finite element problem stiffness matrix needs to be derived. For stiffness matrix
derivation, there are three factors that should be considered, the geometry of elements, the material
properties of elements, and the degrees of freedom allowed for the nodes to move. The solution will give
us stresses and strains at integration points, and displacements at the nodal periods [20]. For federal road
case the geometry of the elements is suggested to be as shown in figure 5. The material properties models
using viscoelastic model for surface layer, and Drucker-Prager model for base and subbase layers with the
suggested boundary condition is shown in figure 6.
Figure 5: Suggested Finite-Element Discretization (Axisymmetric)
352
Figure 6: Suggested pavement cross section geometry and boundary conditions
7.0 TIRE-PAVEMENT CONTACT AREA.
One of the important parameters for pavement analysis is the tire-pavement contact area.Tire-
pavement pressure distribution is usually recognized to be complex and the distribution is affected by the
type of the tire. There are a lot of inconsistencies in the data collected from different experimental studies
evaluating the contact pressure distribution between pavement and tire [21]. However, the area of contact
may be represented by two half circles with a rectangle between them as shown in Figure 7, this shape,
which consists of a rectangle and two half circles, is converted to a rectangle as recommended by Huang,
having an equivalent area of 0.5227 L2 and a width of 0.6 L. Unfortunately this area is not axisymmetric
and cannot be used for axisymmetric model[13]. Since our model is axisymmetry it is suggested to use a
circular contact area having an equivalent area of 0.5227L2 . This assumption has been used for the
multilayer theory. Although it is not correct, the error incurred is believed to be small [13].
Figure 7: tire-pavement contact area [13]
8.0 PRIMARY RESPONSE PARAMETERS
Flexible pavement is a structure that consists of at least three layers; asphalt surface layer,
unbound aggregate base layer, and natural soil subgrade layer. Under usual conditions, the stress
distribution will be similar to the one in Figure 8. Rapid shifting from compression to tension in the top
and bottom of the pavement is caused by moving wheel load. The repetition of this action develops tensile
353
strain will cause fatigue cracks. In accordance with elastic layer theory, the maximum strain located at the
bottom of the surfacing layer [21]. Thus most pavement design models are based on strain at the bottom
of the asphalt layer to forecast the performance of fatigue cracking. In addition, rutting is caused by
cumulative vertical strain on the paving layer and the subgrade [23].
Figure 8: Flexible pavement response affected by uniform load P; successive wheel

positions A and B cause cyclic loading that result in compressive stress and strain
on subgrade, and tensile strain and compressive strain in pavement [24].
After finishing the finite element modeling, there are several mechanistic (primary) response
parameters of the pavements which required for damage prediction. These will assist us to know which
damages produced by heavy axle load. These parameters consist of the tensile strain at the bottom of the
pavement layer, the vertical compressive strain on top of the subgrade layer, and the surface vertical
deflection [21]. Some of these parameters will be used later in the determination of load equivalency
factors.
9.0 LOAD EQUIVALENCY FACTORS
Pavement design is based on traffic loading, which is being determined by the number of
repetitions of standard axle load, generally 18-kip (80 kilonewtons) single axle load applied on the
pavement by two sets of dual tires as shown in Figure 9. This is usually known as the equivalent single
axle load (ESAL) [25].One way of describing the damage caused by certain axle load is by using axle
load factor. The load equivalency factor is described as the damage per pass to the pavement structure by
a certain axle load in question relative to a standard axle load [26].
Figure 9: Standard axle load
354
After completing the finite element models of the pavement structure, it is then possible to
determine the load equivalency factor based on both fatigue and rutting criteria, by using maximum
tensile strain at the underside of asphalt layer under the standard single axle load, maximum tensile strain
at the underside of asphalt layer for the (i) axle load and (j) tire inflation pressure, maximum vertical
compressive strain at the top of subgrade under the standard single axle load, and the maximum vertical
compressive strain at the top of subgrade for the (i) axle load and (j) tire inflation pressure, as shown in
equation 3 and 4.
9.1 LOAD EQUIVALENCY FACTORS BASED ON FATIGUE CRITERIA
The ratio of to with the same material is the Equivalent Axle Load Factor (EALF) on the
basis of fatigue failure [27]:
= number of repetitions to failure of standard load and pressure

= number of repetitions to failure of arbitrary load and pressure
= maximum tensile strain at the underside of asphalt layer under the
standard single axle load of 80 kN and a tire inflation pressure of 80 psi
(550 kPa)
= maximum tensile strain at the underside of asphalt layer for the i axle load
and j tire inflation pressure
= fatigue coefficient.
9.2 LOAD EQUIVALENCY FACTORS BASED ON RUTTING CRITERIA
The ratio of to with the same material is the Equivalent Axle Load Factor (EALF) on
the basis of rutting failure [27]:
= number of standard load and pressure applications

= number of arbitrary load and pressure applications
= maximum vertical compressive strain at the top of subgrade under the
standard single axle load of 80 kN and a tire inflation pressure of 80 psi
(550 kPa)
= is the maximum vertical compressive strain at the top of subgrade for the i
axle load and j tire inflation pressure
= rutting coefficient.
355
A lot of researches have been done to determine the fatigue and rutting coefficients, resulting in
values for m and b ranging from 3 to 6 [2][28]. It is estimated that the values for federal road one should
be between these two values.
10.0 SUMMARY
In order to study the damages in relation with heavy axle load for Federal Road Route One, this
paper introduce a finite element analysis procedures to perform the pavement analysis. Another objective
of this paper is to suggest a method to determine the relationship between axle load, tire contact pressure
and equivalent axle load factor based on fatigue and rutting criteria based on Federal Road Route One
subgrade layer.
This paper started with suggesting the proper models for pavement materials. It suggested
Drucker-Prager model for base and subbase material modeling, and viscoelastic modeling with the use of
dynamic modulus for the asphalt concrete layer. Then, it recommended the methods for obtaining the
important parameters for the analysis. It recommended the usage of falling weight deflectometer for
moduli determination. The friction angle is estimated based on the preceding local geotechnical
experience.
The paper suggested two dimensional axisymmetric models for the finite element analysis since it
is adequate and to reduce analysis time. It also has suggested a certain geometric model with proper
boundary conditions with the use of circular contact area between pavement and tire.
Then it discussed the important response parameters that can be determine from the finite element
model. The response parameters are the maximum horizontal tensile strain at the bottom of the top layer
and the maximum vertical compression strain at the base of the subgrade layer. The maximum strain
located at the bottom of the surfacing layer is responsible for flexible pavement fatigue cracking and the
cumulative vertical strain on the base of the subgrade layer is responsible for rutting of the pavement.
Finally the paper suggests the formulas to be used in order to determine the relationship between
axle load, tire contact pressure and equivalent axle load factor based on fatigue and rutting criteria.
REFERENCES
1- Officials, A. A. o. S. H., & Transportation. (2003). AASHTO: American Association of State

Highway and Transportation Officials.
2- Boussinesq, J. (1885). Application des potentiels l'tude de l'quilibre et du mouvement des
solides lastique. Paris: Gauthier-Villard.
3- Burmister, D. M. (1943). The Theory of Stresses and Displacements in Layered Systems and
Applications to the Design of Airport Runways. Proceedings, Highway Research Board, Vol.
23, 126-149.
4- Burmister, D. M. ( ). The General Theory of Stresses and Displacements in Layered Soil

Systems. Journal of Applied Physics, Vol. 16.
5- Lacey, G., Thenoux, G., & Rodrguez-Roa, F. (2008). THREE-DIMENSIONAL FINITE

ELEMENT MODEL FOR FLEXIBLE PAVEMENT ANALYSES BASED ON FIELD
MODULUS MEASUREMENTS. Arabian Journal for Science and Engineering, Vol. 33, 1B
66-61B66.
356
6- Duncan, J. M., & Chang, C. Y. (1970). Nonlinear Analysis of Stress and Strain in Soils.
ASCE Journal of the Soil Mechanics and Foundations Division, Vol. 96.
7- Rodrguez-Roa, F. (2003). Observed and Predicted Behavior of Maipo River Sand. Soils and
Foundations (Japanese Geotechnical Society), Vol. 43.
8- Drucker, D. C., & Prager, W. (1952). Soil mechanics and plastic analysis for limit design.
Quarterly of Applied Mathematics, vol. 10(no. 2), pp. 157165.
9- Yu, H.-S. (2006). Plasticity and geotechnics: .
10- Ferry, J. D. (1980). Viscoelastic properties of polymers. Wiley, New York.
11- Williams, M. L. (1964). Structural Analysis of the viscoelastic materials. AIAA Journal 2(5),
785808.
12- Punya, P. K., & Mamlouk, M. S. (1995). Tensile versus compressive moduli of asphalt
concrete. TRR Report, No. 1492, Hot mix asphalt design, testing, evaluation and
performance. Washington, p. 144-150.
13- Huang, Y. H. (2004). Pavement analysis and design: Pearson/Prentice Hall.
14- NCHRP. (2004). Guide for Mechanistic-Empirical Design of New and Rehabilitated
Pavement Structures. Final Report, Part 3, Design Analysis.
15- Sukumaran, B., Willis, M., & Chamala, N. (2005). Three Dimensional Finite Element
Modeling of Flexible Pavements. Paper presented at the Advances in Pavement Engineering
(GSP 130), Austin, Texas, USA
16- Poulsen, A., & Stubstad, R. N. (1980). fastlggelse af E-moduler for underbundsmaterialer
ved mlinger i marken. Interne Notater 108,statens vejlaboratorium,vejdirektoratet.
17- Powell, W. D., Potter, J. F., Mayhew, H. C., & Nunn, M. E. (1984). The structural design of
bituminous road Laboratory report 1132, transport and road researsh laboratory, department
of transport.
18- Cho, Y.-H., McCullough, B., & Weissmann, J. (1996). Considerations on Finite-Element
Method Application in Pavement Structural Analysis. Transportation Research Record:
Journal of the Transportation Research Board, 1539(-1), 96-101.
19- Fang, H., Haddock, J. E., White, T. D., & Hand, A. J. (2004). On the characterization of
flexible pavement rutting using creep model-based finite element analysis. Finite Elements in
Analysis and Design, 41(1), 49-73.
20- Hadi, M. N. S., & Bodhinayake, B. C. (2003). Non-linear finite element analysis of flexible
pavements. Advances in Engineering Software, 34(11-12), 657-662.
21- Helwany, S., Dyer, J., & Leidy, J. (1998). Finite-Element Analyses of Flexible Pavements.
Journal of Transportation Engineering, 124(5), 491-499.
357
22- Ullidtz, P. ( ). Pavement analysis. Amsterdam: Elsevier.
23- Lay, M. G. (2009). Handbook of Road Technology: Taylor & Francis.
24- Mulungye, R. M., Owende, P. M. O., & Mellon, K. (2007). Finite element modelling of
flexible pavements on soft soil subgrades. Materials & Design, 28(3), 739-756.
25- Garber, N. J., & Hoel, L. A. (2008). Traffic and Highway Engineering: Cengage Learning.
26- Moussa, G. K. M., & El-Hamrawy, S. (2003). Load equivalence factors as function of axial
loads and tire pressure for design of flexible pavemet. AEJ - Alexandria Engineering Journal,
42(4), 443-451.
27- Kim, O.-K., Bell, C. A., & Wilson, J. E. (1989). Effect of Increased Truck Tire Pressure on
Asphalt Concrete Pavement. Journal of Transportation Engineering, 115(4), 329-350.
28- Hudson, s. w., & Seeds, S. B. (1998). EVALUATION OF INCREASED PAVEMENT

LOADING AND TIRE PRESSURES. Transportation Research Record 1207, 197-206.
358
THE ARRANGEMENT OF SHUTTLE TRAIN TRANSPORTATION MODEL AS AN

INTEGRATED TRANSFER SERVICE MODA OF ADISUTJIPTO AIRPORT, YOGYAKARTA
AND ADISUMARMO AIRPORT, SURAKARTA, CENTRAL JAVA
Lilik Wachid Budi Susilo 1*, Alvin Nur Muhammad 2.
ABSTRACT:
Aviation deregulation policy encourages air transport rapid growth in the last decade and also encourages
local areas to compete having their own airport. However, this infrastructure provision often pays less
attention to the principles of efficiencies (cost and other resources). This condition also occurs to the
implementation of Adisutjipto Airport, Yogyakarta and Adisumarmo Airport Surakarta. Although only 60
km apart, both airports are operated independently (under one management of PT. Angkasa Pura I). With
its current facilities and activities, Adisutjipto Airport can only serve up to 6 million passengers / year
(2025). The development is aimed at optimizing the runway. While for flights using larger aircrafts, the
flights are directed to Adisumarmo Airport because it has a long runway that capable to serve wide-body
aircrafts.
Theoretically, to maximize the two potentials can be reached by combining the two services in both
airports so they can operate as an airport. This research will attempt to make a connecting model of the
two airport services, with the focus on the use of shuttle train) as a means to connect the two airports. The
justifying of the use of shuttle trains provide an opportunity for the business sectors, especially railway
operators and industry as well as the airport management to directly involve in this service. For this
purpose, the researchers cooperate with PT. Railink (airport train operator) as a form of involvement of
the industrial sector. The results of this study are expected to become inputs for PT. Railink (especially)
and operators, the railway industry and the airport management (generally) to be followed up as the initial
indication of profitable business opportunities.
Keywords: Shuttle Train Model, Intermodality, Airport Service .
1
Senior Researcher of Center of Transportation and Logistic Studies, Universitas Gadjah Mada, Yogyakarta, Indonesia
2
Researcher of Center of Transportation and Logistic Studies, Universitas Gadjah Mada, Yogyakarta, Indonesia
*Correspondence Author: Alvin Nur Muhammad, Gadjah Mada University, Yogyakarta, Indonesia. Tel: +6281358034444,
office: +62274 556928 Fax: +62274552229. E-mail: alvin_nm@mail.ugm.ac.id / pustral-ugm@indo.net.id
359
INTRODUCTION
The deregulation of flight policies in Indonesia has been promoting the rapid growth air
transportation during the last 10 years. This rapid growth, in turn, increases public access to various areas
throughout the country and also significantly cuts travel time compared to previous moda (ships). The
travel speed of air transportation moda is also assumed to push the increase of economic. As a result of
the new policy, every region in Indonesia proposes to have its own airport. The provision of adequate and
luxurious airport becomes a pride for the region. However, this infrastructure provision often abandons
efficiency principles (cost and other resources).
This tendency is also experienced by Adisucipto Airport Yogyakarta and Adisumarmo Airport
Surakarta. Although they are managed under the same company, PT. Angkasa Pura (persero) and are
located only 60 km far, both of them are operated independently and disintegrated. This research is based
on two variables. The first one is the not optimal use of Adisucipto and Adisumarmo Airports and the
second one, the limitation of each airport.
Figure 1. Characteristics of Adi Sucipto and Adi Sumarmo Airports
Adi Sucipto Airport Adi Sumarmo Airport

Scope area South parts of Java South parts of Java
Obstacle Natural obstacle (Mount Merapi None
& Boko Hill)
Utilization Base of Air Force & Commercial Commercial
2
Location and Are 07,47 LS/110,26 BT - 88.690 m2 07.31.S - 110,45E - 559.691 m
Distance from City + 9 km + 14 km
Passengers (2006) 3,272,464 494.064
Distance from railway Integrated with railway Relatively far from railway
Reviewing the characteristics of those airports, integrating their services can be performed to
maximize their potential. This research tries to produce an integration model of those airports. This
research focuses on the connecting moda between the airports, i.e. shuttle train service.
Literature Review
Theoretical Concepts
The study on the arrangement of shuttle train model as an integrated connecting transportation
moda between Adisucipto and Adisumarmo airports is supported with strong theoretical literature. There
are three main major concepts which become the theoretical basic of the research. They are intermodality,
concept of airport service, and concept of optimizing the operation of Adisucipto and Adisumarmo
airports.
a. Intermodality Concept
The concept of intermodality is variously defined. In general, however, there is a good connection
and management among various moda (multimoda). Integrated intermodality can be seen as well
managed multimoda which provide efficient transportation service. The advantage of intermodal concept,
both for passenger or freight transportation, can be strengthen by focusing on the improvement of
physical inter-connection, coordination, and operation integration as well as providing better
communication and information system both for the operator and the passenger.
360
Below are definitions of intermodal transportation according to some transportation experts and
institutions:
a. Brad Jones, et. al (2000) in Transportation Law Journal, Vol. 27, 2000 defines intermodal
transportation as the transfer of people or goods using more than one type of transportation moda in
one way travel without any obstacle
b. McKenzie, et. al (2000) in The Book of Intermodal Transportation defines intermodal transportation
as the delivery of certain goods in containers using more than one moda
c. United States Department of Transportation (US-DOT), defines intermodal transportation as:
1. ..containerization, transportation/ piggyback, or other technology which can provide seamless
flow of goods and people using more than one transportation moda
2. ...provides connection between different moda such as road with port or bus to railway transit
3. ...a holistic perception on transportation where all moda corporate with each other according to
its role to provide the best service choice for stakeholders. As a consequence, the policies of all
moda have to be adjusted. This perception is previously known as balanced, integrated or
comprehensive transportation.
d. Dewey, J.F, et. al. (2003) in Summary of Final Report BC-354-44, Part A July 2003 Transportation
Intermodal defines intermodal transportation as transportation flows which employ more than one
moda (e.g. train-motor, motor-airplane, or train-ship). Described as a process of connection,
interaction and movement among transportation moda.
e. Transport for London Integration Department (2001) in Intermodal Transport Interchange for
London defines intermodal interchange as a transfer of passenger/people from one type of moda to
another one such as from bus to train. Meanwhile, one who conducts a travel and shifts between 2
(two) similar moda is not an intermodal transfer, such as shifting from one bus to another one.
f. Jones et al., (2000) in Multimodal Transportation Planning Needs Survey defines intermodal
transportation as the delivery of cargo and flows of people using more than one transportation moda
in a continuous travel.
The aim of intermodality system is optimizing the use of various moda and improving the
connectivity among them. The newest definition of intermodal system is the increase of seamless,
efficient and sustainable transportation which in turn may:
1. reduce the cost and improve the level of demanded service both for freight as well as passenger using
the most appropriate moda,
2. reduce infrastructure burden and improve the total efficiency through shifting to a moda which has
bigger capacity,
3. reduce the cost and time needed as well as the inconveniency of inter moda shifting,
4. improve economic productivity and efficiency which in turn may improve the competitive value of the
products both in regional as well as national level,
5. reduce the level of energy use and improve environment quality.
b. Concept of Airport Service
In order to perform the study on the optimization development of the two airports, a
comprehensive understanding on the concept of airport as a system is essential. Ashford et al (1997)
defines airport as an interaction place of three main air transportation components that is airport, airlines,
and passengers. Airport is a meeting place of three major functions (1) transfer of moda from air moda to
361
land moda and vice versa; (2) processing (document, ticket, including the monitoring of passengers and
goods); and (3) changes of people and goods movement types from land moda to flight groups (in
airplane unit). Further, this interaction can be illustrated in a hierarchy system diagram as follow (Ashford
and Boothby, 1979 in Ashford et al, 1997).
Figure 1. The Hierarchy System of Airport
The figure above shows the interaction among airport, airline, and passengers which results in
several parameters for operational scale, passengers demand, airport capacity and flight capacity.
c. The Optimization Concept of Two Airports Operational.
The integration of two airports services also needs to consider the concept of maximum
operational optimization. Those concepts consider four major aspects i.e. infrastructure, facility, system
and cost.
362
infrastructure Tool
Air side Airlines

Railway liaison service
(land)
liaison service
Land side (train)
optimal
airport
service
flight operations flight rate
liaison service
airport operations rate
system cost
Figure 2. The Optimization Concept of Two Airports Operational
For infrastructure aspect, in order to optimize the integrated service of the two airports there are
some essential infrastructures which need to be considered, both from its quality and quantity aspects.
Those cover: infrastructures which are related to land area, air area, railway route and train station. From
the facility aspect, there are some issues which are related to airline, land interconnection service, and
train interconnection service. Meanwhile, system aspect correlates with the existing flight plan and its
implementation, the existing airport and train operation. Cost aspect as a stimulant which a passenger
must issue before using a service can be used as a standard of comparison for the level of passengers
satisfaction of the given service. This aspect can be reviewed from the cost of flight and cost of
interconnection services. It is hoped that the integration among those four aspects will improve efficiency.
Theoretically, the consideration which becomes a passengers base to determine which airport
that he will choose to perform a flight is significantly related with access time/distance and number of
flight. From this assumption, it can be concluded that the integration of the two airports will be optimum
if the travel time among both of them can be compressed as low as possible and the airports provide
frequent service so that the passengers may have more travel choices.
Figure 3. The Operation System of Two Airports
363
Empirical Literature
Some countries have developed the concept of shuttle train to improve the development of two or
more airports service facilities, such as:
a. Narita International Airport Haneda Airport (Tokyo, Japan)
Haneda and Narita Airports are closely located to each other. It needs 2 hours to reach one other
using train moda. Narita International Airport is managed by Narita International Airport Company
(private), while Haneda Airport runs under the management of Japanese Government with Japan Airport
Terminal Company (governmental company) as the operator of passengers terminal. Although those two
airports are under different management, the existing routes are arranged based on the Regulation of
Department of Infrastructure, Tourism, and Transportation. Generally talk, there is no particular shuttle
train service which connects the airports, but both of them are connected by fast train routes such as
Keikyu, Toei_asakusa and Keisei which operate every 20 minutes.
To connect each airport to the city center, there is a shuttle train with 30 minutes headway. From
the number of flight service, Haneda airport provides less international flight service than Narita airport
(only to Shanghai, Seoul and Hongkong), while Narita airport provides flight almost to all countries in the
world.
b. Heathrow Airport- Gatwick Airport - Stansted Airport (England)

Heatrow, Gatwick and Stansted airports are three airports of England which are closely located to
each other. There is no special inter-connection access among them but bus moda which is connected to
the city center of each airport. Nevertheless, there is a shuttle train which connects each of the airports
with the city center in every 15 minutes. If the passengers want to go to the other airports by train, they
have to take a transit in the city center and take a bus moda to the center city of the intended airport and
finally take a shuttle train to reach the airport.
METHODOLOGY
The choice of the research methodology is generally based on the less optimal utilization of
Adisucipto and Adisumarmo airports and their limitations. The first step of the research is performing an
identification survey on the need and demand characteristics of the airport users. The result of this initial
survey (characteristics, needs and demands) is used as the data for Stated Preference survey arrangement.
364
Stated Preference is an approach of respondents opinion toward several alternative choices. It aims to
reveal the tendency (choice) of each respondent toward the use of shuttle train as an inter-connection
moda between the two airports.
This data analysis phase is the last steps of data processing. Here, the gathered data was analyzed
using quantitative and qualitative descriptive and statistical analyses. In qualitative analysis, the
researcher employed a frequency distribution table to describe the number and dissemination of each
observed variables which cover both information about respondents identity and their demands.
Meanwhile, the researcher used analysis factor, selection process and regression analysis to perform the
quantitative analysis.
The analysis was performed aggregately and non aggregately. The aggregate analysis was
conducted on age, income, sex, level of education, travel aim and travel frequency.
RESULT ANALYSIS
Result of Stated Preference Survey:

In accordance to the survey plan, number of survey forms that were distributed to the respondents
was 150 questionnaires as follows:
a. Yogyakarta = 100 survey forms
b. Solo = 50 survey forms
Each of the form consists of 16 questions. Hence, the total number of choice perception
(respondents) was 150 x 16 (questions) = 24000 respondents.
Generally, the interpretation results of factor and regression analyses can be concluded as follow:
a. The result of aggregate equation shows that future users agree on the integration of the two airport
services using shuttle train.
b. Major aspects which become the main consideration of stakeholders in supporting the integration of
the two airport services suing shuttle train are travel time, cost, frequency, and sterilization of the
service.
c. Each of the future users has different individual and travel characteristics so that they have various
choices elasticity toward the aspects which are used as the basic of the two airport services integration.
In general, however, all respondents have the same sensitivity in determining the service aspects
which should be provided by shuttle train.
d. For the cost aspect, aggregately, the majority of future users (85%) agree on the integration of the two
airports if the shuttle train is free (included in the airport tax).
e. For the travel time, aggregately, in majority (92%), future users agree on the integration of the two
airports service if the travel time by shuttle train is no more than 30 minutes.
f. For the frequency aspect, aggregately, the majority of future users (79%) agree on the integration of
the two airports if the shuttle train provides service every 30 minutes.
g. For the service sterilization aspect, aggregately, the majority of future users (95%) agree on the
integration of the two airports service if the shuttle trains service is performed using a particular
train/not like present public train.
365
e. Aggregately:
1. Based on age criterion: older future users are more sensitive to service sterilization aspect while
younger ones pay more attention to cost, travel time and frequency.
2. Based on education level characteristic: future users with higher education level tend to be more
sensitive to travel time, frequency and service sterilization while those from lower education level
considers the cost aspect more
3. Based on monthly income characteristic: future users with higher income level emphasize on travel
time, frequency, and service sterilization while those who have lower income level tend to consider
the cost aspect of shuttle train.
4. Based on the characteristic of travel aim: future users who travel for business tend to be more
sensitive toward frequency aspect while those who travel for school consider travel time. For social
and tourism trips, people pay more attention to the cost aspect of shuttle train.
5. Based on the frequency characteristics of air moda use: future users who have higher frequency of
air moda use commonly consider travel time and tariff, meanwhile those who have less frequent
use emphasize on frequency and sterilization aspects of shuttle train.
Formulation of Shuttle Train Model
From the result of the analysis, the researcher obtained a model of shuttle train service which
connects Adisucipto Yogyakarta and Adisumarmo airports Surakarta as follows:
a. Travel time : 30 minutes
b. Cost : free (included in the airport service)
c. Frequency : every 30 minutes (during the airport service hours)
d. Sterilization : Using particular train (separated from present public train service)
With the service model above, it is estimated that the number of future users who should support
the integration of the two airports service and are willing to use the shuttle train service range between
79% to 95%.
CONCLUSION
The research concluded that theoretically the integration of Adisucipto Airport Yogyakarta and
Adisumarmo Airport Surakarta will improve the optimization of service provision of the passengers of
both of the airports for they have several differences (positive and negative) which may complete each
other. Those airports also share the same air space and are managed by the same governmental company
(PT. Persero Angkasa Pura I).
The feasibility of using shuttle train to connect those airports is positive because the distance
between them is relatively close enough (only 60 km) and there is already a double railway route
available (near) them. In addition, train moda with double route has an essential advantage compared to
others, i.e. punctuality. Nevertheless, the development of this shuttle train should consider travel time,
cost, frequency and sterilization.
366
REFERENCES
Hole, Arne R. (2004). Forecasting the demand for an employee Park and Ride service using commuters
stated choices. Transport policy 11 (2004) 355-362.
Ashford, N., Stanton., H.P.M., Moore, C.A. (1997). Airport Operations, Second Edition. McGraw-Hill,
New York.
Hair, F.J., Anderson, R E., Tatham, R.L., and Black, W. C. (1998). Multivariate Data Analysis. Prentice
Hall Intemational, Inc, New Jersey.
Johnson, R. A. and Wichern, D. W. (1992). Applied Multivariate Statistical Analysis. 3rd edition, Prentice
Hall, New Jersey.
Pearmain, D., and Kroes, E. (1990). Stated Preference Techniques: A Guide to Practice, Steer Davies
Gleave Ltd. London Hague Consultancy Group, Den Haag Netherlands.
Center for Transportation and Logistic Studies (Pustral) UGM. (2004). Preparation of Adisutjipto
Airport, Yogyakarta Master Plan (1st phase). Final Report, Department of Transportation,
Communications and Informatics (Dishubkominfo) Special Region of Yogyakarta.
Center for Transportation and Logistic Studies (Pustral) UGM. (2005). Preparation of Adisutjipto
Airport, Yogyakarta Master Plan (2nd phase). Final Report, Department of Transportation,
Communications and Informatics (Dishubkominfo) Special Region of Yogyakarta.
Todoroki, T., Hanzawa, Y., Fukuda, A. (1994). Simulation Model for Policy Analysis on Multi-Airport
System in Metropolitan Region. International System Dynamic Conference, Tokyo
Washington, S.P., Karlaftis, M.G., Manerring, F. L. (2003). Statistical and Econometric Methods for
Transportation Data Analisys. Florida Chapman & Hall.
367
368
DEVELOPMENT OF ONELANE ONRAMP MODEL FOR EXPRESSWAYS IN

MALAYSIA
Leong Lee Vien 1 and Tuti Azmalia Azai 2*
ABSTRACT:
Merging activities in the on-ramp of an expressway often cause turbulence and congestion which might
be critical if it does not being handled in accordance to specifications. Till now, not many research works
that has been conducted to access the operational performance of on-ramp junction based on Malaysian
expressway conditions. Guidelines and specifications that are followed in our country are based on US
HCM 2000. These specifications set by HCM 2000 are in accordance to the drivers behaviour,
operational performance and traffic configuration of United States. The model predicted in HCM 2000
must be evaluated and it is vital in order to prevent from any future under or over design of expressway
that are carried out in this country. Hence, models to estimate flow rates in Lane 1 and 2 immediately
upstream before merge were developed. In this study, two types of models of one-lane on-ramp with 3
lanes in the mainstream were developed and discussed. In these models, relationships between parameters
involved are shown by plotting the graphs. It was concluded that the models developed show significant
differences from the models presented in the US HCM 2000 regarding on-ramp junction.
Keywords: On-ramp, expressway, flow rate
1
Lecturer, Universiti Sains Malaysia, Sri Ampangan, Engineering Campus, 14300 Nibong Tebal, Penang, MALAYSIA
2
Student, Universiti Sains Malaysia, Sri Ampangan, Engineering Campus, 14300 Nibong Tebal, Penang, MALAYSIA
*Correspondence Author: Tuti Azmalia Azai, Universiti Sains Malaysia, Malaysia. Tel: +6045996266, Fax:
+6045941009. E-mail: tutiazm_87@yahoo.com.my
369
INTRODUCTION
Expressway has become an integral part of transportation system in Malaysia. According to U.S.
HCM 2000 (Transportation Research Board, 2000), expressway is defined as a divided highway with full
control of access and two or more lanes for the exclusive use of traffic in each direction in which it
provides uninterrupted flow with no signalized intersection and access through and from expressway is
limited to ramp locations. Opposing direction of flows is continuously separated by a raised barrier, at-
grade median or a continuous raised median. According to U.S. HCM 2000, ramp is a length of roadway
providing an exclusive between two highway facilities (expressway, multilane highways, two-lane
highways, suburban streets and urban streets). The expressway network in Malaysia consists of
approximately 1,229.5 km length in total which comprises of many different expressways such as the
North-South Expressways (NSE), Kuala Lumpur-Seremban Expressway, Shah Alam Expressway
(KESAS) and many more. In Malaysia, expressways are defined as high-speed routes with at least four
lanes (two in each direction) and may be either with limited or partial access.
A ramp expressway junction is regarded as an area of competing traffic demand for space and in
truth the vehicles from the merging lane is competing for space with vehicles from the mainline and thus
this creates turbulence in the merging area. As such, merge area represents a critical facility within the
freeway system and should be analyzed and investigated to determine capacity and level of service (LOS)
so that traffic congestion can be predicted (Albanese et al., 2003).
Entrance ramp has become the major concern of many traffic engineers because if the design of entrance
ramp does not comprehend to the available standards, it will cause a serious congestion in the particular
area and hence, it will affect the operational performance of the expressway. The U.S. HCM 2000 defines
that the basic approach in modelling the merge and diverge area is by focusing on the influence area of
450 m including the acceleration lane, Lane 1 and Lane 2 on the freeway. Although other freeway lanes
may experienced some congestion due to merging or diverging operation, but U.S. HCM 2000 only
considered the most critical part of the freeway merging activities to occur in the influence area of within
450 m, which will directly affect the LOS of the ramp. Flows entering ramp junctions on freeways, both
at on-ramps and off-ramps cause turbulence to traffic from mainline.
LITERATURE REVIEW
Reviews on expressway ramp
In U.S. HCM 2000, the procedures may be applied in approximate manner, to completely
uncontrolled ramp terminals on other types of facilities such as multilane and two-lane highways. The
method used to evaluate the flow entering the merging or diverging areas in U.S. HCM 2000 is by using
the ramp level of service (LOS) and capacity. Procedures in the U.S. HCM 2000 allowed for the
identification of likely congestion at ramp freeway terminals with LOS F and for the analysis of
operations at ramp freeway junctions and on-ramp roadways at LOS A through E.
There are 3 major steps in determining the LOS of merge area of on ramp expressway which comprises
of:
a) Determining V 12 (flow rate in Lanes 1 and 2 of freeway immediately upstream of

merge (pcu/hr)
b) Determine the Critical Capacity Values (CCV)
c) Determine the level of Service (LOS)
370
According to U.S. HCM 2000, two types of capacities were needed to be checked. They are the
total flow from merge area and the total flow into merge influence area. The capacity of merge area is
determined by the capacity of downstream of freeway. Hence, the capacity of upstream of freeway
segment plus the capacity of on-ramp of freeway cannot exceed the capacity of departing downstream
freeway.
For an on-ramp, the flow into the ramp influence area includes V 12 and V R . Thus, the total flow
entering the ramp influence area is given in equation (1):
V R12 = V 12 + V R (1)
While the total flows from merge area is determined by the equation (2):
V FO = V F +V R (2)
There will be two conditions which may occur in an analysis. First is the total departing freeway
flow, V FO may exceed the capacity of the downstream freeway segment which will be indicated by LOS
F in which queues will form upstream of the merge segment. The second condition is when the total flow
entering the ramp influence area, V R12 exceeds its maximum desirable level but the total freeway flow,
V FO does not exceed the capacity of the downstream freeway segment. Table 1 shows the lists capacity
flow rates for the total downstream freeway flow (V FO = V F +V R ) and maximum desirable values for the
total flow entering the ramp influence area, V R12 .
Table 1: Capacity values for merge areas (Transportation Research Board, 2000)
Maximum Downstream Freeway Flow,

Freeway Free Max Desirable Flow
V FO (pcu/hr)
Flow Speed Entering Influence Area,
(km/hr) Number of Lanes in One Direction V R12 (pcu/hr)
2 3 4 >4
120 4800 7200 9600 2400/ln 4600
110 4700 7050 9400 2350/ln 4600
100 4600 6900 9200 2300/ln 4600
90 4500 6750 9000 2250/ln 4600
Albanese et al. (2003) stated that a ramp-freeway junction is an area of competing traffic
demands for space. In fact, in merge areas traffic flows in the main line compete for space with ramp
inflows. In particular, ramp vehicles try to find suitable gaps in the main line in order to merge. This
phenomenon creates turbulence in the traffic flow on the main line near junctions. Hence, they studied on
a methodology for estimation of the capacity of the merge area that refers separately to both the main line
and the ramp. The capacity of the competitive streams within the merge area depends on the opposing
traffic volumes. The proposed approach could be considered as a link between the U.S. HCM 2000
hypotheses and those of gap-acceptance theory.
Interaction between ramps and mainstream will cause hesitation from drivers near the ramp
junction. Yi and Mulinazzi (2007) stated that expressway drivers travelling right before the influence
area may decide to accelerate to surpass a ramp influencing driver, seek and accept a gap in the entry
traffic to merge with the ramp flow or slowdown or change lanes to avoid a merging conflict.
Muhammad Akram (2007) has also studied on the development of entrance ramp merging density
model based on urban expressway traffic condition mainly in Federal Highway Shah AlamKuala
371
Lumpur expressway. In his study, he found out that the standard guide for capacity and operational
analysis needs to be justified as it only applies in USA since there are differences in traffic conditions,
geometric design, traffic regulation, vehicle composition and drivers behaviour from other countries. In
his study, he developed two models for estimating flow rates in Lanes 1 and 2 immediately upstream of
merge influence area and another model for prediction of density in merge influence area at entrance
ramp junction. The results showed that the new developed models were more accurate than the U.S.
HCM 2000 entrance ramp model as the models were developed based on local conditions. Based on his
study, flow rates in Lanes 1 and 2 are shown in equation (3):
V 12 = 399 + 0.618V F 0.134V R 0.929L A (3)
In a study conducted by Kittichai (2006), in order to obtain a model that suits local behaviour, a
model is proposed to predict the proportion of vehicles entering the merging influence area based on data
collected in Thailand. In his findings, the U.S. HCM 2000 model was tested on the data of both the
Interstate 25 in the USA and the Srirath Expressway in Thailand. He found out that it is likely that the
proportion of vehicles in Lanes 1 and 2 (PFM) estimated by U.S. HCM 2000 is significantly different
than field data at the 95 percent confidence level. Table 2 shows the list of equations used for predicting
V 12 immediately upstream of the ramp influence area in U.S. HCM 2000. This table does not specify the
number of acceleration lanes involved. These equations apply to six and eight-lane freeways (with three
and four lanes in each direction, respectively). For four-lane freeways (two lanes in each direction), only
Lanes 1 and 2 exist, and V 12 = V F by definition.
Table 2: Models of predicting V 12 on-ramp (Transportation Research Board, 2000)
V 12 = V F * P FM (4)
For 4-lane
freeways P FM = 1.000
(2 lanes each
direction)
For 6-lane P FM = 0.5775 + 0.000092L A (5)
freeways P FM = 0.7289 0.0000135 (V F + V R ) 0.002048S FR + 0.0002L up (6)
(3 lanes each P FM = 0.5487 + 0.0801 V D / L down (7)
direction)
For 8-lane
freeways P FM = 0.2178 0.000125V R + 0.05887L A /S FR (8)
(4 lanes each
direction)
Where:
V 12 = Flow rate in Lanes 1 and 2 of freeway immediately upstream of merge

(pcu/hr)
VF = Freeway demand flow rate immediately upstream of merge (pcu/hr),
VR = On-ramp demand flow rate (pcu/hr),
VD = Demand flow rate on adjacent downstream ramp (pcu/hr),
P FM = Proportion of approaching freeway flow remaining in Lanes 1 and 2 immediately upstream of
merge
LA = Length of acceleration lane (m),
372
SFR = Free-flow speed of ramp (km/hr),

L up = Distance to adjacent upstream ramp (m), and
L down = Distance to adjacent downstream ramp (m).
According to U.S. HCM 2000, the principal influences on flow remaining in Lanes 1 and 2
immediately upstream of the merge influence area are:
Total freeway flow approaching merge area (V F ) (pcu/hr),
Total ramp flow (V R ) (pcu/hr),
Total length of acceleration lane (L A ) (m), and
Free-flow speed of ramp at point of merge area (S FR ) (km/hr)
Considering the analysis on the field data collected from Srirath Expressway, it shows that the
average headway time has a significant effect on PFM values at the 95 percent confidence level.
According to equation (5) from Table 2 (P FM = 0.5775 + 0.000092L A ), the only factor that affects P FM
values is the length of the acceleration lane. Normally, when flow rate increases or headway time
decreases, drivers tend to maneuver to the lane away from a ramp junction to avoid congestion at the
junction. As a result, the P FM values may decrease as the flow rate increases or headway time decrease.
METHODOLOGY
Data collection
This study focuses of on-ramp junction particularly in the rural and suburban areas. The scope of
the study is limited to the cases where one acceleration lane on-ramp merges with 3 lanes dual
carriageway mainline. This study focused on isolated on-ramp in which there is no influence of adjacent
ramps towards the studied ramp. In this study, considerable time is dedicated towards data collection and
data reduction. Traffic conditions where there are occurrences of congestion in the studied areas should
not be considered because this condition will caused different types of driving behaviour and stop and
go conditions were not included in this study.
In this study, data are collected using digital video camera which is being placed in high areas
(condominium, office building and etc) from the expressway where it can capture clearly the merging of
vehicles from acceleration lane with vehicles from mainline. High vantage point from a building roof
will provide best view on the expressway and therefore, it is important that the selected buildings have
flat roof so that a normal tripod for digital video camera can be placed on it. In this study, traffic data
were collected at 4 sites as shown in Table 3.
Table 3: List of sites selected for this study
1. KL-Seremban Expressway 1 KM D3.8

Site (one-lane 2. KL-Seremban Expressway 2 KM D4.1
on-ramp)
3. East-West Link Expressway KM C6.0
4. Kajang Dispersal Link KM 2.5
Expressway
373
Data reduction
In this step, the method used to reduce the data collected is manual counting. Manual counting
were done by video playback method in which vehicles are being counted and separated according to
vehicle classification based on Arahan Teknik (Jalan) 8/86 rural condition and arranged into tabular form.
In this study, the traffic condition of the data was being analyzed in 5 minutes interval. In data reduction,
traffic volumes are the main data to be obtained. Traffic volume can be described as number of vehicles
passing through a given point for a specific time interval. In this study, the unit for traffic volume is
vehicles per hour. In order to ensure the uniformity for all types of vehicles, all different types of vehicles
are converted to one standard by multiplying the vehicles type with passenger car unit (pcu) obtained
from Arahan Teknik (Jalan) 8/86 (Ministry of Works Malaysia, 1986). The conversion factor is shown in
Table 4.
Table 4: Conversion factors (Ministry of Works Malaysia, 1986)
Passenger Car Equivalent (pcu) for rural

