ptq

Q1 2017

PETROLEUM TECHNOLOGY QUARTERLY

REFINING
GAS PROCESSING
PETROCHEMICALS

SPECIAL FEATURES
OUTLOOK FOR 2017
AUTOMATION & CONTROL
ENVIRONMENTAL

cover and spine copy 26 (original).indd 1

15/12/2016 13:56

We create
chemistry
that lets
individual
needs
love global
innovation.

As the global leader in catalysis, BASF draws on the

talent and expertise of more than 1,100 researchers
working in close partnership with our customers.
This collaboration results in innovations that drive
new levels of performance and achievement, today
and over the long term.
When global catalyst innovations help our
customers become more successful, it's
because at BASF, we create chemistry.
www.catalysts.basf.com

basf.indd 1

14/12/2016 10:00

ptq
PETROLEUM TECHNOLOGY QUARTERLY

Q1 (Jan, Feb, Mar) 2017

www.eptq.com

3


5

11

33

This internet thing

Chris Cunningham
Outlook for 2017
ptq&a
Crude oil sourcing: price and opportunity
Misha Gangadharan, S D Pohanekar and M D Pawde
Hindustan Petroleum Corporation Limited

41 Cubas oil: due for development

Amaury Prez Snchez
University of Camagey
49 Control system security
Sinclair Koelemij
Honeywell Process Solutions
55 Data operations transform fuels value
Craig Harclerode
OSIsoft
61 Profiting from plant data
Douglas White
Emerson Automation Solutions
71 Optimal processing scheme for producing pipeline quality gas
Saeid Mokhatab Gas Processing Consultant
Gerrit Bloemendal Jacobs Comprimo Sulfur Solutions
79



83

Correcting vacuum column design flaws

Gary Martin
Recon Management Services, Inc.
Analysing FCC hot spots
Kenneth Fewel
Technip Stone & Webster Process Technology

91 Balanced distillation equipment design

Soun Ho Lee
GTC Technology
101

Firing high sulphur fuel

Adil Rehman, C Steven Lancaster, Sandeepan Ghosh, Om Prakash Sahu
and Pawan Kumar Sharma
KBR Technology

107 Combating reactor pressure drop

Ankit A Jain and Ajay Gupta
Reliance Industries Ltd
113 The design temperature of flare systems
Paul David Paul David Process Ltd
19 Laser scanning with dimensional control
1
Peter Field Warner Surveys
24 Identifying contributors to flaring
1
P Sridhar
Indian Oil Corporation
126 Technology in Action

Cover
Triple flare of the hydrocracking unit at Galp Energias Sines refinery, Portugal.
Photo: Galp Energia

2017. The entire content of this publication is protected by copyright full details of which are available from the publishers. All rights
reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means electronic,
mechanical, photocopying, recording or otherwise without the prior permission of the copyright owner.
The opinions and views expressed by the authors in this publication are not necessarily those of the editor or publisher and while every care
has been taken in the preparation of all material included in Petroleum Technology Quarterly and its supplements the publisher cannot be held
responsible for any statements, opinions or views or for any inaccuracies.

ed com copy 12.indd 1

16/12/2016 15:29

ACTION
If raising the alky feed and distillate from your FCC unit is
a top priority, youve come to the right place.
Thats because with its continuing record of success, Albemarles ACTION is the only commercially proven FCC
catalyst to maximize distillate, butylenes, and octane with minimal gasoline loss. Utilizing its unique zeolite
and matrix technologies, ACTION has been successful cracking all types of feeds, from tight oil to heavy resid.
Your success is too important to risk with unproven alternatives.
Achieve satisfaction and success with the undisputed leader...demand ACTION.

