CRITICAL EVALUATION OF ROADWAY CAPACITY OF MULTILANE HIGH SPEED CORRIDORS UNDER HETEROGENEOUS TRAFFIC CONDITIONS THROUGH TRADITIONAL AND MICROSCOPIC SIMULATION MODELS
235
Paper No. 566
CRITICAL EVALUATION OF ROADWAY CAPACITY OF MULTILANE HIGH SPEED CORRIDORS UNDER HETEROGENEOUS TRAFFIC CONDITIONS THROUGH TRADITIONAL AND MICROSCOPIC SIMULATION MODELS
S. VELMURUGAN*, ERRAMPALLI MADHU**, K. RAVINDER***, K. SITARAMANJANEYULU**** & S. GANGOPADHYAY*****
ABSTRACT
The major factors affecting the Road User Costs (RUC) are the speed coupled with traffic flow characteristics at which vehicles operate on roads, which in turn determines fuel consumption and other cost components per unit distance traveled. Considering this, the Government of India has been involved in roadway capacity augmentation by building multilane divided carriageways to link major cities through the implementation of various projects, like, Golden Quadrilateral, NorthSouth, EastWest and Expressway Corridors during the last decade. These radical changes in road network coupled with radical advancements in vehicle technology have resulted in huge variations in speed  flow characteristics, which necessitated the evolution of exclusive speedflow equations and roadway capacity for multilane highways. Accordingly, an attempt has been made in this Paper to explicitly study the speed  flow characteristics on varying types of multilane highways encompassing fourlane, sixlane and eightlane divided carriageways in plain terrain. From the collected data, free speed profiles and speed  flow equations for different vehicle types for varying widths of multilane highways in the country has been developed based on traditional and microscopic simulation models and subsequently roadway capacity has been estimated. Further, the lane change behavior of different vehicle types has been extensively studied and its impact on roadway capacity has been estimated on multilane highways. Finally, the Design Service Volume for varying types of divided carriageways including fourlane, sixlane and eightlane has been evolved with reasonable degree of authenticity under the prevailing heterogeneous traffic conditions on multilane highways in India.
Key Words: Free Speed, SpeedFlow Equations, Roadway Capacity, Lane Change Behaviour, High Speed Corridors;
1 PREAMBLE
The sustained economic growth in India in recent years has brought opportunities and challenges to the planning and management of the Indian transportation system. Like in other developing countries, the transportation system in India is characterized by limited roadway infrastructure and the lack of operation and management experience. Among the most critical issues in highway planning and management is to determine the roadway capacities of any highway. As such, India has one of the largest road networks in the world hovering around 3.5 million km at present. For the purpose of management and
administration, roads in India are divided into five categories namely, National Highways (NH), State Highways (SH), Major District Roads (MDR), Other District Roads (ODR) and Village Roads (VR). NHs are intended to facilitate medium and long distance intercity passenger and freight traffic across the country and they are also serve as main arterial roads which run through length and breadth of the country connecting sea ports, state capitals, major industrial and tourist centers. Though the NHs constitutes less than 2 per cent of the total road network, but carries 40 per cent of the total road traffic. The road infrastructure and available
* Contact Author and Scientist, TE & TPA
**
***
****
*****
Scientist, TE & TPA
Scientist, TE & TPA
Scientist, PED
Director
} Central Road Research Institute (CRRI) Mathura Road, CRRI (P.O.) New Delhi  110020, India Email: vms_04@yahoo.co.in; vms.crri@nic.in
Journal of the Indian Roads Congress, OctoberDecember 2010
236
VELMURUGAN, MADHU, RAVINDER, SITARAMANJANEYULU & GANGOPADHYAY ON
transport services in the country are highly inadequate for achieving faster movement of passenger and goods movement in comparison with the situation in the developed world. Many sections of the highways are in
need of capacity augmentation, pavement strengthening, rehabilitation of bridges, improvement of riding quality, provision of traffic safety measures, etc. Eventually it
has been realized by the Government of India that the
above mentioned gross inadequacies in the NH network coupled with congestion caused by heterogeneity in traffic
mix is contributing to huge economic losses in terms of
high Road User Cost (RUC), which is also contributing to high rate of road accidents. Realising the present shortcomings in the transport sector, the Government of India has initiated massive construction programmes of highways linking major cities/activity centres. This has led to gradual growth in the quantum of NH network, which was around 22,255 km in 1951 has risen to 70,548 km as of March 2009 (Velmurugan et. al., 2009).
Considering the above mandate of the Government, it
was felt essential to quantify the investment made on the
multilane highways by developing RUC models exclusive to multilane highways. Eventually, a study has been undertaken by the Central Road Research Institute (CRRI), New Delhi under the aegis of Planning Commission, Government of India in 2008 focusing on evolving free speed and speed  flow relationships and finally develop RUC models for high speed corridors. The term 'High Speed Corridors' used in this study implies fourlane, sixlane and eightlane divided intercity highways. The automobile industry and road design standards in India have also undergone tremendous changes in the recent decade. Therefore, it was considered necessary to take a look at the changing trends of prevailing speed  flow characteristics considering the emerging high speed corridors of India. In 1982, speed  flow studies were carried out as a part of Road User Cost Study (RUCS) and they are subsequently updated in the years 1992 and 2001 for varying carriageway widths (CRRI, 1982, Kadiyali, 1992 and CRRI, 2001). From here onwards, these studies will be referred in this Paper as RUCS1982, URUCS1992
and URUCS2001, respectively. Basically, the speed  flow relationships developed in the above studies encompassed only up to four lane divided carriageways as the multilane highways beyond such carriageway widths were not existent at that time in India.
In this Paper, an attempt has been made for the first time to explicitly study the free speed profiles and speed  flow characteristics on varying types of multilane highways covering fourlane, sixlane and eightlane divided carriageways in plain terrain. In order to assess these characteristics, Time Mean Speed (TMS) under free flow conditions and Space Mean Speed (SMS) coupled with traffic flow data was extensively collected spread over different regions of India. From the collected data, free speed profiles of different vehicle types on high speed corridors and speed  flow equations have been developed based on traditional and microscopic simulation models. Subsequently, capacity norms for such high speed corridors were also evolved. The impact of typical Indian driving behavior i.e. how the lane change behavior affects roadway capacity on multilane highways has been assessed through microscopic simulation approach. Finally, based on the results derived in this study, the Design Service Volume (DSV) for varying types of multilane highways including fourlane, sixlane and eightlane divided carriageways has been evolved with reasonable degree of authenticity under the prevailing heterogeneous traffic conditions. The outcome from this study is expected to form an important input for developing RUC models exclusively for varying type of multilane highways. This Paper has been structured as given below.
In the next section, an exhaustive overview of speed  flow studies carried out in India and elsewhere for multilane highways is summarized. The details in respect of free speed and speedflow data collected at various road sections are discussed in Section 3 coupled with the description of data collection methodology. The results derived from free speed analysis are presented in Section 4, whereas speed  flow equations developed for different vehicle types covering varying carriageway widths on multilane highways based on traditional and microscopic simulation models is dealt in Section 5. The
Journal of the Indian Roads Congress, OctoberDecember 2010
CRITICAL EVALUATION OF ROADWAY CAPACITY OF MULTILANE HIGH SPEED CORRIDORS UNDER HETEROGENEOUS TRAFFIC CONDITIONS THROUGH TRADITIONAL AND MICROSCOPIC SIMULATION MODELS
237
procedure adopted for evolving the roadway capacity norms for varying carriageway widths on multilane highways is also discussed in the same section. The impact of lane change behavior on roadway capacity is presented in detail in Section 6. Finally, the conclusions emerging from this study are discussed in Section 7.
2 CAPACITY STUDIES ON MULTILANE HIGHWAYS
This section focuses on the studies accomplished for multilane highways only. As such, the determination of highway capacity is one of the most important applications of any traffic theory (Kerner, 2004). Some previous theories and empirical researches focused on the interrelationships among the influence of capacity, traffic features, geometric elements, environmental conditions and temporal weather factors on interrupted multilane highways (see for example, Hoban, 1987; Iwasaki, 1991; Ibrahim and Hall, 1994 and Shankar and Mannering, 1998). Many years of research has led to the development of theories and methodologies in roadway capacity analysis in the developed countries. For example, the Highway Capacity Manual (HCM) developed in the United States of America describes roadway capacity under ideal conditions and then estimates practical capacities under prevailing conditions in the field. USHCM 2000 (TRB, 2000) suggested that a maximum flow rate that can be achieved on a multilane highway is 2200 PCU/hour/lane. The Danish method is also a modification of U.S. HCM to suit Danish conditions. The adjustment factors in the Danish method cause a steeper capacity reduction than in USHCM 2000 as the conditions become less ideal and therefore, the capacity under ideal conditions on a fourlane highway is 2300 PCU/hour/lane on Denmark highways (Nielsen and Jorgensen, 2008). Similarly, in Finland and Norway too, USHCM 2000 (TRB, 2000) was followed with minor modifications to suit the local conditions and the roadway capacities obtained by the Finnish and Norwegian methods for multilane highways is 2000 PCU/hour/lane The structure of the Swedish method is similar to the USHCM (1995) and it uses the 1995 HCM adjustment factors for the roadway width, whereas other adjustments
factors are mostly omitted. Consequently, the Swedish method yielded higher capacity estimates and the estimated capacity of fourlane divided highways was 4200 PCUs/hour per direction (Luttinen and Innamaa, 2000). The Australian method for analysis of capacity was basically same as that of USHCM method with the basic difference being additional modification has been suggested for specific problems. Under ideal conditions, the average minimum headway of 1.8 seconds was considered and maximum flow of 2000 vehicles per hour per lane was assumed. The succeeding paragraph focuses on the roadway capacity evolved in Asian countries like Indonesia and China for multilane highways wherein largely heterogeneous traffic conditions as experienced on Indian highways is witnessed.
Bang et. al. (1997) in their study for establishing Indonesia HCM mentioned that travel speed (synonymous with journey speed) as the main measure of performance of road segments, since it is easy to understand and to measure, and is an essential input to road user costs in economic analysis. Travel speed is defined in this manual as the space mean speed of light vehicles (LV) over the road segment as given below:


^{L} 

V = 
_{T}_{T} 
where, V 
= 
L 
= 
TT 
= 
(1)
space mean speed (km/h) of Light Vehicles (LVs)
length of segment (km)
mean travel time of LVs over the segment (in hours)
Using this analogy, the capacity of multilane highways has been estimated as 2300 LVs/hour/lane for Indonesian multilane highways. In the case of Chinese conditions, based on the field data collected, VTI highway simulation model was calibrated and validated and this model was used for the determination of Passenger Car Equivalents (PCE) and speedflow relationships for different terrain types in parallel with multiple regression analysis of empirical speedflow data. The results showed that the freeflow speeds of vehicles were substantially low and
Journal of the Indian Roads Congress, OctoberDecember 2010
238
VELMURUGAN, MADHU, RAVINDER, SITARAMANJANEYULU & GANGOPADHYAY ON
that the roadway capacity was also marginally lower (2100 PCEs per lane on fourlane divided carriageways) under Chinese conditions as compared with the values obtained for Indonesian multilane highways. Further, Yang and Zhang (2005) have established based on their extensive field survey of traffic flow on multilane highways in Beijing and subsequent empirical model development that the average roadway capacity per hour per lane on fourlane, sixlane and eightlane divided carriageways is 2104, 1973 and 1848 PCUs, respectively. This is unlike HCM results obtained for many developed countries which prescribe that average capacity per lane on different highways is equal as they assume that highway capacity is constantly proportional to the number of lanes on multilane divided carriageways.
Based on the review of above studies in both developed and developing countries it is obvious that the roadway design and traffic control practices are mostly country or region specific and hence cannot be simply transferred to any country for direct applications. In this context, it is be noted that the roadway capacity and the conditions for adjustment are vastly different on Indian roadways as the local roadway design (i.e. lane width, curves and grades), vehicle size and more importantly, traffic mix and behaviour of a driver especially lane changing and lane discipline phenomenon are entirely different. Further, since there is not a systematic approach to this problem, coupled by a lack of fundamental data, the adjustment factors from say, the US HCM 2000 (TRB, 2000) cannot be easily revised and applied to Indian highways. This is because adherence to lane discipline characterizes homogeneous traffic in the developed nations whereas very loose lane discipline describes heterogeneous traffic which is very much an integral part of all roadways in India including multilane highways. This is due to fast moving vehicles cars, goods vehicles, motorized two wheelers sharing the same road space with bicycles, farm tractors, tractor trailers and other types of slow moving vehicles (like cycle rickshaw, animal drawn vehicles, etc.) on the Indian traffic scene accounting for varying proportion on multilane highways depending on its geographical location.
Ironically, most of the models discussed above developed for homogeneous condition are not applicable for the heterogeneous traffic prevalent on Indian roads. Eventually, the first major research effort in India in this direction was done as part of the RUCS1982 and this was followed by URUCS1992 and URUCS2001. IRC64 (1990) suggested the tentative DSV of 40,000 PCUs for the fourlane divided carriageway in plain terrain, which is significantly lesser than the values evolved in most of the developed and developing countries and therefore the need was felt for revisiting the DSV values evolved under IRC64. Consequently, many research studies (Kadiyali, et. al., 1991, Tiwari, et. al., 2000, Velmurugan et. al., 2002, 4. Chandra S. and Kumar U., 2003, Reddy, et. al., 2003, Chandra, 2004, Errampalli, et. al., 2004, Velmurugan, et. al., 2004, Dey, 2007, Errampalli, et. al., 2009, Velmurugan et. al., 2009) aimed at assessing the roadway capacity for varying carriageway widths including single lane, intermediate lane, twolane bidirectional and fourlane divided carriageway widths covering different terrains have been carried out during the last two decades. URUCS2001 recommended tentative roadway capacity of 70,000 90,000 PCUs/day for a fourlane divided carriageway in plain terrain (17502250 PCU/hr/lane considering 10 per cent peak hour flow). Chandra and Kumar (2003) studied the effect of roadway width on capacity under different volume capacity ratios and varying proportions of vehicles. Shukla (2008) studied the mixed traffic flow behavior on fourlane divided highway for varying conditions of traffic volume and shoulder and developed a simulation model for the observed traffic flow to estimate roadway capacity under these conditions. To understand the traffic flow behavior on fourlane divided highways under mixed traffic condition, the arrival pattern of vehicles, speed characteristics, lateral placement of vehicles and overtaking behavior was analyzed. Shukla (2008) further reported that the roadway capacity of four lane divided carriageways as 4770 vehicles/hour (vph) in each direction is estimated for 'all cars' situation.
This exhaustive look at the literature indicates that no substantial work has been carried out for establishing the
Journal of the Indian Roads Congress, OctoberDecember 2010
CRITICAL EVALUATION OF ROADWAY CAPACITY OF MULTILANE HIGH SPEED CORRIDORS UNDER HETEROGENEOUS TRAFFIC CONDITIONS THROUGH TRADITIONAL AND MICROSCOPIC SIMULATION MODELS
239
roadway capacity and DSV for varying carriageway widths on multilane highways covering fourlane, sixlane and eightlane divided carriageways for the heterogeneous traffic mix prevalent on Indian highways with reasonable degree of confidence and hence this research endeavor can be termed as a significant attempt in this direction. The following section describes the methodology adopted for accomplishing this study.
3 STUDY METHDOLOGY
As mentioned in Section 1, the primary objectives of the present study are to assess the free speeds of different vehicle types on multilane high speed corridors and develop a realistic speed  flow equations for estimating the roadway capacity. From these results, DSV values are proposed to determine for varying types of multilane high speed corridors. In order to achieve the above envisaged objectives, separate methodologies are adopted for deriving free speed profiles and speed  flow relations for varying carriageway widths. To accomplish the above stated objectives, the following studies were conducted:
a) Free Speed studies
b) Speed  Flow studies
3.1 Free Speed Studies
In the present study, free speed data was collected on different NHs and Expressways spread across the length and breadth of the country using ProLaser Instrumentation System (Laser Speed Measurement Gun) and trap length method at 21 selected road sections covering varying carriageway widths on multilane highways. As the selected test sections are divided road segments possessing varying horizontal curvature and traffic conditions in the two directions of the roadway, the above traffic data collection was carried out at each of the locations on both directions of travel as though each direction was a separate oneway road and hence the total number of study sections amounts to 42. The list of test sections considered is given in Table 1. All these study sections possess good riding quality with roughness
ranging around 2500 mm/km. The test sections have been chosen as far as possible away from the urban influence so that free flow conditions can be experienced. As can be inferred from Table 1, the number of sixlane divided carriageway road sections considered is only 3, whereas eight lane divided carriageway considered is only 1. This may be attributed to the relatively lesser number of road sections presently available in these categories without having the urban influence as compared to the four lane divided carriageways on intercity corridors. As the test sections include fourlane, sixlane and eightlane carriageways, separate analysis has been carried out for these carriageways. Further, it can also be noted from Table 1 that some of the selected test sections are lying in curved sections. These sections are specifically selected to incorporate the impact of horizontal curvature on the vehicular speeds and determine the generalized free speed characteristics on these highways. The free speed data was collected by classifying the vehicles into following categories namely Cars (which is further subclassified into two categories namely Small Cars < 1400 cc engine capacity and Big Cars > 1400 cc engine capacity), Two Wheelers (TWs), Auto Rickshaws (Autos), Buses, Light Commercial Vehicles (LCVs), Twoaxle Heavy Commercial Vehicles (HCVs) and Multiaxle Heavy Commercial Vehicles (MAVs).
3.2 Methodology for Free Speed Analysis
The observed free speeds of different vehicle types were classified into suitable intervals generally of 5 kilometers per hour (kmph) to determine the frequency distribution of vehicles as per speed. The mean speed and standard deviation (SD) values were calculated from the frequency distributions. Further, these data were fitted to normal distribution using mean and SD of vehicle speeds. From these distributions, important parameters namely 15 ^{t}^{h} Percentile Speed (V _{1}_{5} ), 50 ^{t}^{h} Percentile Speed (V _{5}_{0} ), 85 ^{t}^{h} Percentile Speed (V _{8}_{5} ), 95 ^{t}^{h} Percentile Speed (V _{9}_{5} ) and Spread Ratio (SR) were calculated to check the validity of the data. V _{1}_{5} is used to determine the lower speed limit whereas V _{8}_{5} is used for upper speed limits and V _{9}_{5} is
Journal of the Indian Roads Congress, OctoberDecember 2010
240
VELMURUGAN, MADHU, RAVINDER, SITARAMANJANEYULU & GANGOPADHYAY ON
used as a design speed for geometric design of highways. The SR is used to explain normality of the observed data and it is defined as,
SR
=
V
85
− V
50
V
50
− V
15
(2)
The estimated frequency curve will be truly normal when SR is unity. It will tend to deviate from the normal distribution as SR deviates from the unity. As can be seen from the fitted normal distributions, the speed data follow
the normal curve only when SR is ranging between 0.69
and 1.35 (Dey, et. al., 2006).
Table 1
Selected Test Sections for Free Speed and Speed  Flow Studies
S. No. 
National Highway (NH) / Expressway 
Location 
Direction 
Number 
Type of Section 
(Chainage) 
of Lanes 