Type of vehicles
conditions
Passenger car 1.00
Motorcycles 1.00
Medium Lorries 2.50
Heavy Lorries 3.00
Busses 3.00
RESULT ANALYSIS
Multiple regressions to develop the models were conducted using the SPSS 16.0 software.
Dependent variable and independent variables for analysis were input into the software and analyses were
conducted by choosing stepwise regression. Variables involved in the analysis are shown in Table 5. By
using stepwise regression, independent variables which are not significant to the model will be removed
based on the significance values generated. The removing of the insignificant variables will increase the
R value of the model.
Table 5: List of independent and dependent variables
1. Flow rate in Lanes 1 and 2 of freeway immediately upstream of

Independent variable
merge influence area, V 12
1. Flow on a ramp, V R
2. Flow entering Lane 3 immediately upstream of the merge
Dependent variables influence area, V 3
3. Flow approaching merge area, V F
4. Length of acceleration lane, L A
Figure 3 shows the relationship between flow entering Lanes 1 and 2 immediately upstream of
the merge influence area, V 12 and total flow approaching merge area, V F . In this relationship, the pattern
of the graph is as expected in which flow V 12 is directly proportional to flow V F .
374
Figure 3: Flow entering Lanes 1 and 2 immediately upstream of the merge influence area, V 12 against
flow approaching merge area, V F
While, Figure 4 shows the relationship between flow rate in Lane 1 on the expressway, V 1 and
flow rate in Lane 2 on the expressway, V 2 and it is found out that is a strong correlation between these
two parameters. Muhammad Akram (2007) stated that the expressway lane in which vehicles from
entrance ramp seeks gaps is Lane 1 and flow rate in Lane 2 are less affected than in Lane 1. Shin (1993)
also stated that among the volumes on three different lanes, the volume in Lane 1 is most affected by
volumes of entering ramp vehicles and the spatial configuration of a junction. However, the findings in
this study indicate otherwise in which this may be due to the reason that flow rate of Lane 1, Lane 2 and
Lane 3 for Site 2 which is the KL- Seremban Expressway are extremely high even during off peak hours
due to its accessibility to few other expressways such as Sungai Besi Expressway, New Pantai
Expressway, Shah Alam Expressway and et cetera. Hence, more vehicles from Lane 1 will seek gaps to
move to Lane 2 to avoid congestion which will occur when merging activities take place.
375
Figure 4: Flow rate in Lane 1, V 1 against flow rate in Lane 2, V 2
Figure 5 shows the relationship between flow entering Lanes 1 and 2 immediately upstream of
the merge influence area, V 12 and on-ramp flow, V R . The results obtained indicated moderate correlation
between the two parameters in which V 12 reduces when V R increases but rationally, V R would not create
much impact to V 12 mainly because only vehicles from Lane 1 will be affected by the merging activity by
changing lane to Lane 2 and hence it would not affect V 12 as a whole.
376
flow on-ramp, V R
In order to have a clear picture of relationship between V 12 and V 3 , a scatter plot between the two
parameters is plotted and is shown in Figure 6. As explained earlier, Site 2 had extremely high flow rate
on mainstream compared to other sites and thus the results shows that there is relationship between the
two parameters. Hence, based on the data collected at the four sites, the results obtained showed that V 3
is moderately affected by merging operation.
377
flow entering Lane 3 immediately upstream of the merge influence area, V 3
length of acceleration lane, L A
In regards of the effect of length of acceleration lane (L A ) towards the flow rate V 12 , a scatter plot
is produced and is shown in Figure 7. From the results obtained, it is found out that there is a weak
positive correlation between the two parameters. This suggests that flow rate V 12 will increase slightly in
regards of L A . According to U.S. HCM 2000, longer acceleration lanes encourage less turbulence as
ramp vehicles enter the expressway traffic stream and therefore lead to higher flow in Lanes 1 and 2.
378
Muhammad Akram (2007) stated that short acceleration lane provides ramp vehicles with restricted
opportunities to accelerate before merging with vehicles from mainstream.
In order to determine a model to predict V 12 multiple regression, a statistical analysis on the

expressway ramp is done by using the software SPSS 16.0. In this analysis, flow entering Lanes 1 and 2
immediately upstream of merge area (V 12 ) is chosen as the independent variable while length of
acceleration lane (L A ), flow on a ramp (V R ) and flow approaching merge area (V F ) are selected as the
dependent variables.
In the first method, a stepwise regression analysis is used to determine the independent variables.
In this method, independent variables which are statistically not significant will be removed to produce
the best multiple regression model indicated by the value of R. The multiple regression equation which
is obtained from this study is as shown in equation (9). The R obtained from this equation is 0.983.
V 12 = 408.524 + 0.551V F + 1.207L A (9)
Table 5 shows that the independent variables of V F and L A are the significant independent
variables in the model to predict flow V 12 . From Table 6, the significant values for all parameter are less
than 0.05 which means that the parameters are statistically significant and can be considered in the model
V 12 . From equation 9, it shows that both flow V F and length of acceleration lane, L A have positive sign
convention. This directly shows that if any of the both parameters increase in value, flow V 12 will also
increase. The parameter V R is removed from this model because it has insignificant contribution towards
the model.
Table 6: Summary of regression analysis for V 12
Independent
Coefficient t-Value Significant value
variables
Constant 408.524 9.004 0.000
VF 0.551 69.064 0.000
LA 1.207 3.926 0.000
R = 0.983
Apart from the regression model, another attempt was made to develop the model using the P FM
which is the proportion of approaching freeway flow remaining in Lanes 1 and 2 immediately upstream of
merge, as the independent variable and L A , as the independent variable, which is similar to the model
presented in the U.S. HCM 2000. The variable, P FM is calculated by dividing V 12 with V F . The result
obtained is as shown in equation (10). However, the value of R obtained from this equation is only
0.047.
P FM = 0.705 + 0.000376L A (10)
Table 7 shows that the significant values are less than 0.05 which means that the parameters are
statistically significant and can be considered in the model.
379
Table 7: Summary of regression analysis for P FM
Independent
Coefficient t-Value Significant value
variables
Constant 0.705 29.483 0.000
LA 0.000376 2.115 0.037
R = 0.047
Initial comparisons based on the values of R2 obtained for both methods indicated that the first
regression model is more suitable to be used to determine the model V 12 . This is because R for first
model is 0.983 while for the second model, the value of R2 obtained is only 0.047, which is much lower
as compared to 0.983. However, the model has yet to be validated. Therefore, in order to develop a more
accurate model to describe such a relationship, more data will be collected and the model has to be
validated.
CONCLUSION
In this study, analyses were conducted and models were developed based on one-lane on-ramp
with 3 lanes on the mainstream. Results from this study showed that there is inconsistency between
models predicted from U.S. HCM 2000 and from models developed in this research. The accuracy of the
data provided by U.S. HCM 2000 needs to be justified as there are differences of operational conditions,
drivers behaviour, geometric design between U.S. and Malaysia. As such, development of models which
suit local traffic conditions is very important.
ACKNOWLEDMENTS
(Inter-Urban). Appreciation also goes to Mr Alex Poi Wai Loong for his contributions in this study.
REFERENCES
Albanese, M., Camus, R., & Longo, G. (2003). Capacity and queue modeling for on-ramp- freeway
junctions. Transportation Research Record 185, 256-264.
Kittichai, T. (2006). Development of a model to predict proportion of vehicles entering merge influence
areas at freeway ramp junctions. Thammasat Int. J. Sc. Tech., 11 (3), 54-60.
Ministry of Works Malaysia. (1986). Arahan Teknik (Jalan) 8/86: A guide on geometric design of roads,
Muhammad Akram Adnan. (2007). Development of entrance ramp merging density model based on
urban expressway traffic condition. Unpublished PhDs thesis, Universiti Sains Malaysia.
Transportation Research Board. (2000). Highway Capacity Manual (2000). National Research Council,
Washington, DC.
Yi, H., & Mulinazzi, T.E., (2007). Urban freeway on-ramps-invasive influence on main-line operations.
California Department of Transportation, 112-119.
380
CUBE DYNASIM AND AASIDRA COMPARISONS IN ANALYZING A SIGNALIZED

INTERSECTION
Azlina, I 1, Intan Suhana, M.R 2,
ABSTRACT:
Due to the complexity task in analysis and design of a signalized intersection, the used of microscopic
simulation model has become very popular for traffic engineers and planners in dealing with such
activities. This is because of great advantage that offers by this model especially on the ability to
realistically model a complex transport system and it provides users with the most desirable statistics and
performance measures of alternative design or existed transport system. However, there is one issue
involving the use of microscopic simulation models which is on the variations of the simulation outputs
results in representing the real life traffic behavior. This study aims to assess the consistency of the results
generated by micro simulation models in analysis the performance of a signalized intersection. In this
case, a micro analytical model will be used alongside micro simulation model. Therefore, the results
produced by both models could be compared in order to assess for the consistency of micro simulation
model. To carry out the modeling of micro simulation and micro analytical, CUBE Dynasim and
aaSIDRA has been used as a tool accordingly. A four legged signalized intersection at South Road and
Henley Beach Road in Adelaide was selected to serve as a case study. The result comparison was based
on the performance measures of the signalized intersection such as the traffic flows, travel speed, delay,
level of service (LOS) and queues. The results demonstrated by CUBE Dynasim microscopic simulation
model was in similar pattern with aaSIDRA microscopic analytical model. It means that in this study the
CUBE Dynasim produced reasonably consistent outputs results in analyzing a signalized intersection.
Keywords: Signalized Intersection, Microscopic Simulation, Microscopic Analytical, Cube Dynasim,

aaSIDRA
1
Lecturer, Universiti Malaysia Pahang, Lebuhraya Tun Razak, 26300 Gambang, Kuantan, Pahang, MALAYSIA
2
Lecturer, Universiti Malaysia Pahang, Lebuhraya Tun Razak, 26300 Gambang, Kuantan, Pahang, MALAYSIA
*Correspondence Author: Azlina Ismail, Universiti Malaysia Pahang, Malaysia. Tel: +60095492952
Fax: +6095492998. E-mail: azlinai@ump.edu.my
381
INTRODUCTION
According to Taylor et al, (2000) and Sadoun (2003), signalized intersections are the most critical
locations in designing a transport system, because it has high tendency to experience traffic problems
such as traffic accidents, congestions and delays. Considering such issues, it is necessary to analyze the
efficiency and the performances of signalized intersections.
Nowadays, micro simulation models have been widely used by traffic engineer in analyzing and
evaluating the performances of signalized intersections. The need of micro simulation modeling has
become more important as they can give traffic engineer an overall picture of the traffic and the ability to
assess current problems and project possible solutions immediately.
However there is still a doubt on the accuracy of micro simulation models in representing the
real life traffic behavior as well as the generated outputs used to analyze signalized intersections. This is
especially true when the models are used without proper calibration and validation. Moreover, it must be
noted that micro simulation modeling are stochastic models whose results vary depending on the random
seed number used. Therefore multiple runs must be performed in order to ensure an accurate estimate of
the various performance measures. Therefore, it is important to quantify the reliability of simulation
model output in analysis the performance of signalized intersections.
OBJECTIVES
The main objective of this thesis is to assess the consistency of outputs generated by microscopic
simulation model in evaluating the performance of a signalized intersection. The assessment is done
based on comparative outputs between microscopic simulation and analytical model that carried out by
using tools of CUBE Dynasim and aaSIDRA accordingly. The primary performances measures selected
for comparison in this study included the traffic flows, travel speed and queues.
LITERATURE REVIEW
Traffic signals offer a great degree of control at intersections compared to the other control
devices (Nigarnjanagool and Dia, 2004). They control the movement of traffic at intersections by
permitting conflicting streams of traffic to share the same intersection by means of time separation. By
alternately assigning right-of-way to various traffic movements, signals provide for the orderly movement
of conflicting flows. However the signals control design is quite complex since it should consider on the
parameters of all the roads that make up the intersection. (Taylor et al, 2000). Considering such issue, it is
important to analyze the performance of the signalized intersection so that the intersection can perform
safely and efficiently.
The need of micro simulation modeling has become more important as they can give traffic
engineer a real overall picture of the traffic and the ability to assess current problems and project possible
solutions immediately. In the modeling process of microscopic simulation model, each vehicle is assigns
into their physical characteristic and operations. Vehicles in the network then will be updated in a second
or sub second basis, thus the interactions between vehicles in the network and between vehicles to the
roadway properties are modeled accurately on each simulation. Therefore, the microscopic simulation
software has become the most popular tool for traffic engineer or planners in analyzing signalized
intersections (Jones and Anderson, 2004; Brian et al, 2006). Furthermore, as been stated by Shaaban and
Radwan (2004) simulation modeling could save users huge amount of time, effort and money if used
correctly in conducting the analysis of transport system. At the same time, the micro simulation also
provides a powerful visual representation and graphical user interface. This is a great advantage
especially where the result of the analysis is to be presented to non-technical persons.
382
Nevertheless, Drew (1968) claimed that Simulation is a powerful tool, and like all powerful tools
it can be dangerous in the wrong hands. Therefore, there is still a doubt on the accuracy of micro
simulation models in representing the real life traffic behavior as well as the generated output used to
analyze the transport system, in particular signalized intersections. Gagnon et al (2008) claimed that
microscopic simulation models must be calibrated first against field data or against other validated
analytical models to ensure the accuracy. Beside, due to the complexity of micro simulation models a
number of pitfalls may exist and this will influence on the imperfect results of the model outputs (Bell,
1998). Some of the existing microscopic traffic simulation models include SimTraffic, VISSIM,
AIMSUN, CUBE Dynasim and others. In this study, the CUBE Dynasim simulation model was utilized.
CUBE Dynasim is a scenario based microscopic simulation model, event driven and stochastic. It
is extremely powerful software which is capable to simulate very quickly a wide variety of transportation
system such as freeway system or combined networks that have a simple or complex control conditions
for instance the signalized intersection. It takes into account any vehicles types as well as the driver
behavior. The output results of CUBE Dynasim could be reported in the form of numerical data and graph
representation. The numerical data output consist of a statistical values such as the maximum, minimum,
mean, standard deviation as well as the 25th, 50th and 74th percentile. The user can define their own
preference time samples to be used to report the output data during the hour simulated for example in
every 15 minutes or others. Then, CUBE Dynasim also provides a powerful visual representation
available in 2D and 3D modes. This is a great advantage especially when the result of the analysis is to be
presented to non-technical persons.
On the other hand, analytical model is based on the Highway Capacity Manual (HCM), 2000
which have been the industry standard for many years. It is developed based on some sort of theories and
empirical data. Traffic analysis tools based on HCM procedures can quickly predict capacity, speed,
delay, and queuing on isolated, small-scale transportation facilities. Some of the existing HCM-based
tools include aaSIDRA, RODEL, EMME 2, HCS and many more. Among all, aaSIDRA appears to be
more a popular and effective analytical tool in modeling and analyzing at micro level. It is micro
analytical software develops by Akcelik & Associates Pty Ltd. It uses detailed analytical traffic models
with an iterative approximation method to analysis the performances of intersection. (Akcelik et al, 2002).
This micro analytical tool uses lane by lane and vehicle drive cycle models together with an iterative
approximation method can be employed to provide estimates of capacity, level of service, delay, queue
length and other measure of performances. It is also can be used to predict and assess the impact of future
traffic growth by using the approaches of flow scale and saturation flow scales parameters. Moreover
aaSIDRA can also provide direct outputs at more sophisticated level such as the operating cost, fuel
consumption and pollutant emissions estimates for the intersection. From the aaSIDRA user guide (2004),
it stated that nowadays at least 1800 sites in 80 countries are using SIDRA.
METHODOLOGY
Study Location
An isolated signalized intersection at South Road and Henley Beach Road in Mile End (TS054)
was selected to serve as a case site study. It is located around two kilometers west from Adelaide CBD.
This intersection is a four-leg signalized intersection which is operated as a fixed time signal control and
has five signal phases. A pedestrian signal was operated at this intersection. However it is being omit for
the modeling and analysis in this study due to very less observed pedestrian movements during peak
periods. The layout diagram of this intersection is shown in figure 1 as in the following.
383
Figure 1: Signalized Intersection Layout
Data Collection
In developing both analytical and simulation models for the intersection the data requirements are
extensive. In this study, following data was collected such as traffic count, intersection geometry and
signal controller settings.
This intersection has vehicles detectors embedded underneath the road surface in each lane and
operated by Sydney Coordinated Adaptive Traffic Systems (SCATS) which this can help to ease the job
in data collection of traffic count. Therefore, the traffic count data were quoted from Transport System
Center, University of South Australia. In intersection geometry survey, the basic geometry features of
road elements such as number of lanes, width of travel lanes, length of slip lanes and short lane for turn
right and others was collected. These elements are necessary as a data input in developing the models for
the intersection. Meanwhile for signal controller settings, the optimum cycle time, operating signal
phasing and timing data was determined. The data gathered is then be used to develop both CUBE
Dynasim and aaSIDRA models. Thus, the signalized intersection plan for both models is identical.
Model Development
1) Micro Analytical Model: aaSIDRA
All the input data that gathered during the field observation was inserted in Road Intersection Data
Editing System (RIDES) that included in the aaSIDRA software packages. RIDES are a graphic based,
highly interactive program that reflects the design process of actual condition of the intersection. In
RIDES operation, it involved several stages of group to insert the data such as basic parameter,
intersection geometry, approaches, lanes, volumes and others. The RIDES program then will run all the
input data to develop the base model. There are some input data that set by default in this program such as
the basic saturation flow, practical degree of saturation, flow parameters and others.
2) Micro Simulation Model: CUBE Dynasim
In developing the CUBE Dynasim model, it involves several stages begin with the development of
network scenarios, traffic flow scenarios, signal scenarios and public transport scenarios. In CUBE
Dynasim package allows of a set of imported background map as an aid for coding. It gives a greater
flexibility by accepting formats in DXF CAD files or BMP image files. As in this study, the selected
signalized intersection has been draw in CAD format and exported to CUBE Dynasim in DXF extension.
384
Model Verification and Calibration
Calibration and validation is the process where the developed model is producing results that are
as close as possible against the actual fields condition. If there is no different made from the results
obtain by the develop model to the actual field condition, imply that the model develop are said to be
calibrated and validated.
Some basic calibration parameters has been selected to be used in this study which were the road
geometry condition, vehicle driving speed and the traffic control system that included the traffic signal
setting and priority management that control vehicles priority. As in this study, give ways control priority
was used.
RESULT ANALYSIS
The outputs generated by both models have been compared to assess for their consistency in
evaluating the performance of a signalized intersection. The comparison was made based on measures of
effectiveness (MOEs) which were the traffic flows, travel speed and queues
1) Traffic Flow
A traffic flow comparison between aaSidra and CUBE Dynasim was conducted for the study site
signalized intersection. This is presented as in table 1 below. The traffic flows for each approach and each
movement were grouped into 1 hour time period parallel to time sample used in this study.
Table 1: Comparison of CUBE Dynasim and aaSIDRA Traffic Flows at the Intersection
CUBE Differences
Approach Movement aaSIDRA GEH
Dynasim (%)
SW 74 73.1 1.21 0.105
South SN 1287 1292.3 -0.41 0.148
SE 187 190.5 -1.87 0.255
ES 82 78.5 4.27 0.391
East EW 454 450.5 0.77 0.165
EN 40 40 0 0.000
NE 112 115.8 -3.39 0.356
North NS 1089 1087.9 0.1 0.033
NW 275 273.8 0.43 0.072
WN 261 259.4 0.61 0.099
West WE 983 984.5 -0.15 0.048
WS 179 177.2 1 0.135
Total 5023 5023.5 -0.009 0.007
It has been noted, the variation of the CUBE Dynasim flows varied to the aaSIDRA flows
by less than 5 percent where most of them were in a range of 1 percent differed as shown in the
table. This indicates a well consistency of the output data between the both model flows. The
only major variation noted was for the east approach on the east to south movement where the
modeled flows differ by 4.27 percent from the observed data. The CUBE Dynasim released only
78 vehicles compared to the aaSIDRA which was 82 vehicles. It has been noted that this major difference
385
was occurred on a left turn movement from east to south where this movement was provided by an
exclusive slip lane that separated from through lane. Therefore it can be said that the entry flows for
CUBE Dynasim did not permit as high capacity as did by the aaSIDRA in representing the actual field
condition. This may due to the differences of vehicles length that specified between both models.
The GEH statistic values were also identified in a way to check for the validation
between CUBE Dynasim and aaSIDRA data. Various GEH statistic values give an indication of
a goodness of fit between the CUBE Dynasim and aaSIDRA data as outlined below:
GEH < 5 Flows can be considered a good fit

5 < GEH < 10 Flows may require further investigation
10 < GEH Flows cannot be considered to be a good fit
Overall, the GEH values that obtained for the entire network of signalized intersection
were less than 5 for the morning peak periods respectively where this means the flows can be
consider a good fit. In fact, the evaluation for the whole traffic flows at the intersection also
indicates a very close match between the CUBE Dynasim and aaSIDRA flows since the GEH
values observed was less than 1. Hence, it can be said that the modeled flow generated by CUBE
Dynasim produced a well consistent and accurate simulations of the observed traffic flows.
Moreover, the comparison of measured traffic flows approaching and exiting at the
intersection were represented as in table 2 below while the figure 2 shows the aaSIDRA outputs
of traffic approaching, circulating and exiting. It can be said that in particular Cube Dynasim
released more vehicles compared to aaSIDRA. However, for overall, the traffic approaching and
exiting Cube Dynasim flows agreed well with the aaSIDRA measurement mostly in a range of
less than 1 percent.
Table 2: CUBE Dynasim and aaSIDRA Approaching and Exiting Traffic Flow
CUBE Differences
Approach Flow aaSidra
Dynasim (%)
Approaching 1548 1555.9 -0.51

South
Exiting 1350 1353 -0.22
Approaching 1476 1477.5 -0.1
North
Exiting 1588 1592 -0.25
Approaching 576 569 1.22
East
Exiting 1282 1276 0.46
Approaching 1423 1421.1 0.13
West
Exiting 803 798 0.62
386
Figure 2: aaSIDRA Approaching, Circulating and Exiting Traffic Flows
2) Travel Speed
Vehicles speed was another performances measure that takes into account in this study. Speed is
the rate of change of distance with respect to time and is usually measured in km/hr. It is conducted to
estimate the distribution of speeds of vehicles in a traffic stream. It may be used to evaluate the
effectiveness of traffic control devices or to determine the speed trends.
In aaSIDRA outputs, the performances measures for speeds include the cruise speed, average
travel speed and average running speed. Meanwhile in CUBE Dynasim, there are two types of travel
speed produced which were the maximum and the average travel speed. However in this study, as the
main objective was to assess the consistency of the performances measures generated by the simulation
model to the observed data, therefore the recognitions on the same parameter used for comparison is very
important. Thus, the model average travel speed demonstrated from CUBE Dynasim has been compared
to the observed average travel speed that generated from aaSIDRA as tabulated in the table 3.
Table 3: Comparisons of Average Travel Speed
Speed (km/hr)
Average Differences
Approach Movement Average
Speed (%)
Speed
(CUBE
(aaSIDRA)
Dynasim)
SW 47.1 48.33 -2.61
South SN 30.1 31.16 -3.52
SE 20.6 21.85 -6.07
ES 43.6 45.73 -4.89
East EW 25 27.04 -8.16
EN 21 23.33 -11.10
NE 41.6 41.95 -0.84
North
NS 32.9 35.1 -6.69
387
NW 19.9 18.82 5.43

WN 29.3 31.78 -8.46
West WE 27.6 29.56 -7.10
WS 28.7 30.76 -7.18
From table 3 above for the comparison between modeled and observed average travel speed, it
was found that the average travel speeds were generally simulated fairly well across all the movements at
the intersection by the CUBE Dynasim simulation model. This is due to small variations of average travel
speeds between them which less than 10 percent.
3) Queues
Basically, the queue in the intersection implies the line of stationary vehicles waiting to proceed
through an intersection. For this analysis, aaSIDRA produce a direct measure for the queue outputs
namely queue distance, queue storage ratio and queue length by means of number of vehicles counted on
the traffic stream. On the other hand, in CUBE Dynasim, the queue data could be obtained by conducted a
direct visualization testing from the simulation animations movie of the CUBE Dynasim. This method
may take a longer time where it requires the user to carefully watching the simulation animations movie
of the CUBE Dynasim in order to count for vehicles on the queue. By conducting this method, the queue
data will be reported in terms of number of vehicles on the queue as well as the queue distance on each
lane manner. However in this study the queue outputs will be reported in approach by approach manner to
make it comparable with the outputs reported in aaSIDRA. The details queue data is tabulated in table 4
below.
Table 4: Comparison of Queue (Number of Vehicles)
Queue (veh)
Approach Movement
aaSIDRA CUBE Dynasim
SW 1 2
South SN 34.5 41
SE 7.2 8
ES 2.1 4
East EW 14.1 19
EN 3.4 4
NE 3.3 7
North NS 25.8 33
NW 10.4 10
WN 21.9 14
West WE 32.9 37
WS 9.2 10
As shown in table 4, the results of CUBE Dynasim queue were found to be matched closely to the
aaSIDRA queue by variations in number of vehicles as small as 1 vehicle which was noted on SW, EN,
and WS movements. Meanwhile the maximum variation was found on the WN as much as 8 where the
longest queue produce by CUBE Dynasim on this movement about 14 vehicles while aaSIDRA produced
longest queue on the same movement about 22 vehicles. Nevertheless, it could be said that overall the
388
CUBE Dynasim queue data tended to be a bit higher than the aaSIDRA queue. This probably is because
of the driver aggressiveness has been counted when they still make a movement even the signal has turn
yellow in the visualization of animations movie.
From the tabulated queue data for both models, it was noted that most of the through lane
movement on each approach produce relatively longest queue of vehicles compared to the other type of
movement. As for examples from the CUBE Dynasim queue outputs, on south to north approach
produced the longest queue of 41 vehicles meanwhile on north to south approach with the longest queue
of 33 vehicles. This condition is related to the traffic flows generated at these approaches which carry
high traffic volumes as these approaches serve as the main route at the intersection. Figure 3 and 4 shows
the vehicles queue on the through lane in the layout form of simulation animations movie in CUBE
Dynasim for south and north respectively.
Figure 3: CUBE Dynasim Queue on South Approach
Figure 4: CUBE Dynasim Queue on North Approach

It same goes to the west approach where the longest queue observed was on the through lanes
(WE) with 37 vehicles compared to left turn (WN) and right turn (WS) with 14 and 10 vehicles on the
queue. This condition is normal for morning peak analysis since this approach provided access for traffic
heading to the CBD. In terms of queue length, it is assumed that the CUBE Dynasim used a shorter queue
space compared to aaSIDRA. As for example for the queue length from south to north approach where
aaSIDRA measure the longest queue about 256m for 35 vehicles while in CUBE Dynasim was 135m for
389
41 vehicles. This is probably due to the different queue space proposed by both models. It should be
noted that vehicle length use in CUBE Dynasim for passenger car was 4.1m. Therefore the queue space
derived in CUBE Dynasim between two consecutive passenger cars was 0.5m. On the other hand,
aaSIDRA used 7m for light vehicles where this could be considered much longer queue space than
proposed by CUBE Dynasim.
CONCLUSION
From the output comparison, it shows that the Cube Dynasim traffic flows output differed to the
aaSIDRA traffic flows by less than 5 percent for all the movement at the intersection. This is supported
by the yielded GEH statistic values which were less than 5 indicating a goodness of fit between Cube
Dynasim models to the aaSIDRA data. It same goes to travel speed outputs where Cube Dynasim
simulation model were generally simulated fairly well across all the movements at the intersection. It was
observed that there is a small variation in average travel speeds between Cube Dynasim models to the
aaSIDRA data such as less than 10 percent.
Finally, the comparison was done corresponding to queue outputs. It was found that both models
demonstrated the same pattern of queue for instance the longest queue was observed on SN approach. The
different of the output generated just as small as 1 vehicle which was noted on SW, EN, and WS
approaches while the maximum different was found on the WN approach as much as 8 where the CUBE
Dynasim queue was 14 and aaSIDRA queue was 22.
As for the conclusion of this study, it can be said that the output results demonstrated by CUBE
Dynasim micro simulation model with basic verification and calibration efforts simulates the signalized
intersection system in a manner that is reasonably consistent with aaSIDRA micro analytical model for
condition where aaSIDRA is valid in this study.
REFERENCES
Nigarnjanagool, S. & Dia, H. (2004) Evaluation Of A Dynamic Signal Optimisation Control Model Using
Traffic Simulation. Iatss Research, Vol.29.
Shaaban, K. S. & Radwan, A. E. (2004) Comparison of Sim Traffic and Vissim Microscopic Traffic
Simulation Tools in Signalized Intersection Modeling Department of Civil Engineering, University of
Central Florida.
Bell, G. (1998) Transportation Network: Recent Methodological Advances. 4th Euro Transportation
Meeting Pergamon The Netherlands.
Brian, B. P., Won, J. & Yun, I. (2006) Application of Microscopic Simulation Model Calibration and
Validation Procedure. Transportation Research Record: Journal of Transportation Research Board
Washington DC, pp 113-122
Akcelik & Associates Pty Ltd. (2002) aaSIDRA User Guide, Greyhorn Victoria Australia.
Gagnon, C., Sadek, A. W., Touchette, A. & Pe, M. S. (2008) Calibration Potential of Common Analytical
and Micro Simulation RoundaboutModels: A New England Case Study. Transportation Research Board
87th Annual Meeting Washington DC.
Taylor, M., B, P. W. & Young, W. (2000) Understanding Traffic System: Data Analysis and Presentation,
Aldershot: Ashgate 2000.
390
A THEORETICAL APPROACH IN EVALUATING THE POTENTIAL IMPACT OF WOMEN

TELECOMMUTERS ON TRAVEL DEMAND
Mootaz Munjid Mustafa 1 and Abdul Azeez Kadar Hamsa 2
ABSTRACT:
A number of Malaysian cities are experiencing severe transport problems resulting from rapidly
increasing vehicle ownership and usage, and the insufficient supply and inferior quality of public
transportation. As a result, supply of transport facilities is constantly lagging behind the level of transport
demand. The trend of congestion, frequent traffic accidents and air pollution is no where more prevalent
than in the Kuala Lumpur metropolitan area. Telecommuting is increasingly being perceived as a viable
travel demand management strategy. Due to a number of travel patterns and characteristics that are unique
to women commuters and the relatively large participation rate of women in the Malaysian work force,
telecommuting by women workers seems a more relevant and potentially more crucial means of
alleviating rush hour traffic in the Kuala Lumpur metropolitan area. The aim of this study is to investigate
both the actual and potential impacts of involving women workers in the practice of telecommuting on
travel demand and rush hour traffic in Kuala Lumpur. However, this paper highlights the literature
background on women workforce in Malaysia, vehicle growth and associated travel pattern, and elements
related to initial stage of this study.
Keywords: telecommuting, working women, travel demand, Kuala Lumpur, rush-hour traffic
1
Lecturer, Centre for Foundation Studies, IIUM, Jln. University, Section 17, Petaling Jaya, Selangor, MALAYSIA.
2
Assistant Professor, International Islamic University Malaysia, Jln. Gombak, 50728 Kuala Lumpur, MALAYSIA.
*Correspondence Author: Mootaz Munjid Mustafa, Centre for Foundation Studies, IIUM. Tel: (+603)-79462400
ext. 2768, Fax: (+603)-79571006. E-mail: mootaz_munjid@iium.edu.my
391
INTRODUCTION
Rapid motorization experienced in Malaysia is posing an enormous strain on the ability of the
existing infrastructure to cope with the increasing traffic volume. During the period from 1986 to 2002,
the number of vehicles in the country increased from 4.43 million to 11.97 million which has grown three
times. No other state in the Malaysia has to handle more traffic than the countrys federal capital of Kuala
Lumpur where almost a third of all registered motor cars in the country are found. This staggering statistic
coupled with the inherent difficulty in upgrading infrastructure at a similar pace, the inadequate public
transport service, and the local car manufacturing industry would without doubt result in traffic
congestion and an almost continuous reduction in the average travel speeds at major routes in the city
especially during rush hours.
High women participation in the workforce is reflected in a proportionately high presence of

vehicles on the road during rush hours. According to data supplied by the statistics department, ministry
of transport, a third (33.6%) of all motorcars and 29% of all vehicles in Malaysia are registered by
women.
The benefits of promoting a telecommuting culture among women workers are threefold: benefits
that are to be gained specifically by the women workers at the personal level, benefits in the form of an
extra reduction in traffic demand and benefits gained by the nation as a whole. Telecommuting is without
doubt more suited to women who when compared to men, experience more work/non-work role conflict
and more career interruptions (Schreiber, 1999). On the other hand, considering the disproportionately
high share of responsibilities shouldered by women, and their tendency to travel more than men when all
other variables are constant, it becomes clear that there is a considerable potential in promoting
telecommuting among women workers as a means of alleviating rush-hour traffic congestion, and that
potentially the possible net reduction in traffic demand as a result of involvement of woman workers in
the Malaysian labor force. Additionally, according to Human Resources Ministry, Malaysia has an
untapped latent women workforce of 1.2 to 1.6 million who are unable to work due to different
constraints.
It is the purpose of this study to verify the aforementioned proposition and explore actual benefits
derived from adopting a telecommuting practice by studying actual women telecommuters. Furthermore,
the study will attempt to investigate the perceptions of potential women telecommuters and consequently
use findings to forecast the impact of telecommuting women on commute travel based on a number of
hypothetical scenarios. However, this paper highlights the literature background on women workforce in
Malaysia, vehicle growth and associated travel pattern, and elements related to initial stage of this study.
LITERATURE REVIEW
An overview of the traffic conditions in the Kuala Lumpur Metropolitan Area
As of 2002, the federal territory of Kuala Lumpur had by far the largest number of motorcars
compared to all other Malaysian states. More than 21% of all registered vehicles and almost a third of all
motor cars (29.3%) in Malaysia are straining the infrastructure of Kuala Lumpur. The total number of cars
in Kuala Lumpur stood at 1.4 million, a figure that is more than twice that of the second and third highest
concentrations of motor cars in the country found in the states of Selangor (0.65 million cars) and Johor
(0.64 million cars) (refer to table 1).
During peak-hour periods, the average travel speed in the CBD of Kuala Lumpur is 28 km/h (refer
to table 2). While Kuala Lumpurs traffic flow conditions have not reached a critical level, the citys
uncontrolled motorization process, the rapidly increasing population as well the limited ability to increase
392
supply of physical resources such as roads and parking facilities especially in the CBD may help create a
traffic situation similar to that of the Bangkok metropolitan area where the average travel speed is a mere
13Km/h (Jamilah, 2007). Table 2 below also illustrates an almost consistent trend of reduced average travel
speed in major routes in Kuala Lumpur over the ten year period from 1986 to 1997.
Table 1: Number of motor vehicles by state as of 31/12/2002
Hire Goods
State Motorcycle Car Bus Taxi Car Veh. Others Total
Perlis 34,114 9,498 160 164 1 1,601 1,180 46,718
Kedah 420,666 154,873 2,845 3,067 272 29,471 15,864 627,058
P.Pinang 735,391 479,160 4,027 2,602 394 40,312 13,415 1,275,301
Perak 733,624 367,863 3,910 3,909 65 49,245 28,416 1,187,032
Selangor 668,789 653,661 4,621 3,828 177 94,764 34,713 1,460,553
Kuala Lumpur 774,850 1,458,030 11,588 19,874 7,269 157,449 110,862 2,539,922
N. Sembilan 294,994 168,887 2,311 1,746 9 30,487 6,559 504,993
Melaka 241,947 143,407 1,750 1,441 44 18,469 4,315 411,373
Johor 882,013 641,171 7,149 10,281 182 88,347 34,640 1,663,783
Pahang 261,839 174,337 1,783 2,483 7 30,953 10,254 481,656
Terengganu 152,545 85,898 912 969 18 16,250 5,562 262,154
Kelantan 222,463 125,470 1,655 1,979 12 20,859 6,742 379,180
Sabah 78,844 221,994 5,701 3,947 1,328 83,862 35,240 430,916
Sarawak 323,881 290,601 2,596 1,630 329 49,669 36,296 705,002
Total 5,825,960 4,974,850 51,008 57,920 10,107 711,738 344,058 11,975,641
Source: Road Transport Department, Federal Territory
Table 2: Average travel speed along selected routes in Kuala Lumpur

No. Route Average travel speed (km/h)
Inbound Outbound
1986 1997 1986 1997
1 Inner ring road* 26.7 15.0 21.2 17.3
2 Middle ring road* 23.6 17.1 33.4 13.6
3 Jln Ipoh, Jln Kuching 32.3 35.6 25.5 67.3
4 Jln Kepong, Jln Ipoh* 25.6 19.3 17.2 28.2
5 Karak Highway, Jln Sentul 23.7 26.4 37.0 45.0
6 Jln Gombak, Jln Pahang* 10.3 12.4 36.6 27.0
7 Jln Ampang* 19.5 10.4 24.9 19.5
8 Jln Cheras, Jln Pudu* 15.8 14.0 34.8 22.8
9 Seremban Highway* 38.8 20.3 68.7 60.6
10 Federal Highway II* 39.9 21.7 45.7 26.9
11 Jln Pantai* 24.3 13.4 32.7 19.6
Source: SMURT-KL study, 1998
* Route with reduced average travel speed
Examining the existing public transportation facilities in Kuala Lumpur further illustrates severity
of the problem. Kuala Lumpur has one of the lowest public transport ridership levels in Asia due in part to
the rising levels of private car ownership and use. The Kuala Lumpur Structure Plan 2020 revealed that
public transport accounted for only 20% of total Kuala Lumpur passenger movements compared to 80%
for private transport. The modal split for person trips at MRR1 using private transport
393
(car/taxi/motorcycle) was 54.3% in 1985, 64.1% in 1997 and 61.1% in 2005 while the share of public
transport (bus and, later, rail from 1997 onwards) was 37.4% in 1985, 25.1% in 1997 and 28.6% in 2005
respectively. Only 20% of Kuala Lumpur is within the transit corridor (i.e. areas within 400 m on each
side of a rail line). It is reported that in 2007, 1.305 million vehicles cross the Middle Ring Road (MRRI)
on a daily basis while 2.125 million vehicles cross the Middle Ring Road (MRRII) daily. About 70% of
vehicular trips crossing MRRI and MRRII during the morning peak hour are single-occupancy vehicles.
The present traffic volume entering the CPA boundary has already exceeded SMURT-KL forecast
(SMURT-KL: A Study on Integrated Urban Transportation Strategies for Environment Improvement in
KL, 1998) of 1,187,000 pcu/day in 2010.
The gender gap in travel behavior
The increasing presence of women in the workforce resulted in a change in the travel patterns
over the past few decades. More and more women in the labor force meant that the income of households
increased while the time available for performing household chores decreased, this phenomenon resulted
in women needing to buy essential services such as childcare and meals from outside and thus increasing
the number of non-work trips in the process (Levinson, 1997).
Two travel surveys conducted 20 years apart support this argument. The first was carried out in
1968 and the second in 1988 in the Washington D.C. metropolitan area. The surveys revealed that buying
such services alone resulted in an increase of 25% in trips per person. Such has been the impact on travel
brought about by the increased number of female workers.
Women in the workforce developed unique travel characteristics that substantially differed from
that of their male counterparts. A number of research papers dwell on the subject of the unique travel
patterns that men and women workers develop (Wachs 1997, Levinson 1997, Mokhtarian 1997 and
1998). It has been noted that men and women use transportation modes differently, and these differences
have persisted over time. The root cause of such differences lies in societys expectation that each gender
will fulfill distinct economic and social roles (Wachs, 1997).
Several studies (e.g. Taylor, 1997 and Prez, 2002; and Bielby and Bielby, 1988) have indicated
that working women still bear a disproportionate share of household responsibilities in addition to their
employment responsibilities that were traditionally shouldered by male employees. In fulfilling both their
employment and household responsibilities, women workers became known to make more linked trips
than men, mainly to drop off children in the morning, and shop during the afternoon (Levinson, 1997).
This dual role is obviously a source of considerable stress.
Many studies have shown that women made far more trips than men fulfilling the needs of
children, the elderly and other household members. In less affluent households, men are expected to use
the family car for the daily commute to work, while women employed or otherwise- relied on public
transportation (Wachs, 1997).
In light of the so called gender gap in travel behavior and the still prevailing trend of the
twofold responsibilities shouldered by working women, they (working women) are more likely than men
to view telecommuting as a potential solution to those extra pressures (Mokhtarian, 1997), and thus tend
to favor telecommuting because it allows them to fulfill household commitments while still remaining in
the labor force. However, studies in the United States have shown that while some women workers
adopted telecommuting for childcare and family reasons, a majority does not (Prez, Carnicer, and
Snchez, 2002).
394
In the local (Malaysian) context however, a more relevant study on the matter revealed that
taking care of the family was one of five most important factors why Malaysian working women are
favourable toward adopting telecommuting working arrangement (Karia and Asaari, 2006). Other studies
highlight even more benefits derived from telecommuting from a womans perspective. It has been
argued that the flexibility offered by telecommuting to working mothers, can potentially liberate women
from male control in the office and provides increased autonomy and creativity for them. For those
worried about the degeneration of family ties, teleworking is perceived as bringing families closer
together as well as bringing down the divorce rates (Ng, 1998).
In addition to the well established fact of women shouldering more household responsibilities
than men, an interesting study compared the travel patterns of women and men of single-adult households
and found that women made on average, 1.7 times as many child-serving trips as men. They also made
1.3 times more grocery shopping trips than men, a finding that indicates that the division of
responsibilities may not be the cause of the variation in travel patterns between men and women but
rather it is the gender that affects the way one travels.
Furthermore, a study found that women were more likely than men to adopt a particular strategy
in order to cope with congestion and long commute trips. The study revealed that working women
especially those with children were more likely to go for solutions such as working part-time, setting up a
home-based business (essentially telecommuting) or stopping work altogether. The study also showed
that women in general regardless of the presence of children and marital status were more likely to
change their work schedules by choosing arrangements like flextime and compressed work weeks.
Working women of households with more than one adult were much more prone to changes than single
women (Mokhtarian, 1997). Another more recent study found that on average women were 21% more
likely to make at least one change to their commute behavior as a response to traffic congestion
(Mokhtarian, Ye, Yun, 2009). According to the study, among the most common changes that were made
in order to reduce vehicle-miles traveled was increasing transit use and increasing telecommuting.
The above discussion illustrates that telecommuting for women workers has the potential of being
a welcomed relief; however some researchers bring forward the notion that women workers who practice
telecommuting will actually face the double burden of work, and end up working longer hours than if they
were to work at the office. Stanworth (2000) noted that at best teleworking offers new work opportunities
for women and at worst, it exploits and reinforces the secondary position of women in the labor market as
housewives.
However, this pessimistic view can be refuted based on empirical evidence gained through a
study of 12 women telecommuters. The study concluded that, among other things, telecommuting is a
work arrangement that does not hinder success in the job, and that it is in fact reduced role conflict and
consequently the likelihood of career interruptions (Schreiber, 1999).
Women in the Malaysian workforce
Malaysia, a country of 27.7 million inhabitants and a working age population of 17.6 million.
Almost half the working age population (49.1%) is comprised of women. However, according to official
figures obtained from the Malaysian department of statistics, women workers represent approximately
36% of all persons currently employed in the Malaysian labor force or 3.8 million employed women
(refer to table 3). This is a substantial increase of 27.5% from the 2.98 million women employed in the
year 1999. Womens participation in the workforce is slightly higher in urban areas where they comprise
approximately 38% of the labor force as compared to 32% in rural areas.
395
Table 3: Statistics of the Malaysian labor force, 2008
Total Male Female