For more information on Albemarle ACTION catalyst or our exceptional portfolio

of products and services, call (281) 480-4747 or visit www.albemarle.com

REFINING SOLUTIONS

ALB15-001_ACTION_prodA4.indd 1

3/18/15 5:29 PM

ptq

This internet thing

PETROLEUM TECHNOLOGY QUARTERLY

Vol 22 No 1

Q1 (Jan, Feb, Mar) 2017

Editor
Chris Cunningham
editor@petroleumtechnology.com
Production Editor
Rachel Storry
production@petroleumtechnology.com
Graphics Editor
Rob Fris
graphics@petroleumtechnology.com
Editorial
tel +44 844 5888 773
fax +44 844 5888 667
Business Development Director
Paul Mason
sales@petroleumtechnology.com
Advertising Sales Office
tel +44 844 5888 771
fax +44 844 5888 662
Publisher
Nic Allen
publisher@petroleumtechnology.com
Circulation
Jacki Watts
Louise Shaw
circulation@petroleumtechnology.com
Crambeth Allen Publishing Ltd
Hopesay, Craven Arms SY7 8HD, UK
tel +44 844 5888 776
fax +44 844 5888 667

PTQ (Petroleum Technology Quarterly) (ISSN

No: 1632-363X, USPS No: 014-781) is published
quarterly plus annual Catalysis edition by Crambeth
Allen Publishing Ltd and is distributed in the US
by SP/Asendia, 17B South Middlesex Avenue,
Monroe NJ 08831. Periodicals postage paid at New
Brunswick, NJ. Postmaster: send address changes to
PTQ (Petroleum Technology Quarterly), 17B South
Middlesex Avenue, Monroe NJ 08831.
Back numbers available from the Publisher
at $30 per copy inc postage.

n the final quarter of 2016 there was a blizzard of announcements about

systems and shared technology deals from the big hitters of plant automation,
all about shaping the future of the industrial internet of things (IIoT). It is a
concept that seems to have arrived at a tipping point, just about where the plain
old internet was in the 1990s, and is set to transform the value of plant data.
PTQ was a guest of the Honeywell Users Group EMEA conference, an
opportunity to hear about some real experience with IIoT, at the time of this
perfect storm of automation news.
In some quarters of the world of refining and petrochemicals the internet
has yet to catch on in any form. The internet is too big a security risk, is it not?
Much bigger issues surely are where to find the next generation of plant engineers and how to avoid losing the experience gained over the years by retiring
senior engineers. An opinion shared at the Honeywell meeting is that the IIoT,
its cloud capture of data and the possibilities of analytics can go a long way to
solving these issues of employment and expertise.
Historically, that expertise is likely to have been acquired over the years in
a single plant. Taking the case of a refiner with a fleet of producing sites, perhaps one involving several regions of the globe, what the IIoT offers (and to
simplify somewhat) is a control system that captures and uploads key operating data from controllers, sensors and other things across those multiple sites
to a corporate cloud resource. This compares with each site normally having only its own plant historian as a private repository of performance data.
Data held in the cloud is then available to engineers across the company to
perform analytic routines that can interpret issues that arise or, more importantly, could arise during plant operations. This cloud historian combined
with data analytics in effect represents expertise of a very different order, even
compared with a senior engineers career-long experience.
So, for instance, a plant engineer could explain, on the basis of company-wide
operations over the years, why flooding occurs in the CDU. Better still, that
engineer can interrogate years of multi-site records to determine which combination of key performance indicators could lead to flooding a matter of envisioning the invisible.
Regarding employment, the aim of any refiner is to hire the best and keep
them. The next-generation plant engineer, after all, wants to run applications
on the internet and wants to know why the tools available at work in the plant
might be inferior to the ones he or she runs at home.
Then there is the more basic question of place of work. Does a young
engineer want to move from five years of education in a city centre with
the prospect of spending some part of the career journey in a remote, neardystopian environment? The world is just not moving in that direction. The
facility to control and analyse plant operations remotely, via a multiple site
cloud resource, provides the opportunity to install workers far from their
refineries, at sites much closer to their personal preferences. And with potential handover of operations across the globe, there is the opportunity for a control room in one continent running a refinerys operations on some distant
land mass. Who, after all, wants to work night shifts?
CHRIS CUNNINGHAM