1 
NH2 
Km 98 
Delhi  Mathura 
Four 
Straight 
2 
Km 98 
Mathura  Delhi 
Four 
Straight 

3 
NH2 
Km 629 
Durgapur  Kolkata 
Four 
Curved 
4 
Km 629 
Kolkata  Durgapur 
Four 
Curved 

5 
NH2 
Km 643 
Durgapur  Kolkata 
Four 
Straight 
6 
Km 643 
Kolkata  Durgapur 
Four 
Straight 

7 
NH4 
Km 1242 
Bangalore  Chennai 
Four 
Straight 
8 
Km 1242 
Chennai  Bangalore 
Four 
Straight 

9 
NH5 
Km 1501 
Chennai  Kolkata 
Four 
Curved 
10 
Km 1501 
Kolkata  Chennai 
Four 
Curved 

11 
Km 44 
Kharagpur  Kolkata 
Four 
Curved 

12 
NH6 
Km 44 
Kolkata  Kharagpur 
Four 
Curved 
13 
Km 47 
Kharagpur  Kolkata 
Four 
Straight 

14 
NH6 
Km 47 
Kolkata  Kharagpur 
Four 
Straight 
15 
Km 29 
Chenglepat  Chennai 
Four 
Straight 

16 
NH45 
Km 29 
Chennai  Chenglepat 
Four 
Straight 
17 
Km 58 
Chennai  Villupuram 
Four 
Curved 

18 
NH45 
Km 58 
Villupuram  Chennai 
Four 
Curved 
19 
Km 98 
Chennai  Villupuram 
Four 
Straight 

20 
NH45 
Km 98 
Villupuram  Chennai 
Four 
Straight 
21 
Km 5 
Vijayawada  Kolkata 
Four 
Straight 

22 
NH5 
Km 5 
Kolkata  Vijayawada 
Four 
Straight 
23 
Km 1069 
Vijayawada  Guntur 
Four 
Straight 

24 
NH5 
Km 1069 
Guntur  Vijayawada 
Four 
Straight 
25 
Km 15 
Hyderabad  Warangal 
Four 
Straight 

26 
NH202 
Km 15 
Warangal  Hyderabad 
Four 
Straight 
27 
Km 25 
Hyderabad  Bangalore 
Four 
Straight 

28 
NH7 
Km 25 
Bangalore  Hyderabad 
Four 
Straight 
29 
Km 462 
Hyderabad  Nagpur 
Four 
Straight 

30 
NH7 
Km 462 
Nagpur  Hyderabad 
Four 
Straight 
(Table 1 Contd
)
Journal of the Indian Roads Congress, OctoberDecember 2010
CRITICAL EVALUATION OF ROADWAY CAPACITY OF MULTILANE HIGH SPEED CORRIDORS UNDER HETEROGENEOUS TRAFFIC CONDITIONS THROUGH TRADITIONAL AND MICROSCOPIC SIMULATION MODELS
241
(Table 1 Contd 
) 

S. No. 
National Highway (NH) / Expressway 
Location 
Direction 
Number 
Type of Section 

(Chainage) 
of Lanes 

31 
NH9 
Km 462 
Hyderabad  Mumbai 
Four 
Straight 

32 
Km 499 
Mumbai  Hyderabad 
Four 
Straight 

33 
NH9 
Km 30 
HyderabadVijayawada 
Four 
Straight 

34 
Km 30 
VijayawadaHyderabad 
Four 
Straight 

35 
Greater Noida 
Near Lotus 
Delhi  Noida 
Six 
Straight 

36 
Expressway 
Valley School 
Noida  Delhi 
Six 
Straight 

37 
Greater Noida 
Near Panchsheel 
Delhi  Greater Noida 
Six 
Curved 

38 
Expressway 
Bal College 
Greater Noida  Delhi 
Six 
Curved 

39 
NH1 
Km 38 
Delhi  Sonepat 
Six 
Straight 

40 
Km 38 
Sonepat  Delhi 
Six 
Straight 

41 
DelhiGurgaon 
Near IFFCO 
Delhi  Gurgaon 
Eight 
Straight 

42 
Expressway 
Chowk 
Gurgaon  Delhi 
Eight 
Straight 
3.3 Speed  Flow Studies
The speed  flow studies were conducted along with free speed studies at the test sections mentioned in Table 1. In the case of speed  flow studies, Registration Plate Survey was conducted for determination of journey speeds and simultaneously Classified Traffic Volume Counts were conducted to estimate flow by synchronizing start time of the traffic volume with that of the Registration Plate survey. Traffic volume counts and Registration Plate Survey were conducted for 8 hrs on a typical normal working day by following the vehicle classification adopted in the free speed studies. Space Mean Speed (SMS) data was extracted out of the Registration Plate survey. This is a conventional procedure used for determining the speeds by recording the vehicle registration number, entry and exit time of vehicles on a defined trap length with the help of two synchronized stop clocks. The trap lengths of the road stretches selected for the mean speed measurements ranged between 400 m to 900 m. By noting the registration number and time of arrival and departure of the vehicles at the entry and exit points, the travel time over the selected trap length was determined and thereby the travel speed was derived. Based on the collected speed and flow data, speed  flow relationships are proposed to develop for different vehicle types for fourlane, sixlane and eightlane divided carriageways separately.
3.4 Methodology for Speed  Flow Equations and Roadway Capacity
In the present study, the traffic flow data was analyzed by typically dividing the traffic volume into two segments corresponding to congested and uncongested traffic conditions as shown in Fig. 1 (Yao, et. al., 2009). The two segments encompass the following:
(i)
Uncongested (Upper Part):Traffic related to Uncongested and Queue Discharge states
(ii)
Congested (Lower Part): Traffic related to Queuing state (Stop and Go)
Fig. 1
Uncongested and Congested Parts of SpeedFlow Curve
Journal of the Indian Roads Congress, OctoberDecember 2010
242
VELMURUGAN, MADHU, RAVINDER, SITARAMANJANEYULU & GANGOPADHYAY ON
It is proposed to analyze these two parts separately and determine the speed  flow relationships separately in the present study. For upper part (Uncongested) of the curve, different models including linear, exponential, polynomial, logarithmic, power, Akcelik and Bureau of Public Roads (BPR) have been attempted whereas for the lower part of the curve, linear, polynomial, logarithmic, power and exponential models were tried to be fitted and the model exhibiting best fit with the field data was adopted. The forms of these models are selected based
on the formulations given in Table 2. These formulations are adopted for developing speedflow equations and selected based on their statistical validity. Using this procedure, the speed  flow equations for different vehicle types on varying carriageway widths on multilane highways namely fourlane, sixlane and eightlane divided carriageways have been developed. In this study, the roadway capacity is considered as the intersecting point of best speed  flow models developed for the upper and lower part as shown in Fig. 2.
Table 2
Functional Form of Candidate Models for SpeedFlow Curves
Name of the Equation 
Functional 
Comments 

Form 

Linear 
v = _{α}  x +β 
Not always advisable; Reaches zero speed at high F/Fcap 

Logarithmic 
v =  _{α} ln x + _{β} 
Not always advisable; Has no value at x = 0 (the logarithm of "x" approaches negative infinity). 