Population (in millions) 27.7 14.1 13.6
Working age population (000) 17,603.8 8,955.7 8,648.1
Labor force (000) 11,028.1 7,074.6 3,953.5
Employed persons (000) 10,659.6 6,851.1 3808.5
Number of unemployed (000) 368.5 223.5 145.0
Outside labor force (000) 6,575.7 1,881.1 4,694.6
Labor force participation rates (%) 62.6 79.0 45.7
Unemployment rates (%) 3.3 3.2 3.7
Source: Labor force statistics, Malaysia, 2008, department of statistics
Classifying the labor force based on gender and level of education attained reveals that women
make up about half of the employed persons under the categories of no formal education (45%) and
more importantly- tertiary level education (46%). Men on the other hand are far more represented in the
categories of primary and secondary level education. Women represented 49.8% and 46% by diploma and
degree holders respectively. These figures are noteworthy since the majority of potential telecommuters
fall under these categories. Women also had a high employed participation rate in STPM holders category
(48%) (refer to table 4).
Table 4: No. of employed persons by educational attainment and gender, Malaysia, 2008
Educational attainment Total (000) Male (000) Female (000)

Total 10,659.6 6,851.1 3,808.5
No formal education 474.7 261.4 213.3
Primary level education 1,978.4 1,378.6 599.9
Secondary level education 5,956.1 3,989.8 1,966.3
Tertiary level education 2,250.3 1,221.3 1,029.0
Breaking down the labor force along the lines of gender and marital status revealed that
women made up 37% of the employed persons under the never married category, and slightly less
(34%) under the married category. This difference, small as it may be, hints at the inability of some
women to remain in the workforce once the new household commitments entailing marriage are
introduced.
The prevalence survey (a high profile study aimed at identifying the level of penetration of the
practice of telecommuting in the country) revealed that the employment sectors with the highest rates of
telecommuters were the finance, insurance, real estate and other business services sectors. The level of
participation of women in the aforementioned employment sectors was one of the highest among the
other employment sectors. Women workers comprised 51.4% of the employed persons under the financial
sector, and 38.1% of those employed under the real estate and other business activities sector. These
figures translate into an average of 42.5% participation rate of women in the two sectors combined (refer
to table 5). The relatively high participation rates of women in the aforementioned employment
subsectors, along with the findings of the telecommuting prevalence survey will be of great importance to
this study, as they will provide the basis for determining the population and consequently the sample
used.
396
Other sectors with exceptionally high rates of participation by women were the sectors of
education and health and social work in which women comprised 66% and 70% of the workforce
respectively.
Table 5: No. of employed persons by industry and gender, Malaysia, 2008
Employment sectors/subsectors Total (000) Male (000) Female (000)

Total 10,659.6 6,851.1 3,808.5
Agriculture, hunting and forestry 1,365.6 1,027.5 338.2
Fishing 122.1 117.4 4.7
Mining and quarrying 54.5 44.6 9.8
Manufacturing 1,944.7 1,182.6 762.2
Electricity, gas and water supply 60.5 50.1 10.3
Construction 998.0 914.7 83.4
Wholesale/retail trade; repair of motor vehicles, 1,729.4 1,119.5 609.9
motorcycles and personal and household goods
Hotels and restaurants 783.6 403.4 380.2
Transport, storage and communications 583.4 487.0 96.3
Financial intermediation 276.0 134.2 141.8
Real estate, renting and business activities 553.2 342.7 210.6
Public administration and defense; social security 751.1 536.4 214.7
Education 656.5 226.1 430.4
Health and social work 252.6 76.1 176.5
Community, social and personal service activities 274.2 158.4 115.8
Private households with employed persons 253.0 29.4 223.6
Extra-territorial organizations and bodies 1.1 0.9 0.2
STUDY OBJECTIVES AND METHODOLOGY
The ongoing study revolves around achieving the following objectives:
1. To investigate the involvement rate of Malaysian working women in the different employment
sectors, as well as their respective involvement in the public/private sectors.
2. To identify the socioeconomic profile of working women and their contribution to household
income in the selected study area.
3. To investigate the current prevalence level of the telecommuting practice among women
workers.
4. To investigate the perceptions towards telecommuting by potential women telecommuters.
5. To identify travel characteristics associated with working women in the study area.
6. To investigate the actual impact of practicing women telecommuters on the reduction of travel
demand.
7. To forecast the impact of telecommuting women on commute travel in Kuala Lumpur.
The first objective of the study would be achieved by relying on secondary source census data
pertaining to the presence of Malaysian women in the public and private employment sectors and the
various subsectors of the Malaysian economy. This will provide an insight into the employment sectors
with the highest participation rates of women, and whether the job scope in such sectors is suited to the
work-from-home arrangement.
397
The aforementioned secondary source data will then be supplemented by primary source data
obtained by an initial survey that aims to verify the level of participation of women in the workforce
employed in Kuala Lumpur. A representative number of private as well as public employers belonging to
the different employment subsectors will be sampled for the purpose of identifying the rate of
participation of women workers.
The same survey will be used to achieve the second and third objectives since it will be used by
the researcher to gather data on womens socioeconomic profile, their contribution to the household
income, and the current prevalence of the telecommuting practice among them as well as the perceptions
of women workers towards telecommuting (refer to table 6).
Data on the perceptions of women workers towards telecommuting will provide along with data
collected from actual telecommuters- the basis necessary for forecasting the potential reduction in travel
demand based on different scenarios.
Achieving the fifth and sixth objectives of the study will involve distributing a two-week travel
dairy to a group of telecommuters of both genders, and another group of non-telecommuters. The sampled
commuters will participate in the study by filling up a travel diary. The travel diary will cover among
other variables, the number and distance of trips travelled by the respondents, the purpose/destination of
trips, the mode of transport used, the type of road used, whether or not the journey was linked (refer to
table 7).
Information obtained by the use of the travel diary will facilitate carrying out a comparison
between the different travel patterns of each gender, and in doing so identifying whether or not there are
unique travel characteristics associated with women as suggested in the literature. In addition, involving
actual women telecommuters will provide an insight into the actual reduction in travel as a result of
adopting a telecommuting practice and is expected to provide a case for promoting telecommuting among
working women as an effective strategy to curb the increasing traffic congestion.
A convenient starting point for this part of the study (involving actual telecommuters) would be
to approach the participants of the Works Ministrys work-from-home programme which is a pioneering
three-month (later extended to nine months) telecommuting initiative that commenced on the 1st of
January 2010.
Achieving the aforementioned objective will contribute directly to achieving the final objective of
the study, which involves forecasting the impact of telecommuting women on commute travel in Kuala
Lumpur under different hypothetical scenarios.
Table 6: Data obtained through the use of the questionnaire survey
Questionnaire survey distributed to sampled private firms & government agencies

Participation rates of women in the surveyed private firm/government agency
Data obtained
Socioeconomic status of sampled women workers & their contribution to household income
Ability to perform duties while telecommuting
Perceptions of women workers towards telecommuting
Prevalence of telecommuting practice among sampled employees
Willingness to participate in the travel dairy exercise
398
Table 7: Data obtained using the travel dairy
Travel dairy distributed to a sample of telecommuters and non-telecommuters of both genders in

order to establish a basis for comparison of travel characteristics
Purpose/destination of trip
Distance travelled
Data obtained
Time trip started

Time destination reached
Type of trip (linked or exclusive)
Mode of transport used for the trip
Type of road used for the trip
CONCLUSIONS
This research derives its basis from the viability of telecommuting as a travel demand
management tool. Telecommuting has been recognized and promoted as one of a number of strategies for
alleviating traffic congestion in urban areas, and a means of reducing rush-hour traffic. Sufficient
evidence in the literature suggests that a more relevant sector or portion of the working population to be
practicing telecommuting in order to reap higher benefits would be women workers.
This is mainly due to womens unique travel behavior and the dual (home and office)
responsibilities they shoulder. Telecommuting for women workers represents more of a convenience or a
welcomed relief, and in addition, will more likely result in a larger reduction in commute travel as
suggested by the available literature. It is this twofold advantage offered to working women by
telecommuting that generated interest in the subject.
This paper attempts to present an overview of the literature pertaining to gender-based differences
in travel behavior and how such differences that are manifested in the unique travel characteristics of
women commuters can increase the efficiency of telecommuting as a travel demand management tool.
Despite the fact that the study is at an initial stage, it is expected to contribute to the understanding and
potential of telecommuting in alleviating if not solving rush hour traffic congestion in the study area.
REFERENCES
Bielby, Denise D., and Bielby, William T. (1988). She works hard for the money: Household
responsibilities and the allocation of work effort. American Journal of Sociology, 93(5), p. 1031-
1059.
Jamilah Mohamad, and T. Amin (2007) The rise of the private car in Kuala Lumpur, Malaysia, assessing
the policy options. IATSS Research, Vol. 31, No. 1, 2007.
Karia, Norliza, and Asaari, Muhammad Hasmi Abu Hassan (2006). Telework and women: Case of
Women Workers in the Northern States of Malaysia. International Academy of Business and
Technology Conference.
Levinson, David (1997). How travel patterns changed from 68 to 88. Review (Journal), Institute of
Transportation Studies, University of California. Vol. 20. No. 2.
Ministry of Works, Highway Planning Unit (2002). Road Traffic Volume, CD-ROM Version.
399
Mokhtarian P. L., Ye Liang, Yun Meiping (2009). The Effects of Gender on Commuter Behavior Changes
in the Context of a Major Freeway Reconstruction. Paper presented at the 4th International
Conference on Research on Womens Issues in Transportation, October 2009.
Mokhtarian P. L., Michael N. Bagley, Lisa Hulse, and Ilan Salomon. The influence of gender and
occupation on individual perceptions of individual perceptions of telecommuting. Transportation
Centre, the University of California, Berkeley, CA 94720.
Mokhtarian P. L., Bagley M. N. and Salomon I. (1998). The impact of gender, occupation and presence of
children on telecommuting motivations and constraints. Journal of the American Society of
Information Science, Vol. 49, no. 12, pp 1115-1134.
Ng, Cecilia (1998). Teleworking and gender in the information age: new opportunities for Malaysian
women? Paper presented at the regional conference on gender and technology in Asia, Bangkok,
Thailand.
Prez Prez, Manuela; Carnicer, Mu Pilar de Luis; and Snchez, Angel Martnez (2002). Differential
effects of gender on perceptions of teleworking by human resources managers. Women in
Management Review Journal, Vol. 17, no. 6, pp 262-275.
Schreiber, Pamela Jane (1999). The Nature of Career Development for Women Who Telecommute.
(Unpublished Doctoral thesis). The University of Georgia.
Stanworth, Celia (2000). Women and Work in the Information Age. Gender, Work and Organization,
Vol. 7, no. 1, January, 2000.
Taylor, Brian (1997). Beyond the Gender Gap, an array of social, ethnic and economic factors influence
travel behavior. Review (Journal) Institute of Transportation Studies, University of California.
Vol. 20, no. 2.
Wachs, Martin (1997). The gender gap, how men and women develop different travel patterns. Review
(Journal) Institute of Transportation Studies, University of California. Vol. 20. No. 2.
Wan Nurul Mardiah Bt. Wan Mohd. Rani (2005). A Perceptual Assessment Towards Telecommuting And
Its Implications On Commute Travel. (Unpublished M.A. thesis) Kulliyyah of Architecture and
Environmental Design, International Islamic University, Malaysia.
400
PALM OIL CLINKERS IN VARIOUS COMPOSITIONS AS COMPOSITE

MATERIALS IN MARSHAL MIX DESIGN
Shuhairy Norhisham 1*, N. Ganesh Rao 2, Herda Yati Katman1, Siti Aliyyah Masjuki4, Fathoni Usman 3,
Megat Asyraff Megat Ibnu Rashad4 and Ean Lee Woen4
ABSTRACT:
The Economic growth has lead to the rapid increase of the demands for the aggregate for use in civil
engineering field. The roads are the main links for all type of transportation to travel from one to another
place. The growth of economy in a country also depends to the system of roads of the country itself.
Besides, the decreasing of natural aggregate in our country is turning out to be a problem as its important
for future generation. Since the road construction in Malaysia really developing rapidly and costs a lot in
a project, an alternatives method or material need to been identified to minimize the cost and gives a good
quality of pavement for long term usage. Malaysia also produce a large volume of palm oil waste and the
usage of palm oil clinker in hot mix asphalt will make benefits for the environment and solid waste
management. Malaysia is known as main manufacturers of palm oil in Asian region. Every oil mill
produces a waste by-product, known as clinker. Clinker was found to be suitable to replace normal gravel
aggregate in concrete mixture and also mix design for pavement. For this project, palm oil clinkers will be
mix together with the natural aggregate in Hot Mix Asphalt with different composition. The proportion
for each category varies as Sample A contains 90% Natural Aggregate and 10% Palm Oil Clinker,
Sample B contains 30% Natural Aggregate and 70% Palm Oil Clinker, Sample C contains 50% Natural
Aggregate and 50% Palm Oil Clinker, Sample D contains 70% Natural Aggregate and 30% Palm Oil
Clinker and Sample E contains 10% Natural Aggregate and 90% Palm Oil Clinker. From the experiment,
the values of Percent Wear and LA Abrasion for natural aggregate are 51.70% and 48.16%. This value
will be compared with the JKR Standard which specifies the value for percent wear should be less than
60% and for Los Abrasion Value less than 20%. For the bitumen Penetration test, the average value for
the total sample is 79.6mm and referring to the penetration reading for grade of Asphalt Cement, the
result indicated that the bitumen sample for this experiment fall in grade 60-80 which suitable according
to the JKR requirement which should be in grade 60-80. The aggregate impact Value for natural
aggregate is 91.06% and for palm oil clinker is 73.54%. The aggregate impact value obtained from the
experiment is higher than required by JKR which is below 15%.
Keywords: Hot Mix Asphalt, Palm Oil Clinker, Composite Materials, Materials Characteristic
1
2
Student, Universiti Tenaga Nasional, KM 7, Jalan Kajang-Puchong, 43009 Kajang, Selangor, MALAYSIA
3
Senior Lecturer, Universiti Tenaga Nasional, KM 7, Jalan Kajang-Puchong, 43009 Kajang, Selangor, MALAYSIA
4
Tutor, Universiti Tenaga Nasional, KM 7, Jalan Kajang-Puchong, 43009 Kajang, Selangor, MALAYSIA
*Correspondence Author: Shuhairy Norhisham, Universiti Tenaga Nasional, Malaysia. Tel: +6003 8921 2020
ext. 6331, Fax: +60389212116. E-mail: shuhairy@uniten.edu.my
401
INTRODUCTION
This project is mainly focused on the Palm Oil Clinker (POC) and Natural Aggregate in different
composition in Hot Mix Asphalt (HMA). The POC is mixed together with the natural aggregate to
evaluate its performance in HMA. Asphalt concrete pavement or hot mix asphalt (HMA) pavement as it is
more commonly called, refers to the bound layers of a flexible pavement structure.
Natural aggregates also had been decreasing in Malaysia due to the deforestation and land use for
construction (Oglesby, 1982). he aggregates produce naturally after the rocks undergo the physical and
chemical process. An alternate material for natural aggregate needed to be obtained for solving up these
upcoming problems (Neville A.M, 1995). There are some research been undergo to produce recycled
aggregate. The applications of recycled aggregate in construction have started since end of World War II
by demolished concrete pavement as recycled aggregate in stabilizing the base course for road
construction (Olorusongo, et.al., 1999).
LITERATURE REVIEW
Hot Mix Asphalt is generally composed of the aggregate bound together into a solid mass by
asphalt cement (Asphalt Institute, 2001). The aggregates total 93% to 97% by weight of the total mixture
are mixed with 3% to 7% asphalt cement (Asphalt Institute, 2001). Mixing normally carried out at 140c
based on Marshall Test procedure. HMA must adequate in providing surfacing for heavily trafficked
roads in hot climate as well (Garber et.al. 2002). Mechanical spreader will be used to complete the paving
mixtures in smooth layers while it is still in hot conditions. HMA will be placed directly on the sub grade
to construct one or more courses pavements (Garber et.al. 2002). Some advantages of using asphalt
mixtures in pavements construction:
1. Economical - HMA is economical to construct rapidly and faster, easy to be recycling, low
maintenance and good performance.
2. HMA will not be affected by the ice control chemicals.
3. Minimize traffic noises.
4. Can be design to handle versatile traffic loading, soil texture and materials.
The aggregates that used in mixtures should accordingly to its specifications for durability,
soundness, hardness and others. The form of the aggregates must be crushed and contains some allowable
percentage of sand composition (Freedy L.Roberts. et. al., 1996). The aggregates used should meet these
five requirements such as it must be strong and tough, it must be durable, easy to crush without many
flaky particles, low in porosity, low in permeability and the particle size and the gradation is suitable for
the pavement type (Yang H.Huang, 2004).
Malaysia for many decades has been known as the main manufacturers of palm oil. The waste
that been produced in palm oil mill known as clinker (Omar .W. et.al. 2001). As the aggregates were
decreasing rapidly, clinker was explored as a suitable material to replace aggregate in Hot Mix Asphalt.
Palm oil clinker can be found easily in our country as Malaysia is the largest manufacturer of palm oil
products. Malaysia is holding the main production for world palm oil production as 51%, 62% for the
world exports and 30% for the oils and fats exportation (Asean Sources.Com, 2004).
402
Combustions of shell and fiber from the palm oil create clinkers as waste material which will be
thrown away from mill [Omar .W. et.al. 2001]. The significant usage of clinker as the artificial
lightweight aggregate in HMA in road pavement will benefit us from the aspect of waste management
from palm oil mill and minimize the demand on natural aggregate in the road construction industry [Omar
.W, et.al., 2001]. Usage of clinkers also will minimize the cost of the road construction due to the rate of
palm oil clinker is cheaper than the natural aggregate [Neville A.M, 1995].
Sterilization Stripping Extraction
Collection Burning Separation
Figure 1.0: Flow Chart for the producing of palm oil clinker
(Source: Omar. W. et.al. 2001)
Gradation (or grain-size analysis) is the one of most common test performed on aggregates to
determine the main property such as stability, stiffness, permeability, durability, workability, fatigue
resistance and frictional resistance (Freedy L.Roberts. et. al., 1996). According to Freedy L.Roberts
(1996), this gradation is a most concern matter in HMA mix design. This gradation usually measured by
sieve analysis. Sieve analysis is carried out by using the sample of dry aggregate of known weight is
separated through a series of sieves with small and tiny openings (Freedy L.Roberts. et. al., 1996). Then
the aggregates separated and mass retained on each sieves measured and compared to the total sample
weight (Freedy L.Roberts. et. al., 1996).
According to Atkins 2003, Surface texture refers to the surface patterns and the smoothness of the
aggregate. Particle shape and surface plays a major role in creating bond between aggregate and material
for cementing. For example, rough surface texture provides something to grip, create a good bond and
produced stronger hot mix asphalt concrete. It also creates higher friction strength if there is a force to
slide over a particle over a particle (Atkins, 2003). Some aggregates with good surface texture but the
smoothness of texture might decay under the heavy traffic loading (Atkins, 2003).
The significant properties of aggregates are the internal pore characteristics. The strength of the
aggregate, surface texture, abrasion resistance, specific gravity, bonding capabilities and resistance to
freezing depends on the pores of an aggregate (Freedy L.Roberts. et. al., 1996). Absorption mainly refers
to the particle's ability to absorb the liquid into it. Porosity is a ratio of the volume of the pores to the total
volume of the particle. Moreover, permeability refers to the particle's ability to allow liquids excess
through it (Yang H.Huang, 2004).
Shapes of the aggregate particles are the main key for other elements such as the asphalt demand
for hot mix asphalt, the workability and the strength of asphalt pavements (Chung Chan Kee, 2007).
Crushed gravel or crushed stone normally used for this asphalt pavements. These crushed aggregates have
various section of shape for instant irregular and angular particles that will tend to interlock when
compacted or consolidated (Yang H.Huang, 2004). Mixes that contain both round and angular particles
tend to result in high workability. Normally, coarse aggregate particles are made of crushed stone or
crushed gravel.
403
According to Freedy L.Roberts (1996), density is measuring of the weight per unit of volume of a
substance. In another hand, Specific gravity is the ratio between the densities of the substance to the water
(Freedy L.Roberts. et. al., 1996). Specific gravity will helps us in determining the amount of asphalt
needed in the hot mix asphalt. Highly absorptive aggregate will absorb asphalt at the starting of the
mixture until the mix cools down. Then the aggregate will undergo the bonding process whereby the
porous aggregate needs more asphalt compared to less porous asphalt (Yang H.Huang, 2004).
Marshall Mix is the most common method widely used producing trial mixes in obtaining the
optimum asphalt content (Garber et.al. 2002). Marshall Method used to check the adequate of the voids in
mixture in HMA design. 50 blows of the Marshall Hammer per side were used to prepare the specimens
[Freedy L.Roberts. et. al., 1996]. 75 compaction blows were not advisable in Marshall Mix design since it
would break down the aggregate into small particles. According to Yang H.Huang (2004), the Marshall
method follows the procedure of heating, mixing and compact the mixture. Then, the sample will be
subjected to a stability-flow test and a density-voids analysis.
The Marshall method conducted using several trial aggregate-asphalt binder blends which is 5
blends with 3 samples for each blend [Freedy L.Roberts. et. al., 1996]. The total samples are 15
specimens for 1 test with asphalt binder content varies from each test. Then, from this trial blends
performance, optimum asphalt binder content can be determined [Freedy L.Roberts. et. al., 1996]. The
trial blends must contain a range of asphalt contents which includes both above and below the optimum
asphalt content. Estimation of optimum asphalt content is necessary in this stage. Trial blend asphalt
contents are then determined from this estimation (Yang H.Huang, 2004).
METHODOLOGY
Marshall Method of Mix Design for HMA (ASTM D1559)
The design of an asphalt mixture including the selection of the best blend of aggregates and the
optimum asphalt content to provide a material that meets the adequate specification and also economic as
well (Ratnasamy Muniandy, 2001). The test specimens of 4 in. (101.6mm) diameter and 2.5 in. (60-
65mm) height are used in this Marshal Method. According to Ratnasamy Muniandy (2001), the
requirement for this method such as the aggregates blended together must meet the gradation limit of JKR
standard specification, Material used for this design also must meet the JKR standard specification, For
obtaining the void analysis and density, the bulk specific gravity of the aggregates used in blend and the
specific gravity of asphalt cement are determined (Ratnasamy Muniandy, 2001).
The weight for each sample is approximately 1200g and the total samples for this project are 75
samples. There are 5 different proportion of mixing of Palm Oil Clinker and Natural Aggregates with the
asphalt content ranges from 5.0% to 70%. The times estimated to complete this Marshal Samples are
around 3 weeks (One week approximately 25 samples). The size for each sample is around 101.6mm
diameter and the height is 76.2mm (Ratnasamy Muniandy, 2001).
The proportion for each category as per below:-
1) Sample A - Mix proportion for 90% Natural Aggregate and 10% Palm Oil Clinker
2) Sample B - Mix proportion for 30% Natural Aggregate and 70% Palm Oil Clinker
3) Sample C - Mix proportion for 50% Natural Aggregate and 50% Palm Oil Clinker
4) Sample D - Mix proportion for 70% Natural Aggregate and 30% Palm Oil Clinker
5) Sample E - Mix proportion for 10% Natural Aggregate and 90% Palm Oil Clinker
404
Marshall Method uses several trial aggregate asphalt binder blends (5 blends with each sample for a
total 15 specimens), each with different asphalt binder content (Ratnasamy Muniandy, 2001). Then the
optimum asphalt binder content can be determined from there. The trial blends must contain a range of
asphalt content both above and below the optimum asphalt content (Ratnasamy Muniandy, 2001).
RESULT ANALYSIS
Los Angeles Abrasion Test Result (ASTM C131)
Table 1: Results of Los Angeles Abrasion test (ASTM C131)
Sample 1 Sample 2 Sample 3 Sample 4 Sample 5 Average

value
Percent Wear
For Natural 50.50 50.07 50.62 53.83 53.52 51.70
Aggregate (%)
Percent Wear
For Palm Oil 63.48 64.19 63.05 73.09 65.38 65.83
Clinker (%)
LA Abrasion
Value For 49.50 49.28 49.38 46.17 46.48 48.16
Natural
Aggregate (%)
LA Abrasion
Value For Palm 36.52 35.81 36.94 35.47 34.62 35.87
Oil Clinker (%)
JKR Requirement For Percent Wear < 60%
JKR Requirement For Los Abrasion Value < 20%
According to the aggregate physical testing required by JKR Malaysia, the requirement value for percent
wear should be less than 60% and for Los Abrasion Value less than 20%. From the experiment, the values
of Percent Wear and LA Abrasion for natural aggregate are 51.70% and 48.16%. Besides, for Palm oil
clinker the values of Percent Wear and LA Abrasion are 65.83% and 41.55%. Thus, both natural
aggregate and palm oil clinker not suitable for road construction due to dissatisfying the JKR requirement.
Aggregate Impact Value Test Result (BS 812: PART 111)
Table 2: Results of Aggregate Impact Value Test (BS 812: Part III)
Test 1 Test 2 Average Value

Aggregate Impact
Value for Natural 90.91% 91.21% 91.06%
Aggregate
Aggregate Impact
Value for Palm Oil 73.07% 74.00% 73.54%
Clinker
JKR Requirement For Aggregate Impact Value(AIV) < 15%
405
The aggregate impact Value for natural aggregate is 91.06% and for palm oil clinker is 73.54%. The AIV
value of natural aggregate shows that the toughness of natural aggregate is more than the palm oil clinker.
The AIV value obtained from the experiment is higher than required by JKR which is below 15%. Thus,
this natural aggregate and palm oil clinker not suitable for road construction.
Bitumen Penetration Test Result (ASTM D5)
Table 3: Results of Bitumen Penetration Test (ASTM D5)
PENETRATION (mm)
SAMPLE
1 2 3 Average
1 70.5 80.5 75.0 75.3
2 70.5 75.0 73.5 73.0
3 85.0 80.0 82.5 82.5
4 86.0 85.0 80.0 83.7
5 75.0 85.5 90.0 83.5
Total Average 79.6
JKR Requirement For Bitumen Penetration Test 60-80 & 80-100
According to the JKR standard, there are only two type of penetration grades are allowed to be used for
pavement construction which are the grade 60-80 and 80-100. From the experiment done in the
laboratory, the average penetration value for the 1st sample is 75.3mm, 2nd sample is 73.0mm, 3rd sample
is 82.5mm, 4th sample is 83.7mm and final sample is 83.5mm. The average value for the total sample is
79.6mm and referring to the penetration reading for grade of Asphalt Cement, the result indicated that the
bitumen sample for this experiment fall in grade 60-80 which suitable according to the JKR requirement
and for the HMA specimen preparation.
Ring and Ball Test Result (ASTM D36)

Table 4: Results of Ring and Ball test (ASTM D36)
Test 1 Test 2 Test 3 Test 4 Test 5

1st Failure 48.5 48.0 49.0 49.0 49.5
(C)
2nd Failure 49.0 48.5 50.0 49.5 50.0
(C)
Average 48.75 48.25 49.5 49.25 49.75
Value
Total Average 49.1C
JKR Requirement For Ring and Ball test Not less than 48C & Not more than 56C
406
Softening Point of bitumen or commonly known as Ring and Ball test is used to measure the
susceptibility of blown asphalt to the temperature change by determining the temperature at which the
material will be adequately softened to allow a standard ball to sink through it. According to the JKR
standard, the requirement is not less than 48C and not more than 56C. The result that obtained from
experiment shows 49.1C which achieve the JKR standard.
Marshall Test Result (ASTM D1559)

Table 5: Results of Marshall Test (ASTM D1559)
Composition Optimum Stability Flow Bulk Voids In Voids In Voids

(Natural Binder (kN) (mm) Density Total Mix Mineral Filled
3
Aggregate Content (g/cm ) (VTM) Aggregate With
Palm Oil (%) (%) (VMA) Binders
Clinker) (%) (VFB) (%)
JKR 3.0% -
75%-85%
Requirement - >500kg >2.0mm - 5.0% -
90% N.A-
5.83 8.78 1.90 1.79 28.20 36.15 22.50
10% POC
70% N.A
5.83 10.13 2.20 1.89 24.60 32.90 25.00
30% POC
50% N.A
6.00 7.97 2.20 1.95 22.50 31.00 28.00
50% POC
30% N.A
6.20 8.62 1.50 2.12 15.30 24.90 37.50
70% POC
10% N.A
6.20 8.63 1.80 1.80 27.60 36.10 23.80
90% POC
From the experiment done, the value for the optimum binder content (OBC) determined from the
average value at maximum stability, maximum density and midpoint of specific air void range (4%
typically). The value for the optimum binder content (OBC) for all the samples as summarize below:
1) HMA sample (Composition: 90% N.A 10% POC) : 5.83%

From these optimum binders content (OBC), the properties such as stability, flow, Voids in total
mix (VTM) and Voids filled with binders (VFB) of each samples determined. JKR Standard Criteria will
be used to compare the properties of the samples. For HMA sample with the composition of 90% N.A
407
10% POC, the stability value which is 8.78kN, flow -1.90mm, Bulk density- 1.79 g/cm3, VTM-28.20%,
VMA-36.15% and VFB 22.50%. The stability satisfies the JKR requirement but the other 3 properties do
not satisfy the JKR requirement. For HMA sample with the composition of 70% N.A 30% POC, the
stability value which is 10.13kN, flow -2.20mm, Bulk density- 1.89 g/cm3, VTM-24.60%, VMA-32.90%
and VFB 25.00%. The stability and flow satisfies the JKR requirement but the other 2 properties do not
satisfy the JKR requirement. For HMA sample with the composition of 50% N.A 50% POC, the
and VFB 28.00%. The stability and flow satisfies the JKR requirement but the other 2 properties
dissatisfy the JKR requirement. For HMA sample with the composition of 30% N.A 70% POC, the
and VFB 37.50%. Only the stability satisfies the JKR requirement but the other 3 properties do not satisfy
the JKR requirement. For HMA sample with the composition of 10% N.A 90% POC, 36.10% and VFB
23.80%. The stability satisfies the JKR requirement but the other 3 properties do not satisfy the JKR
requirement.
CONCLUSION
In conclusion, none one of the samples achieve the JKR Standard Criteria for the Marshall Test
(ASTM D1559). The flow values are not in the JKR range due to the high voids in the total mix (VTM)
rather than specify by the JKR as 3.0% -5.0%. The voids filled with binder value also not in the range by
JKR standard which is 75%-85%. This is due to the low ability of palm oil clinkers absorption. Besides,
the irregular surfaces texture and particle shapes of aggregate and palm oil clinker leads to this dissatisfy
results.
Thus, higher percentage of bitumen or asphalt content needed to cover the palm oil clinker in
order to determine the ideal percentage of mixing composition of palm oil clinker and natural aggregate in
Hot Mix Asphalt.
ACKNOWLEDGEMENT
I would like to thank everyone who involve in this project and make it successful especially to
UNITEN for awarding internal research grant of RM10,000.00.
REFERENCES
Asean Sources.Com (2004), Malaysia Oil Palm Industries, Manufacturers, Producer and Suppliers,
retrieved on 19 Jun 2010, from website http://www. aseansources. com/jsp/ malaysia_oil_palm.jsp
408
ASTM, 1994, ASTM C131-06, Standard Test Method for Resistance to Degradation of Small-Size
Coarse Aggregate by Abrasion and Impact in the Los Angeles Machine, Volume 04.03, American
Society for Testing and Materials, Philadelphia, pp. 202-210.
ASTM, 1994, ASTM D5, Standard Test Method for Penetration of Bituminous Materials, Annual Book
of ASTM Standards, Volume 04.03, American Society for Testing and Materials, Philadelphia, pp.
250-263.
ASTM, 1994, ASTM D36, Standard Test Method for Softening Point of Bitumen (Ring-and-Ball
Apparatus), Annual Book of ASTM Standards, Volume 04.03, American Society for Testing and
Materials, Philadelphia, pp. 402-408.
ASTM (1994), ASTM D1559, Standard Test Method For Resistance To Plastic Flow Of Bituminous
Mixtures Using Marshall Apparatus. 1994 Annual Book of ASTM Standards, Volume 04.03,
American Society for Testing and Materials, Philadelphia, pp. 1103-1187.
Atkins, 2003, Highway Materials, Soils, and Concretes, 3rd ed. Prentice-Hall Inc, pg.548-1138.
Chung Chan Kee, 2007, A Preliminary Study On The Performance Of Artificial Lightweight Aggregate
(POC) As Fine Aggregate In Hot Mix Asphalt, Thesis For Bachelor Of Civil Engineering,Civil
Engineering Department, College Of Engineering, University Tenaga Nasional, Malaysia.
Freddy L. Roberts, Prithvi S. Kandhal, E. Ray Brown, Dah-Yinn Lee, Thomas W. Kennedy,1996, Hot
Mix Asphalt Materials, Mixture Design and Construction, 2nd Edition ,NAPA Education Foundation
Lanham,pg.15-444.
Garber and Hoel , 2002, Traffic & Highway Engineering , Brooks/Cole, 2nd Edition, Pg. 202-355.
Neville A.M, 1995, Thickness of Concrete Pavements with discussion by E.C. Sutherland and
W.Harwood, Transaction, ASCE, Vol.121, pp.1125-1152.
Oglesby,P.K, K.C Mahboub and R.M. Anderson,1982, Flat And Elongated Aggregates In Superpave
Regime, ASCE Journals Of Materials Engineering, Vol.12, No.2, pp.124-130.
Olorusongo F.T and N.Padayachee,1999, Performance of recycled aggregate concrete monitored by

durability indexes, Civil Engineering Discipline, School of Engineering, University of Durban-
Westville, South Africa, Vol.12, No.2, pp.220-236.
Omar.W and Roslli Noor Mohamed, 2001, The Performance Of Pretensioned Prestressed Concrete
Beams Made With Lightweight Concrete, Department of Structure & Material Faculty of Civil
Engineering, Universiti Teknologi Malaysia, volume 14, pg.1-9.
Ratnasamy Muniandy, Radin Umar Radin Sohadi,2001, Highway Materials, A Guide Book For
Beginners, Serdang, Penerbit Universiti Putra Malaysia, Volume 1,pp.1-73.
Yang H. Huang, 2004, Pavement Analysis and Design, Pearson Prentice Hall, Upper Saddle River, 2nd
Edition, pg.1-692.
409
410
EVALUATION OF THE LABORATORY COMPACTING EFFORT ON STONE MASTIC