PTQ Q1 2017

ed com copy 12.indd 2

15/12/2016 11:28

FLEXIBILITY
FOR PROFITABILITY
DIESEL

MID DISTILLATE

MID DISTILLATE/NAPHTHA NAPHTHA

MID DISTILLATE SELECTIVITY

Z-HD10
Z-503

Z-2513
Z-513

Z-MD10

Z-603

Z-2723
Z-2623

Z-FX10

Z-MD20

Z-623

Z-2723
Z-673

Z-FX20

Z-FX10
Z-FX20
Z-723

Classic

Tailored

Next Generation

Z-NP10
Z-NP20

Z-733
Z-803
Z-853
Z-863

ACTIVITY
Criterion and ZEOLYSTTM have developed another generation of specialized Advanced Trilobe
(ATX) cracking catalysts that perform with improved saturation, higher activity and higher selectivity,
which is already driving value across the globe. ZEOLYST has been producing market-leading
catalysts for 25 years, and our scientists have been diligently working to keep raising the bar.
NEXT GENERATION
Z-2723

Z-FX10

Z-FX20

High middle distillate yield High middle distillate and

Paraffinic feedstock flexibility
Naphtha yield
Improved hydrowax quality
Max middle distillate yield
Heavier feedstocks
High diesel quality
Improved diesel cold flow
Max liquid/low gas production Max liquid/low gas production

This new generation is setting the standard in hydrocracking to improve refiners profitability.
The ZEOLYSTTM FX Series is the ultimate choice for converting even the most challenging
feedstocks into high-quality products that adapt to seasonal refining demands in hydrocracking.
With Criterion Catalysts, you gain our innovative science backed by leading technical service
for the quality you need.

Leading minds. Advanced technologies.

www.CRITERIONCatalysts.com

criterion.indd 1

14/12/2016 11:35

Outlook for 2017

What are the key trends affecting the downstream processing industry in 2017? Senior
executives from leading technology developers discuss the challenges and prospects facing
refiners. To begin this review of the prospects for petroleum refining, we examine a longrange view of trends in the International Energy Agencys latest World Energy Outlook.

f there is a key conclusion to be drawn from the

International Energy Agencys (IEA) latest predictions
for the oil industry, it is that the demise of petroleum
processing is some distance beyond the forecasters
horizon. The 2016 World Energy Outlook (WEO) points
out that the governmental promises on tackling
greenhouse gas emissions made within the Paris
Agreement will not lead to a peak in oil demand before
2040. In particular, the likelihood of finding sufficient
alternative feedstocks for major consuming sectors
including road freight, aviation and petrochemicals
means that expected growth in just those three
activities will far outweigh expected growth in oil
demand globally during the coming quarter century.
Of more immediate interest, the IEA draws up a
changing pattern of global demand for oil. For instance,
India will become the leading source of future growth,
while China will surpass the US as the largest national
consumer after 2030.
Returning to the major competition for oil refining
in its key markets, the number of electric cars worldwide is set to grow rapidly, from 1.3 million in 2015 to
over 30 million by 2025 and to more than 150 million by
2040. That latter figure displaces 1.3 million b/d of oil
processing. Although battery costs will dip to less than
half todays level, the time required to recoup the added
cost of investment in electric cars is sufficiently higher
than the equivalent for conventional cars that more
rapid growth in electric car ownership is held back.
The worst case scenario, from refinings point of view,
is that new support policy emerges globally, leading to
faster roll-out, with as many as 710 million electric cars
on the roads in 2040, a number that would displace
around 6 million b/d of petroleum processing.
Predictions for the supply side of oil processing
revolve largely around the worlds newest swing
resource, tight oil. The dip in oil prices of the last two
years led to predictions of mayhem in production in
the US tight oil industry, in particular through 2016.
But the fracking industry proved to be far more resilient than many expected. In turn, this means that a
stable recovery in global oil prices will not have any
dramatic impact on tight oils development. According
to the WEO, output from tight oilfields will reach
a high point in the late 2020s, peaking at a little over
6 million b/d. With greater fuel efficiency and fuel
switching on the nations roads, the US will effectively

www.eptq.com

outlook copy 8.indd 1

eliminate net imports of oil by the end of the quarter

century in view.