Exponential 
v 
= _{α} v _{f} exp( β x) 
Has all the required traits for equilibrium assignment 
Power 
v = _{α} /x ^{β} 
Not always advisable; It goes to infinity at F/F _{c}_{a}_{p} at x = 0. 

Polynomial 
v = _{}_{α} x ^{2} βx + γ 
Not always advisable; It reaches zero speed at high F/F cap 

Bureau of Public Roads (BPR) 
v 
= v _{f} /(1+ _{α} (x) β ) 
Has all the required traits for equilibrium assignment 
Akcelik 
V = L/[L / v _{f} + 0.25{(x  1) + SQRT{(x  1) ^{2} + _{α} x}}] 
Has all the required traits for equilibrium assignment. 
Note
v = Speed; _{α} , β and γ = Global Parameters for Equation; x = F/F _{c}_{a}_{p} ratio; v _{f} = Free  Flow Speed;
F = Flow; F _{c}_{a}_{p} = Capacity Flow; L = Link Length;
Fig. 2
Flow (PCU/hr) Capacity Estimation from SpeedFlow Curves
3.5 Design Service Volume (DSV)
Design Service Volume (DSV) is defined as the maximum hourly volume at which vehicles can reasonably be expected to traverse a point or uniform section of a lane or roadway during a given time period under the prevailing roadway, traffic and control conditions while maintaining a designated Level of Service (LOS). From the view point of smooth traffic flow, it is not advisable to design the width of carriageway (or for determining the number of lanes) for a traffic volume equal to its capacity which is available at LOSE. At this level, the speeds are low (typically half the free speed) and freedom to maneuver
Journal of the Indian Roads Congress, OctoberDecember 2010
CRITICAL EVALUATION OF ROADWAY CAPACITY OF MULTILANE HIGH SPEED CORRIDORS UNDER HETEROGENEOUS TRAFFIC CONDITIONS THROUGH TRADITIONAL AND MICROSCOPIC SIMULATION MODELS
243
within the traffic stream is extremely restricted. Besides, at this level of service, even a small increase in volume would lead to forced flow situation and breakdowns within the traffic stream. Even the flow conditions at LOSC and LOSD involve significant vehicle interaction leading to lower level of comfort and convenience. In contrast, LOSB represents a stable flow zone which affords reasonable freedom to drivers in terms of speed selection and maneuvers within the traffic stream. Under normal circumstances, therefore, the use of LOSB is considered desirable for the design of rural highways. At this level, volume of traffic will be around 0.5 times the roadway capacity and this is taken as the DSV for the purpose of determining the carriageway width.It is recommended that for major arterial routes, LOSB should be adopted for design purposes. On other roads under exceptional circumstances, LOSC could also be adopted for design. Under these conditions, traffic will experience congestion and inconvenience during some of the peak hours, which may be acceptable. This planning decision should be taken in each case specially after carefully considering factors, like, suburban conditions, economic feasibility, etc. For LOSC, DSV can be taken as 40 percent higher than those for LOS B.
4 FREE SPEED ANALYSIS
The analysis of collected free speed data was carried out as per the methodology explained in Section 3. As mentioned in Section 3.1, the data collected for all the 42 test sections have been utilized. These sections include fourlane, sixlane and eightlane divided carriageway, but separate analysis has been carried out for these carriageways. The observed free speed data was fitted through normal distribution and relevant parameters namely average speed, standard deviation, percentile speeds and SR were estimated. The typical normal distribution and cumulative distribution curves for free speeds are given in Fig. 3.
From the normal distribution curves, free speeds of vehicles on various selected sections of the multilane high speed corridors are estimated and presented in Table 3, 4 and 5 for fourlane, sixlane and eightlane divided carriageways respectively. These tables also present the various percentile speeds and the spread ratio of different type of vehicles. Since the free speed analysis mainly focuses on freeflow conditions, the vehicles travelling with higher speeds are considered while arriving at the average free speeds. For this purpose, the vehicle
travelling below a specified speed range was ignored as they have outliers based on the scatter plot of the data. Hence, the speed data considered for Two Wheelers, Auto Rickshaws, Buses, Cars, LCVs/Two Axle Heavy Commercial Vehicles and MultiAxle heavy Commercial Vehicles are more than 65 kmph, 50 kmph, 60 kmph, 80 kmph, 60 kmph and 55 kmph, respectively.
From Table 3, 4 and 5, it can be observed that the normal distribution curve described the speed distributions satisfactorily in most of the vehicle types, since the SR value is ranging around 1.0 (from 0.950 to 1.157) demonstrating that SR is well within the limits. A critical evaluation of the free speed studies on fourlane, six lane and eightlane divided carriageway reveals the following:
a) The free speed of both small and big cars is much higher when compared with other vehicle types implying the rapid advancements in car manufacturing technologies and superiority of these engines.
b) The mean free speed of HCVs and LCVs are more or less same.
c) The mean free speed of TW is marginally higher than that of LCVs and Buses.
A summary of the free speeds in fourlane, sixlane and eightlane divided carriageway is presented in Fig. 4. The growing speeds of different vehicle types can be easily understood from Fig. 4 and the following inferences have been drawn:
a) Generally, the mean free speeds of different vehicle types on eightlane are higher when compared to fourlane and sixlane divided carriageways.
b) Free speeds of two wheelers and cars marginally increased from fourlane to sixlane while the increase is somewhat significant from sixlane to eightlane. This can be attributed to the achievement of their desired speeds on fourlane divided carriageway itself; hence, there is insignificant improvement of speeds on sixlane carriageway. However, the addition of one more lane on eightlane divided carriageway offering higher freedom for vehicular movements might have aided in attaining substantial increase in desired speeds and thus resulting in enhanced free speeds.
Journal of the Indian Roads Congress, OctoberDecember 2010
244
VELMURUGAN, MADHU, RAVINDER, SITARAMANJANEYULU & GANGOPADHYAY ON
Fig. 3 Typical Distribution of Free Speed on FourLane Divided Carriageways (NH45 at Km 98)
Table 3
Free Speed Statistics of Different Vehicles on FourLane Divided Carriageways
Vehicle 
Sample 
Avg. 
^{V} 15 ^{*}^{V} 50 ^{*}^{V} 85 ^{*}^{V} 95 ^{*} 
Max. 
SD* 
SR 

Type 
Size 
Speed * 
Speed* 

TW 
1191 
74.2 
63.5 
71.7 
80.1 
85.0 
120 
7.8 
1.027 
Auto 
753 
54.2 
46.3 
51.6 
56.9 
60.5 
79 
5.0 
0.995 
Small Car 
2688 
92.4 
80.7 
89.9 
100.6 
106.7 
161 
10.1 
1.157 
Big Car 
4137 
93.0 
79.7 
90.1 
100.4 
106.3 
149 
9.8 
1.000 
Bus 
2138 
71.1 
59.8 
68.6 
77.1 
82.1 
108 
8.2 
0.961 
LCV 
1614 
68.6 
58.5 
66.1 
73.9 
78.1 
113 
7.3 
1.019 
HCV 
504 
68.5 
58.0 
65.9 
74.2 
78.9 
103 
7.7 
1.038 
MAV 
1924 
64.0 
58.2 
64.6 
71.2 
75.2 
97 
6.1 
1.019 
*kmph 
Journal of the Indian Roads Congress, OctoberDecember 2010
CRITICAL EVALUATION OF ROADWAY CAPACITY OF MULTILANE HIGH SPEED CORRIDORS UNDER HETEROGENEOUS TRAFFIC CONDITIONS THROUGH TRADITIONAL AND MICROSCOPIC SIMULATION MODELS
245
Table 4
Free Speed Statistics of Different Vehicles on SixLane Divided Carriageways
Vehicle 
Sample 
Avg. 
^{V} 15 ^{*}^{V} 50 ^{*}^{V} 85 ^{*}^{V} 95 ^{*} 
Max. 
SD* 
SR 