ASPHALT MIXTURES
1
Pourtahmasb, M.S , Karim, M.R
ABSTRACT
Recent laboratory studies have shown that, the compaction can highly affect the performance of the
Stone Mastic Asphalt (SMA) mixtures. The goal of compaction is to achieve the optimum air void
content and compressing the coated stones together by increasing the density of the mix to the
considered level of compaction with a minimum change in the gradation and structure. Inappropriate
compaction may draw the binder to the surface of SMA causing flushing of the surface and loss of
texture or aggregate segregation. California kneading compactor, Gyratory compactor and Marshall
Hammer are being used as SMA compactors due to mix design method. But according to the
performance of the SMA compacted specimens, none of them could simulate the field compaction
100%. Breaking down of the aggregates during the compaction is one of the most current problems
and if that become excessive, the mixture may not be able to meet the minimum VMA (void in
mineral aggregate) and VCA (void in coarse aggregate) requirements. Also, all the laboratory
specimens are confined within a mould during the compaction, so aggregate segregation which is one
of the most common problems in the field has not been considered as a laboratory study. Correcting
these errors is expected to improve the laboratory compactors such as moving roller in semi confined
moulds, like heavy steel rollers in road construction to be more comparable to the field compaction.
Keywords: SMA mixture, laboratory compactors, compaction energy.
*Correspondence Author: Mohammad Saeed Pourtahmasb, University of Malaya (UM), 50603, Kuala Lumpur,
MALAYSIA. Tel: +60126869472. E-mail: msp_5500@yahoo.com
411
INTRODUCTION
SMA was originally developed in Germany in the 1960s to resist the pavement damage and
loadings of studded tires. It is now a surfacing that is used in many countries of the world. Not all
surfacing called SMA are identical in their composition or their performance [1]. SMA is a gap
graded mixture, which consists of a coarse aggregate skeleton bound with mastic consisting of
bitumen, crushed rock fines, filler and cellulose fibers to prevent drain down of bitumen. The essential
requirement for a successful SMA is the gap grading (above the 2.36mm fraction) with minimum 70%
coarse aggregate.
The stone to stone contact of coarse aggregate ensures a very durable matrix capable of high resistance
to deformation. Typical SMA composition consists of80% 70 coa rse aggregate, 812% filler,
6.07.0% binder, and 0.3% fiber [2]. The deformation resistant capacity of SMA stems from a coarse
stone skeleton providing more stone-on-stone contact than with conventional dense graded asphalt
(DGA) mixes. The primary difference in placing SMA, compared to DGA is in compaction
procedures. Multi-tiered rollers are not used due to the possible working of binder-rich material to the
surface of the asphalt and consequent flushing and pick-up. The preferred method of compaction is to
use heavy, non-vibrating, steel-wheeled rollers. If these are not available, vibrating rollers may be
used, but vibration should be kept to a minimum to avoid fracture of coarse aggregate particles, or
drawing of binder to the surface of the mix and also temperature control is the other consideration in
these situations [3].
SMA is very stiff and less workable than other mixtures. The use of polymer modified binder may
decreases mix workability and necessitate increased compactive effort to achieve high standards of
compacted density and low field air voids has been identified as an important factor in the
performance of all SMA works in compare with other hot mix asphalts so compaction play the most
important role in SMA performance. Attention to asphalt mix compaction is especially crucial in cold
weather conditions, when air voids after compaction can be as high as 16 percent.
Pavements with this level of air voids have shown signs of deterioration after two years [4]. The
different mixes need the different methods and equipments of compaction and one of the most
considerations of researchers is to try to make laboratory activities similar to field. The objective of
this study is to show the importance of improving laboratory compactors, mostly for new generation of
mixtures including Stone Mastic Asphalt mixes.
LITERATURE REVIEW
Considerations in SMA Compaction
Purpose of compaction in all type of asphalt mixtures is to achieve optimum density and Void in
Total Mix (VTM). Optimum Asphalt Contents (OAC) in SMA mixtures and percentage of coarse
aggregates are higher than other types of hot mixes so, it is highly dependent on method of compaction
in compare with Hot Mix Asphalt (HMA) mixtures [5]. The other important factors in SMA
performance are Void in Mineral Aggregates (VMA) and Void in Coarse Aggregates (VCA) [6].
Inadequate Compaction can affect all of these parameters, and changing in VTM, VMA and VCA
during compaction will influence the performance of SMA. High quality compaction is an essential
factor in the design and production of high quality pavements. Achieving consistency in compaction,
both in the laboratory and on site is necessary, if accurate correlation is to occur between laboratory
performance and the observed site behaviour.
It has been demonstrated that differing laboratory compaction methods can produce volumetrically
identical specimens but with widely varying mechanical performance. Laboratory studies show that
412
California Kneading Compactor, Marshall Impact Compactor and Gyratory shear compactor have
some advantages and some disadvantages for HMA compaction [7]. However, using these equipments
for SMA mixtures have not shown comparable result, with field compaction. Using Impact
compactors for SMA even will show worse results, due to breaking down the aggregate and changing
the gradation, VCA and VTM. Some common laboratory compactors, their backgrounds, advantages
and disadvantages of them are briefly explained in this study.
Common Laboratory Compactors
1- California Kneading Compactor
In the late 1920s, the California Division of Highways had come to use an asphalt-aggregate
blend commonly known as an "oil mix" on many of their rural roads. An oil mix was a compromise
between the more expensive high performance HMA used on major urban streets and highways and
the cheaper low performance penetrative method (asphalt oil sprayed on a roadway surface of
unbound particles) used on low-volume rural highways. An oil mix consisted of a combination of
aggregate and asphaltic oil that was mixed either in a plant or on the road itself (called a "road mix),
spread by blade, and then compacted by traffic. Unfortunately, there was no method available for
designing these oil mixes.
Francis N. Hveem developed a method for determining the correct amount of oil based on aggregate
surface area, which could be determined from gradation. It also became evident that even given the
right oil content, roads containing aggregates with "hard, glassy surface texture" tended to deform
excessively under load while roads containing aggregates with a "rough, irregular surface texture"
were more stable. Therefore, Hveem worked to develop a device that would measure stability, which
eventually became the Hveem Stabilometer. The other problem was, specimens compacted in the
laboratory for the Stabilometer did not produce the same readings as those taken from field cores.
Therefore, a new compaction machine, which eventually became the California Kneading Compactor,
was developed to more closely simulate the compaction produced by rollers in the field [7, 8]. Key
parameters of this compactor are:
- Sample size = 102 mm (4-inch) diameter cylinder approximately 64 mm (2.5 inches) in height
(corrections can be made for different sample heights)
- Tamper foot = Shield-shaped with an area of 20 cm2 (3.1 in2)
- Compaction pressure = Ranges from 2.4 to 3.4 MPa (350 to 500 psi)
- Number of blows = 150 (plus any preparatory blows at 1.7 MPa (250 psi) )
- Simulation method = the tamper foot strikes the sample on the top near the edge. The base
rotates 1/6 of a revolution after each blow. This helps achieve a sample particle orientation
that is somewhat like that achieved in the field after roller compaction.
California kneading compactors have been used as a laboratory compactor in Hveem mix design
method and using this compactor is not recommended for SMA.
2- Marshall Impact Compactor
During World War II, the U.S. Army Corps of Engineers (USCOE) began evaluating various
HMA mix design methods for use in airfield pavement design. Motivation for this search came from
the ever-increasing wheel loads and tire pressures produced by larger and larger military aircraft. The
most promising method eventually proved to be the Marshall Stability Method developed by Bruce G.
413
Marshall at the Mississippi Highway Department in 1939. Waterways Experiment Station (WES) took
the original Marshall Stability Test and added a deformation measurement (using a flow meter) that
was reasoned to assist in detecting excessively high asphalt contents. This appended test was
eventually recommended for adoption by the U.S. Army because:
1- It was designed to stress the entire sample rather than just a portion of it.
2- It facilitated rapid testing with minimal effort.
3- It was compact, light and portable.
4- It produced densities reasonably close to field densities.
WES continued to refine the Marshall method through the 1950s with various tests on materials,
traffic loading and weather variables. Today the Marshall method, despite its shortcomings, is
probably the most widely used mix design method in the world. It has probably become so widely
used for compacting hot asphalt mixtures, because It was adopted and used by the U.S. military all
over the world during and after World War II and it is simple and inexpensive. Key parameters of the
compactor are:
- Sample size = 102 mm (4-inch) diameter cylinder 64 mm (2.5 inches) in height (corrections
can be made for different sample heights)
- Tamper foot = Flat and circular with a diameter of 98.4 mm (3.875 inches) corresponding to
an area of 76 cm2 (11.8 in2).
- Compaction pressure = Specified as a 457.2 mm (18 inches) free fall drop distance of a
hammer assembly with a 4536 g (10 lb.) sliding weight.
- Number of blows = Typically 35, 50 or 75 on each side depending upon anticipated traffic
loading.
- Simulation method = the tamper foot strikes the sample on the top and covers almost the entire
sample top area. After a specified number of blows, the sample is turned over and the
procedure repeated [7, 9].
Marshall Impact Compactor is being used for SMA mixtures so far, but according to NAPA, it has
been recommended to decrease the number of blows from 75 blows per each side for HMA to 50
blows for SMA to prevent aggregate braking but recent researches show that, the expected density and
air void cannot be reached with 50 blows and it is one of the negative point of using Marshall hammer
as a SMA laboratory compactor. On the other hand, one of the factors for density control is thickness
control which cannot be done with this method of compaction and the result is same materials with
variation in density and air voids due to thickness of specimens.
3- Gyratory Shear Compactor
Superpave is an acronym for Superior Performing Asphalt Pavements. It is the product of the
Strategic Highway Research Program. Superpave includes a new mixture design and analysis system
based on performance characteristics of the pavement. It is a multi-faceted system with a tiered
approach to designing asphalt mixtures based on desired performance. Superpave includes some old
rules of thumb and some new and mechanistic based features. The Superpave mix design system is
quickly becoming the standard system used in the United States (US). The US was looking for a new
414
system to overcome pavement problems such as rutting and low temperature cracking that had become
common with the use of design systems such as Marshall and Hveem. The Superpave system offers
solutions to these problems through a rational approach.
Gyratory compaction has been used in asphalt mixture design since the 1930's when a procedure was
developed by Texas Highway Department (Department of Transportation today).The original gyrator
compaction procedure was done manually. A mold (constructed from a section of 4 inch inside
diameter pipe) was placed between two parallel plates. The plates were spaced one half inch further
apart than the mold height which allowed the mold to be tilted approximately 6 degree until the
diagonal corners contacted the upper and lower plate. [10]
The compaction procedure used a hydraulic jack mounted upside down to apply a vertical load. The
mold loaded with hot mix was placed between the parallel plates and hydraulic jack was pumped until
a pressure of 50 psi was applied to the specimen. The mold was grasped using two long leveraging
rods and tilted until opposite corners of the mold contacted the plate. Then the mold was gyrated three
times such that opposite corners remained in contact with the parallel plates. Before applying the angle
the pressure is 50 psi. When the angle is applied, the pressure increases initially and then decreases as
the three gyrations are applied.
The angle was removed from the mold by allowing it to sit on the lower plate and the pressure in the
hydraulic jack was readjusted to 50 psi and the process was repeated. Sets of triplicate gyrations were
continued until the endpoint was reached. The endpoint was defined as an increase in pressure to 150
psi with one full stroke of the hydraulic jack. When the endpoint was reached a "level up" load of
2500 psi was applied to complete the compaction. In the late 1950's early 1960's a mechanized
compactor was developed to simulate the manual process and this compactor remains in use today in
the Texas method of mix design.
A second gyratory compactor (6 inches in diameter) was developed in Texas to design mixtures with
large aggregates. The Texas 6 inch protocol is significantly different from the 4 inch method.
Gyrations are applied continuously and the vertical pressure is held constant. Gyrations continue until
the rate o height change per revolution decreases below a specified limit [11].
In this method each sample has to be heated to the anticipated mixing temperature, aged for a short
time (up to 4 hours) and compacted with the gyratory compactor, a device that applies pressure to a
sample through a hydraulically or mechanically operated load. Mixing and compaction temperatures
are chosen according to asphalt binder properties so that compaction occurs at the same viscosity level
for different mixes. Key parameters of the gyratory compactor are:
- Sample size = 150 mm (6-inch) diameter cylinder approximately 115 mm (4.5 inches) in
height (corrections can be made for different sample heights). Note that this sample size is
larger than those used for the Hveem and Marshall methods.
- Load = Flat and circular with a diameter of 149.5 mm (5.89 inches) corresponding to an area
of 175.5 cm2 (27.24 in2)
- Compaction pressure = Typically 600 kPa (87 psi)
- Number of blows = varies
- Simulation method = the load is applied to the sample top and covers almost the entire sample
top area. The sample is inclined at 1.25 and rotates at 30 revolutions per minute as the load is
continuously applied. This helps achieve a sample particle orientation that is somewhat like
that achieved in the field after roller compaction.
415
This compactor was produced to make prepared laboratory samples comparable with tire load pressure
of in-service compacted mixture after traffic compaction, but still it is not possible to achieve the field
density with gyratory compactors in laboratory samples [12]. The gyratory testing machine is a
combination of a kneading compactor and a shear testing machine. It is a realistic simulator of the
abrasion effects caused by repetitive stress and inter granular movement of the mass of material within
a flexible pavement structure [13]. According to recent changes in number and angle of gyration for
SMA mixtures, Gyratory compactor specimens have shown better results in compare with Marshall
Impact Compactor specimens.
A summary of the air voids, bulk density, Marshall Stability, and flow for all the field and laboratory
compacted specimens for Hot Mix Asphalts, from previous researchers is presented in Table 1 to
clarify the difference between laboratory compactors and field compaction.[14] (Gradation, asphalt
type and aggregates are the same in all the mixes). According to table 1, the test results of Gyratory

compactor with gyration angle of 1.25 are more similar to field cores.
Table 1- Summary of Stability, Flow, Air void and Bulk Density for Laboratory and Field Specimens
Test California Kneading Marshall Impact Gyratory Shear Field Cores

Compactor
Compactor Compactor(1.25 )
Stability (KN) 17.3 20.1 31.00 30.1

Flow (mm) 4.26 4.61 3.27 2.91
Air void (%) 5.85 4.31 5.10 5.04
Bulk Density (gm/cm 3 ) 2.315 2.353 2.31 2.335
METHODOLOGY
Laboratory Rolling Compactors
Traditional laboratory mixes have been compacted using the impact method i.e. Marshall
Hammer. Currently, greater use is being made of the gyratory and roller compactor methods to reach
desired density and prepare samples for mechanical testing. Both the roller compactor and gyratory
compactor simulate the action of a roller and the material compacted is more representative of the
material laid and compacted on site. Roller compaction is generally considered to be the best method
of compacting asphaltic paving material in the laboratory. Slabs compacted with the roller compactors
are in a large size, so it will give researchers a chance to have more core samples for laboratory tests
and in some cases which asphaltic beam is needed, it can be prepared easily with cutting the laboratory
slabs to the expected beam size.
Also steel roller compactors are the real simulation of heavy steel rollers on site and results can be
comparable with taken cores from site. Few numbers of rolling compactors were introduced so far and
their performances were acceptable in compare with site compaction. European standard asphalt
compactor (Figure 1) is one of the laboratory roller compactors, which was originally designed to
provide a solution to the problem of making homogeneous laboratory specimens large enough to be
used for wheel-tracking and other laboratory tests. It is used to compact to either a target mixture
density or to provide a standard compactive effort to a range of different asphaltic paving materials.
416
Figure 1: European Standard Rolling Asphalt Compactor
In recent years, pavement researchers have discovered the importance of laboratory compactors and
their influence on compacted specimens performance. On the other hand using roller compactors
instead of previous laboratory compactors are similar with site compaction, and in both methods all
the specimens are prepared with coring. Figure 2 is the simple modeling of site compaction in the
laboratory, which was developed in Worcester Polytechnic Institute in United States. The mix is
placed in a wooden mould and then compacted with a hand driven heavy steel roller up to expected
level but load cannot be adjusted in this compactor and it is an initial modeling, for laboratory roller
compactors.
Figure 2: Worcester Polytechnic Institute Roller Compactor
According to collected information in this study, California Kneading Compactors is not

recommended for Stone Mastic Asphalts because the Hveem mix design method must be used for
sample preparation, and this method is not suitable for SMA.
417
Marshall Impact Compactors are being used as a SMA compactor, and according to NAPA number of
blows must be 50 per each side to prevent aggregate braking. But compacted specimens test results
show:
- Different specimens heights

- Aggregate breaking
- Different densities and air voids
- Degradation and changing in VCA mostly in SMA with large nominal aggregate size.
- Expected air void cannot be reached with 50 blows and increasing the number of blows cause,
degradation and unexpected air voids due to aggregate breaking and changing in VCA
But these problems in Gyratory shear compactors are less than Marshall Impact Compactors
Because:
- Height of specimens can be adjusted, so achieved air void is closer to expected

- Aggregate braking is less than Marshall, so degradation and changing in VCA is also lower
- Method of compaction is more comparable with field compaction
Gyratory shear compactor seems to be the best choice for compacting Stone Mastic Asphalt in
laboratory. All the three types of laboratory which are mentioned above can just prepare one specimen
but numbers of core specimens which can be taken from roller compactor slabs are more than them.
Specifications of Marshall Impact Compactor, Gyratory Shear Compactor and roller laboratory
compactor are given in table 2.
Table 2- Specifications of Laboratory Compactors in Compare with Roller Compactor

Specification Marshall Gyratory Roller Compactor
Confined mould YES YES NO
Surface Aggregate breaking YES YES NO
Density& Void control NO YES YES
Height control NO YES YES
Similar to site NO NO YES

Time consuming due to YES YES NO
preparing samples one by one
According to table 2, Roller compactor is the real simulation of site heavy steel compactor in smaller
scale and could meet all the site specifications. Also, previous studies on Roller compactor slab
properties have shown the acceptable results in compare with site [15].
CONCLUSION
Based upon given information in this study, it was found that Stone Mastic Asphalts (SMA) slab
properties such as Density and Air voids, highly dependent on the method of compaction so laboratory
methods of compaction must be approximately similar to the site compaction. California Kneading
Compactor was not recommended to be used as a SMA compactor due to Hveem mix design method,
418
which is not being used anymore and also seems not to be suitable for SMA mixture design. Also it
was found reducing the number of blows from 75 blows for HMA to 50 Blows for SMA in Marshall
Impact Compactor to prevent breaking of aggregates was not performed well, because expected Air
void cannot be reached with 50 blows.
Gyratory Shear Compactor was recognized as a better laboratory compactor in compare with Marshall
due to possibility of specimens height control and method of compaction. In this study using Roller
compactor instead of previous laboratory compactors is highly recommended for SMA, because of
real simulation of field compactors. Possibility of controlling the slab thickness, the adjustable level of
pressure, better quality of taken core specimens and capability of producing many cores (100mm
diameter for ordinary tests or even 200mm diameter for Rutting and wheel tracking tests) from each
slab are good enough to be suggested as a future SMA laboratory compactor.
REFERENCES
1- ARRB Transport Research (2004). Stone Mastic Asphalt, Austroads Technical Note 16 April
2004.
2- Austroads (2003). Selection and Design of Asphalt Mixes, Australian Provisional Guide, APRG
Report 18, * ARRB-Transport-Research.
3- AAPA (2000). Stone Mastic Asphalt Design and Application Guide.
4- McLeod, N.W., (1967). "Influence of Viscosity of Asphalt Cements on Compaction of Paving
Mixtures in the Field," Highway Research Record No.158 Highway Research Board, National
Academy of Sciences, National Research Council, Washington, D.C., 1967, pp. 76-115.
5- A.E.Hunter, G.D.Airey and A.C.Collop, (2004). Effect of asphalt mixtures compaction on
aggregate orientation and mechanical performance, (CAPSA 2004).
6- E.R.Brown & Rajib B.Mallick. (2002). Stone Matrix Asphalt properties related to mixture
design, NCAT report 94-02.
7- Linden, Fons and Van Der Heide, Jos. (1987). Some Aspects of the Compaction of Asphalt
Mixes and Its Influence on Mix Properties. V.56-87, p.408.
8- Roberts, F.L.; Kandhal, P.S.; Brown, E.R.; Lee, D.Y. and Kennedy, T.W. (1996). Hot Mix
Asphalt Materials, Mixture Design, and Construction, National Asphalt Pavement Association
Education Foundation, Lanham, MD.
9- Asphalt Institute (1997). Mix Design Methods for Asphalt, 6th ed., MS-02, Asphalt Institute.
Lexington, KY.
10- D'Angelo, John A., Paugh, Charles, Harman, Thomas P. and Bukowski, John (1995).
Comparison of the Superpave Gyratory Compactor to the Marshall for Field Quality Control,
V.64-95, p.611.
11- Ziauddin A. KhanU et al.(1998). Comparative study of asphalt concrete laboratory compaction
methods to simulate field compaction, Construction and Building Materials 12 (1998) 373-384.
12- Asphalt Institute. (2001). Superpave Mix Design, Superpave Series No. 2 (SP-02), Asphalt
Institute, Lexington, KY.
13- Kumar A, Goetz WH. (1974). The gyratory testing machines a design tool and an instrument for
bituminous mixture evaluation, Proc Assoc Asphalt Paving Technology 1974; 43:350-371.
419
14- National Asphalt Pavement Association. (1982). Development of Marshall Procedures for
Designing Asphalt Paving Mixtures, Information Series 84, National Asphalt Pavement
Association, Lanham, MD.
15- Muniandy, R. (2007). Thickness Analysis of Stone Mastic Asphalt (SMA) Slab Compacted
Using a Newly Developed Roller Compactor. American Journal of Applied Sciences 4 (4): 233-
236, 2007.ISSN 1546-9239.
420
TYPES OF DAMAGES ON FLEXIBLE PAVEMENT FOR MALAYSIAN FEDERAL ROAD
Nurul Elma Kordi 1*, Intan Rohani Endut1,2, Bahardin Baharom1,2
ABSTRACT:
Flexible pavements in Malaysia are designed until ten to fifteen years design life to support load.
However, sometimes these roads not able to carry the load with the specified design life especially in
industry areas because heavy lorries always using these road to move their goods. A lot of distress on the
flexible pavement and it become uncomfortable feelings to the driver. A lot of factors contribute the
problem of failure to flexible pavement either internal or external factors. Among the heavy trucks'
operators, they want to raise the axle loads limit for all trucks to avoid delivery delays and meet higher
storage chargers; in fact Road Transport Development (RTD) had given special permission to some 1000
lorries since August 2007 to carry the new maximum load. In order to quantify flexible pavement
damages, the impacts of increase axle load on Federal road must identify and evaluate. This paper
discusses a type of damages on the flexible pavement, causes of failure to flexible pavement and
appropriate structural requirements for pavement performance that can carry the load on a routine basis.
With the best understanding of the causes of failure and process to rehabilitation, the responsibilities'
authorities will know well the key to proper maintenance of flexible pavement in order to design, manage
and maintain the highways.
Keywords: Axle Load, Distress, Flexible Pavement, Rehabilitation
1
Malaysia Institute of Transport (MITRANS), Universiti Teknologi MARA, 40450 Shah Alam, Selangor, MALAYSIA
2
Faculty of Civil Engineering, Universiti Teknologi MARA, 40450 Shah Alam Selangor, MALAYSIA
* Correspondence Author: Nurul Elma Kordi, Malaysia Institute of Transport, Universiti Teknologi MARA, Malaysia.
Tel: +6003 5521 1492, Fax: +603 5544 2344. E-mail: ema_nuelko@yahoo.com
421
INTRODUCTION
Road has played an important role in the trade and transportation system throughout the world,
and it become rapid increase in the pavement infrastructure development in Malaysia. Gazette under
Federal Roads Ordinance is usually roads linking the state capitals, airports, railway stations and ports.
Currently, Malaysia has more than 80,300km roads. The road is divided into three main categories
namely toll expressway (1,700km), federal roads (17,500km) and state roads (61,100km) and the life
spans are between 10 to 15 years (Zakaria and Hassan, 2005). Local authority road (city mall, municipal
or local council) or kampong (district office) road is depending upon jurisdiction and normally maintained
by the responsibility local authority (Haron, 2004).
In Malaysia, Federal road use the flexible pavement that constructed with asphaltic cement and
aggregates and consist of several of layers with the lowest called the subgred which natural soil itself. The
next layer is the subbase, which consist of crushed aggregate. The next layer is called road base, which
can be made of crushed aggregates with a cementing material. The top layer is called surfacing layer. It is
usually made of asphaltic concrete. The structural strength depends on individual material strength
characteristic and thickness layer. Some form of deflection within the elastic limit is allowed. Surface
readability is good but relatively less durable or susceptible to high temperature (Haron, 2004).
Federal road at Malaysia almost used the flexible pavement compare to rigid pavement because
flexible pavement more comfortable to the user. Furthermore, flexible pavement can be use by users once
ready. Table 1 shows the differences between flexible and rigid pavement. With best design and proper
construction method, adequate routine and periodic maintenance are necessary to achieve the determined
design life spans and even beyond. The main factors of pavement deterioration due to increasing the
number of vehicle and load traffic by the run which is five percent per annum.
Table 1: The differences between flexible and rigid pavement
No. Flexible Rigid

1 Last for 20 years Last for more than 40 years
2 High cost and schedule maintenance More economic maintenance, no schedule
3 Cheap and easy to get the material Material shortage problem always occurs
4 Low initial cost Higher initial cost
5 Easy to upgrade / stage construction Cannot upgrade / no stage construction
6 Rutting and potholes may occur Free from rutting, potholes and corrugation
7 Less economic for the long duration More economical for the long duration
8 Can be used once ready Traffic disturbance, wait for maximum
strength (28 days)
(Source: Haron, 2004)
As a part of the study by research, this paper discusses on failure on flexible pavement, sources
failure of flexible pavement and appropriate structural requirements for pavement performance that can
carry the loads on a routine basis. This is a preliminary study before conduct the actual study which
mapping all the damages along Federal road number one after this.
TYPES OF DAMAGES
According to Highway Research Board, 1970 cited by Pavement Interactive, 2010, pavement
distress is "any indication of poor or unfavourable pavement performance or signs of impending failure;
422
any unsatisfactory performance of a pavement short of failure". The causes of pavement failures are
divided into two categories, which are an internal and external failure. Internal failures of pavement are
usually because of the lack pavement mixture, weaknesses of component materials and poor construction.
Meanwhile, external failures are due to overloading, diesel spillage, flooding, sink holes and other
unforeseen cause such as earthquake, volcanoes and others. The failures of flexible pavement are divided
into four categories, which are surface deformation, surface defects, cracking and patching and potholes
(Rani, 2007). Tables 2, 3, 4 and 5 below shows the lists the pavement distress groups, possible causes,
and the probable treatments that summarized by Rani, 2007 and PFI, 2010.
Table 2: Surface deformation
No. Possible Causes Probable Treatments

Shoving
1 Instability mixture of asphalt (course or fine Remove surface and base as necessary and
aggregate too rounded or too smooth textured) replace with a more stable material
2 Low modulus bases course Reconstruction of base
3 Thin wearing course Bituminous overlay
4 Dragging of pavers during laying when Remove surface and base as necessary and
bituminous mix temperatures were low replace with a more stable material
5 High stress due to braking and acceleration Bituminous overlay with the stiffer mix or use
movements the high compaction mix.
Corrugations
1 Inadequate stability of bituminous Replace bituminous surface
surface (excessive moisture)
2 Compaction of base in wave form Base reconstruction
3 Faulty paver behaviour with some Replace the faulty mixes and correct the
mixes (contamination caused by oil spillage) faulty behaviour
4 Heavy traffic on steep downgrade or upgrade Mill off the corrugated surface and replace
with the stiffer mix or use hot mix asphalt
5 Stopping at intersection stop lights or Mill off the corrugated surface and replace
roundabout with the stiffer mix or use hot mix asphalt
6 Inadequate stability of base course Base construction.
Rutting
1 Inadequate pavement thickness Strengthening overlay or reconstruction
2 Inadequate compaction of structural layers Reconstruction
3 Unstable bituminous mixes Replace or recycle bituminous surfacing or
use the stiffer mix
4 Unstable shoulder material which does not Shoulder improvement and overlay rutted
provide adequate lateral support area with bituminous surfacing
5 Overstressed subgrade which deforms Reconstruction
permanently
6 Unstable granular bases or subbases Base or sub base strengthening
Depression
1 Differential settlement of Subgrade or base reconstruction
subgrade or base materials
2 Settlement of services and / or Reconstruction of services and/or widening
widening trenches trenches
3 Volume change of subgrade Improve sub-soil drainage and
due to environmental influences reconstruct
4 Settlement due to instability of embankment Embankment stabilization
423
(Source: Rani, 2007 and PFI, 2010)
Table 3: Surface defect

Bleeding / Flushing
1 Excessive asphalt in the surface layer. On hot Apply hot sand to blot up the excess binder to
days the binder expands into air voids; if the the stone size.
volume of air voids is too low, continued
expansion results in lower stability of the mix
with the consequence that traffic will force out
excess binder to the surface.
2 Paving over flushed surfaces. The excess Apply hot sand or aggregate seal coat.
bitumen on the old surface may be pumped up
through the new paving over period of time.
3 Paving over excessively primed surfaces Apply hot sand
4 Lack of proper rolling during placement Resurfacing the asphalt
5 Failure to protect a newly constructed surface Resurfacing the asphalt
from traffic until the asphalt cured sufficiently.
Ravelling
1 Insufficient bitumen content Thin bituminous overlay
2 Poor adhesion of bitumen binder to aggregate Thin bituminous overlay
particles due to wet aggregate
3 Inadequate compaction or construction during Thin bituminous overlay
wet weather
4 Deterioration of binder and/or aggregate Thin bituminous overlay
Polishing
1 Inadequate resistance to polishing of surface The bituminous overlay of use of stiffer mix
aggregates particularly in areas of heavy traffic
movements or where high stresses are
developed between surface and tyres
2 Use of naturally smooth uncrushed aggregates Thin bituminous overlay.
Delimitation / Surface Lifting /Seal Break
1 Inadequate cleaning or inadequate tack cot Mill off and re-lay upper layers.
before placement of upper layers
2 Seepage of water through asphalt, especially in Replace wearing course or thin bituminous
cracks, to break the bond between surface and overlay
lower layers.
3 Weak, loose layer immediately underlying seal Reconstruction of weak layers
4 Adhesion of surface binder to vehicle tyres Thin bituminous overlay
424
Table 4: Cracking crack

Fatigue (alligator/crocodile) crack
1 Inadequate pavement thickness Strengthen the pavement or reconstruction
2 Low modulus base Strengthen the base or reconstruction
3 Brittle base Base recycling or reconstruction
4 Poor base drainage Improve the drainage and reconstruct
5 Brittle wearing course Replace or treat wearing course
Block cracking
1 Joints in underlying layer Crushed aggregate overlay
2 Shrinkage and fatigue of underlying Replace underlying cemented
cemented materials materials
3 Shrinkage cracks (due to bitumen Seal cracks or replace bituminous surfacing
hardening) in bituminous surfacing
4 Fatigue cracks in embrittled Cut and patch or crushed aggregate overlay
bituminous wearing course
Longitudinal cracking
1 Reflection of shrinkage cracks Cut and patch
2 Poorly Constructed paving lane in bituminous Replace bituminous surfacing
surfacing
3 Displacement of joints at pavement widening Reconstruction of joints
4 Differential settlement between cut and fill Crushed aggregate overlay or reconstruction
of joints.
Transverse (Thermal) Crack
1 Reflection of shrinkage cracks Cut and patch
2 Construction joint in bituminous surfacing Crack sealant
3 Structural failure of Portland Cement Reconstruction of base
4 Shrinkage crack bituminous surfacing Seal cracks or replace bituminous surfacing
5 Reflection of joints in the underlying base Crushed aggregate overlay or reconstruction
of joints
Edge cracks
1 Excessive traffic loading at the pavement Widen the pavement or strengthen the
edge pavement edge
2 Poor drainage at pavement edge and shoulder Improve drainage and shoulder
3 Inadequate pavement width which forces Widen treatment
traffic too close to the pavement edge
4 Insufficient bearing support Reconstruction
Edge breaks
1 Inadequate pavement width Widen the pavement
2 Alignment which encourages drivers to Pavement widening and realignment
travel on the pavement edge.
3 Inadequate edge support Shoulder strengthening
4 Edge drop-off Strengthening and leveling of shoulder with
road surface
5 Loss of a adhesion to base Cut and patch or bituminous overlay.
Edge drop-offs
1 Inadequate pavement width Widen the pavement
2 Shoulder material with inadequate resistance Replace shoulder material and reconstruct
425
to erosion and abrasion.

3 Resurfacing of pavement without Leveling of shoulder with road surface
resurfacing of shoulder
Crescent shaped cracks / parabolic / slippage / shear cracks
1 Lack of bond between wearing course and Cut and patch
the underlying layers
2 Low modulus bases course Reconstruction of Reconstruction of base
base
3 Thin wearing course Bituminous overlay
4 Dragging of pavers during laying when Cut and patch
bituminous mix temperatures were low
5 High stress due to braking and acceleration Bituminous overlay with the stiffer mix or use
movements the high compaction mix.
Table 5: Patching and Potholes

Patch/Patch Deterioration
1 Inadequate cleaning or inadequate tack cot Mill off and re-lay upper layers.
before placement of upper layers
2 Seepage of water through asphalt, especially in Replace wearing course or thin bituminous
cracks, to break the bond between surface and overlay
lower layers.
3 Weak, loose layer immediately underlying seal Reconstruction of weak layers
Pothole
1 Loss of surface course Patching
2 Moisture entry to base course through a Cut and patch
cracked pavement surface
3 Load associated disintegration of base Base reconstruction
From the table, it shows most of the possible causes damages of pavement are due to of
imperfection during construction. This failure will give uncomfortable to the users during their driving.
Either responsibility contractor does their work perfectly or not during construction of the road need to
investigate.
CHARACTERISTICS EFFECTING FAILURE OF PAVEMENT
Truck factors
Most comment from the users of the road about the vehicle that carries the heavy load that makes
the pavement damage. On behalf of heavy truck operators, they want to increase the axle load limit to
avoid delivery delays and meet higher storage chargers. Table 6 below shows the comparison weight limit
of the truck among various countries.
426
Table 6: Comparison Weight Limit of Truck among Various Countries
Country Type (load+vehicle) Weight

limit (maximum)
Malaysia Two-axle-fixed truck 16 tons
Three-axle-fixed truck 21 tons
Four-axle-fixed truck 25 tons
Thailand Two-axle-fixed truck 12 tons

Five-axle-fixed truck 39 tons
Singapore Two-axle-fixed truck 16 tons

Three- axle-fixed truck 24 tons
Four-axle-fixed truck -
Articulated truck 24 tons
ASEAN Standard Two-axle-fixed truck 16 tons

Four-axle-fixed truck -
Articulated truck 38 tons
(Source: MRTV3, 2010)
Table 7 show the comparison of axle load among various countries. Axle load and total truck load
limits are also an important factor, which can determine the efficiency of the road system and
competitiveness of the countrys logistics system. Thailands maximum axle load limit is 8.2 tons while
the truck limit had increased to 25 tons from 21 tons in 2006. Thailands axle load limit is lower than
other countries. Truck load limit and axle load limit will have impact on Thailands long-run freight
transport cost and road maintenance cost, which need to be balanced. Different load limits among
Thailand and neighboring countries might also cause problems for transshipment of goods across national
borders and complicate international trade negotiations (Transport Sector, 2008).
Table 7: Comparison of Axle Load among Various Countries
Country Axle Load Limit (Ton)

Malaysia 12
Thailand 8.2
Singapore 10
Japan 11
Europe 10-13
The Peoples Republic of China 10
ASEAN member countries 10
(Source: Transport Sector, 2008)
427
Heavy vehicle also contributed to deterioration of the road, there were 19.3 million registered
vehicles on the Malaysias road, and the government spent RM5 billion between 2001 and 2010 to keep
sustaining all the Federal roads (New Strait Times, June 2010). Heavy vehicle has been banned during
morning peak hours started 2 August 2010 from certain stretches to avoid traffic congestion on the North-
South Expressway. Federation of Malaysian Manufacturer presidents, Tan Sri Mustafa Mansur (2010)
said the move would not cause much impact as lorries could still use other alternative roads to get their
destination. This rule is applied after investigations have been done. The result shows the total of heavy
vehicle breakdown three times more than light vehicles and this situation will give big impact on traffic
congestion (Berita Harian, 2010). However, Pan Malaysia Lorry Drivers Association president, Er Sui
See, unhappy with the ban because it unfair and absurd to expect them apply for temporary permits each
time they need to use the highway during the ban time (The Star, July 2010).
Pavement and environmental factors
The other factor of pavement damages is due to extra ordinary weather such as heavy rain and
flooding. It is bringing on water absorption to pavement and weakened the structural pavement base. The
pavement will weaken and fail due to increasing moisture content of the subgrade. Therefore, drainage is
very important in the roads' system to keep the low water table. In Malaysia, RM220 million was spent
every year on repairing road damages by rains and floods (New Strait Times, 2010). In general,
conventional asphalt easy to damage when exposed to water and high incidences of flooding. The used
less quality of material, non effective construction, construction procedure without control and
overweight commercial transports also as sources of failure pavement.
Tolled expressways are better maintained than Federal roads because government has to take care
of more roads, which are 16,500km of federal roads compared only 1,360km that managed by private
highway concessionaires (The Star, 2010).
Contact condition and configuration of tire
The consensus is clear which the details of the contact conditions such as the exact area, pressure
and pressure distribution, effect stresses and strains near to the surface of the pavement, whereas the
response in the lower layers depends mainly on the overall load. Haas and Papagiannakis, 1986 (as cited
in Cole & Cebon, 1997) showed that increasing the tyre inflation (contact) pressure from 415 kPa to 830
kPa at constant load will increase the theoretical vertical compressive strain near to the surface of a
200mm thick asphalt layer by up to a factor of eight, but hardly affected the strain at the bottom of the
layer. In 1963, Zube and Forsyth (as cited in Cole & Cebon, 1997) performed an experimental
comparison of the vertical deflections and transverse strains of a flexible pavement surface, to wide-base
single tyres and dual wheels. Their results indicated that pavement deflection was equivalent to 27 kN
carried on a single tyre or 40 kN carried on a dual pair.
Cole and Cebon, 1997, conclude that fatigue cracking failed due to wide single tyres cause up to
seven times more damage than dual tyres carrying the same total load for the relatively thin asphalt
pavements (minor roads). For thicker pavements, where permanent deformation is the main mode of
failure, wide single tyres are likely to cause one and half to two times more damage than dual tyres.
Dadoo and Thorpe (2005) identify the main factors responsible for pavement damage caused by heavy
vehicle such as dynamic axle load, number and type of axles (e.g. single, tandem), tyre properties (e.g.
wide-base, dual) and pavement properties (e.g. pavement type, thickness, temperature and roughness).
This research will discover which these factors have the greatest influence on pavement response.
Marshek et al. (1986) lists of the possible damage or defect on the pavement are shown in Table 8. It can
be concluded that, pavement damage will be increased due to less of contact area of tires to road. The
higher pressure of tires is one of the factor's pavement damages because small of contact area between
tires and pavement.
428
Table 8: Summary of characteristics influencing pavement damage
Characteristic Decrease in Pavement Damage Increase in Pavement Damage

Fewer Tires per Axle
More Tires per Axle (Tandem Axle)
Greater Load per Tire
Less Load per Tire
No. of Tires per Axle Greater Load per Unit Contact
Less Load per Unit Contact Area
Area
Decrease in Pavement Damage
Increase in Pavement Damage
Fewer Axles
More Axles
Greater in Load per Axle
Less Load per Axle
No. of Axles Greater Load per Unit Contact
Less Load per Unit Contact Area
Area
Decrease in Pavement Damage
Increase in Pavement Damage
Load Distributed Evenly Among Load Not Distributed Evenly
Axle Among Axles
Load Distribution Less Load for the Most Heavily Greater Load for the Most
Loaded Axle Group Heavily Loaded Axle Group
Decrease in Pavement Damage Increase in Pavement Damage
Higher Speeds Lower Speeds
Speed / Congestion Decrease in Load Duration Greater Load Distribution
Decrease in Pavement Damage Increased Pavement Damage
Lower Tire Pressure Higher Tire Pressure
Greater Surface Contact Area Smaller Surface Contact Area
Tire Pressure between Tire and Pavement between Tire and Pavement
Less Load per Unit Contact Area Greater Load per Unit Area
Decrease in Pavement Damage Increase in Pavement Damage
EFFECTS OF INCREASED LEGAL LOAD LIMITS
Studies of the effect of increasing the legal load limit from 18,000 to 20,000 pounds for single
axles and from 32,000 to 36,000 pounds tandem axles were made by the California Transportation
Department. According to a study by Smith (1973) cited by Matthews and Baumeister (1976), if half load
trucks which hauled legal limit loads were to increase axle loads to the proposed limits, the decrease in
fatigue life of existing pavements is predicted to be between 20 and 25%. The effect of this decrease in
pavement life will be to increase maintenance costs on existing pavements considerably over what they
have been in the past and require reconstruction and rehabilitation at an earlier date. Findings from the
study indicate that an 11% increase in legal load limits appears to reduce service life in years
approximately 20% (Matthews and Baumeister, 1976).
Jarviss (2008) opinion the increasing of truck Gross Vehicle Weights (GVW) on the federal
highway system, while keeping individual axle weight limits at the current level, will improve
productivity, fuel conservation, air quality, infrastructure conservation, and public safety, while reducing
carbon emissions and traffic congestion. While Eason and Greene (2008) explained by providing fuel-cost
savings, increasing the allowable GVW also would streamline trucking across national borders. Current
maximum GVWs allowed by Mexico and Canada are 106,920 pounds and 95,900 pounds, respectively.
According to the Americans for Safe and Efficient Transportation (ASET), the payload increase would
make trade more parallel and obtain $14.5 billion of potential savings in shipping costs. For an individual
truck, ASET also states that tractor-trailers (with six axles rather than five) can reduce vehicle miles
429
travelled (VMT) by 11% and reduce fuel usage by 6%. Accidents would decrease as well, as fewer trucks
driven by a smaller pool of more highly qualified drivers would be on the road. With the increasing
prices of fuel, the haulier needs to improve their trucking system because fuel costs will reduce the profit
margins on every industry in the economy, and logging is certainly no exception.
APPROPRIATE STRUCTURAL REQUIREMENT
The rehabilitation of flexible pavements encompasses a broad range of activities, which could be
grouped into three categories namely; restoration, resurfacing (structural) and reconstruction (Caltrans,
2000). In order to prevent failure flexible pavement, the polymer-modified asphaltic (PMA) concrete has
started used in several project road upgrading by Work Ministry. The National Economic Action Council
had agreed that PMA be used on five per cent of new projects as a part of the 10th Malaysia plan. PMA is
a mixture of natural or synthetic polymer materials with conventional bitumen. Their rubberlike surfaces
are highly resistant against rutting and cracking and absorb little water, which is the cause of cracks in
roads. PMA has a lower life cycle cost and can admit the higher axle load, also, with using PMA it can
save until RM200,000 for every kilometer of the road in yearly maintenance work, although the cost
PMA is 1.7 times more than conventional (New Strait Times, 2010). A study was completed for the
Affiliate Committee of the Asphalt Institute on the use of PMA for reducing distress in flexible
pavements and hot mix asphalt (HMA) overlays. The results from that study found that the use of PMA
reduced pavement distress and increased the life of flexible pavements by two to ten years (Asphalt
Institute IS-215, 2005).
Figure 1 below illustrates the alternatives of pavement maintenance and rehabilitation. This
figure is helpful in understanding the aim for which a given treatment.
Pavement maintenance and

Rehabilitation
Maintenance Rehabilitation
Surface Subsurface Reconstruction Recycling
- Preventive Overlays
- Corrective
Combination of
recycling and overlays
Figure 1: Alternatives of Pavement Maintenance and Rehabilitation