Prospects for growth

The rest of the producing world offers a less gung-ho
future. Prospects for growth among other significant
producers will depend on investment decisions that
pre-date the tight oil boom. The drop in oil prices hit
the producing industry in Brazil and Canada particularly hard, but the IEA posits some reward for investment in those countries over the coming five years.
In the OPEC camp, production in Iran has risen by
around 700 000 b/d as a result of the lifting of international sanctions at the start of 2016, although official
estimates of further growth in production may prove to
be ambitious. Iraq has shown the biggest recent gain in
output among OPEC members, but declining oil revenues and continuing security issues may also affect the
nations future prospects for production. Even Saudi
Arabia has its economic woes and is looking for opportunities to develop away from an oil-reliant structure.
Towards the end of 2016, OPEC finally conceded that
it could cut back production to hold up crude prices.
But the organisation appeared to require that nonOPEC members, in particular Russia, should share
the burden and march in step by cutting production.
However, a more natural decline in production from
existing conventional crude oil fields is likely to have a
more lasting impact than any gathering of energy ministers. The IEA points out that the current decline in
output is equivalent to losing the current oil output of
Iraq from the global balance every two years.
The volume of conventional crude oil resources which
received the go-ahead for development in 2015 fell to
its lowest level since the 1950s and there is no current
prospect for a rebound in output. If this decline were to
continue deep into 2017, the IEA warns, there is a risk
of a return to volatile oil prices: There is no reason to
assume widespread stranding of upstream assets for
oil, even in a drive to decarbonise the energy sector
provided governments pursue unambiguous policies
to that end. Investment in oil and gas, albeit at reduced
pace, remains an essential component of an orderly
transition to a low-carbon future. But abrupt changes
to climate policymaking or misjudgements of future oil
demand by the oil industry could lead to more severe
financial losses.

PTQ Q1 2017 5

14/12/2016 11:58

Refining capacity and runs by region (million b/d)

Capacity Net capacity Refinery runs

2015 change to 2040 2015 2025
2040
21.3
-0.3
19.3
18.2
16.0
North America
Europe
16.5
-1.2
13.7
11.8
10.0
Asia
OECD Asia Oceania 7.6
-0.9
6.8
5.8
4.8
China
12.8
4.9
10.8
12.6
14.8
India
4.4
3.4
4.6
5.3
7.6
Southeast Asia
5.0
2.6
4.0
4.8
6.6
Russia
6.2
0.1
5.5
4.9
4.6
Middle East
8.8
4.3
6.4
10.1
11.6
Africa
3.3
1.7
2.1
3.1
4.1
Brazil
2.1
0.8
2.0
2.3
2.7
Other
4.9
0.5
3.4
3.7
3.5
World
94.8
16.1
79.7
84.1
87.6
Source: International Energy Agency

Table 1

Better for refiners

Although upstream operations have felt the pain of collapsing oil prices, the news has been far from disastrous
for the refining industry. Margins have reached their
highest levels for some years as refiners benefited from
cheaper feedstocks while at the same time managing
to prop up product prices. Even the European refining
industry, where falling local demand and competition
from exporting refiners mean that the region has been
on a steady decline for the past decade and more, has
enjoyed a relatively good two years. For all that talk of
European decline in refining, it is as well to note that
about a third of the global increase in refining throughput during 2015, amounting to about 2 million b/d,
came from refineries in Europe.
However, the IEA warns, higher margins and higher
throughput are all very well if supply is in balance with
demand. Even though recent growth in demand for oil
products was at its highest for the past 10 years, excluding a post-recession bounce, it still was not enough to
absorb all of the extra output of refined product in
Europe. Added to this, some of the increased demand
in recent years for petrochemical feedstocks has been
met by natural gas liquids products including ethane
and LPG. These are materials which are for the most
part outside the scope of refining. Refined product
stocks increased throughout 2014 and 2015 and started
putting pressure on the margins at the beginning of
2016. As a result, refiners in Europe have no option
other than to cut back throughput to below their 2015
levels.
This pull-back in throughput has not been limited to
European refiners. Such is the inflated over-capacity in
global refining, a slump in crude oil prices brought on
by over-production upstream, coupled with a period
of strong demand pushing up product prices, cannot deliver upward movement in refining margins for
ever. The WEO also states that the headline growth in
demand for oil distorts the picture for refiners since natural gas liquids processing and other non-refined products claim increasingly higher market shares.