Type 
Size 
Speed * 
Speed* 

TW 
723 
75.0 
59.3 
72.5 
85.7 
92.5 
109 
12.6 
1.000 
Auto 
95 
56.3 
44.8 
54.1 
63.8 
68.6 
88 
8.7 
1.043 
Small Car 
749 
93.1 
80.2 
90.6 
100.9 
106.8 
136 
9.8 
0.998 
Big Car 
1132 
95.5 
82.1 
93.0 
104.0 
110.6 
135 
10.5 
1.014 
Bus 
283 
74.4 
65.1 
71.8 
78.4 
82.1 
102 
6.2 
0.997 
LCV 
93 
73.6 
62.9 
71.0 
79.5 
84.4 
105 
7.9 
1.041 
HCV 
83 
70.7 
58.6 
68.2 
78.3 
84.4 
101 
9.7 
1.049 
MAV 
109 
70.5 
59.0 
68.0 
76.8 
81.9 
100 
8.5 
0.980 
*kmph 
Table 5
Free Speed Statistics of Different Vehicles on EightLane Divided Carriageways
Vehicle 
Sample 
Avg. 
^{V} 15 ^{*}^{V} 50 ^{*}^{V} 85 ^{*}^{V} 95 ^{*} 
Max. 
SD* 
SR 

Type 
Size 
Speed * 
Speed* 

TW 
343 
77.5 
63.7 
75.0 
86.3 
92.5 
90.0 
10.8 
1.000 
Auto 
11 
56.7 
48.8 
53.9 
58.8 
61.7 
64.0 
4.5 
0.950 
Small Car 
165 
98.0 
83.2 
95.5 
107.5 
115.2 
97.0 
11.8 
0.980 
Big Car 
180 
101.4 
84.7 
98.9 
113.7 
122.5 
104.0 
14.2 
1.050 
Bus 
246 
75.7 
64.9 
73.2 
81.3 
86.2 
72.0 
7.7 
0.980 
LCV 
127 
74.1 
61.9 
71.6 
107.5 
86.9 
74.0 
9.3 
1.020 
HCV 
24 
73.2 
64.1 
70.7 
77.0 
81.2 
56.0 
6.1 
0.970 
MAV 
13 
72.0 
62.7 
69.6 
76.6 
81.1 
57.0 
6.7 
1.030 
*kmph 
Fig. 4
Comparison of Average Free Speeds on FourLane, SixLane and EightLane Divided Carriageways for Different Vehicle Types
c) Free speeds of heavy vehicles and autos significantly increased from fourlane to sixlane while marginally increased from sixlane to eightlane. This can be attributed to the above vehicle types not able to attain the desired speed levels on fourlanes whereas on sixlane divided carriageway the presence of additional lane is helping in achieving significant increase in free speed from fourlane to sixlane. However, there is insignificant improvement in free speeds on eightlane as compared to six lane carriageways. As auto and heavy vehicles has achieved their desired speed levels on sixlane itself, there is no impact of eightlane divided carriageway though it offers higher LOS for vehicle movements
Journal of the Indian Roads Congress, OctoberDecember 2010
246
VELMURUGAN, MADHU, RAVINDER, SITARAMANJANEYULU & GANGOPADHYAY ON
compared
carriageways.
to
six
and
four
lane
divided
5 
SPEED  FLOW ANALYSIS 
5.1 
Traditional Model 
5.1.1 
Development of Speed  Flow Equations 
As mentioned in Section 3.3, the traffic data in respect of journey speed, free speed and flow were collected at
the test sections given in Table 1 through registration plate method, Laser Gun and classified volume count surveys through manual means was collected for a period of 8  12 hrs on different types of multi lane highways spread over the country. The observed traffic volume on these road sections was analyzed and the ranges of traffic composition of various vehicle types are presented in Table 6.
Table 6
Observed Traffic Composition on Varying Types of Divided Carriageways
Vehicle Type 
Traffic Composition on Divided Carriageways (in per cent) 

FourLane 
SixLane 
EightLane 

Two Wheelers 
4 
 59 (24) 
9 
 50 (28) 
9 
 29 (18) 
Autos 
0 
 23 (6) 
0 
 12 (4) 
0 
 1 (0.5) 
Small Cars 
1 
 36 (14) 
7 
 55 (33) 
23 
 53 (37) 
Big Cars 
1 
 43(15) 
14  33(23) 
27 
 48 (37) 

Buses 
1 
 45 (10) 
1 
 11(3) 
1 
 9 (3) 
LCVs 
1 
 40 (7) 
1 
 6 (2) 
1  18 (4) 

HCVs 
1 
 39 (11) 
1 
 6 (2) 
1 
 9 (1) 
MAVs 
1 
 32 (7) 
1 
 11 (4) 
1 
 9 (1) 
Cycles & other Slow Moving Vehicles (SMVs) 
0 
 28 (5) 
0 
 4 (1) 
0 
 1 (0.5) 
Note: Value in the parenthesis indicates the average share (in per cent) of the specific vehicle type in the traffic stream
From Table 6, it can be inferred that the two wheelers contribute for the major proportion of traffic on fourlane and sixlane divided carriageways compared to eightlane divided carriageway, whereas, the share of auto rickshaws is very insignificant on eightlane carriageways. Cars dominate the proportion of traffic in all types of multi lane carriageways and share of both two wheelers and cars together constitutes more than 80 per cent on sixlane and eightlane carriageways. In contrast, the heavy vehicles share in total volume is more in case of fourlane compared to sixlane and eightlane carriageways. This phenomenon of higher passenger traffic on six and eight lane divided carriageways and
relatively less share of goods traffic may be due to the selection of test sections comparatively nearer to the urban center (i.e. 10  40 km away from the city center). The share of cycles and SMVs including tractors, animal carts etc constitute less than 5 per cent on fourlane whereas, it is negligible on sixlane and eightlane carriageways. In order to develop speed flow equations and estimate roadway capacity, it is necessary to convert these observed traffic volume into a common unit, which is termed as Passenger Car Unit (PCU). In the present study, the PCU factors as per Table 7 given in IRC:64 (1990) has been used for converting the total volume into PCUs.
Journal of the Indian Roads Congress, OctoberDecember 2010
CRITICAL EVALUATION OF ROADWAY CAPACITY OF MULTILANE HIGH SPEED CORRIDORS UNDER HETEROGENEOUS TRAFFIC CONDITIONS THROUGH TRADITIONAL AND MICROSCOPIC SIMULATION MODELS
247
The observed traffic volume of different vehicle types is converted into PCUs based on the PCU factors given in Table 7. Using the analogy explained in Section 3.4, the first ever attempt was done in India in the present study by segregating speed  flow data into uncongested and congested conditions. Subsequently speed  flow
Table 7
PCU Factors adopted based on IRC Specifications (IRC: 641990)
Vehicle Type 
PCU 
Factor 

Motor Cycles (MC) 
0.5 
Scooters (SC) 
0.5 
Autos (A) 
1 
Cycle Rick. & Other Slow Vehicles (OT) 
1.5 
Small Cars (<1400 cc) (CS) 
1.0 
Big Cars (CB) 
1.0 
Cycles (CY) 
0.5 
Buses (B) 
3.0 
Mini Buses (MB) 
3.0 
Tractors and Tractor Trailers (TT) 
3.0 
Light Commercial Vehicles (LCV) 
1.0 
Two Axle Commercial Vehicles (HCV) 
3.0 
Multi Axle Commercial Vehicles (MAV) 
3.0 
relationships were developed for different vehicle types using both nonlinear and linear formulations considering uncongested and congested areas of speed  flow data separately. On critical examination of the statistical validity of each of the developed speed  flow equations, the BPR model and linear model were considered for developing the speedflow equations. For upper curve (Uncongested) BPR equations was considered whereas for lower curve (Congested), linear formulations was considered in the case of fourlane divided carriageway considering statistical validity of the equations. In case of sixlane and eightlane divided carriageways, linear models were considered for both upper and lower curve as they are showing higher statistical validity compared to other models. A summary of developed speedflow equations for different vehicle types are given in Table 8, 9 and 10
for fourlane, sixlane and eightlane divided carriageway respectively. For the estimation of goodnessoffit in terms of R ^{2} and other statistical estimates in nonlinear form of the BPR equation, the software called Statistical Packages for Social Studies (SPSS) has been used.
From Table 8, 9 and 10, it can be seen that the developed linear and BPR speed  flow equations exhibit good statistical validity in terms of good R ^{2} values. Hence, the developed equations are considered appropriate for estimating speeds under varying traffic conditions and can be explored for evolving roadway capacity.
5.1.2 Roadway Capacity from Traditional Model
Roadway capacity is the maximum number of vehicles which has a reasonable expectation of passing over a given section of a lane or a roadway in one direction (or in both directions for a twolane highway) during a given period of time under prevailing roadway and traffic conditions. The capacity is usually expressed as an hourly volume. The theoretical speed  flow curve which is the fundamental diagram of traffic flow is parabolic in shape. The maximum speed is the free speed. The parabola starts from the free speed and as the volume increases, the speed generally falls down. At a point, known as the maximum capacity the parabola takes an invert turn as already shown in Fig. 1. In the present study, roadway capacity was estimated from the intersecting point of upper curve and lower curves. The estimated roadway capacity of fourlane, sixlane and eightlane divided carriageways through traditional model is presented in Fig. 5. From the Fig. 5, it can be observed that the estimated roadway capacity i.e. intersecting point of upper and lower curve for fourlane, sixlane and eightlane divided carriageway is 6050, 6400 and 10500 PCUs/hour/ direction.
5.1.3 Comparison of Observed Free Speeds and Intercept of Speed  Flow Equation
To demonstrate the suitability of developed speed  flow equations through traditional model, the intercept of the equations are compared with the free speeds (refer Table 3, 4 and 5) and the comparison is presented in Table 11.
Journal of the Indian Roads Congress, OctoberDecember 2010
248
VELMURUGAN, MADHU, RAVINDER, SITARAMANJANEYULU & GANGOPADHYAY ON
Table 8 SpeedFlow Equations Derived from Traditional Models for Different Vehicles on Four Lane Divided Carriageways
S. 
No. 
Vehicle Type 
Uncongested (Upper) Curve NonLinear (BPR) Equation 
Congested (Lower) Curve Linear Equation 