(Source: Monismith, 1979)
430
There are differentiated between preventive and corrective approaches preventive strategies for
pavement surface include fog-seal asphalt, rejuvenators, joint sealing, seal coat (with aggregate) and thin
blanket. For surface corrective there are patching, crack filling, joint sealing, seal coat, friction course,
grooving, thin blanket, surface recycling, cold planning and drainage. For subsurface, there have drainage
for preventive maintenance and sub-sealing, in-situ stabilization and drainage for corrective maintenance.
CONCLUSSION
The failures of pavement have increased significantly over the year. The effect of axle load plays
an important role in the deterioration of flexible pavements. Although the Association of Malaysia
Hauliers (AMH) and truck operators want to raise the lorry load limit for all as similar size vehicles have
been transporting heavier loads at overseas, the responsibilities' authority still needs to investigate
carefully the effect to the road and users. Damaged roads need to repair immediately as they can cause
accidents and traffic congestion. The study that was conducted and discussed in this paper show that the
possible causes of flexible damages mostly did not cause by heavy truck. Therefore, the further research
needs to conduct to check whether the responsible contractor done their work properly or not and the
other factor of flexible pavement easy to damage need to consider and investigate. New road construction
will involve big investment. Therefore, to ensure life span of the road that constructed more durable,
comfortable and safe to the user, construction and maintained of the road need to be done effectively.
REFERENCES
Asphalt Institute, (2005). Quantifying the Effects of PMA for Reducing Pavement Distress,
Asphalt Institute Information Series 215.
Berita Harian. (2010, July 30). Tiada Kompromi Lori 6 Tayar Masuk Lebuhraya, pp. 2.
Caltrans, S. L. (2000). Maintenance Technical Advisory Guide (TAG). American Concrete Pavement
Association.
Dodoo, N. A. & Thorpe, N. (2005). New Approach for Allocating Pavement Damage between Heavy
Good Vehicles for Road-user Charging. School of Civil Engineering and Geosciences, University of
Newcastle Tyne, pp. 419-430.
Eason, K. & Greene, D. (2008), Potential Impact of 97,000-Gvw on Logging Costs. Retrieved from
Agricultural Transportation Efficiency Coalition website: http://ag-haul.org/resources/
multimedia/PDF/ AgTEC-GVWstatement.pdf.
Haas, R. C. G. & Papagiannakis, A. T. (1986). Understanding Pavement Rutting. Special Workshop on

Rutting in Asphalt Pavements, Toronto, Roads and Transport Association of Canada.
Haron, S. (2004). Highway and Traffic Engineering (1st ed.). Universiti Teknologi MARA, Perlis.
Jarvis, S. (2008). Increase Truck Weight Limits For Agricultural Transportation from 80,000 to 97,000
Pounds. Retrieved from Agricultural Transportation Efficiency Coalition website: http://ag-
haul.org/resources/ multimedia/PDF/ AgTEC-GVWstatement.pdf.
431
John, S. M. & William, Y. B. (2003). Distress Identification Manual for the Long-Term Pavement
Performance Program (4th revised ed.). U.S. Department of Transportation, Federal Highway
Administration, 6300 Georgetown Pike.
Marshek, K. M., Chen, H. H., Connell, R. B. & Saraf, C. L. (1986). Effect of Truck Tire Inflation
Pressure and Axle Load on Flexible and Rigid Pavement Performance. Transportation Research
Record N1070, Transportation Research Board, Washington DC.
Matthews, J. A. & Baumeister, K. L. (1976). Damage to Pavement due to Axle Load. Technical Report
Document Page, California Department of Transportation, Sacramento, California 95807.
Monismith, C. L. (1979). Pavement Evaluation and Overlay Design Summary of Methods. Transportation
Research Record No. 700, Transportation Research Board, National Research Council, Washington,
D.C.
New Strait Times. (2010, June 22). Fewer Potholes with New Type of Asphalt, pp. 14.
Nicholas, J. G. & Lester, A. H. (2002). Traffic and Highway Engineering (3rd ed.). University of
Virginia, United States: Thomson Learning.
Pavement Failure Identification, PFI (2010). Retrieved from http://www.apai.net/cmdocs/apai/

designguide/Appendix_A-B.pdf
Pavement Interactive (2010). Pavement Distress. Retrieved from http://pavementinteractive

.org/index.php?title=Surface_Distress
Quintus, H. L. & Mallela, P. E. (2005). Reducing Flexible Pavement Distress in Colorado Through the
Use of PMA Mixtures. Final Report No. 16729.1/1 (prepared for the Asphalt Institute), Applied
Research Associates, Inc., Round Rock, Texas.
Rani, O. A. (2007). The Effectiveness of Pavement Rehabilitation at Kuala Lumpur Karak

Highway. Master thesis, Universiti Teknologi Malaysia.
Thailand Infrastructure Annual Report (2008) Retrieved on 12 April 2010 from

http://siteresources.worldbank.org/INTTHAILAND/Resources/333200-1177475763598/3714275-
1234408023295/5826366-1234408105311/chapter2-transport-sector.pdf
The Star. (2010, May 5). Tolled Roads are Better Maintained, pp 11.
The Star. (2010, July 27). Heavy Vehicle Banned, pp. 12.
Vehicles of today and motor ways in Myanmar (4). Retrieved on 16 August 2010 from
http://www.mrtv3.net.mm/open6/230909for1.html
Zakaria, S. & Hasan, M. Z. (2005). Data Collection Strategies in Malaysia, Public Work Department
Malaysia.
432
THE EFFECTIVENESS OF A ROAD TRANSPORTATION NETWORK SYSTEM

IN A PORT CITY: TOWARDS GREEN LOGISTICS IN MALAYSIA
Noraini Anor 1*, Zakaria Ahmad 2
ABSTRACT:
The road transportation network system is part of the economic generation of Malaysia. This system has a
significant impact to the transportation industry and a major player of logistics system. Hence, this
network system needs further improvement in the port city. To date, little research has been carried out on
the environmental sustainability of Malaysias port cities and how Green elements can be introduced
into their road transportation networks. Currently, the road transportation network system is not integrated
to other infrastructure, facilities and other required services. An unplanned road transportation network
system will disrupt the transportation movement, environmental sustainable create network system
problems and affect other logistics factors. The main purpose of this study is to examine what extent the
effectiveness of transportation network system in port cities contributes towards green logistics in
Malaysia. The objectives of the research study are: first, to identify the impact of the road transportation
network system in the port city on green logistics; secondly, to examine the relationship between the road
transportation networks system and a port city and finally, to identify the structure and patterns was
current practice of the road transportation network system. This study was based on a case study that
analyzed the structure and patterns of the road transportation network system at Port Klang, Malaysia.
The research process included collection of primary data such as site investigation and observation while
secondary data from a literature review to support the research. The findings of research will help develop
utilization of rational land for the sustainable development of a road transportation network system in the
port city.
Keywords: Network; Port city; Green Logistics
1
Postgraduate Student, Malaysia Institute of Transport (MITRANS), Universiti Teknologi MARA, 40450 Shah Alam, Selangor,
MALAYSIA
2
Lecturer, Universiti Teknologi MARA, 40450 Shah Alam, Selangor, MALAYSIA
*Correspondence Author: Noraini Anor, Malaysia Institute of Transport (MITRANS), Universiti Teknologi MARA,
Malaysia. Tel: +603 55442638, Fax: +60355442344. E-mail: ainianor@yahoo.com
433
INTRODUCTION
In todays rapidly changing world, the road transportation network system has become very
crucial to boost economic growth, physical transformation, and the emergence of an outward-looking,
pluralistic population linked to a transportation network system of maritime connections engendered by
international trade. As mentioned by Taaffe (1973), geographers considering a road transportation
network system as an aspect of spatial organization will first look specifically at the structure of the
network. The basic structural elements of the network are its linkages and nodes. According to Lambert
(1997), traffic and transportation are a key logistics activity that provide for the movement of materials
and goods from point of origin to point of consumption, and perhaps to their ultimate point of disposal as
well. Transportation involves several selections of modes (e.g., air, rail, water, truck, or pipeline), the
routing of shipments, assuring compliance with regulations in the region of the country where shipment is
occurring, and selection of the carrier.
LITERATURE REVIEW
In the port, transportation is the main activities since it plays a role to enable transfer of goods
from ships to other transportation modes and vice versa. Other port related activities such as wholesale,
warehousing, logistics and transport are attracted to the direct proximity of ports. Over the years, the roles
of port have been expanded. Consequently the relationships between port and city changed considerably
during the second half of the twentieth century (Levinson, 2006). As a result, ports have become
increasingly disconnected from cities. The increased intensity of port industrial activity, in combination
with urban growth, lack of available land for further expansion, and environmental constraints have led to
the move of port facilities away from city centres (Hoyle, 1989). Thus, transport players are willing to
insert the port cities within the global transport chain to cope with normalised logistics systems which are
managed by an ever-reducing number of powerful global companies, local and regional specifically in
terms of economic development and spatial planning (Ducruet, 2007).
Therefore, port city have been seen as a vital component to generate the global economy, and also
increasingly important concentrations of population and asset value. Thirteen out of the twenty most
populated cities in the world in year 2005 were port cities. In addition, their economic importance in
terms of international trade has grown markedly, particularly in developing countries, in line with
globalisation and the rapid development of the newly industrialised countries (Morlot, 2007). The general
level of port development and urban growth is a reflection of the land situation, which includes the nature
and extent of economic development in the tributary area and the character of the transport network. Most
of the land site, whether artificially created or largely due to natural growth, constitutes a direct control
over the pattern of urban growth and often determines the general layout of a port city. Decisions
concerning port expansion have frequently entailed important consequences for the location, character,
and comparative development of urban areas (Hoyle, 1981).
In this study, the term transportation network refers specifically to road transportation. Road
transportation network system is the most important infrastructure in any country as it provides the means
to move people and goods. A road transportation network system manifests properties such as the
hierarchy of roads, connectivity, accessibility, linkages and movement. On the other hand, the road
network system also influences port city to grow depending on the functions of the effective integration of
land use and road network system. A routeway concept is defined by Becth (1971) as a way or road for
passage or travel. A highway, or railroad, or pipeline, or waterway, or airway is, by definition, a route
between any two points located on it.
434
As mentioned earlier, the road transportation network system in port city is very important to
achieve sustainable development. Port cities are places where the mode of transportation changes from
land to water-borne systems; it is primarily important as a central place for economic and cultural
interchange. According to Banga (1992), port cities are centres of exchange where different cultures and
different environments meet, at the boundary between land and sea.
Transport integration is an essential port function of a modern seaport node in which a

multimodal transport system develops. As stated by Hoyle (1980), the growth of a port city is
fundamentally affected by four sets of separate factors: the land situation, the water situation, the land
site, and the water site. The balance between these factors influences port city growth which depends on
the functions of the effective integration of land use and the road network system. For that reason, the
integration of land use and a transportation network system deserves particular attention and research.
From here can be seen, the importance of a road transportation network system in port city to generate the
logistics industry and supply chain.
To create sustainable development, green logistics must be implemented to improve the quality of
environment and to improve a road transportation network system be more efficient and effective. Green
Logistics is a modern transportation system which may improve the cost, efficiency and reliability of
freight and passenger transport systems with environmental concerns. In other words, it refers to an
environmentally-friendly and efficient transport and distribution system (Rodrigue, 2001). Green
logistics adaptation in Malaysia is sustainable logistics to achieve a more sustainable balance between the
economic, environmental and social elements related to the transportation network system. The
implementation of green logistics in port cities to achieve sustainable development in Malaysia is still in
the early stages. The current transportation network system does not connect well to the port and its
hinterland and therefore does not meet the requirement of green logistics in Malaysia. Therefore, the
study has looked into the role of a road transportation network system to generate the development of
green logistics in the port city of Port Klang.
OVERVIEW OF PORT KLANG
In this paper, we examine the effectiveness of a road transportation network system in a port city,
specifically Port Klang, Malaysia. We have narrowed down the scope to focus only on the effectiveness
of a road transportation network, since it is the main transportation network system in the Port Klang area.
Port Klang is Malaysias major gateway and has been recognized as the busiest port in Asia. It is situated
on the west coast of Peninsular Malaysia, near national capital city, Kuala Lumpur, at the north end of the
Straits of Malacca. It is about 6 kilometres southwest of Klang town and 38 kilometres southwest of
Kuala Lumpur. This port has been well-linked to the others part of Malaysia by road and rail network
connections, and is located about 70 kilometres from Kuala Lumpur International Airport (KLIA).
Furthermore, Port Klang is served by three major gateways which are Northport, Westport and Southport
(Southpoint). Figure 1 shows the location of Port Klang.
435
Figure 1: Study Area

Source : www.gettingaround.net/pages/poc-...ysia.php
436
Northport
Westport
Southport
Figure 2: Location of Port Klang

Source: www.pka.gov.my/Background.asp
Port Klang has sufficient capacity to go further in the years ahead, like others port cities in the
world. Port Klang has been selected as the area of interest in this study based on several criteria, including
urban population (Brinkhoff, 2006), freight forwarders and logistics agents (International Transport
Journal, 2006. Several indicators that exist at a global scale but have not been included in the study to
avoid redundant values are ton traffic, total quay, administrative and suburban population and surface of
the metropolitan urbanised area (Ducruet, 2007). Based on these criteria, Port Klang has been identified
as a potential world class port city of the future. The table in Appendix 1, from Ducruet (2007) shows the
concentration of the transport chain is based on logistics activities, port throughputs and infrastructures.
According to Ducruet (2007) Port Klang is ranked 34 of 100 port cities in the world.
437
PROBLEM STATEMENT
Several problems have been identified that limit the effectiveness of a road transportation
network system in meeting the standard of green logistics in Malaysia. The major problems that have
been identified are the connectivity and accessibility of road network system in the study area. Port Klang
is last mile connectivity for freight and cargo movement to transfer the goods because the connectivity
and accessibility of the road network system in Port Klang is not efficient and would create traffic
congestion in Port Klang.
Furthermore, criteria for road design in the study area such as the width of lanes, load bearing
standards and junctions have not been established. In terms of width of lane, the existing narrow widths of
lane in Port Klang are not suitable to cater for the capacity of freight and private vehicle movement to or
from the port city. The size and width of current road lanes cannot accommodate the capacity of vehicles
based on traffic volumes in Port Klang. Data on traffic volume shows that the volume has continued to
increase year by year. The lane width of roads that lead to Port Klang are Federal Route 2 (11.0 metres),
Persiaran Raja Muda Musa 11.0 metres, and Jalan Negeri (B10) (6.0 metres). Meanwhile the Federal
Route 2 to Northport is 7.0 metres road wide. The narrow width of lanes slows down the movement of the
freight and delays freight forwarding, which indirectly increases monetary costs and reduces time.
One of the aspects of road design is their load bearing standard. Currently, existing roads cannot
accommodate the load of goods carried by freight. Most, existing roads can only accommodate container
loads of 20 to 40 tons. Freight transporters passing through the Port Klang road have no choice because
the existing road leading to Port Klang has narrow lane. Furthermore, the physical situation of the main
road leading to Port Klang is not satisfactory. The surface of road is uneven as a result of load bearing
movement of cargo and freight. This problem must be solved immediately to enable the freight to be
moved smoothly and efficiently. In addition, traffic levels on the road exceed its capacity, resulting in
traffic congestion, especially during peak hours.
The physical conditions of junctions in the study area are not satisfactory, to determine the Level
of Service, the ratio is calculated by comparing the traffic volume with the road capacity. Table 1 shows
the level of service of road from highway capacity manual. A ratio over 1.0 (LOS F) indicates that the
road can no longer cater to the traffic volume because there are too many junctions such as junction at
Jalan Negeri B1/ North Klang Straits Highway from Klang to Northport level of services (D) stable flow
to congested flow and Northport to Klang Level of service (E) congested flow. This situation can cause
the delay of freight and private vehicles, and at the same time traffic would increase. Junction at Persiaran
Raja Muda Musa/Jalan Negeri B8 from Klang to Port Klang and Port Klang to Klang level of service (F)
is shows that these junctions have congested flow. This situation creates conflict in the study area and
especially causes problems for transport movement, including a traffic flow bottleneck at Port Klang.
Table 3 : Level of Service of Road

LOS A Free Flow (0.0<0.60)
LOS B Free Flow but close to Stable (0.60-0.69)
Flow
LOS C Stable Flow (0.70-0.79)
LOS D Stable Flow but close to (0.80-0.89)
congested
LOS E Congested (0.90-0.99)
LOS F Very congested ( >1.00)
Source : Highway Capacity Manual,Ministry of Works, (2006)
438
FINDINGS
The transportation network system is an important element for ensuring efficiency in the function
of a port citys development. It acts as an indicator of economic growth provides infrastructure, and it also
influences the development pattern of an area. It plays the role as a connector between geographical
locations on a local, regional and global scale.
a) Network
The current road network system in Port Klang consists of federal road and city roads.
Maintenance of federal roads is the responsibility of Public Works Department through the provision of
the Federal Government. Persiaran Raja Muda Musa and Jalan Kem are part of the alignment of Federal
Road 2 that connects the study area to nearby cities such as Klang, Shah Alam, Petaling Jaya, and Kuala
Lumpur.
b) Hierarchy of Roads
Roads in the study area may be divided into hierarchical namely:
i) Highways
Kuala Lumpur Klang Federal Highway (FR2)
The highway is an existing network that connects Kuala Lumpur city centre with western
areas such as Petaling Jaya and Shah Alam. Before the construction of other highways in the
surrounding area, this road was the main transportation network between Kuala Lumpur and the
Klang region.
New Klang Valley Expressway (NKVE)
The New Klang Valley Expressway was built in the north Klang Valley area and links
the western side of the Klang Valley, including Shah Alam and Klang and surrounding areas to
the east of Kuala Lumpur. The highway has implemented a closed toll system. It is an alternative
route for traffic instead of the Federal Highway traffic using the route Kuala Lumpur Klang
Federal Highway (FR2).
Shah Alam Expressway (KESAS)
Shah Alam Expressway is complementary to the existing road network system in the
Klang Valley and the existing Federal Highway. This highway links the study area to Sungai
Besi.
North South Expressway (PLUS)
The North-South Expressway is the longest expressway in Malaysia, starting at Bukit

Kayu Hitam in the far north and extending to Johor Bahru in the far south of the country. This
highway crosses the Klang Valley region and Selangor state, as well as the eastern part of study
area. It also connects to the Kuala Lumpur International Airport (KLIA) in Sepang and also to the
Sultan Salahuddin Abdul Aziz Shah Airport (LTSAAS) in Subang via the New Klang Valley
Expressway (NKVE).
439
An elevated Highway passing along the northern end of Jalan Kem connects the study
area to the Shah Alam Highway (KESAS) in the south and to the Shahpadu Highway in the north.
The elevated highway is also the main route connecting the Northport and the Westport.
Currently, to access the elevated highway, road users must go through Jalan Kem.
ii) Main Distributor Road
The main distributor roads are the road that connects Port Klang with the outside areas.
This road covers most of the entry and exit road trips, as well as direct travel for long distances.
Persiaran Raja Muda Musa serves as the main distributor road linking the study area to Klang,
Shah Alam, and Kuala Lumpur. It is a four-lane route with a divider in the middle.
Figure 3: Existing of Road Network System

Source : Klang Municipal Council (2002)
440
c) Traffic Management
i) Junctions Control
Elevated junctions, traffic lights junction, and priority junctions are examples of
intersection control or junction control. In the study area, junctions controlled by traffic lights are
5 junctions only.
d) Patterns and Traffic Characteristics
Patterns and characteristics of the traffic from the traffic census have been carried out and the
following section contains information obtained from the traffic volume report issued by the Malaysian
highway planning unit.
Daily Traffic Volume
Daily traffic volumes of the main roads in study area show that the Federal Highway
(FR2) has the highest traffic volume which is 110,600 vehicles per day, followed by the North
Klang Straits Highway with 91,300 vehicles per day. Meanwhile, the state road B10 (Jalan
Pandamaran) has the lowest volume of traffic which is 21, 800 vehicles per day of the roads
included in the study.
Peak Hour Traffic Volume
Peak hour traffic volume refers to the period of highest traffic volume in a day. The
volume of traffic for every hour in a day is different; however, the volume of traffic during peak
hours is used for the assessment of road operations and planning the design of the road. The
volume of traffic during peak hours is around 9 percent to 11 percent of total traffic volume for a
day. Federal Highway 2 (Klang-Kuala Lumpur) registers higher traffic from peak with 8,600
vehicles per hour, while the B10 road (Pandamaran-Sentosa) has the lower peak traffic volume
with flow 1,700 vehicles per hour. North Klang Straits Highway also has a high peak traffic
volume of 7,800 vehicles per hour. Other segments recorded peak a volume of between 3,000 to
6,000 vehicles per hour (Klang Municipal Council, 2002).
e) Operational Evaluation of Existing Roads
Operational assessments of the existing roads are based on the comparison between volumes of
traffic during peak hours and the capacity of the existing road. The capacity of an existing road is
determined by the width of road lanes, road width, traffic distributions, the composition of traffic,
obstructions in the road side and the slope of the road. Table 1 shows the operation of main roads in the
study area. Federal Highway 2 (Klang-Shah alam) operates more than 54% above the capacity of existing
roads. Similarly, North Klang Straits Expressway operates more than 40% above the capacity of existing
roads, while Federal Route 2 (Port Klang-Northport) has reached the capacity of existing roads.
441
Table 2: Main Road Operation in Study Area
Peak Hours Existing Road The Ratio of

Road Segment Volume Capacity Volume/Capacity
(Vehicle/Hour) (Vehicle/Hour)
Federal Road 2 (FR2) 8,600 5,600 1.54
(Klang-Shah Alam)
Federal Route 2 3,300 5,600 0.59
(Persiaran Raja Muda Musa)
Federal Route 2 5,600 5,600 1.00
(Port Klang-Northport)
North Klang Straits 7,800 5,600 1.40
Highway
(Klang-Northport)
Source: Klang Municipal Council (2002)
f) Road Accidents
Road accident statistics were obtained from the Royal Malaysian Police Traffic Branch Klang
District, Selangor. Accident statistics shows the accident rates in 2001 as compared to the year 2000
declined to a total of 1,303 cases from 2,176 cases in the same period. From 1998 to 2001 for the Klang
district statistical data, the accident or potential road accidents occurred more frequently on the federal
roads. Among the factors that caused the accident was the composition of heavy vehicles, such as high-lift
on the main streets, especially on Federal Route 2 to or from Port Klang.
RECOMMENDATION AND CONCLUSION
This paper has focused on infrastructure development in a port city; it is undeniable that the
planning of infrastructure development should be embodied in sustainable planning. Thus, for the
infrastructure transport, each country seeks to promote a system of sustainable transport which can reduce
energy consumption, air pollution, minimize traffic congestion, reduce accidents, and increase the use and
diversification in order to ensure consumer safety (United Nations, 1992). Therefore the effectiveness of a
road transportation network system in port city must be increased in order to achieve sustainable
development, specifically towards achieving green logistics.
The road network system is an essential requirement for infrastructure development that is
necessary to create sustainable land transport and to enhance economic activities and services, thereby
helping to improve the quality of life. Thus there is a need to develop an efficient road transportation
network system which can cater the needs of population growth and cover the population area, business
and industries. Furthermore, for infrastructure development, planning and development of highways and
roads, it is important to take into account the needs of local communities and it also requires a strategy in
order to ensure that the design of highways and roads make use of technology and resources in more
efficient manner.
Planning and development of land transport and the port system need to prioritize conservation of
energy and resources. Even though land use in the port area is still relatively low density, it is important
to take into account the needs of local communities, as several elements that lead to sustainable
development must be implemented for future development and future generations. In order to encourage
442
the design of surface transportation systems that are environmentally friendly and prioritize the safety of
consumers, public facilities must be efficient and meet the needs of local and global community as well as
support economic and social development.
The successful planning of a road transportation network system in port city planning must
consider the demands of freight movement such as a pickup and delivery, intermodal connections, and
regional consolidation in the context of urban passenger transportation issues such as congestion and
accessibility. To plan for an efficient road network system, the economic drivers of freight elements,
which are the individual practices of shippers, carriers, and the public in shaping the need for freight
movement, must be taken into account.
Transportation network systems, ports and urban growth are changing rapidly due to the
efficiency of road network systems. The close associations between ports and cities are vital for the urban
growth and an industrial expansion plan need an efficient transportation system. Planning for road
network system in port city that can achieve green logistics should be investigated further and be revised
in ways that enhance environments of living and working in maritime zones.
REFERENCES
Banga, I. (1992).Ports and their hinterlands in india 1700-1950, New Delhi, Manohar.
Banister, D. (1995). Transport and urban development. Oxford, Alexandrine Press.
Basu, D.K. (1985). The rise and growth of the colonial port cities in Asia. Lanham,University of
California.
Becth, J. E. (1971). A geography of transportation and business logistics. Wm.C.Brown.
Hoyle, B., R. K. (1998). Modern Transport Geography. Chichester: John Wiley & Sons.
Bobzin, H. (2006). Principles of Network Economics, Berlin Heidelberg: Springer-Verlag.
Coyle, J.J., Bardi, E.J. & Langley, C.J. (2003). The Management of Business Logistics: A supply chain
perspective, Cincinnati, OH: Thomson Learning.
Coyle, J.J., E. J. (2000). Transportation. Ohio: South-Western College Publishing.
Department of Public Works. A Guide On Geometric Design of Roads: Arahan Teknik Jalan. Kuala
Lumpur
Lambert, D.M., J. R. (1997). Fundamentals of Logistics Management. Irwin McGraw-Hill.
Ducruet, C. (n.d.) (2007). A metageography of port-city relationships. Ports, cities, and global supply
chains, Ashgate (Ed.) 157-172.
Taaffe, E.J., H. L. (1973). Geography of transportation. London: Prentice-Hall.
Frain, J. (1970). Transportation and Distribution for European Markets. London: Butterworth&Co.Ltd.
Fujita, M. & Mori, T (1996), The role of ports in the making of major cities: self-agglomeration and hub-
effect. Journal of Development Economics, vol.49 (1). pp.93-120.
443
Hall, R. W. (1987). Consolidation Strategy: Inventory, Vehicles and Terminals. Journal of Business
Logistics, 8, 57-73.
Hoyle, B.S. (1989), The port-city interface: trends, problems, and examples, Geoforum, vol.20, pp. 429-
435.
Johnson, S. (1993). The Earth Summit: The United Nations Conference on Environment and Development
(UNCED). Graham & Trotman, London.
Ortuzar, J.d.D, L. G. (1994). Modelling Transport. London: John Wiley & Sons Ltd.
Kerajaan Negeri Selangor (1999). Strategi Pembangunan Mampan dan Agenda 21 Selangor: Kawasan
Sensitif Alam Sekitar. LESTARI, Universiti Kebangsaan Malaysia, Bangi.
Klang Municipal Council (2002). Study of Klang Local Plan Technical Report 2002-2015.
Lowe, J. C. (1975). The Geography of Movement. Washington: Houghton Mifflin Company.
Lumsden, K.R. (1998) Logistikens Grunder, Studentlitteratur, Lund.
Pinder, h. &. (1981). CITYPORT INDUSTRIALIZATION AND REGIONAL DEVELOPMENT: Spatial

Analysis and Planning Strategies. BRITAIN: PERGAMON PRESS.
Sadler, I. (2007). Logistics and Supply Chain Integration. Britain: Athenaeum Press, Gateshead, Tyne &
Wear.
Tan, T.-Y. (2007). Port Cities and hinterlands: A Comparative study of Singapore and Calcutta. Journal
Of Political Geography, 851-865.
Thalmann, P. (2004). The Dynamics of Freight Transport Development A UK and Swiss Comparison.
Great Britain: MPG Books Ltd, Bodmin, Cornwall.
Thomas, I. (2002). Transportation Networks and the Optimal Location of Human Activities.
Massachusetts: Edward Elgar Publishing Limited.
Jean-Paul Rodrigue, B. S. (2001). Green Logistics (The Paradoxes of).
Town and Country Planning Peninsular Malaysia. Klang Port Cities Action Area Plan 2015
444
ECONOMIC IMPACTS OF THE SHIFTING OF FUEL CONSUMPTION IN INDONESIA
Heru Sutomo 1* and Joewono Soemardjito 2*
ABSTRACT:
The growth of motorized vehicles in Indonesia after 2000 increased sharply at more than 15% annually.
Meanwhile, the crude oil production from 1990 to 2010 tends to decrease on an average of 5.8% per year.
These conditions give important implications on fuel consumption for transportation sector which
increased about 6% annually during 1990-2005. Further, the increasing fuel consumption will give rise to
the emissions that are of potential hazard to human health. This report presents an investigation using
empirical approach through simulations of two conditions. The first condition is the baseline which
illustrates the composition of gasoline and diesel consumption i.e. 20%:80% with EURO II emission
standard. While the second condition is based on projected scenario that illustrates composition of
gasoline and diesel consumption i.e. 80%:20% with the EURO IV emission standards after 2014. The
simulation shows that the scenario provides substantial economic benefits from fuel consumption, social
costs, and global emissions. Carbon dioxide (CO 2 ) emission is reduced by 2,874 million tons during 2014
to 2030 and total economic benefit is 1.73% of Indonesias gross domestic product (GDP) 2009 (current
value). Further, there is also saving in governments budget for fuel subsidy amounting 96,933 million
USD in 17 years. The economic impacts will be much higher when EURO IV is implemented earlier than
2014.
Keywords: transportation sector emissions, fuel consumption, green diesel, economic impacts.
1
Lecturer, Civil Engineering Department, Universitas Gadjah Mada, Bulaksumur, 55281 Yogyakarta, INDONESIA
2
Researcher, Center for Transport and Logistics Studies, Universitas Gadjah Mada, Bulaksumur E-9, 55281, Yogyakarta,
INDONESIA
*Correspondence Author: Heru Sutomo, Universitas Gadjah Mada, Bulaksumur, 55281 Yogyakarta, INDONESIA,
Tel: +62274556928, Fax: +622746491076. E-mail: hsutomo@mstt.ugm.ac.id
445
INTRODUCTION
In todays modern age, air pollution is caused by the consumption of energy sources such as oil,
gas, and coal. As we consume energy, we also emit high amount of pollutants into the ambient air. CO 2 is
the most common greenhouse gas (GHG). In Indonesia, the production of CO 2 until 2007 was still
dominated by the electricity sector, followed by the transportation sector. CO 2 emissions from power
plants in 2007 were estimated at 90 million tons, while CO 2 emissions from the transportation sector were
estimated at 70 million tons (State Ministry of Environment of Republic of Indonesia, 2007).
Transportation is the biggest and most rapidly growing primary energy consuming sector in
Indonesia. Approximately 48% of Indonesian primary energy is consumed by the transportation sector,
with an estimated of 67 million tons of CO 2 (National Development Planning Agency, 2009). Currently,
the large energy usage by transportation sector means that this sector produces the highest concentration
of (air) pollution, especially urban areas which have a high number of vehicles. Emission of local
pollutants from road transport is growing at an annual average rate of 8-12%.
The rapidly growing vehicle population gives impact to higher mobility, especially in urban areas
in Indonesia. Moreover, with the increased of economy growth, the number of private vehicles is
increasing rapidly. Data from Ministry of Transportation (2010) stated that in 2009 there were about 81
million vehicles in Indonesia. Out of 27% of total motorized vehicles are passenger cars (14%), trucks
(5%), and buses (8%). The rest are motorcycles. From 1990-1999, the growth of passenger cars is lower
than 10% annually, as well as trucks and buses. This figure increased to more than 15% annually after
2000.
High growth in motor vehicle gives a consequence of increased fuel consumption. Transport
sector fuel demand increased about 6% annually during 1990-2005. During such period, the portion of
fuel use by type is 40% diesel fuel, 49% petrol, and 11% others (Department of Energy and Mineral
Resources, 2006). The increase in transportation sector fuel consumption is expected to affect the GHG
emissions level. If no action taken, the impact on public health could be very serious.
In the long term policy, as stated in Indonesia Climate Change Sectoral Roadmap/ICCR (National
Development Planning Agency, 2009), government states that mitigation of transport emissions is
needed. Strategy to achieve the goal of sustainable development in transport sector is improving the
energy and carbon efficiency of motorized vehicles through technological improvements on engine and
fuels. Regarding to the government policy, this research develops a conceptual framework to identify the
advantage of implementation of the green diesel technology in passenger cars in Indonesia in contributing
to fuel efficiency improvement and creating the lower emission transport in Indonesia and at the end is to
achieve a sustainable transportation system. This research presents economic benefits from
implementation of such technology due to the CO 2 emission reduction and its impact to social cost, fuel
consumption, and global warming damage.
LITERATURE REVIEW
Efficiency on Vehicle Technology and Fuel Consumption
The diesel engine is the most fuel efficient of all internal combustion engines. Along with great
reliability and durability, these are some of the well-established reasons why so many European
consumers choose diesel engines over petrol (gasoline). The popularity of diesel passenger cars varies
from region to region around the world, largely in function of local economic factors, with diesel
acceptance highest in Europe. In the heavy-duty road transport sector, however, diesel is the engine of
choice throughout the world, due to its greater torque and efficiency. In Indonesia, the popularity of diesel
446
passenger cars is way below its competitor, the gasoline engine. However, research on such engine
funded by the US Department of Energy, under its Light Truck Clean Diesel Engine Program proves the
potential role of diesel engine to be a leader in fuel efficiency. A clean diesel engine market penetration
trajectory is developed, representing an industry response to meet the Corporate Average Fuel Economy
(CAFE) standards. Research indicates that the conventional gasoline engine's efficiency is in the range of
27.31%, while the current turbocharged diesel engine's efficiency is 44% (Teotia et al., 1999).
According to the U.S. Environmental Protection Agency (EPA, 2005), the transportation sector is
responsible for just over one-quarter of total domestic GHG emissions. Almost all of these CO 2 emissions
come from the consumption of petroleum products: gasoline (60%), middle distillates (diesel fuel 22%),
jet fuel (12%) and residual oil (mostly marine 3.1%). When viewed by mode of travel, 62% of U.S.
transportation-related GHG emissions came from light-duty vehicles used for personal transport:
(1) passenger cars 35%;
(2) light-duty trucks 27% (including SUVs, minivans and pickup trucks); and
(3) motorcycles less than 1%.
The level of fuel efficiency in the use of types in terms of vehicle mileage for different vehicle
types have varying characteristics, as described in Table 1. For passenger cars, utility passenger and
freight, and pick-ups, the use of diesel is more efficient.
Table 1: Assumption of Fuel Economy
EURO II EURO IV
Type of Vehicle Type of Fuel Source
liter/km liter/km
Car premium gasoline 0.11 0.078
Utility Freight gasoline 0.11 0.078 Analysis from US Environment Impact
Agency, 2007:
Utility Passenger gasoline 0.11 0.078 http://www.epa.gov/orcdizux/rfgecon.htm
Pick up gasoline 0.11 0.078
Car diesel diesel 0.09 0.0625
Utility Freight diesel 0.09 0.0625 What Green Car?
Utility Passenger diesel 0.09 0.0625 http://www.whatgreencar.com/petdiesel.php
Pick up diesel 0.09 0.0625
Bus diesel 0.33 0.256 http://www.iges.or.jp/en/ue/activity/mega-
Truck/Trailer diesel 0.25 0.192 city/article/htm/far43.htm
GHG Emission from Transport Sector
In addition to legislated emissions, an additional non-legislative target has been agreed between
the European Community and l'Association des Constructeurs Europens d'Automobiles (ACEA). The
association aims to achieve 140 g/km average CO 2 emissions for all new passenger cars sold in 2008.
Also a further improvement to 120 g/km CO 2 is promised by 2012 (EU Commission Environment, 2004).
Figure 1 shows the results published in a joint report by ACEA and the EU Commission in October 2004.
From 1995 to 2001, there was a progressive improvement in CO 2 . This was achieved by improvements in
both gasoline and diesel technology and increases in diesel engine penetration.
447
Source: European Commission Environment (2004)
Figure 1. EU Trends of ACEA Members Fleet, Average Specific Emissions of CO 2
The GHG emission can be calculated by multiplying the fuel consumption with CO 2 emission
standard of Intergovernmental Panel on Climate Change (IPPC, 1996) as followed.
Table 2. Assumption of CO 2 Coefficient

Emission Specification Standard (ton/liter)
Type of Vehicles Type of Fuel
EURO II EURO IV
Gasoline Passenger Cars Gasoline 0.0023 0.00161
Pick up Gasoline 0.0023 0.00161
Diesel Passenger Cars diesel 0.0027 0.00189
Pick up diesel 0.0027 0.00189
Bus diesel 0.0027 0.00189
Truck/Trailer diesel 0.0027 0.00189
Source: IPCC, 1996
Table 2 explains that one liter of gasoline with EURO II emission standard will emit 0.0023 ton
CO 2 , and with EURO IV will emit 0.0161 ton CO 2 . Whereas, diesel fuel with EURO II emission
standard will emit 0.0027 ton CO 2 , and with EURO IV emission standard will emit 0.0189 ton CO 2 . This
means that one liter diesel emit higher CO 2 than gasoline, although 1 liter diesel achieves higher km
travelled.
The Impact of Global and Local Environment
There are two benefits derived from the use the better quality of fuel: global and local benefits.
Global benefits are derived from the reduction of GHG emission resulted from fuel savings and fuel
switching from petrol to green diesel. The reductions of GHG emission are expected can reduce global
warming that lead to the climate change. Meanwhile, local benefits are savings from health cost and
productivity lost derived from the reduction of local pollution (PM10 and Pb). Air pollution at local level
due to transportation activities will cause negative impact both to human health and environment.
Achmadi (1981) stated that respiratory illness claims the sixth cause of death after traffic accident,
diarrhea, cardiovascular illness, tuberculosis, and measles.
Social impacts of pollutants from mobile sources are a key element in urban design and traffic
planning. One of the most relevant impacts is health effects associated with high pollution periods
(Monzon and Guerrer, 2004). Ostro et al. (1999) researched that a 10 g/m3 change in daily PM10 was
448
associated with an approximately 1% increase in mortality, which is of similar magnitude to particle-

associated impacts identified in urban areas.
As an illustration, the study of Asian Development Bank (ADB) in 2002 estimated the economic
impacts of Jakartas air pollution (as the biggest and a metropolitan city in Indonesia) to be Rp1.8 trillion
(about 200,000 million USD), projected to reach Rp4.3 trillion by 2015. The total cost of health problems
associated with PM10, NO 2 and SO 2 for the whole of Jakarta in 1998 is approximately 181 million US$.
This is approximately only 1 percent of Jakartas GDP, however it is as much as approximately 100
percent of the Jakarta Governments total revenue for 1998. Those numbers will be increased
approximately 2.4 times in 2015.
METHODOLOGY
Scenario Development of Fuel Consumption
The scenario development aimed to identify the economic impact of fuel usage based on EURO
emission standard compared with the baseline condition. The baseline condition is based on the fuel use
comparison between gasoline and diesel with the proportion of fuel consumption 20%:80% under the
EURO II and EURO IV emission standards. The projected scenarios development in this research is
based on the use of fuel-based EURO II emission standard in the base year until the year 2014, with the
proportion of fuel use in gasoline and diesel is 20%: 80%, and experienced a shift in fuel use based on
EURO IV emission standards until the year 2030 with the same proportion of fuel use.
Baseline scenario Projected scenario
Figure 2. Scenario Development
From each scenario, it will be calculated to indicate the economic benefits derived from the use of
fuel during the period of 17 years (2014-2030), in terms of its impact on social costs, fuel use, and
environmental damage (the effect of global warming).
449
Valuation of Fuel Consumption and CO 2 Emission
Based on the km travel, fuel consumption is calculated using the following formula:
FCbaseline, project = Mileage x K ......................(1)
where:
FC = daily fuel consumption (litre)