PTQ Q1 2017

outlook copy 8.indd 2

Global growth
For the long term outlook, global refining capacity will
grow by over 16 million b/d by 2040, but global figures
do not paint a picture of the winners and losers who
will contribute to the net figure. Growth will be in large
measure down to expanding refinery throughput in the
Middle East which is forecast to double as the region
seeks to match rapidly growing domestic demand for
refined product (set to grow by 1.3% compared with a
global average of 0.4%) as well as the export markets it
aims to win through lower-cost production and refining
structures. Despite their rapid expansion of domestic
refining, both India and China are expected to become
net refined importers of oil product, even though the
two nations will account for 3 million b/d and 4 million
b/d respectively of the 16 million b/d global figure (see
Table 1)
Companies both from oil exporting regions such as
the Middle East and Russia and from significant centres of refining such as Japan and Korea are expected
to start looking for downstream partnerships to
expand their businesses. This search for markets leads
to sub-Saharan Africa, where a hefty deficit in transport fuels demands extra refining capacity. At the same
time, refining capacity elsewhere will become surplus
to needs in view of lower demand and refinery runs.
Almost 15 million b/d of capacity is at risk of closure
during the next 25 years, with Europe accounting for a
third of the total.
Chris Cunningham is the Editor of PTQ

Eric Benazzi

Vice President Marketing & External

Communication
Axens
In 2016, global growth should fall
to 3.1% before rising back up to
3.4% in 2017 according to IMFs
latest estimations. Long-term perspectives in industrialised countries are moderate and a firming up
of growth is forecasted at mid-term
in emerging and developing countries. These forecasts rely on several
important hypotheses including, in particular, the gradual slowdown of the Chinese economy with growth
rate close to 6% at mid-term. The Chinese government will have to manage the transition from a growth
driven by investments and exports to a model based
on domestic consumption and development of services
without leading to a major economic slowdown. India,
another heavyweight in terms of global demography,
should continue growing at a sustained rate of 7.6% in
2017.
Growth in demand for crude oil and products associated to its extraction should be close to 1.2% to reach
97.5 million b/d in 2017. Uncertainty on crude oil oversupply should remain and will equally depend on the

www.eptq.com

14/12/2016 11:58

COMPLETE SOLUTIONS FOR

YOUR REFINERY CHALLENGES
Todays Refinery Challenges

Processing tight oil

Managing stringent sulfur limits
Monetizing orphan streams
Upgrading residuals

CB&Is Comprehensive Solutions

CB&Is broad portfolio of both refining and petrochemical technologies,
combined with our execution expertise, will help you maximize processing
flexibility and achieve margin benefits in the widest range of scenarios.
We are with you through every stage of the process plant life cycle, from
feasibility studies to identifying plant optimization and upgrade solutions,
through technology selection, full-scope EPFC, commissioning and start-up.
PROCESS PLANNING AND DEVELOPMENT
LICENSED TECHNOLOGIES AND CATALYSTS
FULL-SCOPE EPFC SERVICES
PROJECT MANAGEMENT AND CONSULTING
AFTERMARKET SERVICES

A World of Solutions

Visit www.CBI.com
19M112016H

cbi.indd 1
cbi_ptq_q1_2017_ad.indd
1

14/12/2016
10:32
12/2/2016
4:35:21
PM

OPECs policy about its level of production and the

tight oil level of extraction itself related to crude oil
prices. As a final result, crude oil prices should continue
to fluctuate between $40 and $60/bbl during 2017. This
will contribute to maintain investments in exploration
and production at low levels and could translate to a
production shortfall as we approach 2018-2020, leading
to a new oil crisis.
For more than 20 years, gasoline and diesel have
been gradually reformulated and their sulphur content reduced to comply with legislation mandating
decreased emissions and lower levels of airborne pol-