1 
Auto 
y=66.731/(1+2.320*(x/6000.165)^¹) 
y 
= 0.011*x + 14.20 R² = 0.952 

R² = 0.863 

2 
TW 
y=99.49/(1+2.585*(x/7000.385)^¹) 
y 
= 0.008*x + 19.52 R² = 0.814 

R² = 0.880 

3 
Cars 
y=110.761/(1+1.564*(x/6999.968)^1) 
y 
= 0.004*x + 21.29 R² = = 0.861 

R² = 0.887 

4 
Bus 
y=94.080/(1+2.794*(x/6998.148)^1.544) 
y 
= 0.006*x + 22.23 R² = 0.674 

R² = 0.885 

5 
LCV 
y=87.345/(1+2.083*(x/6998.876)^1) 
y 
= 0.005*x + 20.83 R² = 0.900 

R² = 0.766 

6 
HCV 
y=84.230/(1+2.056*(x/6999.049^1.089) 
y 
= 0.002*x + 24.69 R² = 0.713 

R² = 0.713 

7 
MAV 
y=67.709/(1+3.466*(x/6995.476)^2.208) 
y 
= 0.002*x + 24.67 R² = 0.701 

R² = 0.690 
Note: y = speed (kmph); x = Flow (PCU/hr/Dir)
Table 9 SpeedFlow Equations from Traditional Model for Different Vehicles on Six Lane Divided Carriageways
S. 
No. 
Vehicle Type 
Uncongested (Upper) Curve NonLinear (BPR) Equation 
Congested (Lower) Curve Linear Equation 

1 
Auto 
y 
= 0.004x + 59.39 R² = 0.525 
y 
= 0.009x + 19.64 R² = 0.570 

2 
TW 
y 
= 0.009x + 77.50 R² = 0.638 
y 
= 0.009x + 5.971 R² = 0.525 

3 
Cars 
y 
= 0.011x + 112.3 R² = 0.493 
y 
= 0.004x + 15.52 R² = 0.646 

4 
Bus 
y 
= 0.007x + 92.47 R² = 0.663 
y 
= 0.013x + 3.790 R² = 0.528 

5 
LCV 
y 
= 0.005x + 97.66 R² = 0.362 
y 
= 0.012x + 3.273 R² = 0.777 

6 
HCV 
y 
= 0.011x +82.60 R² = 0.559 
y 
= 0.011x + 2.227 R² = 0.739 

7 
MAV 
y 
= 0.011x + 92.38 R² = 0.635 
y 
= 0.005x + 12.04 R² = 0.538 
Note: y = speed (kmph); x = Flow (PCU/hr/Dir)
Journal of the Indian Roads Congress, OctoberDecember 2010
CRITICAL EVALUATION OF ROADWAY CAPACITY OF MULTILANE HIGH SPEED CORRIDORS UNDER HETEROGENEOUS TRAFFIC CONDITIONS THROUGH TRADITIONAL AND MICROSCOPIC SIMULATION MODELS
249
Table 10
SpeedFlow Equations from Traditional Model for Different Vehicles on Eight Lane Divided Carriageways
S. No. 
Vehicle Type 
Uncongested (Upper) Curve NonLinear (BPR) Equation 
Congested (Lower) Curve Linear Equation 

1 Auto 
y 
= 0.001x + 53.45 R² = 0.606 
y 
= 0.001x + 31.20 R² = 0.709 

2 TW 
y 
= 0.004x + 93.49 R² = 0.823 
y 
= 0.001x + 31.20 R² = 0.603 

3 Cars 
y 
= 0.002x + 86.28 R² = 0.726 
y 
= 0.001x + 33.84 R² = 0.495 

4 Bus 
y 
= 0.002x + 72.69 R² = 0.830 
y 
= 0.002x + 24.42 R² = 0.704 

5 LCV 
y 
= 0.002x + 70.31 R² = 0.786 
y 
= 0.002x + 22.87 R² =0.812 

6 HCV 
y 
= 0.001x +66.55 R² = 0.589 
y 
= 0.001x + 32.33 R² = 0.553 

7 MAV 
y 
= 0.003x + 74.39 R² = 0.481 
y 
= 0.001x + 29.31 R² = 0.643 
Note: y = speed (kmph); x = Flow (PCU/hr/Dir)
Fig. 5 Roadway Capacity of FourLane, SixLane and EightLane Divided Carriageways Evolved through Traditional Models
Journal of the Indian Roads Congress, OctoberDecember 2010
250
VELMURUGAN, MADHU, RAVINDER, SITARAMANJANEYULU & GANGOPADHYAY ON
Table 11
Comparison of Observed Free Speed with Intercept on SpeedFlow Equation
Carriageway 
Vehicle Type 
Observed Mean Free Speed (kmph) 
Intercept of Speed Flow Equation (kmph) 
Error 
(per cent) 