Mileage = km travel of vehicles (km) see Table 3.
K = kilometer per litre of vehicles (km/litre) see Table 1.
K is fuel economy with the assumptions as shown in Table 1.
Average annual travel is used to calculate km travel from each vehicles type. For bus for
instance the average annual travel is 36,000 km with assumption the operational day is 15 days per month
with daily travel of 200 km (Netherland Government - SSN Project, 2002).
Table 3. Assumption of Average Km Travel Annually
Number of Days per month Average Km Travel per Day Average Km Travel Annually
Type of Vehicles
(day) (km) (km)
Car 28 55 18,480
Utility Freight 25 80 24,000
Utility Passenger 25 200 60,000
Pick Up 28 55 18,480
Bus 15 200 36,000
Truck 20 150 36,000
Source: SSN Project, 2002
Based on the above assumption and formula, valuation of fuel consumption can be estimated by
multiplying the fuel consumption with fuel price. For non subsidy gasoline and diesel fuel is based on
The Decree of Director of Marketing and Sales of PT Pertamina (Persero) No. Kpts - 638/F00000/2007-
S0 on Pertamina Oil Fuel Economy Sales Price dated 29 August 2007 which is Rp 5877 (gasoline) and
Rp 6689 (diesel fuel). This price is used to calculate fuel price for EURO II standard. Meanwhile for
EURO IV standard is using current fuel price in Europe (www.theaa.com) which is Rp 16,300 (gasoline)
and Rp 14,000 (diesel fuel). But the price is assumed tax included, therefore this study is using 60% of
the current price i.e Rp 9,780 for gasoline and Rp 8,400 for diesel fuel. The current price is then estimated
for the next year price until 2030 with using inflation rate of 6% (year inflation in June 2007 is 5.77%,
Coordinating Ministry of Economy Affairs).
After GHG reduction is calculated, the number is valuated into monetary value using Delucchis
equation (Delucchi, 1998), which states that 1 ton decrease of eCO 2 in the atmosphere is equivalent to
USD 10. This amount is called Global Warming Damages (GWD). But in this research the value of 1 ton
of eCO 2 referred to the average carbon price, i.e. USD 5-6 per ton (Pelangi, 2006).
Measuring the Economic Benefit
Economic benefit from the improved environment condition expected to be able to indicate a
potential savings from the fuel use. In this research, economic benefit emphasized on the followings:
450
(1) Global warming damage (GWD) that is based on the volume of CO 2 (tons) which is reduced;
(2) fuel consumption that is based on the value of fuel subsidies might be saved by the Government; and
(3) social cost that is based on the saving of health costs due to GHG emission reduction which is
potential to the decline of public health.
Total savings in these costs above are an indication of national economic benefits as a result of fuel
switching from gasoline to diesel with the specifications EURO IV emission standards. Method of
measuring economic benefits is shown in Figure 3.
Identification of all potential environment

impact due to transport activities
Global The Nature Local

of the impact Review of similar
cases in other places
No
Green House Impact to human health
Gases (eCO2) (Parameter PM10 and Pb)
Any
ACEA Standard available
Yes
data ?
GHG calculation health cost for medical
baseline condition treatment.........(C)
GHG calculation Production Lost = non working

with project days x GNP/capita/day (D) Its estimated that the
GWD Formula
percentage of GHG
ton eCO2 equivalent to
reduction equivalent to
US$ 5/ (B)
the percentage of
Number of cases and patients PM10 & Pb reduction
GHG Reduction ( ton).....(A) reduction due to project
implementation.........(F)
Valuation of GHG Reduction

= (A) ton X (B) USD Cost Saving due to project implementation:
(C + D) x (F)
Figure 3. Method of measuring economic benefit
RESULT ANALYSIS
Measuring and Valuation of Fuel Consumption
Analysis conducted on data on the number of vehicles and total travel of the vehicle (Veh-km) in
the base year 2005, and then make forecasts up to 2030 with growth scenario of 5%, 7% and 9% (data can
be found in Appendices). With data based on these forecasts, the next estimate of fuel consumption by
type of vehicle and type of fuel used technology. The result of the calculation described below.
451
Resume of data on fuel consumption at baseline and projected scenario conditions is described in
Table 4. Valuations of fuel consumption at baseline and projected scenario conditions are described in
Table 5.
Table 4. Fuel Consumption at Baseline and Projected Scenario (Million Liters)

Type of 2014 2015 2020 2025 2030
Vehicles Fuel Baseline Projected Baseline Projected Baseline Projected Baseline Projected Baseline Projected
Total gasoline 42,494 28,505 45,016 28,947 60,416 31,642 89,042 36,652 132,679 44,288
Total diesel 47,421 36,883 48,510 38,750 54,654 49,759 64,830 69,265 78,483 97,569
Total All 89,915 65,388 93,526 67,697 115,070 81,401 153,872 105,917 211,162 141,857
In 2014, there will be total fuels savings amounted to 24,527 million liters or 27.3% and increase
to 69,305 million liters or 44.5% in 2030.
Table 5. Valuation of Fuel Consumption (2014-2030)

Condition Fuel Consumption in Saving in Million liters Benefit for Government Benefit for Public
Million liters (2014 2030) (from subsidy in USD Million) (in USD Million)
(2014 2030) Gasoline Diesel Gasoline Diesel
Gasoline 1,308,714
Baseline Diesel 1,022,517 38,965,545,490
Total 2,331,231
721,510.31 (10,124.62) 99,352 (2,419)
Gasoline 587,203
Project Diesel 1,032,642 38,411,582,487
total 1,619,845
TOTAL 711,386 96,933 55,396
Current Value (USD Million) 152,329
Present Value (USD Million) 113,372
Note: Present value is calculated with annual interest of USD at 4%
Table 5 explains that in 17 years, fuel consumption saving is about 711.4 billion liters. The
economic benefit from government subsidy that can be saved is about 96.9 billion USD. There is also
public benefit from the fuel price which is 55.4 billion USD. Therefore the total economic benefit of the
fuel consumption is 152.3 billion USD (current value) or 113.4 billion USD (present value).
Based on Table 5, the global warming damage (GWD) can be calculated, as shown in Table 6. In
17 years, the reduction of CO 2 emission is about 2.87 billion ton or 169 million ton annually. This is
equal to the economic benefit at 14.4 billion USD (current value) or around 10.7 billion USD (present
value).
Table 6. Global Warming Damage

Economic Benefit from CO 2
CO 2 Emission in Million Ton Saving
Condition (Million USD)
(2014 2030) (Million Ton)
Current Value Present Value
Gasoline 3,010
Baseline Diesel 2,761
Total 5,771
2,874 14,369 10,694
Gasoline 945
Project Diesel 1,952
Total 2,897
Note: Present price is calculated with annual interest of USD at 4%
Economic benefit of fuel use on the basis of projected scenario condition can be calculated as
shown in Table 7 and Figure 3. The results of the simulation show that the projected scenario gives high
452
economic benefit from the saving of fuel consumption, social cost and CO 2 emission. Total economy
benefits are up to 2.97% of Indonesias GDP 2006 at current value or equal to 2.21% of Indonesias GDP
2006 at present value. This value is equivalent to 1.73% of Indonesia's GDP 2009 at current value or
1.29% Indonesia's GDP 2009 at present value.
Table 7. Economic Benefit of Projected Scenario Condition (2014 2030)

Criteria Current Value (Million USD) Present Value (Million USD)
Global Warming Damage (GWD) 14,369 10,694
Fuel Consumption 152,329 113,372
Social Cost 1,911 1,422
TOTAL (2014 2030) 14,369 10,694
Average Annually 9,918 7,381
% from 2006 GDP 2.97% 2.21%
% from 2009 GDP 1.73% 1.29%
Note: 1. Indonesias GDP 2006 = 333.82 billion USD; Indonesias GDP 2009 = 572.79 billion USD
2. Present value is calculated with annual interest of USD at 4%
Figure 3. CO 2 Emission Reduction at Scenario Condition (million ton)
CONCLUDING REMARKS
The results showed that the shift in the use of diesel fuel with EURO IV emission standards, with
characteristics that are more environmentally friendly, will provide significant economic benefits in terms
of government subsidy savings, reduced use of fossil-based fuels, and reducing public health costs. To
address the goal of air quality improvement, regulatory reform is needed on fuel specification and policy
to support market mechanism; The regulation concerning the specification of fuel should adopting the
standard of Euro IV and therefore this type of fuel can be competitive in Indonesias market. The second,
is the issuance of policies promoting market mechanism, that is by giving facilities for the players in
green technology based industries (automotive and oil fuel industries).
Until now, the government has developed a series of regulations and policies in order to achieve
sustainable transportation development and the green economic development, but there are still other
policies and regulations needed to be formulated and implemented. In 2004, through the issuance of Law
453
Number 17, the government ratified Kyoto Protocol as an agreement of the countries in the world to
reduce the global emission (CO 2 ). Indonesia is actively involved in the conference of the countries
ratifying Kyoto Protocol or Conference of the Parties (COP). Indonesia has also established the National
Commission for Clean Development Mechanism acted as the Designated National Authority or the
national institution evaluating whether a program has supported sustainable development in its countries
or not and then it can be listed in the CDM projects. The result of this research is expected to become
inputs for the Government in an effort to accelerate the realization of the implementation of the ICCR in
Indonesia.
Appropriate financial framework related to fuel subsidy and investment must be formulated. The
use of fuel for green fuel technology such as diesel gives the implication on the revision of the
governments budget for fuel subsidy. In order to address such a problem, the financial framework must
be carefully formulated. The second is the funding of investment; how to create attractive business
climate for private sectors, such as giving fiscal incentives to companies promoting green technology.
REFERENCES
Achmadi, U.F. 1981. The Impact of Pollutant Pb (Led) for Preliminary School Students in Jakarta.
Faculty of Public Health, University of Indonesia.
Andres Monzn and Mara-Jos Guerrer. 2004. Valuation of social and health effects of transport-related
air pollution in Madrid (Spain). Science of the Total Environment. Volumes 334-335. Pages 427-434.
Arvind Teotia, Anant Vyas, Rolando Cuenca, Frank Stodolsky, James Eberhardt. 1999. CAFE
Compliance By Light Trucks: Economic Impacts of Clean Diesel Engine. Energy Policy 27. Pages
889-900.
Asian Development Bank. 2002. Study on Air Quality in Jakarta, Indonesia: Future Trends, Health
Impacts, Economic Value and Policy Options. Technical Report.
Bart D. Ostro, Susan Hurley and Michael J. Lipsettc. 1999. Air Pollution and Daily Mortality in the
Coachella Valley, California: A Study of PM10 Dominated by Coarse Particles. Environmental
Research. Volume 81. Issue 3. Pages 231-238.
Delucchi. 1998. Marikina Bikeway Project. Project Brief. page 2a-9.
Department of Energy and Mineral Resources. 2006. Indonesia Energy Economy Statistic.
EU Commission Environment. 2004. Reducing CO 2 emissions from light-duty vehicles. Retrieved from
http://ec.europa.eu/environment/air/transport/co2/co2_home.htm
National Development Planning Agency. 2009. Indonesia Climate Change Sectoral Roadmap (ICCSR),
Synthesis Report.
Netherland Government in associated with PELANGI Indonesia (NGO). 2002. The Green House Gases
Emission Reduction Program for Urban Buses in Yogyakarta (South South North Project).
Pelangi. 2006. Satu Tahun Ratifikasi, Gaung Protokol Kyoto Tak Terdengar. Retrieved from
http://www.pelangi.or.id/media.php?mid=164
State Ministry of Environment. 2007. Status of Indonesian Environment (SLHI).
The AA Public Affairs Fuel Price Report. 2007. Fuel Price Report August 2007. Retrieved from
www.theaa.com/onlinenews/allaboutcars/fuel/2007/August2007.doc
U.S. Environmental Protection Agency. 2005. Emission Reduction Incentives for Off-Road Diesel
Equipment Used in the Port and Construction Sectors. Final Report.
454
APPENDICES
Data of The Original Number of Vehicles

Type of Vehicles 2001 2002 2003 2004 2005
Passenger Cars* 3,261,807 3,862,579 5,133,746 6,748,762 7,355,154
Buses 687,570 731,990 1,270,020 2,013,176 2,804,198
Trucks** 1,759,747 2,015,347 3,058,218 4,260,889 4,631,544
Total 5,709,124 6,609,916 9,461,984 13,022,827 14,790,896
Noters;
* Passengers cars are consisting of private cars, utility freight, utility passengers both gasoline and diesel fuel
** Trucks are consisting of trucks (diesel fuel) and pick up (diesel and gasoline fuel).
Source: Directorate of Road Transport and Traffic, Directorate General of Land Transportation, 2005
Forecasted Number of Vehicles

Type of Vehicles 2005 2010 2015 2020 2025 2030
Passenger Cars 7,355,154 9,387,247.4 13,166,100.1 18,466,136.5 28,412,440 43,716,061
Buses 2,804,198 2,947,240.3 3,175,014.8 3,420,392.7 3,776,390 4,169,440
Trucks 4,631,544 4,867,799.3 5,244,002.3 5,649,279.8 6,237,261 6,886,441
Total 14,790,896 17,202,287 21,585,117 27,535,809 38,426,091 54,771,941
Source: Directorate of Road Transport and Traffic, Directorate General of Land Transportation, 2005
Additional Vehicles
Type of Vehicles 2005*) 2010 2015 2020 2025 2030
Passenger Cars 606,392 447,012 861,334 1,208,065 2,345,981 3,609,583
Buses 791,022 29,181 46,921 50,548 74,047 81,754
Trucks 370,655 48,196 77,498 83,487 122,299 135,028
Sub Total 1,768,069 524,388 985,753 1,342,100 2,542,327 3,826,365
Notes: number of additional vehicles in 2005 is the difference between the total number of vehicles in 2005 and 2004, and so on for
subsequent years
Source: Directorate of Road Transport and Traffic, Directorate General of Land Transportation, 2005 (recompiled and processed by
PUSTRAL, 2007)
Additional Vehicles Based on Engine Type (Forecasted)

Type of Vehicles 2005 2010 2015 2020 2025 2030
Private Car 166,654 122,852 236,719 332,011 644,743 992,016
Gasoline Utility Freight 194,301 143,232 275,990 387,091 751,704 1,156,590
Passenger
Utility Passenger 124,158 91,525 176,357 247,351 480,338 739,060
Cars
Pick up 287,937 37,440 60,203 64,855 95,006 104,894
Sub Total 1 773,051 395,050 749,270 1,031,307 1,971,791 2,992,561
Private Car 41,663 30,713 59,180 83,003 161,186 248,004
Diesel Utility Freight 48,575 35,808 68,998 96,773 187,926 289,148
Passenger
Utility Passenger 31,040 22,881 44,089 61,838 120,085 184,765
Cars
Pick up 71,984 9,360 15,051 16,214 23,752 26,224
Sub Total 2 193,263 98,762 187,317 257,827 492,948 748,140
Buses (Diesel) 791,022 29,181 46,921 50,548 74,047 81,754
Trucks (Diesel) 10,734 1,396 2,244 2,418 3,542 3,910
Sub Total 3 801,756 30,576 49,166 52,965 77,588 85,664
TOTAL GASOLINE 773,051 395,050 749,270 1,031,307 1,971,791 2,992,561
TOTAL DIESEL 995,018 129,339 236,483 310,792 570,536 833,804
455
Number of Vehicles in Indonesia at baseline condition (Forecasted)

Type of Vehicles 2005 2010 2015 2020 2025 2030
Private Car 2,021,406 2,579,883 3,618,420 5,075,021 7,808,548 12,014,419
Gasoline Utility Freight 2,356,754 3,007,882 4,218,710 5,916,959 9,103,976 14,007,595
Passenger
Utility Passenger 1,505,963 1,922,033 2,695,750 3,780,929 5,817,428 8,950,835
Cars
Pick up 3,597,937 3,781,467 4,073,714 4,388,547 4,845,311 5,349,615
Sub Total 1 9,482,060 11,291,265 14,606,595 19,161,457 27,575,263 40,322,463
Private Car 505,351 644,971 904,605 1,268,755 1,952,137 3,003,605
Diesel Utility Freight 589,189 751,971 1,054,678 1,479,240 2,275,994 3,501,899
Passenger
Utility Passenger 376,491 480,508 673,938 945,232 1,454,357 2,237,709
Cars
Pick up 899,484 945,367 1,018,429 1,097,137 1,211,328 1,337,404
Sub Total 2 2,370,515 2,822,816 3,651,649 4,790,364 6,893,816 10,080,616
Buses (Diesel) 2,804,198 2,947,240 3,175,015 3,420,393 3,776,390 4,169,440
Trucks (Diesel) 134,123 140,965 151,859 163,596 180,623 199,422
Sub Total 3 2,938,321 3,088,205 3,326,874 3,583,988 3,957,013 4,368,862
SUB TOTAL GASOLINE 9,482,060 11,291,265 14,606,595 19,161,457 27,575,263 40,322,463
SUB TOTAL DIESEL 5,308,836 5,911,022 6,978,523 8,374,352 10,850,828 14,449,478
TOTAL ALL 14,790,896 17,202,287 21,585,117 27,535,809 38,426,091 54,771,941
Total Km Travel (in Million Km) (at baseline condition)

Type of Vehicles 2014 2015 2020 2025 2030
Private Car 62,494 66,868 93,786 144,302 222,026
Gasoline Utility Freight 94,625 101,249 142,007 218,495 336,182
Passenger
Utility Passenger 151,164 161,745 226,856 349,046 537,050
Cars
Pick up 74,170 75,282 81,100 89,541 98,861
Sub Total 1 382,452 405,145 543,750 801,384 1,194,120
Private Car 15,623 16,717 23,447 36,075 55,507
Diesel Utility Freight 23,656 25,312 35,502 54,624 84,046
Passenger
Cars
Pick up 18,542 18,821 20,275 22,385 24,715
Sub Total 2 95,613 101,286 135,937 200,346 298,530
Buses (Diesel) 112,611 114,301 123,134 135,950 150,100
Trucks (Diesel) 5,386 5,467 5,889 6,502 7,179
Sub Total 3 117,998 119,767 129,024 142,452 157,279
SUB TOTAL GASOLINE 382,452 405,145 543,750 801,384 1,194,120
SUB TOTAL DIESEL 213,611 221,054 264,961 342,799 455,809
TOTAL ALL 596,063 626,198 808,710 1,144,183 1,649,929
456
Number of Vehicles in Indonesia (at projected scenario condition)

Type of Vehicles 2005 2010 2015 2020 2025 2030
Private Car 2,021,406 2,579,883 3,274,956 3,639,106 4,322,488 5,373,956
Gasoline Utility Freight 2,356,754 3,007,882 3,818,266 4,242,829 5,039,583 6,265,487
Passenger
Utility Passenger 1,505,963 1,922,033 2,439,867 2,711,162 3,220,287 4,003,638
Cars
Pick up 3,597,937 3,781,467 3,984,078 4,062,786 4,176,977 4,303,053
Sub Total 1 9,482,060 11,291,265 13,517,167 14,655,882 16,759,334 19,946,134
Private Car 505,351 644,971 1,248,069 2,704,670 5,438,197 9,644,068
Diesel Utility Freight 589,189 751,971 1,455,122 3,153,371 6,340,387 11,244,006
Passenger
Utility Passenger 376,491 480,508 929,821 2,015,000 4,051,499 7,184,905
Cars
Pick up 899,484 945,367 1,108,065 1,422,898 1,879,662 2,383,966
Sub Total 2 2,370,515 2,822,816 4,741,076 9,295,938 17,709,745 30,456,945
Buses (Diesel) 2,804,198 2,947,240 3,175,015 3,420,393 3,776,390 4,169,440
Trucks (Diesel) 134,123 140,965 151,859 163,596 180,623 199,422
Sub Total 3 2,938,321 3,088,205 3,326,874 3,583,988 3,957,013 4,368,862
SUB TOTAL GASOLINE 9,482,060 11,291,265 13,517,167 14,655,882 16,759,334 19,946,134
SUB TOTAL DIESEL 5,308,836 5,911,022 8,067,950 12,879,927 21,666,757 34,825,807
TOTAL ALL 14,790,896 17,202,287 21,585,117 27,535,809 38,426,091 54,771,941
Total Km Travel (in Million Km) (at projected scenario condition)

Type of Vehicles 2014 2015 2020 2025 2030
Private Car 59,428 60,521 67,251 79,880 99,311
Gasoline Utility Freight 89,982 91,638 101,828 120,950 150,372
Passenger
Cars
Pick up 73,348 73,626 75,080 77,191 79,520
Sub Total 1 366,504 372,177 406,829 471,237 569,421
Private Car 18,690 23,064 49,982 100,498 178,222
Diesel Utility Freight 28,299 34,923 75,681 152,169 269,856
Passenger
Cars
Pick up 19,365 20,477 26,295 34,736 44,056
Sub Total 2 111,561 134,254 272,858 530,493 923,229
Buses (Diesel) 112,611 114,301 123,134 135,950 150,100
Trucks (Diesel) 5,386 5,467 5,889 6,502 7,179
Sub Total 3 117,998 119,767 129,024 142,452 157,279
SUB TOTAL GASOLINE 366,504 372,177 406,829 471,237 569,421
SUB TOTAL DIESEL 229,559 254,021 401,882 672,946 1,080,508
TOTAL ALL 596,063 626,198 808,710 1,144,183 1,649,929
457
458
AN ALTERNATIVE METHOD IN EVALUATION APPROACH OF
AIRPORT PAVEMENT PERFORMANCE
Ervina Ahyudanari 1*, Nasir Shafiq 2, Ibrahim Kamaruddin 3
ABSTRACT:
Airport pavement needs to be maintained regularly in order to guarantee the safety in airport
operation. The time schedule for airport pavement maintenance is based on analysis of data
monitoring of airport pavement condition. The monitoring system is usually conducted on daily or in
certain frequent basis. For busy airport, pavement inspection during daytime is difficult to be
implemented. This paper proposes an alternative method in predicting the pavement condition based
on recorded data of traffic and pavement maintenance history. The method starts with comparison of
airport pavement strength and all aircraft types operated; continues with checking adequate design of
pavement structure. The pavements that meet these two preliminary evaluations, i.e. pavement
strength and structure, are unlikely to have structural defect but are considered to have surface defects.
The second comparison is between the operational characteristics and type of pavement functional
distress type. The agreement of the second comparison will be considered as the preliminary judgment
of the existing pavement performance.
Keywords: Airport Pavement, Operational Characteristics, Pavement Performance
1
PhD Student, Universiti Teknologi PETRONAS, Tronoh 31750, Perak Darul Ridzuan, MALAYSIA
2
Associate Proffesor, Universiti Teknologi PETRONAS, Tronoh 31750, Perak Darul Ridzuan. MALAYSIA
3
Associate Proffesor, Universiti Teknologi PETRONAS, Tronoh 31750, Perak Darul Ridzuan. MALAYSIA
*Correspondence Author: Ervina Ahyudanari, Universiti Teknologi Petronas, Malaysia (lect Institut Teknologi
Sepuluh Nopember, Indonesia).
E-mail: ervina@ce.its.ac.id
459
INTRODUCTION
The cost of pavement maintenance may be reduced if the right maintenance timing can be
captured based on periodic pavement evaluation. The periodic pavement evaluation requires good
management in time and adequate experienced staff. Time management become crucial in busy
airport since there is difficulty in conducting pavement evaluation.
Pavement evaluation consists of two types; those are evaluation on functional of pavement
and evaluation on structural condition of pavements (Ullidtz, 1987). The functional condition of
pavement relates to comfort (Woods, 2008), safety (FAA, 1997) and user cost. The structural
condition of pavements could be evaluated through laboratory methods, in situ methods and falling
weight deflectometer (Gopalakrishnan & Thompson, 2007). The structural condition also may be
approached by deriving elastic parameters (Wardle & Rodway, 1998), residual life (Garg, Guo, &
McQueen, 2004) and overlay design, also use of ACN/PCN method (FAA, 2006) and (Loizos &
Charonitis, 2004).
Both pavement evaluation types require field inspections that necessitate daylight schedule in
conjunction with skilled and experienced staff as field inspector and data interpreter. Usually, airport
management will hire consultant to conduct the field inspection and data analysis of pavement
performance.
This paper attempts to present an alternative approach in conducting pavement evaluation that
may be used as an alternative if the field activity is not viable. The alternative approach presented here
is in conceptual format based on historical data of traffic and pavement maintenance available. The
objective of this alternative pavement evaluation is to provide a conceptual method to evaluate
pavement condition for airports which have limited experienced staffs.
This paper caters the proposed method in the following sequence. The objective and the
reason behind the proposed method is presented in Introduction Section. The next section is Literature
Review that covers the previous works related to the topic in this paper. The step by step method in
developing the proposed method is briefly described in Methodology Section. Results and Analysis
Section presents the result of data analysis and additional analysis to support the concept of the
proposed method. The last section is Conclusion and Further Works as the closure of this paper.
LITERATURE REVIEW
Airport Pavement Strength
Airport pavement strength is denoted as Pavement Classification Number (PCN) that

indicates the load carrying capacity of a pavement without specifying any particular airplane or
providing information of pavement structure (FAA, 2006). PCN is the reference number to manage
operational characteristics of particular aircraft on particular airport. To express the effect of an
individual airplane on different pavements structure, International Civil Aviation Organization
(ICAO) adopted Aircraft Classification Number (ACN) that was introduced in 1978 (ICAO, 2004).
The value of ACN should be less or equal to PCN value to operate without any restriction. The airport
operator uses ACN-PCN method to evaluate the acceptable operation of an aircraft. This ACN-PCN is
the only structural classification procedure for airport pavement that internationally adopted (Loizos
& Charonitis, 2004).
The value of ACN is determined at a frequency of 10,000 coverages. This means that the
particular aircraft with ACN value greater than PCN value is possible to operate in the corresponding
pavement as long as the coverage frequency is less than 10,000 (FAA, 2006). The frequency of
aircraft coverage is not the same as the frequency of aircraft movement. To have the coverage value of
an aircraft, there is a need to have mathematical derivation based on pass to coverage ratio of each
aircraft (FAA, 2006).
460
Airport Pavement Design
Airport pavement design in Indonesia follows the design guidelines published by Federal
Aviation Administration (FAA). Most of the first establishment of airports pavements in Indonesia is
around some decades ago and some rehabilitation of pavements, such as overlay, has been conducted
in order to guarantee the safety of airport operational. Therefore, airport operator needs to evaluate the
adequacy of pavement structure. FAAs Advisory Circular related to airport pavement design (FAA,
2006b) also provides back calculation process to check the adequate existing structure in carrying the
existing traffic.
Aircraft Take Off/Landing Process
Almost all aircrafts operated in hub airport will land and take off in the same day, only in base
airport that aircrafts land and take off in the following day. In the take off phase, aircraft moves from
stationary state, ground maneuver; roll up then into airborne state. Similarly, in landing phase, the
aircraft from airborne state maneuver to ground borne state. Both aircraft maneuvers impose certain
load on different area on pavement.
The area imposed by the load from aircraft movements depends on the aircraft weight, wheel
track, frequency of movements, take off and landing direction. Take off and landing direction depends
on prevailing wind when aircraft does ground maneuver or approach the runway. Technology of jet
engines and wings of each aircraft influence take off and landing distance required in a particular
airport.
According to Horenjeff (Horonjeff, McKelvey, Sproule, & Young, 2010), the range of
runway required for each aircraft varies from 340m (DA-42, Twin Star) to 3000m (A-380).
Additionally, aircraft is not always landing in the same lateral path. The variation forms normal
distribution as discovered by HoSang (HoSang, 1975).
METHODOLOGY
Data Collection
The alternative in monitoring method on airport pavement performance that is proposed in

this paper uses data from Juanda International Airport, Surabaya, Indonesia. This airport is operated
under PT Angkasa Pura 1. The Juanda Airport is denoted as SUB following International Air
Transport Association (IATA) code with 3000m in runway length and has runway direction 10/28. The
runway direction represents runway designation of magnetic azimuth, which is measured clockwise
from the magnetic declination. Pavement strength (PCN) for SUB is 83/F/D/X/T. This means that the
load carrying capacity of the pavement is 83, F is flexible pavement, D shows the strength of subgrade
based on CBR value that is less than 4; X illustrates that the allowable tire pressure for the particular
pavement is limited to 218 psi (1.5 MPa); and T indicates that the pavement strength value (PCN) is
evaluated by technical method.
Data related to traffic, such as aircraft movements as well as type of aircraft also utilized from
operational characteristics of SUB. Data traffic used is data from year 2007. Other data related to
pavement structure and pavement maintenance history also available from the particular airport.
Assumptions
SUB is selected as the location of the study and is assumed that there is no significant change
in the environmental condition. Therefore, the environmental factors will not be considered.
461
The aircraft type information available in data of traffic is only indicates the series type of
particular aircraft. In aircraft manual provided by aircraft manufacturer, each series has variation on
take off weight, tire pressure, type of jet engine and dimension of wheels configuration. Since there is
no detail information related to variation mentioned, the aircraft selected is assumed to be have the
same characteristics. For example, B737 (http://www.boeing.com) has variation on weight and
engine types since weight and jet engine types in aircraft industry is customized depends on the
function of the aircraft to the airline (passenger or cargo). Therefore, in analysis only one type of
weight and engine that will be considered.
Methodology
The methodology used in proposing the alternative method to monitor airport pavement
performance incorporate the operational characteristics of the selected airport, statistics of airport
pavement maintenance, evaluation on airport pavement structure and reported airport pavement
strength. Figure 1 presents the methodology of analysis in presenting the proposed of airport
pavement evaluation approach.
Pavement Operational Pavement

Strength Character. Structure
N Y
ACN> Adequate Maint.
PCN struct record
Y N
Possible Y
Freq< Y Surface
surface
10,000 distress
defects
Possible
structural
defects
Figure 1: Flow chart of the methodology of analysis
Figure 1 shows the process of three comparisons analysis of the available data. The first
comparison is between data of pavement strength (PCN) of selected airport and ACN of all aircrafts
operated in corresponding airport as listed in operational characteristics data. If the ACN value greater
than PCN value, then proceed to the next evaluation regarding frequency of coverage. If frequency of
coverage of particular aircraft is less than 10,000, this means that the aircraft will not cause structural
damage on airport pavement.
The second comparison analysis is between the selected airport pavement structures with the
traffic as listed in the operational characteristics. The pavement thickness required is obtained by
utilizing back calculation procedure of FAA (FAA, 2006b). The obtained pavement thickness is
compared to the existing pavement thickness. If the required thickness is less than the existing one,
there is most likely to have pavement surface defects.
The last evaluation is on the pavement distress pattern as recorded in pavement maintenance
history in the selected airport. If data of pavement maintenance history shows only pavement surface
defects occurred, that means there is an agreement with the previous two evaluations.
The proposed method gives an idea of how to reduce field survey in busy airport. Based on
the evaluation results, the evaluation regarding pavement structural condition may be omitted.
462
RESULT ANALYSIS
Operational Characteristics
Operational characteristics relate to fleet mix, take off/landing direction and time of
departure/arrival. Fleet mix is composition of aircraft types operated that represents load magnitude
distribution onto the pavement. Take off/Landing direction indicates the runway end that is most
frequently used. Time of departure or arrival represents the period when the pavement receives
intermittent loads.
Figure 2: Runway end usage for SUB in 2007
Figure 2 shows that the activities in runway are not always occurred in the same direction.
This is for the reason that aircraft movements depend on head wind direction. The runway usage in
Figure 2 also indicates that most of the movements are from Runway 10 and for a couple of months
no movement from Runway 28.
The number of aircraft movements in Figure 2 consists of different types of aircraft. This
means that the load and the position of load stress from runway centerline are vary. The position of
load stress depends on gear type and wheel track for the corresponded aircraft types. Table 1
illustrates gear types of aircrafts operated in SUB with the number of aircraft types that utilized the
related gear types.
Table 1: Fleet mix in SUB
Number of aircraft types

Gear Type
2007
Double dual tandem 11
Dual tandem 17
Dual wheel 60
Single tandem wheel 3
Single wheel 19
Tridem 10
Helicopters 43
Total 163
463
The number of aircraft types in Table 1 indicates the different aircraft manufacturers and
aircraft series, for example, A319, B732 and F100 have similar gear type but different weight to be
distributed to pavement and distance between wheels. Figure 3 illustrates the example mentioned.
Figure 3: Variation on gross weight and wheel track for dual wheels types
The illustration shows that the load distribution on pavement has variation on distance from
runway centerline. The assumption taken in this case is that the lateral distribution of aircraft wheel
paths on runway is represented by normal distribution (HoSang, 1975). The trend on load distribution
represents the area of pavement that receive highly frequent load. Figure 4 illustrate the intensity of
load repetition received by airport pavement based on the distance of the wheels from runway
centerline and the frequency of related aircraft movements.
B737
MD80
A319
Figure 4: Load distribution based on aircraft wheel position from runway centerline (in meter)
Figure 4 shows that there are a number of wheel positions of aircrafts operated annually in
SUB for Runway 10 movements. This Figure represents the load distribution laterally. However, only
three wheel positions that can be considered in the analysis due to those frequencies. The wheels
positions as shown in Figure 4 belong to A319, B732 and MD80 at 3.8m, 2.6m and 2.5m respectively.
Those three aircrafts have frequency of movements around 5000, 44000 and 12000 respectively,
while the other wheels positions from runway centerline have less than 1000 pass annually.
464
The next analysis is the length of take off distance to see the load distribution longitudinally
from each runway end. The analysis of load distribution longitudinally only consider A319, B732 and
MD80 which have take off distance 1750m, 1830m and 2052m respectively.
From both lateral and longitudinal load distribution, the wheel paths of B732 and MD80 can
be combined. This come to the result that at 2.55m from runway centerline, the pavement surface
received high intensity load for approximately 56000 pass for 2000m long instead of 3000m (the
length of runway). This means that at this area, pavement has high severity compare to other area.
Airport Pavement Strength
The objective of evaluation on pavement strength is to check whether the traffic as presented
in subsection Operational Characteristics does not over burden the pavement structure. The value of
ACN of operated aircrafts is compared to PCN value of airport under study. Since there are three
different approach in obtaining the ACN value, this study adopts all those approaches. The methods in
determining ACN value follow COMFAA (FAA, 2006a), Transport Canada (Transport Canada,
2004) and aircraft manufacturer. Figure 5 shows the ACN of some aircrafts exceed the PCN value.
The exceeding value is up to 45% beyond the required value. However, those aircrafts with higher
ACN value have frequency less than 5000 movements and those mean that the coverage far less than
10,000 as required.
Figure 5: Number of aircrafts with ACN value higher than PCN value
Evaluation on Pavement Structure
Pavement structure is designed based on design guidelines provided by FAA (FAA, 2006b).
Data of the CBR value and existing traffic is utilized to determine the pavement thickness required to
carry the exixting load.. The result is compared to the existing pavement thickness. The thickness
available of pavement in SUB airport is 142.5cm above the stabilized subgrade since the CBR value
for the subgrade is around 3.5%. The required thickness of pavement based on calculation of dual
wheel equivalent is 105cm and shows that the thickness is able to carry the existing load.
Pavement Maintenance Record
Most of airports under PT Angkasa Pura 1 were built in some decades ago. SUB, for example,
was built in year 1960. However, some airports are new due to relocation case. For old airports, some
overlays have been conducted to meet the increasing demand. Table 2 presents the pavement
maintenance record for SUB.
The record of pavement maintenance for SUB indicates that overlays were conducted
annually in the second decade after construction (it was constructed around 1960s) since there is no
record before those years. The data shown until 1992 provides information regarding the length of
overlay which are less then the runway length (3000m).
465
The area rehabilitated in Table 2 shows an agreement to load distribution as presented in

Figure 4 and different length of take off distance as stated previously. This shows the agreement with
overlap area caused by aircraft movements, either laterally or longitudinally.
Table 2: Pavement maintenance record for SUB
Type of Dimension
Work/Pavement W Thickness
Airport Year Area Condition L (m) (m) (cm) Type
SUB 1974/1975 overlay 10 AC
1977/1978 R10 overlay 800 8 AC
R28 overlay 1000 8 AC
1978/1979 R10 overlay 800 8 AC
1979/1980 R10 overlay 300 8 AC
Central overlay 600 8 AC
1980/1981 R10 overlay 584 8 AC
R28-landing
1981/1982 zone overlay 1119 8 AC
1989/1990 R10 overlay 800 8 AC
1991/1992 Central overlay 1400 8 AC
2005 R10 Weak spot 3 locations
Skid resistance test
rubber deposits
removal
2006 R10 raveling 5 locations
R28 raveling 1 location
rubber deposits
removal
2007 R28 raveling 2 locations
R10 Weak spot 1 location
rubber deposits
removal
The type of pavement distress shown in Table 2 indicates that the distress occurred is only
functional not structural. In addition, the type of distress supports the temperature as possible cause
of raveling problem. Raveling is the type of surface defect that is caused by stripping of bituminous
film from the aggregates or asphalt hardening due to aging (FAA, 2004).
CONCLUSIONS AND FURTHER WORKS
Conclusions
This paper presents the alternative in evaluation approach of airport pavement performance.
The methodology to conduct the alternative approach has been presented and covers three-comparison
466
analysis. Those comparison analyses require data related to airport pavement strength, operational
characteristics, airport pavement structure and airport pavement maintenance record.
The comparison analysis conducted shows that the load pattern as represented by ACN value
and load distribution on pavement, both laterally and longitudinally, has an agreement to the
pavement distress occurred as recorded in pavement maintenance history. The analysis results indicate
that the airport pavement under study needs maintenance for surface condition and not structural.
The proposed method may use as an alternative in monitoring airport pavement performance
for busy airport. Therefore, the safety operational can be guaranteed. However, some justifications are
required to determine the degree of the pavement surface defects predicted.
Further Works
As stated previously that the concept of this method needs further analysis. The analysis that
would support the concept relates to determination of loading area, and relationship between the
frequency of aircraft movements and rate of pavement aging. Determination of loading area requires
further analysis of wheel configuration and dimension, frequency of movements, and aircraft speed.
To be able to analyze the rate of pavement aging, there is a need to incorporate the bituminous
behavior in receiving the repetition load at certain temperature due to solar radiation.
ACKNOWLEDGEMENT
This research is sponsored by University Teknologi Petronas, Malaysia with support data
from PT Angkasa Pura 1 Indonesia. Appreciation is also addressed to Assoc.Prof. Dr.Madzlan Napiah
and Dr.Syed Baharom for the feedback to improve this paper.
REFERENCES
FAA. (1997). Measurement, construction, and maintenance of skid-resistant airport pavement

surfaces. Advisory Circular 150/5320-1 (pp. 1-45). Washington D.C.: Federal Aviation
Administration. Retrieved from www.faa.gov.
FAA. (2004). Airfield pavement surface evaluation and rating manual. In L. Entine (Ed.), Advisory
Circular 150/5320-17 (p. 57). Federal Aviation Administration.
FAA. (2006a). Standardize Method of Reporting Airport Pavement Strength - PCN. In D. L. Bennett
(Ed.), Advisory Circular 150/5335-5A (p. 63). Wahington, D.C.: U.S. Department of
Transportation Federal Aviation Administration. Retrieved from www.faa.gov.
FAA. (2006b). Airport Pavement Design and Evaluation. Advisory Circular 150/5320-6D. Federal
Aviation Administration.
Garg, N., Guo, E., & McQueen, R. (2004). Operational Life of Airport Pavements. Security
(DOT/FAA/AR., pp. 1-117). Washington, D.C.: FAA, US Department of Transportation.
Gopalakrishnan, K., & Thompson, M. R. (2007). Use of nondestructive test deflection data for
predicting airport pavement performance. Journal of Transportation Engineering, 133(6), 389.
doi: 10.1061/(ASCE)0733-947X(2007)133:6(389).
Horonjeff, R., McKelvey, F. X., Sproule, W. J., & Young, S. B. (2010). Planning and Design of
Airports. (L. S. Hager, Ed.) (Fifth Edit., pp. 51-55). McGraw-Hill.
467
HoSang, V. A. (1975). Field Survey and Analysis of Aircraft Distribution on Airport Pavements (p.
286). Washington D.C. Retrieved from www.airporttech.tc.faa.gov.
ICAO. (2004). Aerodromes. Annex 14. International Civil Aviation Organization.
Loizos, A., & Charonitis, G. (2004). Bearing Capacity and Structural Classification of Flexible
Airport Pavements. Journal of Transportation Engineering, 130(I), 34-42. doi:
10.1061/(ASCE)0733-947X(2004)130:1(34).
Transport Canada. (2004). Aircraft Classification Numbers ( ACN s ). Transport Canada Technical
Program (pp. 0-18). Ottawa, Ontario: Transport Canada. Retrieved from
http://www.tc.gc.ca/CivilAviation/International/Technical/menu.htm.
Ullidtz, P. (1987). Pavement Analysis (pp. 228-284). Lyngby, Denmark: Elsevier Science Publisher
B.V.
Wardle, L. J., & Rodway, B. (1998). Layered Elastic Pavement Design- Recent Developments.
Transport 98, 19th ARRB Conference (pp. 7-11). Sydney. Retrieved from
mincad.com.au//Arrb98Paper.htm.
Woods, J. E. (2008). A study of airport pavement-aircraft interaction using wavelet analysis.

gradworks.umi.com. THE UNIVERSITY OF TEXAS AT SAN ANTONIO. Retrieved from
http://gradworks.umi.com/14/50/1450630.html.
468
POPULATION BASED VEHICLE ROUTING PROBLEM - CASE STUDY : SCHOOL BUS FOR
NEEDY STUDENT IN SURABAYA, INDONESIA
Daniel Hary Prasetyo 1*, Jamilah Muhamad 2,

Rosmadi Fauzi2
ABSTRACT:
Vehicle Routing Problem methods were widely used in the transportation research. This research has a
purpose for finding the best route of school bus for covering the most number of needy students. This
research uses needy student of Surabaya, Indonesia as the passengers. Surabaya government has freed the
school tuition for helping the needy students, but a number of needy still cannot go to school. It is because
the school cost is not just the tuition fee. This research tries to modeling free school bus for helping them
in the transportation cost. Because of the lack of data, the needy students house location is not mapped. It
will replace with street load points that were influenced by the population of needy student in the sub-sub-
district area. This needy population data will accompany street map, school map, and bus depot in the
VRP process. The preparation process, the VRP process, and the analyst process in this research
conducted in the GIS environment. The result of this research indicates that the method proposed can be
used for finding the optimal routes using population of needy.
Keywords: VRP, School bus, gis, spatial analyst
1
Student, Universiti Of Malaya, 50603 Kuala Lumpur, MALAYSIA
2
Lecturer, Universiti Of Malaya, 50603 Kuala Lumpur, MALAYSIA
*Correspondence Author: Daniel Hary Prasetyo, University of Surabaya, INDONESIA, Tel:+62312981395