Auto 
54.2 
66.7 
23 

Two Wheeler 
74.2 
99.5 
34 

Big Car 
93.0 
110.8 
19 

FourLane 
Small Car 
92.4 

Bus 
71.1 
94.1 
32 

LCV 
68.6 
87.3 
27 

HCV 
68.5 
84.2 
23 

MAV 
64.0 
67.7 
6 

Auto 
56.6 
59.4 
5 

Two Wheeler 
75.0 
77.5 
3 

Big Car 
95.5 
112.3 
19 

SixLane 
Small Car 
93.1 

Bus 
74.4 
92.5 
24 

LCV 
73.6 
97.7 
33 

HCV 
70.7 
82.6 
17 

MAV 
70.5 
92.4 
31 

Auto 
56.7 
53.5 
6 

Two Wheeler 
77.5 
93.5 
21 

Big Car 
98.0 
86.3 
13 

EightLane 
Small Car 
101.4 

Bus 
75.7 
72.7 
4 

LCV 
74.1 
70.3 
5 

HCV 
73.2 
66.6 
9 

MAV 
72.0 
74.4 
3 
From Table 11, it can be observed that the error between observed free speed and intercept of speed flow equations of different vehicle types on fourlane divided carriageway is ranging from 6 to 34 per cent, whereas in the case of sixlane and eightlane divided carriageway, the range is 3 to 31 per cent and 3 to 21 per cent respectively. From the above results, it can be inferred that even though the developed speed  flow equations are exhibiting good statistical validity, the intercept derived from the traditional models has not appropriately
represented the field conditions as the error difference between the observed free speed and intercept of speed  flow equations of different vehicle types is high. This phenomenon can be attributed to the dataset considered for developing the speed  flow relationships encompasses aggregated traffic data (i.e. which includes traffic flow and average free speed over a specified time interval) and also not accounting of the typical random lane change behaviour experienced on Indian highways in all the traditional models derived in this study. To overcome this
Journal of the Indian Roads Congress, OctoberDecember 2010
CRITICAL EVALUATION OF ROADWAY CAPACITY OF MULTILANE HIGH SPEED CORRIDORS UNDER HETEROGENEOUS TRAFFIC CONDITIONS THROUGH TRADITIONAL AND MICROSCOPIC SIMULATION MODELS
251
limitation, the microscopic simulation approach has been attempted in this study which is capable of representing individual vehicles on road section and estimate the driver behaviour more realistically. The microscopic simulation model has been developed to estimate speed and flow characteristics for varying carriageway widths and the details are furnished in the succeeding section.
5.2 
Microscopic Simulation Model 
5.2.1 
Need for Microscopic Simulation 
The traditional capacity estimation methods assume homogeneous conditions and lane discipline, however, it is not applicable for Indian conditions. In such circumstances, roadway capacities could be either underestimated or overestimated. Capacity estimation is primarily depends on vehicular movements on the road stretch and in this regard, lane change behaviour can severely affect the movements. On Indian roads, vehicles seldom observe lane discipline and make their own virtual lanes instead of the demarcated physical lanes. The conventional methods ignore vehicle movements and interactions and these behaviours cannot be explained which has great impact on speed  flow relationships and capacity estimation. In the absence of accounting for such situations, the output might be far from reality. As described earlier, microscopic simulation considers each and every vehicle movement on a roadway and hence such a lane change behaviour and vehicle interactions can be described. More realistic estimation of speed  flow relationships can be achieved through microscopic simulation system, which can lead to the estimation of capacity with reasonable degree of accuracy. This is because tremendous advancements that have been brought forth with by deploying microscopic simulation techniques for modelling transportation systems. Such microscopic simulations are able to model individual vehicles and pedestrians in a large area and it is possible to estimate realistic speed  flow characteristics and capacity considering all possible lane change behaviour even under heterogeneous traffic conditions. Further, these techniques are highly useful in estimating the traffic characteristics under different traffic flow and driver behaviour conditions, which cannot be observed
on the field. However, it is to be borne in mind that the data collection task needed to develop the microscopic simulation can be a bit tedious and cumbersome. To arrive at speed  flow characteristics and establish capacity norms through microscopic simulation, one has to model the flow of individual vehicles in a detailed manner for which established simulation packages can be used. The data collection and methodology followed for this phase is explained in the succeeding sections.
5.2.2 Data Collection
In order to develop a microscopic simulation model, the traffic data was again collected on Delhi  Mathura section of NH2 near Hodal which is a fourlane divided carriageway considered to develop speed  flow equations through traditional method (S.No. 1 and 2 in Table 1). This section was specifically chosen to check the suitability of these two models namely microscopic simulation and traditional models. The reconnaissance survey was conducted on 23 ^{r}^{d} March 2010 and the videography survey was eventually conducted from 9.30 am to 2:00 pm on 25 ^{t}^{h} March 2010 by capturing the traffic plying during the morning and afternoon time periods on both directions of travel.
5.2.3 
Development 
of 
Microscopic Simulation 
Model 
The methodology followed for the microscopic simulation is shown in the form of flow chart in Fig. 6. From the Fig. 6, it can be observed that the data collection is the first and foremost requirement for understanding speedflow characteristics on multilane highways. To capture lane change behavior on these multilane high speed corridors, videography method was adopted for data collection. The recorded film was replayed on television screen in the laboratory of CRRI, New Delhi and the required data were decoded through manual method. The vehicles were divided into ten categories and the data extracted from video recording were Volume, Space Mean Speed (SMS) and number of lane changes by individual vehicles during every five minute time interval. The video data on classified traffic volume counts, SMS and lane change behavior were decoded in a
Journal of the Indian Roads Congress, OctoberDecember 2010
252
VELMURUGAN, MADHU, RAVINDER, SITARAMANJANEYULU & GANGOPADHYAY ON
synchronized fashion. Using this data, a model is developed in VISSIM 4.10, a microscopic simulation software. Then the model is appropriately calibrated and validated using the observed data considering volume, speed and number lane changes. Using the validated simulation model, speed  flow relationships have been developed under two scenarios namely 'with' random lane change behavior (which is a common phenomenon on multilane highways in India) and also by assuming 'without' lane change conditions. The roadway capacity estimated under these two scenarios is used to assess the impact of lane change as shown in Fig. 6.
Fig. 6
Methodology for Estimating Capacity Considering Impact of Lane Change Behaviour
In microscopic simulation, a model which accurately represents the existing situation is known as the 'Base Model'. The base model is constructed by representing the network area that was defined in the model scope and using actual, observed traffic flow data. The validated base model is used to develop a 'future year base model' against which scenarios and design options can be
compared. The base model development can be summarized in the following steps:
1. Developing base network.
2. Defining model parameters.
3. Calibrating the network.
4. Validating the model.
Development of a network that accurately determines the constraints of a road network is an important stage in the modelling process. The basic key network building components are: Links and Connectors. In the present simulation model, links are created spanning for 130 m representing the test section near Hodal on NH2 for both directions. However, a buffer link is provided for buffering process of the network which is taken 100 m. Both test section link and buffer links are appropriately connected by connectors. Fig. 7 shows the links created separately for Delhi  Hodal and Hodal  Delhi directions in VISSIM.
As mentioned earlier, the test section selected on NH2 is a fourlane divided carriageway with approximately 2.0 m paved shoulder and 0.5 m earthen shoulders. Accordingly, the links are created in VISSIM with total of four lanes on each link including two lanes of main carriageway, one lane of paved shoulder and one lane of earthen shoulder as shown in Fig. 7. During the reconnaissance survey at the site, it was observed that the majority of fast moving vehicles movements are using the main carriageway and major proportion of slow moving vehicles and some proportion of the two wheelers are using the paved shoulders. By considering this phenomenon, road links are created as shown in Fig. 7.
Fig. 7
Created Links with Main Carriageway and Shoulders for FourLane Divided Carriageway in VISSIM
Journal of the Indian Roads Congress, OctoberDecember 2010
CRITICAL EVALUATION OF ROADWAY CAPACITY OF MULTILANE HIGH SPEED CORRIDORS UNDER HETEROGENEOUS TRAFFIC CONDITIONS THROUGH TRADITIONAL AND MICROSCOPIC SIMULATION MODELS
253
5.2.3 Calibration of Microscopic Simulation Model
Calibration is a process of adjusting the model parameters, network and vehicle demand to reflect and represent observed data and/or observed site conditions to a
sufficient level to satisfy the model objectives. The calibration process is explained in the form of flow chart as shown in Fig. 8.
Fig. 8 Calibration Procedure Adopted in Development of Simulation Model
By giving the parameters listed in Fig. 8 as an input to simulation model, simulation runs were carried out in order to estimate the output. In this simulation model, the outputs obtained are volume, speed of vehicles and number of lane changes. Since the observed data on these parameters were collected in the field for validation of the developed simulation model. The comparison of estimated values with observed values is carried out and error is estimated. This iterative process of simulation model calibration was carried out through the modification of the various model parameters and simulation runs were performed till the error is within the satisfactory level.
5.2.4 Validation of Microscopic Simulation Model
Validation is the process of checking the developed simulation model in terms of predicted traffic performance for road system against field measurements of traffic performance such as traffic volumes, travel times, average speeds, and lane changes. In the present study, the calibration and validation process was carried out by trial and error method. After carrying out many trials, the prediction error in volume, speed and lane changes is
reduced to satisfactory level. The final validation results for volume speed and lane change criteria are estimated for Delhi to Hodal and Hodal to Delhi directions separately. Fig. 9 and 10 shows the validation results of traffic volume, speed and lane changes for Hodal to Delhi and Delhi to Hodal directions, respectively.
From the Fig. 9, it can be observed that the error in estimation of traffic volume is less than 10 per cent for different vehicle types except in the case of buses and bicycles on Hodal to Delhi direction whereas the overall error in the estimation of traffic volume is almost zero which represents the accuracy of the developed simulation model. The comparison of observed and estimated data of different vehicle speeds shows that the error in vehicular speeds is ranging from 2 per cent to 20 per cent for different vehicle types except in the case of trucks (due to large variation in observed speeds of trucks) which represents the developed simulation model is reasonably accurate. The simulated lane changes during each 5minute time interval is also compared with the observed lane change data and it can be observed
Journal of the Indian Roads Congress, OctoberDecember 2010
254
VELMURUGAN, MADHU, RAVINDER, SITARAMANJANEYULU & GANGOPADHYAY ON
from Fig. 9 that the overall error in predicting number of lane changes made by different vehicle types is about 20 per cent in Hodal to Delhi direction which can also be regarded to be reasonably accurate in reflecting the real world conditions considering the traffic mix being simulated is heterogeneous in nature and resorting to random lane changes.
Fig. 9
Comparison of Observed and Estimated Traffic Volume, Speed and Lane Changes (Hodal to Delhi Direction)
From Fig. 10, it can be seen that the error in estimation of traffic volume is less than 10 per cent for different vehicle types on Delhi to Hodal direction while the overall error is about 1 per cent implying the effectiveness of the calibrated model in replicating the ground conditions. At the same time, it can be seen that the error in speed prediction on Hodal to Delhi direction of NH2 is found to be ranging between 1  19 per cent except in the case of cars and two wheelers which is about 30 per cent. The high error in cars and two wheelers may be attributed to the high influence of local conditions (such as median
Fig. 10
Comparison of Observed and Estimated Traffic Volume, Speed and Lane Changes (Delhi to Hodal Direction)
Journal of the Indian Roads Congress, OctoberDecember 2010
CRITICAL EVALUATION OF ROADWAY CAPACITY OF MULTILANE HIGH SPEED CORRIDORS UNDER HETEROGENEOUS TRAFFIC CONDITIONS THROUGH TRADITIONAL AND MICROSCOPIC SIMULATION MODELS
255
gap and roadside friction.) on Delhi to Hodal direction acting as a major impediment, which is causing significant reduction of observed speeds of cars and two wheelers as compared to other direction.
On the other hand, the predictive capability of the model in terms of lane changes in the case of different vehicle types like two wheeler, auto rickshaw, small car, big car and multi axle truck is less than 20 per cent error, whereas the overall lane change error is only about 2 per cent on Hodal to Delhi direction. From the above calibrated and validated results, it can be inferred that the developed simulation models are able to predict the vehicular movements (i.e. flow, speed and lane changes) with reasonable degree of accuracy under heterogeneous traffic conditions for fourlane divided carriageways. Based on the developed simulation models, the evolution of speed  flow relationships is attempted and using the same, the roadway capacity can be estimated.
5.2.5 Development of Speed  Flow Equations and Roadway Capacity through Simulation
Using the developed simulation model, the speed data for different vehicle is estimated for different traffic volume conditions for fourlane divided carriageway. The simulation runs are carried for following scenarios of traffic volumes for estimating capacity:
a) Observed Flow (ranging from 1000 to 1500 Vehicles/hr)
b) Flow of 2000 Vehicles/hr
c) Flow of 4000 Vehicles/hr
d) Flow of 6000 Vehicles/hr
e) Flow of 8000 Vehicles/hr
In the same way, the developed simulation model is applied to estimate speeds of the vehicle for different types of carriageway namely sixlane and eightlane divided carriageways. However, traffic flow up to 10000 vehicles/hr was considered for sixlane and eightlane divided carriageways. For this purpose, separate network has been created by introducing extra lanes so as to formulate sixlane and eightlane divided carriageway scenarios. The lane restrictions are also considered for
these carriageways same as that of fourlane divided carriageways. However, the driving behaviour is kept same as fourlane divided carriageway for sixlane and eightlane divided carriageway assuming that it will not drastically change in spite of increase in the number of lanes. These aspects would be further investigated by observing real data on these carriageways as the desired speed characteristics might be different on these carriageways compared to fourlane divided carriageways. This may be regarded as the limitation of the present model and it is worthwhile to study this aspect in future scope of the study. Considering the above flow conditions, the simulation runs are made to estimate speeds of different vehicles on fourlane, sixlane and eightlane divided carriageways. The estimated speed data of the cars are plotted against given traffic flow with linear equations. The developed linear speed  flow equations speedflow equations are having high goodness offit as the R ^{2} values are more than 0.9 for all the carriageways, however, the intercept of the speedflow equation, which is also considered as free speed of the vehicle is 99.96 km/hr, 97.0 km/hr and 96.35 km/hr, which is slightly decreasing as the number of lanes increases from fourlane to eightlane divided carriageways respectively. Further, the capacity of these carriageways is calculated from these linear speedflow equations by assuming the fact that capacity would be occurring at half of the free speed. Accordingly, half of the free speed is substituted in the speedflow equation to estimate roadway capacity. From this exercise, the capacity is estimated as 5,553 PCU/hour/Dir, 9,700 PCU/hour/Dir and 14,160 PCU/hour/Dir for fourlane, sixlane and eightlane divided carriageways, respectively. Though the fit of the speedflow equation is very good, the estimated free speed and capacities are not realistic especially in the case of sixlane and eightlane divided carriageways. Free speeds are under predicted and capacities are over predicted as the slope of the equation line is very mild which shows insignificant impact of traffic volume on vehicle speeds. Since the linear method had produced unrealistic values of capacities and free speeds, nonlinear method has been subsequently attempted. A nonlinear equation has been formulated from the second order
Journal of the Indian Roads Congress, OctoberDecember 2010
256
VELMURUGAN, MADHU, RAVINDER, SITARAMANJANEYULU & GANGOPADHYAY ON
polynomial equation and the final form of the equation under nonlinear form is as given below:
V = a _{1} + (a _{1} ^{2} + a _{2} * F) ^{0}^{.}^{5}
where, V is speed in km/hr,
F is flow in PCU/hr/dir
(3)
a _{1} , a _{2} are parameters to be estimated
Using the estimated speed data of the cars and traffic flows from the different simulation runs, the equation shown in Eqn. (3) has been developed for fourlane, sixlane and eightlane divided carriageways. A comparison of speedflow equations evolved for cars through traditional and microscopic simulation approaches are given in Table 12. To demonstrate the validity of developed speed  flow equations through microscopic simulation model, the estimated free speed which is the intercept of the equations (at flow almost equal to 0 Vehicles/hr) are compared with the observed free speeds (refer Table 3, 4 and 5) and presented in Fig. 11. The above fig. illustrated the error ranges obtained from microscopic simulation model by comparing with the traditional model:
a) Fourlane: 0.3 to 10 per cent (which is 6 to 34 per cent in the case of traditional model)
b) Sixlane: 0.1 to 16 per cent (which is 3 to 31 per cent in case of traditional model)
c) Eightlane: 2 to 18 percent (which is 3 to 21 per cent in case of traditional model)
From the above results, it can be concluded that developed speed  flow equations through simulation model has significantly reduced the prediction error in free speeds compared to traditional model except in the case of LCVs and MAVs on eightlane divided carriageway. The occurrence of relative larger error for LCVs and MAVs can be examined by conducting traffic studies and calibrating the simulation model for eightlane using
observed data on the field. From these results, it can be noted that the developed microscopic simulation model is able to predict the traffic phenomenon on multilane highways more realistically compared to traditional model. Thus the evolved roadway capacity through simulation approach can be adjudged to be realistic for the heterogeneous traffic conditions observed on multilane divided carriageways. Using this evolved speed  flow relationships, the roadway capacity is estimated as shown in Fig. 12.
From the Fig. 12, it can be observed that the speedflow equations are having high goodnessoffit as the R ^{2} values are about 0.77 for all the carriageways and from this it can be said that that the developed speedflow equations can be used to predict the speed of cars for given flow conditions with reasonable degree of accuracy. Further, the capacities of the carriageways are calculated based on the nonlinear speedflow equations and from this exercise, the roadway capacity is estimated as 5574 PCU/hour/dir, 7733 PCU/hour/dir and 9796 PCU/hour/dir for fourlane, sixlane and eightlane divided carriageways, respectively.
Based on the detailed analysis, it can be inferred that both traditional method and microscopic simulation method are estimating the comparable results in the case of fourlane and eightlane divided carriageway, however traditional method is under predicting the capacity in case of sixlane divided carriageway. This can be attributed to the paucity of data used for model development in the case of sixlane carriageways, whereas in the case of microscopic simulation model, the speeds can be estimated just by substituting for any flow conditions and thus estimate the capacity thereafter. This is the biggest advantage of the simulation model over the traditional method. Hence, it can be concluded from this detailed evaluation, that the microscopic simulation model is able to predict the speeds and flow conditions and thereafter roadway capacities were estimated with good degree of statistical validity.
Journal of the Indian Roads Congress, OctoberDecember 2010
CRITICAL EVALUATION OF ROADWAY CAPACITY OF MULTILANE HIGH SPEED CORRIDORS UNDER HETEROGENEOUS TRAFFIC CONDITIONS THROUGH TRADITIONAL AND MICROSCOPIC SIMULATION MODELS
257
Table 12
Comparison of SpeedFlow Equations of Cars for different Multilane Carriageways
S. 
Carriageway 
Traditional Method Uncongested 
Congested 
Microscopic Simulation Method (Nonlinear) 

No. 

110.761 
= 47.633+(2268.931 

1 Fourlane divided 
y = 
1 + 
1.564* 

Viel mehr als nur Dokumente.
Entdecken, was Scribd alles zu bieten hat, inklusive Bücher und Hörbücher von großen Verlagen.
Jederzeit kündbar.