Fax: +62312981394. E-mail: daniel@ubaya.ac.id
469
INTRODUCTION
Surabaya resides in East Java at 07 21' south latitudes and 112 36' - 112 54' east longitudes.
As being the second biggest city in Indonesia, with 33.306 KM2 areas wide, approximately 3 million
people live there. They spread in 31 districts and 163 sub districts. Surabaya has a total of 492.495
school-aged citizens, with 270.076 at the elementary school age, 114.733 at the lower secondary school
age, and 107.686 at the upper secondary school age. There are 1.622 elementary schools consist of 564
public and 1.058 private. There are 342 secondary schools consist of 42 public and 300 private. There are
257 high schools consist of 33 public and 224 private. With the number of elementary schools about 5
times that of secondary schools and 7 times that of high schools, students must travel greater distances as
they advance to higher levels of study.
The government of Surabaya has calculated the participation rate of each education level. The
participation rate is a comparison between the numbers of students in a certain education level with all
citizens at the respective age level. In early 2008, in Surabaya, the elementary school participation rate
was 92.92%, the secondary school participation rate was 79.85%, and in the high school participation rate
was 83.53%. These figures indicate that there are more than 19,000 citizens of elementary school age,
more than 23,000 of secondary school age, and more than 17,000 of high school age who are not
participating in school. The primary reason these citizens are unschooled is that they cannot pay schooling
costs because they live in needy families. The Surabaya city government has taken action by waiving
tuition costs in many public schools. In early 2008, there were 544 elementary schools and 58 secondary
schools that did not collect admission costs and monthly costs from their students. The schools which still
collect costs from their students can waive costs for needy students by the letter from sub district
government. The government has not yet made some high schools free because the government is still
focusing on the national education target nine years of compulsory study. This target dictates that all
Indonesian people should study for a minimum of nine years at the elementary and secondary levels.
However, even with free admission and no monthly costs, there are still numerous children who cannot go
to school because they have no extra money for transportation or buying uniforms, shoes, books, and
other school supplies. The purpose of the project addressed in this paper is to provide help in
transportation costs.
This study has a goal to discover a model that can be use to design an appropriate school bus
routes which favoring the needy student. In order to achieve this goal, the following objectives have been
identified: - To investigate an appropriate method in finding bus routes with the population data
- To investigate an appropriate analyst method to measure how well the proposed
routes
This research focused on the north area of Surabaya, due to internal limitations, which has the
largest number of needy students. All the spatial and tabular data used in this project represent this area.
This research also limits the scope to needy secondary school students. However, the model developed
here is not dependent on these limitations. The model can be easily used for surveys of other areas in
Surabaya, for the needy in elementary or high school, and for any other places.
LITERATURE REVIEW
One of the important GIS areas is the transportation. There are many transportation model and
analysts using GIS as the application. Bus is the most commonly use in mass transportation model and
has been attracted many researchers. Some studies focus in reviewing the existing system and the other
proposing to the new one. In reviewing the existing system, some researcher assessing the transportation
performance and its efficiency (Thevenin, 2001. Huang and Ian, 2002). The most commonly used for
accessing performance is using the bus stop coverage and it accessibility. Bus stop can be review and
470
grading (Athanasiou and Photis, 2006) and calculating its utilizing (Sankar, 2003) and its walking time
and waiting time (Gray and Bunker, 2005). While accessibility of the bus stop also have been considered
as the aspect for successful transportation system and reviewed by GIS approach (Munshi and Brussel ,
2005. Kesten and Tezcan, 2006. Kimpel, 2007). Accessibility can be measured by calculating the distance
from the dwelling, and the public area like school, office, market, etc. Jennifer Rogalsky underlining that
the accessibility of the poor is crucial in the public transit (Rogalsky, 2009). There are several concepts in
the Vehicle Routing problem. Depend on the characteristic of the busses transportation, the fit in concept
is the multi vehicle PDPTW approach. The Pickup and Delivery Problem with Time Windows (PDPTW)
is the problem of designing a number of routes for a fleet of M identical vehicles to service the
transportation requests of N customers. Each customer request asks that a given load is transported from a
pickup location to an associated delivery location.
Several researchers have projects related to bus routing. Some of them focus-on the use of a new
algorithm or advancing an existing algorithm, while others are implements existing algorithm to real
world problems (Dorronsoro Daz, 2007. Toth and Vigo, 2002. Bruce Golden et al, 2008). In advancing
algorithm, Robert Bowerman introduces a multi-objective approach to modeling the urban school bus
routing problem (Robert Bowerman, 1995). The first task is grouping students into clusters using a multi-
objective districting algorithm. Second is generating a school bus route and bus stops for each cluster. A
heuristic algorithm based on their formula tested with data from a sample school board location in
Wellington County, Ontario, Canada. They have defined several optimization criteria to evaluate the
desirability of a set of school bus routes. These are:
1. Number of routes. Because the capital cost is significantly larger per bus than the incremental cost
over the year, the number of routes generated should be held to a minimum.
2. Total bus route length. This criterion reduces the total length of the school bus routes.
3. Load balancing. Load balancing involves minimizing the variation in the number of students
transported along each route.
4. Length balancing. This criterion involves reducing the variation in route lengths.
5. Student walking distance. This criterion balances the total distance that students walk from home to
and from their bus stops against route length.
Based on the number of school, bus routing can be divide into many-to-one and many-to-several
(Spada et al, 2005). An Example of many-to-one can be viewed in the work of M Fatih Demiral and
friend (Demiral et al, 2008) and Nayati Mohammed (Nayati, 2008). Both use one school location as a
depot and student house location for the customer location for generating bus routes. They work with
study area at Isparta, Turki and Hyderabad, India respectively. The others are Li and Fu (Li and Fu, 2002)
and Bektas and friends (Bektas, 2007). Li and Fu implement a heuristic algorithm for an existing data of a
kindergarten in Hongkong, while Bektas using integer programming for elementary school in central
Ankara, Turki. They both saving 29% and 26% respectively for the generated new route compared to the
current implementation.
Based on the location or environment of the data, bus routing can be divided into urban and rural
area. In the urban area, the many-to-one from Bektas and friend and Li and Fu can be a robust example.
In the rural area, Armin Fu genschuh take five counties in German for the student location (Fugenschuh,
2009). While the destination is multiple school, Instead of sending the bus back to the depot after having
served a trip, he push to re-use the bus to serve other trips, if it is possible. He integrates optimization of
school start times with the optimization in the school bus transportation. He believes with this integration
for a single county can save up to 1 Mio Euro year by year. Based on the location or environment of the
data, bus routing can be divided into urban and rural area. In the urban area, the many-to-one from Bektas
and friend and Li and Fu can be a robust example. In the rural area, Armin Fu genschuh take five
counties in German for the student location (Fugenschuh, 2009). While the destination is multiple
school, Instead of sending the bus back to the depot after having served a trip, he push to re-use the bus to
471
serve other trips, if it is possible. He integrates optimization of school start times with the optimization in
the school bus transportation. He believes with this integration for a single county can save up to 1 Mio
Euro year by year.
METHODOLOGY
Process in this research divides into 3 stages, first is the initial stage, second is the routing stage
and the last is the analyst stage. Figure 1 shows this 3 stages equipped with its steps and data. The initial
step produces appropriate data for the routing step. The routing step use ArcGIS Vehicle Routing Problem
analyst in finding the optimal routes based on data and setting from the initial steps. The second step
produces a proposed Routes class. This Routes class then analyzed in the analyst process.
needy
Region map School location Bus Depot Street Map
citizen
database
Mapping the Extract Public Refining Street

needy School Map
Needy map
and Needy
Density map
Network
Dataset
Refining
Needy Area
Calculate
Street load
Refined needy
map Street Load
Map
INITIAL STAGE
Designing
Designing
School
Route Layer
Distribution
Orders Layer Routes Layer Depots Layer

Class Class Class
ROUTING STAGE
Referenced as next
design
ArcGIS VRP
Analyst
Bus routes
ANALYSIS STAGE
Time Consume, Road Balance,
and Covered Area Calculation
Load Analyst
load of passenger
in 3D Visualization
Figure 1. Stages of the methodology in this research
472
The Basic map
Several collected data use in the initial stage. There is needy citizen database, region map, school
location, bus depot location, and the soul of VRP, the street map. The Surabaya city government has
collected needy citizen data. In 2007, 550,783 people of the 119,219 families registered lived below the
poverty line. This needy people survey recorded name, birthplace, birthday, address, sex, and occupation.
Meanwhile, The Surabaya city government only has a map with sub-district detail. Consequently, a
survey for mapping the sub-sub-district boundary of the projects study area had to be conducted. The
total number of sub-sub-districts in this study area is 274. The needy citizen data then filtered to select
people from 13 to 15 years old. The selected data were converted to DBF before assimilated and copied to
the sub-sub-district map. A total of 8,579 needy students were spread throughout this274 sub-sub-district.
In this study area, there are 57 schools, consist of 8 public schools and 49 are private schools.
Public schools are schools owned by the government. These schools serve both purposes because they
commonly have appropriate buildings and surrounding areas. Most importantly, the government has a
right to manage these areas. So they are used not only as places for delivering and picking up students but
also as bus depots. Completing the real two bus depot reside in the center and west area.
The city government of Surabaya already maps their streets. There are Highway Street, Primary
Street, Secondary Street, and Tertiary Street in this study area. This research mainly uses Primary and
Secondary Street. Highway Street is eliminated because it cannot use for picking up passengers. Some of
Tertiary Street that can be passed by bus is also included. However, this existing street map cannot use
directly for this research. Street map used in the ArcGIS Network Analyst requires a specific format. It
needs to conduct some digitizing work for fulfill this format. This special requirement sometimes
demands extra attention and time. This map also has to be equipped with drivetime field. It determined
how much time is needed for a bus for traverse each street. Lets call this street map and previous map as
basic maps. Figure 2 shows these basic maps in one view.
Figure 2. The basic map
473
Refining needy area and calculating street load
There are two process need before above data to be taken into routing process. First is Refining
the needy area and second is generating the street load. In the previous paragraph, it is determined that the
number of school is 57 while the number of needy student is 8,579. Each school can be set to provide 155
seats in order to cover all needy students. Let call this number as needy capacity. However, not all
students will need to ride the bus. Government can eliminate the number of needy passenger with a rule
that forces the needy to choose a school near their home. With this rule, the needy capacity of each school
will be reduced by the number of needy students surrounding it. Some schools may be eliminated from
the bus route if the needy capacity is filled because they lie in poor neighborhoods. This step will refine
the raw needy area to become the must covered needy area.
The process for this refinement is as follows. First, making a neighborhood area for each school,
then calculate how many needy are in these areas. Second, comparing the neighborhoods needy number
with the needy capacity of each school. This step will determine how many needy fulfill the capacities in
each school and how many needy are still left in the surrounding areas. If any school has its capacity
filled by the surrounding needy, then this school will be eliminated because a school bus will not need to
deliver needy student to it. After this process, the needy students are reduced to 5.265, and the schools are
reduced to 48. The reduced needy layer map and reduced school layer map, equipped with the removed
schools, was shown in Figure 3.
Figure 3. The new needy layer and the removed schools.
In the vehicle routing, some of the process needs to accumulate a load of passengers while
traveling. The load value is calculated from the computation of the number of the needy surrounding each
road which traveled. For this calculation, it needs to prepare a field, which can represent the load value.
The process is a little bit similar with the process of calculating needy surrounding schools, with the
school points changed with the road part polyline. This street load will become points in orders class. The
474
points are in the middle of the roads. This is the point that actually replacing the student residence in the
other VRP researches. Figure 4 shows the street load in gradation color for symbolizing the different
value.
Figure 4. Street load in different value
Designing school distribution and VRP datasets
The main resource for building the network dataset is a street map. As long as it has standard
field names, the process will automatically detect and use data to generate the network dataset. After the
dataset has been built, the process continues with determining the settings of the VRP class. In a
minimum, the VRP class must set 3 sub-classes, which are Orders, Depots, and Routes. This step also has
an optional setting sub-class called Breaks. First, the Orders are defined. Orders are places that must be
visited by the vehicle. These can be locations for delivering something, picking up something, or just
visiting. After the refining process, the total number of schools was 48. All 48 schools were imported as
Orders, with the service time set to 2 minutes and the time window set from 6:00 am to 7:00 am. Two
different school distributions made for several route designs. These two different school distributions can
be seen in Figure 5. Another layer to be imported to Orders class is the street load. Street load imported
with time window 5:30 am to 7:00 am.
Figure 5. Two different school distributions. Left is the first distribution, right is the second one.
475
Next is Depots. In school bus routing, the bus starts from the bus depot, travels to and visits
schools, and ends up at another depot. The bus may come back to the starting depot, or it may stay in the
destination depot until the end of school and then reverse the same route. As mentioned earlier, there are
two bus terminals that can be used as depots. There are also other available places that can act as depots,
such as the public schools. The depot has a service time and time window parameter. The service time is
set to 2 minutes, and the service time is set to 5:30:00 am to 7:00:00 am. This setting means the depot can
start loading the bus at 5:30:00 AM and have enough time to travel and pick up passengers before visiting
the first school at 6:00:00 AM.
Sometimes a bus on a route needs directions to pass over a certain street and sometimes not to
pass. Barriers are used to omit streets from traveling route. Some streets in the map lie outside the study
area (northern Surabaya), while some streets lie in east and center areas. This project does not focus on
streets outside the study area, and including these streets will make the route unproductive and will waste
time. Therefore, barriers are needed here.
Now the Routes class can be defined. The less number of routes the more number of choices of
schools. For example, if all schools can be visited in one route, then the needy can choose to go to any
school in that route. If there are two routes, the schools will be divided into two paths, and the student
can only choose about half of all the schools. It was found that the minimum number is 4. These 4 routes
named with west route, east route, north route, and south route. The capacity of each route is limited by
the student capacity in its direction. After the route has been generated, the process continues with
analyzing the covered area. Because there are several Depots available, the generated route and analysis
of the covered area are held at several times with some combination of depots used.
The steps continue by running the VRP Analyst process. The ArcGIS VRP engine finds the most
optimal routes by implementing the Djikstra method. Four different routes direction will be produced that
start from the center depot, travel to selected schools and end at each depot direction. Each route will
collect a number of student in the picking up points, which in this research were represented by the street
load points, accumulate this number to get most optimal number under its capacity.
Analyzing the routes
Next step after routes have been proposed is analyzing the routes. The aspects to be analyzed as follows:
Balance of the school capacity and the covered area of the route
To calculate this balance, each route must be separated from the others and their loads
calculated individually. This number is then compared with the total school capacity in this
route.
The number of covered needy
The more the needy can cover the better the route. To find how many needy can be covered, a
covered area analysis must be performed. The Analyst uses a buffer with a reference distance
of 300 m, as mentioned earlier. This buffer clips the sub-sub-district areas to make sub-sub-
district areas that are near the route. After continuing with recalculating the total number
based on the needy density field, the total needy that live near the route can be found.
Travel time
The less travel time consumed the better the route. After the VRP process is done, the travel
time of each route can be easily read.
476
How many needy are in the sharing area

The sharing area is an area that is near two or more routes. The greater the number of needy
covered in a sharing area, the better the route. To find this area, each route destination must
be calculated independently. After finding the covered area of each destination, this step
continues with combining two different routes and calculating their intersections. Intersection
areas will be produced that are west-east, west-north, west-south, east-north, east-south, and
north-south. All intersection areas will then be joined with Union process. This process will
produce an area with two or more routes around it.
Above criteria is in sequence. It means the load balance is the primary criteria, followed by
covered needy, travel time, and the sharing area. The travel time is placed in the third criteria after the
number of covered needy because the object of this research is a free bus school. For another object,
maybe this criteria is different in sequence.
The best route will be chosen and continue with load analyst. In general, there are 3 big processes
in load analysis. Those are dividing each route into small part, calculating the dynamical of the number of
passenger in it, and then showing this changing in good presentation. In the first step, the street will
divided into parts. Every part will have a sequential id number that increment from starting depot to end
depot. Next, every part needs to be calculated how many needy surround them. A series of buffer,
identify, select, and dissolve process need to conduct. The number of surrounded needy then copied in the
Load field of the parts. The process is similar with the process in calculating street load. While that
process calculating covered needy in each street, this time it calculates the needy covered for each 300
meter part. The second step is calculating this parts load with summing up this value when travel from
the depot to destination, and subtracted with the school capacity when it meet a school. If there is some
passenger left when arrived in the destination depot, a reverse calculation is needed.
RESULT ANALYSIS
The VRP outputs
The initial stage is done in one time, the routing stage is done in several times. The analyst have two
different type, the covered analyst done in several time while the load analyst is one time after choosing
the best routes. The VRP output of the first school distribution is in appendix A, and for the second school
distribution is in appendix B.
In appendix A, the generated routes in the first school distribution, in differences end depot
locations, always have a route with lack of needy, the west route. The design 2, 3, and 4 even have the
significant excessive of needy in the other route direction. Therefore, the best route design in the first
school distribution is the design 1. In appendix B, the generated routes in the second school distributions
have the load balance better than the first one. All design has no significant excessive of needy. Some
needy covered by each route may have a little excess than the capacity, but it is not a problem, since the
routes have sharing areas. Needy in these areas counted two or three times for different routes. It can be
adjusted by the lucky area. Some covered needy has to be divided among the routes. This will make the
actual covered needy less than the covered in each route. With about the same rate of load balance, the
choice is now depended on the next criteria, the number of covered needy. The best choice in the number
of covered needy is the design 2 with total 4.306 covered needy. Figure 6 shows this chosen route.
477
Figure 6. The chosen output
Each route direction then detailed to discover its characteristic. For example in the east route,
Figure 7, The Bus is going to the east. In one third of its journey, it visits first school to fifth school in the
contiguous time, because the schools lie in very short distances. After the fifth school, in the second third,
it visits the sixth school. Continue going to South East to visit the seventh to tenth school, and then ending
its journey.
Figure 7. Detail of the east route
478
This detail route will be explained clearer after the load analyst. The load analyst will produce the
changes of passenger time to time while travel from center depot to each direction. This load will show in
3D in order to make it easy to understand. Figure 8 shows the 3D view of the East route.
Figure 8. The 3D view of passenger load in the East Route
East route has its crowd at the end of the journey. In the first third of the journey, the passenger is
filling in groups of schools. In the second third of the journey, it has an excessive passenger than the sixth
school capacity. Therefore, a lot of passengers still participate in the last third route, and it become a
crowd in this part before arrive in the east group of schools. The maximum passenger, when it travels, is
761. East route just needs one way because a group of schools reside in the end of the route and fill in the
rest passengers.
With the same way, it is discovered that north route has a two-way bus route for covering all its
needy and delivering to all their schools. The main direction has a maximum number of passengers of
475, while the opposite flow has a maximum passenger number of 248. The south route needs the most
number of bus fleets for the main direction because the maximum is 1.115, while the opposite flow needs
few because it has a maximum of 53. Finally, the west routes main direction has a maximum number of
632, while the opposite flow is 269.
The load analyst result can predict the passengers load in each route. With this prediction, the
government can make a decision of how many buses that needs to be provided. Say, the bus can carry
about 70 needy (50 sitting and 20 standing). The west route will need 9 fleets to the main direction and 4
fleets for the inverse. The east route needs just for its direction, with the number of the fleet is 11 buses.
The north route needs more inverse than the others. For the main direction, it needs 7 fleets, and for the
inverse, it needs 4 fleets. The south route has to provide the most fleets in the main direction. It needs 16
fleets. For accommodate a little number passenger, in the inverse, 1 fleet is enough.
CONCLUSION
Author found that the population based routing in research can not use the arcgis 9.3 VRP function
without any direction. It needs directed distribution of depots equipped with capacity in each distribution
479
area. The covered analyst methods proposed in this research can be use to discover the pattern and
correlation between needy and schools. The load analyst method can predict how many buses need to
provide to covers all orders completely. In this case study, the numbers of buses to be provided may be a
great number for the government. However, if cannot provide all the bus for covering all needy, the
government still can use this calculation number for a reference to distribute their limited number of buses
wisely.
REFERENCES
Thomas Thevenin (2001). The performance of the public transport system in time and space.
www.geocomputation.org/2001/papers/thevenin.pdf
Melissa Reese (2007). A GIS Analysis of the Bus Rapid Transit System in Curitiba, Brazil . PLAN 512 GIS for
Planning Fall 2007.
Zhengdong Huang; Masser Ian (2002). Extending bus line representation for transport applications. AGILE
Conference on Geographic Information Science, Palma (Balearic Islands, Spain) April 25th-27th 2002.
Fotini Athanasiou, Yorgos N. Photis (2006). Evaluation of Public Service Networks Efficeincy and Accessibility
Level in Urban Areas. European Regional Science Association in its series ERSA conference papers with number
ersa06p947 2006.
R.Sankar, J.Kavitha, S.Karthi (2003). Optimization of Bus stop locations using GIS as a tool for Chennai city- A
case study. Map India Conference 2003.
Matthew Gray and Dr Jonathan Bunker (2005). Kelvin Grove Urban Village Regional Accessibility by Active
Transport http://eprints.qut.edu.au/2782 2005
Munshi T, Brussel M (2005). Use of Geo-information to determine work place accessibility using public transport in
Ahmedabad City, India .Computers in Urban Planning and Urban Management (CUPUM), 2005 conference,
UCL, London
Kesten, A. S., Tezcan, H. O (2006). Bus and Subway Stations Ubiquity: A Case Study for Stations Along Istambul
Metro Line. First International Congress of Citizens and Mobility Management, Madrid, Spain. 2006.
Thomas J. Kimpel, Kenneth J. Dueker, and Ahmed M. El-Geneidy (2006). Using GIS to Measure the Effect of
Service Areas on Passenger Boardings at Bus Stops. URISA Journal Vol. 19 No. 1 2007.
Dorronsoro Daz. (2007). The VRP Web. Languages and Computation Sciences department of the University of
Mlaga. http://neo.lcc.uma.es/radi-aeb/WebVRP. 2007.
P. Toth and D. Vigo (2002). The Vehicle Routing Problem. Monographs on Discrete Mathematics and Applications.
SIAM, Philadelphia,PA.2002.
Bruce Golden, S. Raghavan, and Edward Wasil (2008). The Vehicle Routing Problem: Latest Advances and New
Challenges. Springer Science+Business Media, LLC 2008.
480
Robert Bowerman, Brent Hall, and Paul Calamai (1995). A Multi-Objective Optimization Approach to Urban
School Bus Routing Formulation and Solution Method. Elsevier Science Transportation Research Part A: Policy
and Practice Volume 29, Issue 2, March 1995, Pages 107-123.
Spada, M., Bierlaire, M., Liebling, Th.M. (2005). Decision-aiding methodology for the school bus routing and
scheduling problem. Transportation Science 39, 477490. 2005.
M.Fatih Demiral, Ibrahim Gungor, Kenan Oguzhan Oruc. (2008). Optimization at Service Vehicle Routing and A
Case Study of Isparta, Turkey.First Inernational Conference on Management and Economic (ICME 2008)
Tirana, Albania 2. 2008.
Nayati Mohammed Abdul Khadir. (2008). School Bus Routing and Scheduling usng GIS. Master thesis in
Geomatics, University of Gavle. 2008.
L. Y. O. Li and Z. Fu. (2002). The School Bus Routing Problem: A Case Study. The Journal of the Operational
Research Society, Vol. 53, No. 5 (May, 2002), pp. 552-558
Bektas T., Elmastas S. (2007). Solving school bus routing problems through integer programming. Journal of the
Operational Research Society 58 (12)(2007), 15991604.
Armin Fugenschuh (2009). Solving a school bus scheduling problem with integer programming. European Journal
of Operational Research 193 (2009) 867884.
481
Appendix A. Route Design, Output, and its covered area of school distribution 1
ROUTE DESIGN OUTPUT

Design 1
Route Start-Stop Time Needy covered School covered School capacity

West 5:49-07:00 71 min 1045 14 1581
East 5:58-07:00 62 min 1484 10 1286
North 5:38-07:00 82 min 1641 14 1302
South 5:30-07:00 90 min 1499 10 1311
Total 305 min 48 5480
Access 1+: 4096 needy, Access 2+: 1119 needy , Access 3+: 220 needy
Design 2

West 5:44-07:00 76 min 1073 14 1581
East 5:45-07:00 75 min 1982 10 1286
North 5:33-07:00 87 min 2368 14 1302
South 5:30-07:00 90 min 1476 10 1311
Total 328 min 48 5480
482
ROUTE DESIGN OUTPUT

Design 3

West 5:44-07:00 76 min 1052 14 1581
East 5:57-07:00 63 min 1508 10 1286
North 5:37-07:00 83 min 1869 14 1302
South 5:30-07:00 90 min 1326 10 1311
Total 312 min 48 5480
Design 4

West 5:40-07:00 80 min 1348 14 1581
East 5:59-07:00 61 min 1511 10 1286
North 5:31-07:00 89 min 2035 14 1302
South 5:31-07:00 89 min 1415 10 1311
Total 319 min 48 5480
483
Appendix B. Route Design, Output, and its covered area of school distribution 2
ROUTE DESIGN OUTPUT

Design 1

West 5:46-07:00 74 min 1498 13 1502
East 6:01-07:00 59 min 1525 10 1286
North 6:12-07:00 48 min 907 10 854
South 5:48-07:00 72 min 2144 15 1838
Total 253 min 48 5480
Design 2

West 5:44-07:00 76 min 1532 13 1502
East 6:05-07:00 55 min 1475 10 1286
North 6:00-07:00 60 min 1043 10 854
South 5:45-07:00 75 min 2095 15 1838
Total 266 min 48 5480
484
Design 3

West 5:50-07:00 70 min 1707 13 1502
East 6:10-07:00 50 min 1411 10 1286
North 6:12-07:00 48 min 950 10 854
South 5:59-07:00 61 min 1738 15 1838
Total 229 min 48 5480
Design 4

West 5:50-07:00 70 min 1713 13 1502
East 6:09-07:00 51 min 1373 10 1286
North 6:12-07:00 48 min 950 10 854
South 6:04-07:00 56 min 1701 15 1838
Total 225 min 48 5480
485
486
PERFORMANCE MEASUREMENT OF MALAYSIAN AIRPORTS USING DEA METHOD
Syamsul Rizal Razali, 1,Muhammad Zaly Shah 2*
ABSTRACT:
This paper presents the proposal performance indices implementation in determining the benchmark of
Malaysian airports operation.Airport efficiency and performance intensively being measured using Data
Envelopment Analysis (DEA) for many years. Previous research had been done on the evaluationof the
airport efficiency around the globe. As for examples, airport from the United States, United Kingdom,
Spain, Italy,Argentina, Mexico and Brazil were extensively being studied by this technique due to the
skies liberalizationin Americas and Europe. In this region,members of the
AssociationofSoutheastAsianNations (ASEAN) are continuing with the
negotiationsforthefullliberalizationofthe air transportation industry that targeted for 2015. In response to
the current situation, this study is focusing only on the existing twenty airports that manage by Malaysia
Airports Holdings Berhad (MAHB) over a five-year period. The primary task is to investigate their
performance level and provide an alternative tool as a benchmark. The implications of the research are
needed for managerial purpose to continuously monitor their performancetoward the ASEAN target for
the open skies policy.
Keywords:Airport efficiency, Malaysian airports, Data envelopment analysis
1
Ph.D. candidate, Universiti Teknologi Malaysia, Skudai, 81310 Johor Bahru, Johor, MALAYSIA
2
Senior Lecturer, Universiti Teknologi Malaysia, Skudai, 81310 Johor Bahru, Johor, MALAYSIA
*Correspondence Author: Muhammad Zaly Shah, Universiti Teknologi Malaysia, Malaysia.

. E-mail:zaly@live.utm.my
487
INTRODUCTION
The evolution of the airport sector from a public utility to a commercialized, and in some cases
privatized, industry has given airports greater freedom, expertise and motivation to exploit the
commercial opportunities that exist. There has been increasing pressure from the airline industry for
airports to control and to upgrade their efficiency relative to their competitors. The increase of
competition among airlines resulting from widespread market liberalization has placed airports in much
more competitive environment.
Airports, air traffic control facilities, and government airlines were increasingly being put in a
more commercial orientation. The genesis of the reform was produced in 1978 when the United States
deregulated its domestic market. An overall positive impact on the sector has been reported in many
different studies that analyze the effects of the deregulation policies introduced in the air industry in the
United States. Productive and allocative efficiency were improved, load factors raised and traffic grew
substantially more than it would have done in the absence of the deregulation (Caves et al., 1987;
Morrison and Winston, 1995; and Baltagi et al., 1995).
The process initiated in the United States provoked a chain reaction effect for the rest of the
world. Thus, new and more liberal bilateral agreements between the United States and some European
countries (UK, Netherlands, Belgium, Ireland, and Germany) were signed after the Airline Deregulation
Act. Domestic markets were deregulated to some degree in Canada and Mexico during the 80s and the
European Union took three steps on a gradual process to liberalize its air industry.
Because regulation may deliver more disadvantages than advantages, there was a tendency
towards deregulation in air transport services. The Association of South-East Asian Nations (ASEAN)
also involved in the process to liberalize its air industry. In 1995, the first moved was made when ASEAN
Leaders adopted the Agenda for Greater Economic Integration in Bangkok which included the
development of an Open Sky Policy as an area of cooperation in the Plan of Action for Transport and
Communications (19941996) have made a policy decision to move to open skies within their region.
This move will essentially be a negotiated move, with individual countries agreeing to policy changes
which will bring open skies into effect. The ASEAN Memorandum of Understanding (MOU) on Air
Freight Services signed in September 2002 is a first step towards the full liberalization of air freight
services in ASEAN.
The most visible effect of deregulation was the emergence of hub-spoke networks. Airlines were
now free to determine the optimal network type especially in the USA and Europe, and in most cases this
was the hub-spoke network. Another effect in consideration was the emerging of the low-cost airline
revolution. Low-cost airline strategy was to minimize its cost. Most low-cost airlines avoid congested
and expensive primary airports, because this reduces the expenditure on airport charges and allows for
short turn around times. The challenges faced by the airports were to enhance their efficiency and
performance as to be included in the hub-spoke network and to gain competitive advantage over its
competitor airports.
The objectives of the study are to propose a method which can be used to assess the performance
of Malaysian airports using Data Envelopment Analysis (DEA). This research intends to explore the use
of DEA as an instrument for investigating the performance indicator of Malaysian airports. The selection
of the model types will be determined after a comprehensive overview of the available and suitable DEA
models.
488
AIRPORT PERFORMANCE MEASUREMENT
Most of the business in air industries, whether in the public or private sector need to continuously
monitor their performance. Many research papers try to compare the operational efficiency, effectiveness
and the relative competitive position of airports, airlines and air traffic control. Airports performance has
been usually assessed based on some financial and/or operational efficiency throughout a set of partial
indicators. Doganis and Graham (1987) found that most airports only use partial financial indicators to
study their performance. Humphrey and Francis (2002) highlighted the importance of airport performance
information for various stakeholders such airport managers, government, airline, shareholders and
passengers.
Research on airport performance has grown considerably due to its wide range of services and
facilities to passenger, shippers, airline and others. These including runway and taxi services, apron
services, loading and unloading of baggage/freight hold, concessions, car parking and etc. Measuring
airport performance is a tricky business due to the diversity of the industry. Therefore, previous studies
carefully identify the measurement based on their different ownership and regulatory, different mixes of
services and operating characteristics, as well as the diversity and heterogeneous quality differentiation.
Some different methodologies have been developed and applied in the sector (Francis et al., 2002;
Humphreys and Francis, 2002; Oum et al., 2003). For example Yoshida and Fujimoto (2004) apply DEA
and TFP (Total Factor Productivity) to evaluate the benchmarking of 67 Japanese airports. They
considered three outputs (passenger loading, aircraft movement and cargo handling) and four inputs
(runway length, terminal size, access cost and labour). Gillen and Lall (2001) investigate the efficiency of
22 major US airports for the period 1989 to 1993 using Mamlquist TFP index and DEA. They include
four outputs and eight inputs which are divided to terminal and movement services model; the four
outputs are the number of passengers, pounds, air carrier and commuter movements passing through an
airport. The eight inputs are the number of runways, gates, employees, baggage collection belts, public
parking places, area of the airport, runways and terminal.
Pels et al. (2001) used the DEA techniques to study the performance of 34 European airports over
the period of 1995-1997. By using the DEA-BBC model in an effort to identify the effects of various
explanatory variables on airlines inefficiency. Identifying inefficiencies and detecting their sources via
DEA-BBC model serves a variety of policy purposes, wherever applicable. Airports often look
favourably to maximize benefits of economies of scope and densities. Parker (1999) conducted a research
into airport efficiency before and prior privatization of 32 UK regulated airports using DEA-BBC
compared over the period of 1979/1980 to 1995/1996 and DEA-CCR of 22 airports from 1988/1989 to
1996/1997 with outputs of turnover, passengers handle, cargo and mail business, and outputs are
considered: number of employees, capital, operating cost and others.
Oum and Yu (2004) summarized the 2003 ATRS (Air Transport Research Society) Global
Airport Benchmarking Report, and it presents the results on various measures of airport productivity and
efficiency, unit costs and cost competitiveness index, and financial performance, for up to 76 airports of
various sizes and ownership forms in North America, Europe and Asia Pacific. In this study, the
following four output categories are included: number of passengers handled, tonnes of air cargo handled,
number of aircraft movements (landings and takeoffs) handled, the amount of non-aeronautical service
outputs, including concessions and other rental revenues, car parking, and numerous other services from
which airports generates revenues . On the inputs side two inputs are considered: labour and cost
input.
Pestana Barros and Dieke (2008) used the two-stage procedure of Simar and Wilson (2007) to
measure the efficiency of Italian airports. The Simar and Wilson technique was applied to find the most
489
efficient airports in terms of the total productivity by using the airports relative efficiency estimated
earlier by DEA form 2001 to 2003. Sarkis (2000) concluded the operational efficiency of US airports
using DEA. The input set choose were operational costs, number of airport employees, gates and runways
while the selected outputs were operational revenue, number of passengers, aircraft movements and
cargo. Hooper and Hensher (1997) studied the performance of six Australian airports over a 4-year period
using the TFP method. They considered deflated revenue index as an output, and three inputs considered
were labor, capital, and other. Diana (2010) studied the airport performance from selected New York
airports using a SFM (Stochastic Frontier Model) from June to August 2008 compared with the summers
of 2000 and 2007. The variable in this study were scheduled arrivals and departures, taxi, airborne, block
and gate arrivals. She found that the airport efficiency did not depend only on airport operational factors,
but also were more likely affected by the traffic management initiatives.
DATA ENVELOPMENT ANALYSIS (DEA)
Multiple inputs and outputs are essential to develop a strategic performance evaluation. In DEA, a
scalar measure of relative efficiency by comparing the efficiency achieves under study is called a DMU
(Decision Making Unit). The definition of DMU is to allow flexibility in obtaining well-define relation
between outputs and inputs. Its application is proposed to be applied to evaluate relative efficiency of
each considered airports in Malaysia. The basic DEA model as shown in Charnes et al. (1978), the
efficiency of an airport is measured as follows:

M
uy
i =1 i i 0
max

N
ui ,v j
vx
j =1 j j 0

M
uy
s.t. i =1 i il
1 l = 1,..., L

N
vx
j =1 j jl
ui , v j 0 i = 1,..., M , j = 1,..., N (1)
where L is the number of reference airports, y il and x jl are the ith output and jth input of the lth airport
respectively, and M and N are the number of outputs and inputs respectively. By controlling nonnegative
parameter u and v, subjects to the constraint that such ratio mush be less than or equal to unity for every
airports, the efficiency of an airport is measured as the maximum ratio of linear combination of outputs to
linear combination of inputs.
The solution of the linear programming will reveal how efficiently inputs are used for the given
output level. Note that this measure of DEA efficiency will focus on input-oriented and assumes constant-
return to scale (CRS) for airport operation activity.
The suggested extended DEA models were: the DEA-CCR (or Charnes, Cooper and Rhodes)
model (Charnes et al., 1978); the DEA-BCC (or Banker, Charnes and Cooper) model (Banker et al.,
1984); the Cross-Efficiency DEA model (Sexton et al., 1986; Doyle and Green, 1994); and the Super-
Efficiency DEA model (Andersen and Petersen, 1993). The study will utilize panel data for 5 years
(20052009), to examine the relative roles of dimension, managerial status and workload unit (WLU) in
determining the proximity of airports to the frontier of best practices.
The analysis is based on DEA model that allows for the incorporation of multiple inputs and
output in determining relative efficiency. Research will measure several important facets of airports
490
management such as productivities, efficiencies and financial result to current benchmark. Twenty
airports were selected for the study in order to evaluate their productivity and efficiency, unit cost, and
financial results respectively and airport characteristics to explain differences in performance and try to
adopt a modified frontier model to look at the efficiency of these airports.
DATA
Our sample includes 20 airports selected with different sizes which owned and governed by
Malaysia Airports Holdings Berhad (MAHB) located in several states of Malaysia. MAHB was licensed
by the Minister of Transport Malaysia to carry out its function as the airport operator since November
1992. The core activities include the management, operation and maintenance as well as development of
airports, with primary importance being placed on the operational efficiency, safety and security of
passengers, cargo and aircraft operations. MAB owned twenty airports in Malaysia and three overseas
operations with a range of aviation and non-aviation services. The data is compiled from various sources
including the International Civil Aviation Organization (ICAO), Airport Council International (ACI),
Malaysia Department of Civil Aviation (DCA), International Air Transport Association (IATA), direct
communication with airports and airport annual reports for which balance sheet data allowing the
combination of financial and operational related data are available to be examined here using DEA.
To measure airport productivity, one must first identify outputs that an airport produces and the
inputs it uses in producing these outputs. In this study, we divided the measurement into to two models
which separate the classes of services. The terminal services modeled as having two outputs tonnes of air
cargo handled and number of aircraft movements (landings and takeoffs) handled and the four
inputnumber of employees, runways, gates and checking desk. Movements have selected one
outputaircraft movement and three inputsnumber of runways, employees and runway area.
Movements are assumed to be produced under CRS (Gillen, 1994).
The identification of inputs and outputs are crucial requirement for frontier models. Several
criteria can be used in the selection. The first empirical criterion is the availability of inputs and outputs.
Second, a literature review can provide supporting evidence as can the professional opinion of the airport
managers.
EXPECTED FINDINGS AND SUMMARY
This paper will use the Data Envelopment Analysis (DEA) ranking methods to analyze the
relative performance of each individual Malaysia airport, and to fully rank both efficient, as well as
inefficient airports to provide a good understanding of the overall performance of the airports. For this
reason, research will try out different ranking approaches in order to fully rank the performance of
Malaysian airports by utilizing the DEA methodology. The approaches to rank the airports performance
through the exclusion of the airport being analyzed in the DEA parlance will be considered. Finally,
benchmarks are provided to help managers for improving the operations of poorly performing airports.
CONCLUSION
There have been considerable developments within the area of airport benchmarking in recent
years. The knowledge and development in this sector not far behind to the other sector for the practical
use of performance indicators techniques.
However, there are still yet to be decided and resolved effectively on globally scale for the
difficulties in dealing with problems of the variables which arising largely from the diversity of input and
outputs. Further research is needed to improve the airports performance measurement. Other
491
organizations, such as regulatory authorities may also participate to further improve the current practice in
this area.
REFERENCES
Andersen P.,&Petersen N.C. (1993). A procedure for ranking efficient units in data envelopment analysis.
Management Science39, 12611294.
Barros, C. P. (2008). Technical efficiency of UK airports. Journal of Air Transport Management 14(4),
175-178.
Banker, R.D., Charnes, A., &Cooper, W.W. (1984). Some models for estimating technical and scale
efficiency in data envelopment analysis. Management Science 30, 1078-1092.
Diana, T. (2006). Benchmarking airport efficiency: an application of data envelopment analysis. Air
Traffic Control Quarterly 14, 183202.
Diana, T. (2010). Can we explain airport performance? A case study of selected New York airports using
a stochastic frontier model. Journal of Air Transport Management 16(6), 310-314.
Doyle J.R., &Green R. (1994). Efficiency and cross-efficiency in data envelopment analysis: derivatives,
meanings and uses. Journal of Operational Research Society 45(5), 567578.
Francis, G., I. Humphreys, et al. (2002). The benchmarking of airport performance.Journal of Air
Transport Management 8(4), 239-247.
Gillen, D., &Lall.A.(1997). Developing measures of airport productivity and performance: An application
of data envelopment analysis. Transportation Research Part E-Logistics and Transportation
Review 33(4), 261-273.
Hooper, P. G.,& Hensher D. A. (1997). Measuring total factor productivity of airports-- an index number
approach.Transportation Research Part E: Logistics and Transportation Review 33(4), 249-259.
Martin, J. C.,& RomanC. (2001). An application of DEA to measure the efficiency of Spanish airports
prior to privatization. Journal of Air Transport Management 7(3), 149-157.
Oum, T. H., & YuC. (2004). Measuring airports' operating efficiency: a summary of the 2003 ATRS
global airport benchmarking report. Transportation Research Part E: Logistics and Transportation
Review 40(6), 515-532.
Parker, D. (1999). The performance of BAA before and after privatisation - A DEA study. Journal of
Transport Economics and Policy 33, 133-145.
Pels, E., P. Nijkamp, et al. (2003). Inefficiencies and scale economies of European airport operations.
Transportation Research Part E-Logistics and Transportation Review 39(5), 341-361.
Pestana Barros, C., & DiekeP. U. C. (2007). Performance evaluation of Italian airports: A data
envelopment analysis. Journal of Air Transport Management 13(4), 184-191.
Sarkis J. (2000). An analysis of the operational efficiency of major airports in the United States. Journalof
Operational Management 18, 335353.
492
GREEN SHIPPING IN MARITIME HUMAN CAPITAL DEVELOPMENT
Muhammad Zaly Shah, Lee Ghim Teck
ABSTRACT:
The gradual melting of the polar caps proves to benefit the shipping especially the oil tankers. The
thinning of the polar cap especially the North Poles resulting the opening of the NorthWest Passage on
the Canadian northern coast and Northern Passage on the Russian northern coast which connect the
Atlantic Ocean with the Pacific Ocean. The opening of these two passages allow shorter duration of
travelling time for oil tankers transit both ocean where these areas are known for their oil rich reserves in
the environment where the market price of crude oil at the all time high. Taking that into account, the
International Maritime Organization (IMO) in the recent adopted revision into the Standard Training and
Certification of Watchkeepers and Seafarers (STCW) 1978 Convention has incorporated in Part B
Chapter V of the Convention, the elements of navigation in polar caps water. The sea still remain as the
main medium of transportation of goods and for the ship or more specifically the cargo ships which act as
the carriers and responsible to transportation of 85% of the world trading goods being it in the dry or
liquid forms. It still remain as the most cost effective mode of transport, it is vital for the maritime
industry to ensure the sustainability of this mode of transport. The incorporation in the Chapter V of the
STCW Convention on ice navigation is the testimonial to the importance of human capital development
in ensuring sustainability of the maritime transportation. The paper is to share the areas of human capital
development in ensuring sustainability of this mode of transportation.
Keywords: Maritime, education and training, human capital development, assessment, continuous
professional development
493
INTRODUCTION
At the launching ceremony for World Maritime Day (WMD) 2009, the Honourable Admiral Efthimios E.
Mitro, the Secretary General of International Maritime Organization (IMO) mentioned Mankind is on
the horns of a dilemma. For, whether we like it or not, our collective way of life has become
unsustainable and we need to do something about it and soon. The choices we have made about the way
we lead our lives have been slowly eating away at the very support system that enable us to live and
breathe. This cannot, and should not, go on. We need to make some tough decisions, we need to make
them now and we need to act on them as one, with total and undivided commitment today and in the
future. The theme for the WMD 2009 was Climate change: a challenge for IMO too! As the main
cargo transportation mode which responsible for the movement of 85% world goods, the cargo ships
could not exonerate itself from the degradation of the marine environment and the world environmental
calamities. Various technological advancement and introduction of more stringent operating and design
rules and conventions already place into the standard operating procedures and ship design and
development, but it is still highly visible in term of natural disasters that reported today where the main
contributor are the human errors. The need to goes to the root cause of the matter is critical if the shipping
to remain a sustainable mean of transport as the most cost effective mode of vehicle.
LITERATURE REVIEW
After 150 years of industrialization, the planet are experiencing some of the worse natural
calamities such as greenhouse emission, global warming, rise of CO 2 , ozone depletion, and many more
that are visible today than two or three decade ago. The effect the are experience from these calamities are
such as draught, flood, melting of the polar caps, extreme low or high temperature, abnormal temperature
variation and many more. Through the years, IMO and the shipping communities and players have
initiated various activities to ensure the sustainability of the sea borne trade. Such initiatives as Marpol
73/78, Ballast Water Management, Double Hull, International Safety Management (ISM) Code by IMO
and Tanker Management & Self Assessment (TMSA) and Tanker Officer Training Standard (TOTS) by
INTERTANKO are few of the examples and initiatives to ensure ships operations do not contribute to the
environmental degradation. INTERTANKO is an organization which comprises of independence tanker
owners responsible for the management of oil and chemical tankers and their main concerned in ensuring
that their members ensure the highest standard of tanker operation are maintain as these tankers
contributed the highest CO 2 emission in comparison with their counterpart as reflect in Figure 1 below.
At sea, the effect of these natural calamities do raised concerns for the shipping industry which
depend on the mercy of the sea as a medium for transportation of goods by sea. On the other hand the
global warming and the melting of the polar caps do benefitted the shipping industry on another level.
Ships which carried nearly 85% of the world trade covered approximately 33,000 billion tonne-miles
contributed 870 million tones of CO 2 in 2007 which corresponding to 2.7% of the global CO 2 emissions.
The primary source of this emission is from exhaust gases and it is estimated that by 2050, ship emissions
may go up by between 150% to 250%. Even though in comparison with other mode of transportation,
shipping in general do seem to be an energy efficient means of transportation as demonstrated in Figure 2
below, but through increase efficiency, this form of emission could be reduced by further 25% to 75%
from the current level (2nd IMO GHG Study 2009). The table 1 below reflect the various studies made on
the fuel consumption, CO 2 emissions and projected growth as can be seem from the table, the percentage
of the world fuel combustion is remain at a very low level.
494
495
For the above purpose, human capital development is critical to ensure the reduction of GHG
initiative is achievable as this ship are manned and maintain by the same group of people who are running
the whole operation of the ships on the ground or at sea for this context. As through the years, various
initiatives and technological advancement have already put into place and impressive result have
demonstrated the effectiveness of these initiatives and enhancement. In 1970s the energy efficiency of
ship engines oil consumption is at about 215 gram/kw/hour and in 2000s the figure goes down to about
160 gram/kw/hour. Today effort are in place to reduce the CO 2 emission via various initiative and this
include human capital development in combating pollution prevention as the shipping community
realized that the 80% of the incident and accident occurred at sea are due to human errors but the
development of human capital is only about 20% compared with 80% of the technological advancement
in preventing pollution. This trend took a sharp turn in 1989 after the incident of Exxon Valdez, where a
crude oil tanker went aground off the coast of Prince William Sound in Alaska and suddenly the
international community realized that little progress was in place for human capital development and the
International Safety Management (ISM) Code came into placed.
Realizing the criticality of human capital in shipping, the IMO and the shipping community took
several initiatives to review and enhanced the existing conventions and codes relating to human capital
development in maritime industry to ensure it continual relevancy and bridging the training gaps and
industry demand. The recent concluded Manila Amendment which saw the complete review of the IMOs
Standard Training and Certification of Watchkeepers and Seafarer (STCW) Conventions 1978. The
revised Convention took into consideration on both the national and industry human capital need into the
emphasis of training and education for seafarers including the requirement of continuous professional
development. This paper will also be discussing on areas relating to protection of environment and
pollution prevention.
496
Green Shipping
Various efforts have been put into place to ensure the greening of the maritime transportation and
some of the mitigation options are shown in the Table 2 below. The options available were sparsely
divided into three areas namely Technology and Energy, Operational and finally Market-based. On the
Technology and energy portion, the reduction of sulfur content has already come into place as a new
Annex VI of the Marpol 73/78 Convention. The third option if very much link to incentives which could
be introduced to ship operators to encourage them moving toward application of green technology and
practice but this may involved high cost in term of investment and changes in the performance output
which may proves a failure taking into account the present economic return. The second option will be
leaning toward ensuring efficient management of the shipboard operation as well as optimization of ships
utilization to ensure maximum output is achieve based on the measured input. This is where the efficiency
of the personnel manning the ships as well as managing the ship operation plays a crucial roles to ensure
sustainability of the shipping.
Some of the international conventions and codes that were put into place with the objectives of
ensuring the system and procedures onboard the ships contribute to ensuring safer sea and cleaner ocean
are as below:
497
1. International Convention on Civil Liability for Oil Pollution (Brussel 1969 Protocal 76 & 92)
2. International Convention on Establishment of an International Fund for Compensation for Oil
Pollution Damage (Brussel 1971 Protocol 76 & 92)
3. International Convention for the Prevention of Pollution from Ships (Marpol 73/78)
4. International Convention Relating to Intervention on the High Sea in Cases of Oil Pollution
Casualties (Brussel 1987)
5. International Convention on Prevention of Marine Pollution by Dumping of Waste and Other
Matters (London 1972)
6. International Convention on Prevention of Pollution of the Sea by Oil (London 1954)
7. International Convention on Oil Pollution Preparedness, Response and Co-operation (London
1990)
8. International Convention on Liability and Compensation for Damage in Connection with the
Carriage of Hazardous and Noxious Substances by Sea (London 1996)
9. Safety of Life at Sea (SOLAS) 1974
10. Standard Training and Certification of Watchkeepers and Seafarers (STCW) 1978
11. Control of Harmful Anti-Fouling Systems on Ship (AFS) 2001
12. Control and Management of Ships Ballast Water and Sediments (BWM) 2004
13. International Safety Management (ISM) Code
14. International Code on Construction and Equipment for Ships Carrying Dangerous Chemical in
Bulk (IBC) Code
15. International Code on Construction and Equipment for Ships Carrying Liquefied Gases in Bulk
(IGC) Code
16. International Code for Safe Carriage of Packaged of Irradiated Nuclear Fuels , Plutonium, and
High Level Radioactive Waste on Board Ships(INF) Code
17. Code on Noise Levels on Board Ships
18. NOx Technical Code Technical Code on Control of Emission of Nitrous Oxides from Marine
Diesel Engine
19. Procedures for Port State Control
Standard Training and Certification of Watchkeepers and Seafarers (STCW) 1978
The ISM Code specified the need to ensure the seafarers employed onboard must complied with
the minimum standard of training as stipulated by the STCW 78 Convention and it is the company
responsibility to ensure that the seafarers are qualified. The process of ensuring the compliance is both
flag state and port state responsibility and on top of that the industry players such as OCIMF which
represent oil majors and charterers have their own vetting system in place to monitor operational
efficiency will always in place. The following sections of the STCW govern the above requirement:
Section A-I/4 Control Procedures linking assessment requirement to the requirement for ISM
and SOLAS
Section A-I/11 Revalidation of Certificate continued professional competence
498
Section A-I/14 Responsibilities of companies

The areas that were involved are as follows:
Chapter II and III of the STCW 78 -- Table A-II/1 and A-III/1
Bridge Resource Management
Passage Planning + ECDIS
Prevention of pollution of the marine environment and anti-pollution procedures, and equipment
Monitor compliance with legislative requirements
Oily water separator
Chapter II and III of STCW 78 -- Table A-II/2 and A-III/2
Forecast weather and oceanographic condition
Plan and ensure safe loading, stowage, securing, care during the voyage and unloading cargoes
Monitor and control compliance with legislative requirement
Chapter V Special Training Requirements for Personnel in Certain Types of Ships
Tanker Familiarization
Basic Training for oil and chemical tanker (A-V/1-1-1)
Advanced training for oil tanker operation (A-V/1-1-2)
Advanced training for chemical tanker operation (A-V/1-1-3)
Basic training for liquefied gas tanker cargo operation (A-V/1-2-1)
Advanced training for liquefied gas tanker cargo operation (A-V/1-2-2)
To ensure continuous professional competency, the seafarers concerned must revalidate their
professional certificate which is known as Certificate of Competency from the issuance body on five
yearly basis and documentary evidence in term of service record must be produced to ensure only genuine
practitioners are allow to continue the practice. The reasons for such a strict requirement are to prevent
possibility of accident occurred due to lack of competency and experience. All seafarers pursuing career
in the maritime industry must undergoes minimum standard of education, training and assessment to
ensure that they posses the right level of knowledge, understanding and proficiency for their respective
roles and responsibility onboard. For every level of competency acquired, they must fulfilled specific
shipboard service requirement and duration. This is to ensure that the knowledge and understanding
gained at the training institution can be translated to practice, thus improving on the efficiency and
proficiency of an individual seafarer. The assessment to ensure evidence in term of proficiency is carry
out in multiple levels and by different agencies including industry participations to ensure the seafarer
possess the right knowledge, understanding and attitude in ensuring the roles and responsibilities
entrusted on them can be carry out in a safe and efficient manner.
Shipboard continuous assessment are carried out by the shipping company through collaboration
with the industry partners and training institutions to ensure that continuous competency development are
in place and latest technological know-how is internalized so as possibility of human errors in operational
of these vessels is minimize. Shipping companies through the requirement from the industry have to
ensure that continuous professional development are in placed for the seafarers and it is monitored
through a system of audit and vetting done by government agencies as well as the industry players. Any
failure to ensure continuous professional competency may result in the operation license to be withdrawn
499
by the government agency or being barred from carry the cargo or even loss the chartering contract from
the client which are represented by the industry players. Shipping companies today are held responsible
for the training for their seafarers and documentary evidence of training being carryout either onboard or
ashore must be kept up to date for audit and vetting purposes. As the industry realized the weak link in
between safe operation and accident laid on the human element.
The operational efficiency are being carryout in non-mandatory manners, but for the industry that
is known to only complied with the mandatory requirement, shipping companies today invest a lot of their
resources in improving efficiency. Trainings related to operation efficiency such as energy conservation,
risk assessment, marine engine diagnostic, marine accident investigation and cargo handling simulation
are a norm for the shipping companies staff as they contributed to the saving of the ever increasing cost of
operation but at the same time reduces the green house gas emission.
SUMMARY
In ensuring green shipping be able to support sustainable transport especially maritime transportation ,
two prone approach have to be put into place, that is one on the ensuring that the opportunities for marine
pollution from taking place is minimize and eliminate altogether and second approach is the enhancement
of technological and operational efficiency. Technological development do ensure the reduction of
possible marine environmental disasters but technological enhancement alone will not be able to ensure
complete elimination as human errors are the major contributory factor in marine pollution. Thus the
closing of these gaps have to be look into in a more structure and organize manner through international
legislation and performance standard as shipping is international in nature. Human capital development is
an important capital investment that cannot be ignored if the industry is to achieve 75% reduction of CO 2
emission.
REFERENCES
IMO., Second IMO Greenhouse Gas Study 2009, IMO, London, 2009;
IMO, Chapter II-1 - Construction - Structure, subdivision and stability, machinery and electrical
installations, Safety of Life At Sea (SOLAS) Convention 1974, IMO
IMO, Annex VI - Regulations for the Prevention of Air Pollution from Ships, MARPOL - International
Convention for the Prevention of Pollution from Ships 1978, IMO,
IMO, Standard Training and Certification of Watchkeepers and Seafarer 1978 Convention, 2001 Edn.,
IMO London;
IMO, Polar Water Guidelines agreed by DE Sub-Committee, IMO News, Issue 2 2009, IMO, London;
ICS, Shipping, World Trade and The Reduction of CO2 Emissions, International Chamber of Shipping,
London;
IMO, Climate change: a challenge for IMO too!, IMO News, Issue 3 2009, IMO, London;
UNCTAD, IMO 2000 Update Study on Greenhouse Gas Emissions for ship, 2008, UNCTAD,
Lee G.T., Developing Human Capital to Operate in a Green Shipping, MIMA Shipping Seminar, Kuala
Lumpur, 2010
500
CASE STUDY BASED ON VISUAL INSPECTION TOWARDS BRIDGE DISTRESS AT PLUS

EXPRESSWAY CONNECTING SUNGAI MERAB KAWASAN PERINDUSTRIAN BANGI
Siti Aliyyah Masjuki3*, Shuhairy Norhisham1, Herda Yati Katman1, Fathoni Usman2,
Megat Asyraf Megat Ibnu Rashad3
ABSTRACT:
In recent two decades, rapid deterioration of existing bridge structures has become a serious technical and
economical problem in many countries, including highly developed ones. Bridge problem are often relate
with distress, defects, damage, deterioration and also faultiness during bridge constructions. These can
relate towards failure in existing bridge and shorten bridge life span. This report is mainly to highlight on
bridge problem encountered in Bangi, Selangor area and some preventive actions recommended in order
to ensure bridge in this is safe to sustain serviceability load and durable.
1
2
Senior Lecturer, Universiti Tenaga Nasional, KM 7, Jalan Kajang-Puchong, 43009 Kajang, Selangor, MALAYSIA
3
Tutor, Universiti Tenaga Nasional, KM 7, Jalan Kajang-Puchong, 43009 Kajang, Selangor, MALAYSIA
*Correspondence Author : Siti Aliyyah Masjuki, Universiti Tenaga Nasional, Malaysia

Email : Aliyyah@uniten.edu.my
501
BACKGROUND
Public Works Department of Malaysia (PWD or JKR as locally known) is the custodian of
majority of the nations road bridges. Based on recent bridges inspection the total number of JKR bridges
along Federal roads is over 7000. As in JKR database has 7002 bridges along the Federal roads in
Peninsular Malaysia. These bridges include culvert of span more than 0.5m. It is noted that more than
68% of the bridge stock are culverts meanwhile 83% are simple girder bridges. Based on JKR record in
terms of the construction of superstructure, it has shown that 81% of the structures are made up of
concrete. (1)
In present, problem in the existing bridges has become a serious technical crisis in our country.
Bridge problem are often relate with distress, defects, damage, deterioration and also faultiness during
bridge constructions. These can relate towards failure in existing bridge and shorten bridge life span.
Distress may be classified into some categories which are distress due to deficient design, construction
fault, unintended loading, and deterioration due to weathering effects. (2). Despite of these effects, high
impact due to vehicles collision towards some parts of bridges may also cause damage and continuously
deteriorate if left untreated. Therefore, preventive maintenance is required to preserve the structure, retard
further deterioration and maintain or upgrade functional condition. The method used for the maintenance
works cover some of rehabilitation, repairing works and reconstruction.
This report is mainly to highlight some of the bridge problem encountered in Bangi, Selangor
area and some preventive actions recommended in order to ensure bridge is safe, able to sustain
serviceability load and durable. Bridge selected is located at PLUS Expressway connecting Sungai Merab
Kawasan Perindustrian Bangi. It is approximately 15 km from Universiti Putra Malaysia (UPM). All the
defects and main constituents towards deteriorations were identified during the inspections.
Figure 1 shown the location of Bridge A (PLUS EXPRESSWAY) connecting Sungai Merab Kawasan
Perindustrian Bangi
Thorough inspection towards material defect, structural defect, geotechnical defects and
hydrological defects were done based on visual inspection. The results shown that there were several
common deficiency distinguished. Indeed, these problems also occurred else ware in Malaysia.
Preventive actions will be taken to attained designed objectives of ensuring safety quality, serviceability,
durability and economy. (3)
502
INTRODUCTION
Bridge rehabilitation related not only to technical and economical aspects, but also to social ones. For
instance, when old bridges with historical values demand to be restored and modernized to meet the
requirement of new traffic conditions, aesthetical aspect should be taken into account and the selection of
the materials and structural solutions in the bridge rehabilitation process itself is crucial. It should be
emphasized that in many situations, the rehabilitation process is more difficult and complex than
construction of new bridges, as for instance in the case of severely deteriorated large concrete structures.
Road Engineering Association Malaysia (REAM) has categorized and describes the problem encountered
as a basis for condition rating indicates as below:
i. Problems in concrete members

ii. Problem in steel members
iii. Bearing problems
iv. Joint problems
v. Excessive vibrations
vi. Impact of vehicles
vii. Vegetation growth.
Consequently bridge rehabilitation process should be preceded by assessment and evaluation of the
structure to determine its actual technical condition and to select the proper rehabilitation techniques and
materials. The anticipated period of the future service of the bridge and the traffic conditions during this
period should also be determined prior to the rehabilitation project and operations.
All the above general remarks show that bridge rehabilitation covers many technical and economical
problems, including inspection system. Bridge rehabilitation needs to be performed in every stage from
inspection before rehabilitation effectiveness after the works are completed and by highly experience
engineering staff.
GENERAL DISTRESS IN BRIDGE COMPONENT
Bridge component is cascaded into two constituent parts namely superstructure which consists of
parapet, deck, approach slab and cable, and superstructure which consists of wingwall, bearing, abutment,
pier, diaphragm and foundations. Deck slab, girders, abutments and piers are primary members
transmitting and bringing the loads from the bridge deck down through the foundation. On the other hand,
secondary members play as the important role i.e. expansion joints and bearing in order to allow
movement for displacement due to temperature changes (4). Parapet serves as protection to users,
carriageways and railways and it is designed based on usage type for pedestrian, crash barriers and
vehicular barriers.
Distress in bridge generally subdivided into 3 main categories which are structural defect, non-
structural defects and defects due to faulty installation of bridge equipments. Structural defects usually
caused by insufficient detailing and some of the constructions which is not following the specified
structural plan and detailing. Types of defects visible to naked eyes are honeycomb at some part of RC
beam or wall, and cracks at almost parts of the bridge depending on distress manner. Cracks at some part
of bridge area may lead to several further deterioration for instance efflorescence, discoloration, spalling,
sulphate attack on cement concrete, alkali aggregate reaction and corrosion of reinforced bars.
503
Non-structural defects on the other hand, commonly occurred due to supervision not within
specification, insufficient detailing as well as no practicality with the environment site, untidy work and
inexpert site officer and contractor. The examples of non-structural defects mainly govern with
settlement at bridge approach, ripping and crack to road pavement, unclean debris at site, erosion and
scouring, and inlet & outlet of culvert chocked. Lastly is failure in installation of bridge equipment. The
main causes of this failure commonly due to the bridge component installation were not within
specification and poor quality of workmanship.
Cracking, corrosion and other damage can all affect a bridge's load carrying capacity. Therefore,
all of the elements that directly affect performance of the bridge including the footing, substructure, deck,
and superstructure must be periodically inspected or monitored. Visual inspection is the primary method
used to evaluate the condition of the majority of the nation's highway bridges (5)
VISUAL INSPECTIONS CONDUCTED AT PLUS EXPRESSWAY CONNECTING SUNGAI

MERAB KAWASAN PERINDUSTRIAN BANGI
Bridge inspected is concrete prestressed M beam. By conceptual principle, concrete is durable

nevertheless required a frequent maintenance to ensure sufficient strength provided to transmit loading.
Conversely, concrete is susceptible towards some deteriorating agent which can affect its behavior.
Figures below explained the finding based on the inspection conducted.
i. STRUCTURAL DEFECTS:
1. FRP peeled of Improper fixing of FRP at the soffit of concrete beam.
504
Based on the figure shown above, can be seen that soffit of the beam has already been done with repairing
work, since there was FRP covered the surface. The basic FRP strengthening techniques, which is most
widely applied involves manual application of either wet lay-up or cold cured adhesive bonding.
Commonly this technique is applied as the external reinforcement is bonded onto the concrete surface. By
observations, some part of the FRP was peeled off due to improper fixing. Although FRP has vastly used
as part of repairing purpose; nonetheless, the behavior of FRP which is brittle materials is not advisable
for long term assessment.
2. Concrete spalling along the alignment of reinforcement.
Figures shown above are concrete spalling occurred along the alignment of reinforcement. Based on the
inspection, due to vehicular impact at the underside of beam concrete may weaken the strength and
spalling occurred. Apart of vehicular impact, insufficient concrete cover will lead to reinforcement
exposed eventually accelerate corrosion at reinforcement by the action of moisture and oxygen from the
air.
3. Hollow part visible at affected area caused by spalling
505
Due to vehicular impact and insufficient cover which caused to reinforcement exposed, there was a
portion of the spalled area contained hollow underneath the beam. By its mean, there are neither
aggregate at that area nor cement matrix. This will directly positioned the beam in jeopardy since will
directly reduce the strength of the beam.
4. Faulty installation of reinforcement resulted to reinforcement beam anchorage shown.
Figure shown improper installation of reinforcement at prestressed beam during concrete fabrication
which lead to protruding reinforcement between beam gaps. Exposed reinforcement will accelerate
corrosion process and deteriorate concrete surface.
5. Honeycomb to soffit of parapet wall resulted to expose of aggregate without cement paste.
Honeycomb will caused voids appearance as there were no cements paste at the area. This will influence
to water or any chemicals sipping through the concrete and affect the strength of it. This may cause due to
poor compaction of concrete during mixing as well as congested reinforcement at that particular area.
506
6. Scaling to soffit along beam caused to loss of surface mortar and some parts loosen the
aggregate.
Scaling will lead to loss of surface mortar in concrete exposing aggregates and in severe cases, resulting
to the aggregate. Scaling will further disrupt the performance of concrete strength.
7. Poor installation of reinforcement (Reinforcement swirling)
507
Based on the figure shown above, there was reinforcement swirled in the concrete. This might be from the
defect during manufacturing process. This problem should be taken seriously as reinforcement plays as
important role to provide strength in the concrete.
8. Concrete popped-out in between to soffit of parapet wall attached with beams connecting to
carriageway
Concrete pop out might be due to vehicular impact. The concrete has loosened the strength affected by
water from the deck percolating to the concrete. The reaction will also deteriorate the surface of concrete
lead to efflorescence and staining.
9. Crack along end of beam to parapet wall. Due to crack, water easily sipped through the
opening area, affecting the concrete.
508
10. Crack at joint between two beams, on top of pier.
Crack found at this area was still in early stage. However, if it is not being treated, it will further prolong
and cause distress in concrete especially towards the strength on the ability to sustain load.
11. Poor plastering work at the end beam connecting parapet wall and beam
Although poor plastering doesnt really contribute towards major distress, somehow it can influences
water to move into the concrete through crack opening and further deteriorate the concrete by weathering
action.
509
12. FRP peeled off
FRP peeled off seen almost around the affected surface. For this particular case, maintenance is highly
required in order to revamp this problem since the used of FRP is not catered for a long term remedial
action.
13. Joint not properly installed causing gap between beam and joint.
Water can be easily percolating through the opening area and deteriorate the concrete. This resulted from
the deck water moving into the expansion joint installed at the bridge approach. In addition to this case, it
has extended to the end diaphragm. The process will slowly continue and simultaneously affected the
reinforcement inside the concrete. In case there were crack occurred, it will facilitate to corrosion.
510
14. Efflorescence, Discoloration and Staining to soffit of beam and pier.
Efflorescence occurred as in white deposit appeared on the surface of concrete. This will be due
to the reaction of Calcium Hydroxide from cement react with Carbon Dioxide to form Calcium
Carbonate. This will have no significance towards the strength of concrete however, as reaction continues
it resulted to corrosion. Similar to discoloration and staining, they appeared as in patchiness on the surface
as well as brownish rust stains. Rust colored stain may be due to the pyrite react with water and oxygen
form sulfuric acids which concurrently emerge in rust staining.
Non destructive Rebound Hammer Test has been conducted on the pier surface to measure the
strength of the concrete at the affected area. It is based on the principle that the rebound of an elastic
mass depends on the hardness of the surface against which the mass impinges.
Results obtain shown that concrete strength at the affected area range in 35MPa to 40MPa
meanwhile for non-affected area strength obtained was in between 40MPa to 45MPa. Results have
revealed that affected area at pier consequence towards low concrete strength towards the unaffected area.
511
15. Crack along gabion, due to settlement of soil.
Settlement resulted to crack at the gabion and also at the end of diaphragm. This is due to the action of
soft soil underground settled over times.
ii. NON-STRUCTURAL DEFECTS & POOR INSTALLATION OF BRIDGE EQUIPMENT
1. Outlet of drain pipe chocked and inadequate extension drain pipe protruded length to soffit
of paparet wall connected to beam.
512
Water mark can be seen due to the choked inside the pipe. In order to overcome the problem,
drain pipe has to be cleaned immediately to allow water to flow easily. Inadequate protruding of drain
pipe is also some part of defect due to faulty installation of bridge component.
CONCLUDING REMARKS
Based on the inspection, the major problem occurred lead to structural defects were crack to soffit
of beam and water percolating through the deck of bridge and affected the end diaphragm as well as the
beam. Besides, there were also some minor defects occurred. However, if there are no remedial actions
taken, it will further prolong the deterioration and more defects will arise over period of time. Corrosion
found almost at every part of bridge which contain reinforcements. Due to corrosion it may lead to further
deterioration for instance carbonation, spalling of concrete, chemical attacks and eventually reduce the
strength of concrete.
As for remedial works, there were some preventive actions suggested to overcome distress in concrete,
stated as below:
1. Concrete cover should be sufficient to avoid expose reinforcement. Pressure grouting is most
recommended to avoid this problem.
2. Strengthening is highly required to avoid reoccurrence of cracks. Based from inspection,
strengthening (FRP) has been implied to the bridge however there were poor in maintenance
work. Maintenance should be done frequently since the use of FRP is not mending for long term
strengthening. Besides FRP strengthening.
3. Epoxy Resin Injection which is highly required for remedial action towards crack especially at
bridge substructure.
4. Used of crack sealant at the affected area to minimize water percolating through the slab.
5. Further corrosion can be avoided by cathodic prevention.
In conclusion, distress in bridge can be minimized if all the remedial actions are implemented. This will
subjected to long bridge live span and ensure bridge in this area is capable to sustain load.
REFERENCES
1. BRIDGE MANAGEMENT IN MALAYSIA, Mohamad Razali bin Othman Chin Wei Choong &
Leow Choon Heng
2. BRIDGE ASSESSMENT AND REHABBILITATION , Prof Mohd Saleh Jaafar, Universiti

Putra Malaysia
3. BRIDGE MANAGEMENT , M.J. Ryall
4. BRIDGE PROBLEM IN MALAYSIA, Ir. Dr. Ng See King, Ir. Ku Mohd Dani Ku Muhamud,
Evenfit Consultant Sdn. Bhd.
5. http://www.ndted.org/AboutNDT/SelectedApplications/Bridge_Inspection/
513
514
Sustainable Urban Transport Research Centre (SUTRA)
PENGENALAN
Pengangkutan bandar mapan merupakan bidang yang melibatkan kepakaran dari perbagai bidang
dan disiplin. Pusat ini mengumpul dan menggembleng tenaga-tenaga pakar ini untuk membentuk
satu pasukan yang bersepadu. Pusat seperti ini amat diperlukan di Malaysia untuk memainkan
peranan yang aktif dan menerajui usaha untuk memajukan sistem pengangkutan pintar bagi
membina bandar mapan.
OBJEKTIF SUTRA, antaranya
1) Membangun keupayaan teknologi pengangkutan pintar melalui aktiviti penyelidikan dan

pembangunan dalam bidang asas dan gunaan seperti kajian binaan fizikal (physical
architecture), binaan logik (logical architecture), kajian sistem komunikasi, sensor,
bahan-bahan pintar dan termaju serta berbagai-bagai aplikasi dan penyebaran data serta
maklumat.
2) Membina tenaga manusia yang berkeupayaan dalam bidang-bidang berkaitan dengan

pengangkutan pintar dan bandar mapan melalui latihan ikhtisas, program penyelidikan,
kursus pendek dan aktiviti jangkauan. Keupayaan yang dimaksudkan ialah dari segi
teknologi, kaedah-kaedah perlaksanaan, analisis ekonomi perlaksanaan pengangkutan
pintar dan kesannya kepada pengguna.
3) Merangka strategi pembangunan dan pemasangan sistem pengangkutan pintar dan bandar
mapan melalui penyelidikan dan kajian pengangkutan.
Untuk keterangan lanjut sila layari
http://pkukmweb.ukm.my/~sutra/index.html
LYCORPOLE
PRODUCTS
Street Lighting Poles
Decorative Lighting Poles
Mid-Hinged Columns
Park Lighting Poles
Power Transmission Poles
Traffic Sign Gantry
Stadium Floodlighting Masts
CCTV Poles
Traffic Light Poles
(102382-M)
Power Distribution Poles
Flag Masts / Flag Poles
Floodlighting Poles
High Masts
Antenna Poles
Mono Poles
Traffic Signal Masts
Telecommunication Poles
Advertising High Masts
Standards
Lycorpole Hot Dip Galvanized columns, masts and poles are designed and manufactured in accordance with
BS EN 40, BS 6399-2, BS EN 10025, BS EN 1011-1, 2, Institution of Lighting Engineers Technical Report No. 7, AS
4100, AS 1170.2, AS 1798, AS 2979, AWS D1.1 and AASHTO Standard Specifications for Structural Supports for
Highway Signs, Luminaires and Traffic Signals. Lloyds Register
ISO 9001
Hot Dip Galvanized to BS 729, BS EN ISO 1461, ASTM A123, ASTM A153 or AS/NZS 4680 is a standard feature,
Where specified upon request, our Hot Dip Galvanized products can be enhanced with additional Powder Coating
with oven baked PE-F polyester based powder to BS 6497 to withstand weather and UV resistance for anti-chalking
and colour stability.
No. 11 Jalan Majistret U1/26, Seksyen U1, Hicom-Glenmarie Ind. Park, 40150 Shah Alam, Selangor Darul Ehsan, Malaysia.
Tel: (603) 7880 3750 Fax: (603) 7880 3720 Email: lysaght_sales@lysaghtmarketing.com.my
IPOH SARAWAK SABAH SINGAPORE
Tel: (605) 546 8278 Tel: (082) 245 145 Tel: (603) 7880 3750 Tel: (65) 6861 7122
Fax: (605) 546 3589 Fax: (082) 237 311 Fax: (603) 7880 3720 Fax: (65) 6862 1242
Email: lysaght8@streamyx.com Email: lysaght@singnet.com.sg
Website: www.lysaghtmarketing.com.my
Giving is a gesture of showing how much you care
Bingkisan Hamper
Baby Hamper
Festive Hamper
Corporate Gift
Food Hamper Door Gifts
Occassional - Sports Day, Graduation,
Family Day Hamper Customised door gifts for :
Birthday Party
Family Day
Annual dinner
Birthday Farewell Party
& Wedding Favours Reunion Day / Gathering
Aqiqah
Party Pack
Open house
Gift Packaging
House warming
Cake / Cup Cake
Clown
Balloon decor
www.facebook.com/bingkisan Party Planner
hello@bingkisan.my
Door Gifts
Boxes/Baskets
Wedding Tray Rental
Hantaran
Invitation Card
Gift Bag/Paper bag
Green Transportation
for Future Generation
Malaysian Universities Transportation Research Forum and Conference (MUTRFC) is jointly organized by
the Department of Civil Engineering, UNITEN, in collaboration with the Transportation Science Society of
Malaysia (TSSM) and supported by the Institution of Engineers, Malaysia (IEM) Transportation and
Highway Technical Division. This conference provides an interdisciplinary forum for the presentation of
the current and latest in research, development and commercialization of any transportation products
including consultation activities.
The topics of interest include, but are not limited to Transportation energy issues, Green Transportation,
Intelligent Transport Systems, Road and Rail Transport Modes, Public Transportation System,
Motorcycles and Non-motorized Transport, Maritime Transport and Ports , Airports and Aviation, Low
Cost Carrier (LCC), Transportation Demand and Modeling, Traffic Control System and Policy, Transport
Demand Management, Transportation Information System and GIS, Physical Distribution and Logistics,
Transport and the Environment, Traffic Accident and Road Safety, Transportation land use and Spatial
Analysis, Privatization and Financing of Transport, Issues in Transport Economics, Roads and Highway,
Road Infrastructure and Maintenance, Highway Planning, Pavement Engineering and Materials and
other related topics.
Official Website: www.uniten.edu.my/go/mutrfc2010

MUTRFC2010 Full Version Proceeding

Hochgeladen von

Dokumentinformationen

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

MUTRFC2010 Full Version Proceeding

Hochgeladen von

Copyright:

Verfügbare Formate

Green Transportation for

Engr. Mohd Sufian Abdul Karim

Engr. Shuhairy Norhisham

Engr. Herda Yati Katman

Engr. Mohd Amiruddin Bin Zainol

Perpustakaan Negara Malaysia : Cataloguing-in-Publication Data

Cover Design : Fathoni Usman

Printed and bound in Malaysia by Universiti Tenaga Nasional

3) To promote a knowledge and research environment in transportation engineering

Universiti Tenaga Nasional Mohd Sufian Abdul Karim

1) Professor Dr Ahmad Farhan Mohd Sadullah,

2) Prof. Ir. Dr. Ruslan Hassan

3) Ir. Looi Hip Peu

12.30pm 1.45pm Lunch

Professor Dr Ahmad Farhan Mohd Sadullah, Director General,

Professor Farhan is presently the Chair of the World Road Association

Synopsis Presentation: Transportation Research in Malaysia: Are we making the difference?

Ir. Dr. Ruslan Hassan

T1-A19-3C,Science & Technology Complex

Synopsis Presentation: Transportation Sustainability in a Climate-Constrained World

Ir. Looi Hip Peu

Vice President, Malaysian Green Building Confederation (2010/11)

Organizing Committee vii

Proceeding Committee vii

Invited Keynote Speakers viii

001 PG TRIP PRODUCTIVITY EVALUATION OF BUS SERVICE: MEDAN KIDD 1

002 G PASS-BY NOISE OF MALAYSIAN ARMED FORCES (MAF) VEHICLES 11

003 G AIRCRAFT NOISE LEVEL IMPACT AT AIRPORT AND ITS 19

004 PG INCREASED USE OF OWN CAR IN TRIPOLI LIBYA 33

005 PG INFLUENCE OF BLENDING INTERACTION OF CRUMB RUBBER 43

006 PG PRELIMINARY ASSESSMENT ON RELIABILITY OF PUBLIC BUS 49

007 G APPLICATION OF PROBABILISTIC DYNAMIC PROGRAMMING IN 59

009 PG USING THE BEHAVIORAL SCIENCES THEORY AND STRUCTURAL 67

010 PG EFFECT OF THERMOPLASTIC COPOLYMERS ON 79

012 PG REGIONAL FREIGHT TRANSPORTATION PLANNING: AN OVERVIEW 99

013 PG EFFECT OF RAINFALL ON MICROSCOPIC TRAFFIC FLOW 117

014 UG DESIGN OF BATTERY PACK FOR ELECTRIC VEHICLE BASED ON 127

016 UG STUDIES OF ELECTRIC MOTORS FOR LIGHT-WEIGHT ELECTRIC 135

017 PG QUICKBIRD IMAGE-BASED ESTIMATION ON NUMBER OF 149

018 G TEST PROTOCOL FOR MALAYSIAN ARMED FORCES (MAF) 161

019 PG IMPROVISED ENGINEERING SPECIFICATION DESIGN FOR ROAD 169

020 UG CASE STUDY OF EV CONTROLLER AND POWER MANAGEMENT 175

022 PG EXPLORATION OF 85th PERCENTILE OPERATING SPEED MODEL ON 203

023 UG BUS TRANSIT SYSTEM MODELING USING PARAMICS: A CASE 213

024 PG KEEPING CITIES MOVING: A LITERATURE REVIEW ON GREEN 221

026 PG E-ASSIST: A TRANSPORT DEMAND MANAGEMENT EXPERT 229

027 G TIMES SPENT IN DRIVING TAXI A CASE STUDY IN PUTRAJAYA 241

028 PG ESTIMATION OF DISPERSION OF CARBON MONOXIDE (CO), 249

029 PG INSTITUTIONAL RELATIONSHIP IN PORT CITY, PENANG MALAYSIA 259

030 PG THE PHYSICAL DEVELOPMENT OF PORT CITY 269

031 PG THE HEADWAY PATTERNS AND POTENTIAL PARAMETERS OF BUS 279

033 PG EXPERT SYSTEM FOR PAVEMENT CONDITION RATING 303

034 PG FREE-FLOW SPEED AT TWO-LANE HIGHWAYS IN MALAYSIA 311

035 PG TRAFFIC ACCIDENTS ANALYSIS ALONG EAST COAST 323

036 PG PROPERTIES OF BITUMINOUS BINDER MODIFIED WITH WASTE 333

037 PG FINITE ELEMENT ANALYSIS OF FLEXIBLE PAVEMENT SUBJECTED 345

038 PG THE ARRANGEMENT OF SHUTTLE TRAIN TRANSPORTATION 359

039 PG DEVELOPMENT OF ONELANE ONRAMP MODEL FOR 369

040 G CUBE DYNASIM AND AASIDRA COMPARISONS IN ANALYZING A 381

041 G A THEORETICAL APPROACH IN EVALUATING THE POTENTIAL 391

043 PG EVALUATION OF THE LABORATORY COMPACTING EFFORT ON 411

044 PG TYPES OF DAMAGES ON FLEXIBLE PAVEMENT FOR MALAYSIAN 421

045 PG THE EFFECTIVENESS OF A ROAD TRANSPORTATION NETWORK 433

046 G ECONOMIC IMPACTS OF THE SHIFTING OF FUEL CONSUMPTION IN 445