Friday 4th November 2016 will

remain a historic day for people
and the planet with the Paris
Agreement entering into force
lutants. In 2017, this trend will continue. The US government is enforcing the Tier 3 regulation standards
and large US refineries must comply by 2017. To help
curb air pollution, which has reached worrying levels in Chinese cities, the government is aggressively
implementing China V fuel quality standards. By 2017,
on-road fuels should contain less than 10 ppm sulphur
and non-automotive diesel will have to comply with
China V standards by 2018. Air quality is also an issue
in many Indian cities. The Ministry of Road Transport
announced that Indian refineries will have to produce
Bharat Stage 6 standard fuels, an equivalent of Euro 6,
by 2020.
Recently the IMO confirmed that the sulphur content
of marine fuels will be capped at 0.5 wt% globally with
effect from 1st January 2020.
Those new regulations will require investment in
additional hydroprocessing capacities.
Dieselgate, which is witnessing new developments
in late 2016, could lead to a partial balancing of the
on-road diesel/gasoline demand ratio in Europe which
experiences the highest rate of dieselisation worldwide at about 2.3 (v/v). In this ever-changing situation, improving the refining assets flexibility enabling
it to adapt to market trends virtually in real-time will
become a must. Technologies, advanced catalysts and
digital solutions will be the successful combination for
a higher performance and profitability.
In addition and more than ever integration of refining and petrochemical sites will be required to mitigate
risks related to raw material and product price variations, to overcome market trend changes, and to benefit
from the stronger dynamic of the petrochemical market
and improve asset profitability.
Friday 4th November 2016 will remain a historic day
for people and the planet with the Paris Agreement
entering into force. With ever-tightening environmental
regulations, new significant challenges that will drive
innovation are emerging from the energy transition and
climate change.

PTQ Q1 2017

outlook copy 8.indd 3

Energy consumption represents 34-69% of the refinerys total operating costs depending on its location
and the local energy cost. Therefore, every fraction of
energy savings in refining or petrochemical units is not
only a way to reduce greenhouse gas emissions but also
to create value.
Our experience as a licensor has taught us that
during projects, tight schedule and lack of time mean
energy efficiency is not sufficiently optimised. To take
full advantage of our strong knowledge in process and
technology, we developed the CEED methodology,
standing for Custom and Efficient Early Design. During
the early stage of the project, in close collaboration
with the refining and petrochemicals operators, alternate schemes are defined, carefully evaluated and the
most appropriate option is selected taking into account:
energy constraints, capex, opex and plant operability.
The decisions made by the next US President and its
administration will also be key factors that will have an
influence on the course of events in 2017 and after.

Chet Thompson
President
AFPM

The first days of the Trump administration are about to get under way,
and we are cautiously optimistic on
the outlook for the fossil fuels industry. The contrasts between President
Obama and President-Elect Trump
are stark. While Obama has been
hostile to fossil fuels, Trump has
made positive comments about
supporting and promoting energy growth through a
sensible regulatory framework. For the refining and petrochemical industries, this can only be received as a sea
change in our ability to work constructively with the
White House and executive agencies.
The 2016 elections also resulted in a Republican controlled Congress, creating a newly unified Republican
government that could lead to pro-manufacturing and
pro-energy policies.
Already, the incoming administration and Republican
controlled Congress are talking about an issue that
AFPM has been promoting for years: the need to
improve our nations infrastructure, which is aging and
reaching a desperate state of disrepair. A safe and modern infrastructure allows our refineries and petrochemical manufacturers to get feedstock to our facilities and
to distribute our products to consumers. Unfortunately,
for the last several years, infrastructure projects have
been politicised, and often in reckless and dangerous
fashion. Under the new administration, we are optimistic that the keep it in the ground fringe will be sidelined and the potential for a comprehensive bi-partisan
infrastructure bill is real.
Of course, optimism throughout the oil and natural gas industries must be tempered with the reality

www.eptq.com

14/12/2016 11:58

that government is slow to move, bureaucracies can be

US Congress, among a growing and diverse group of
3self-powered
www.gartner.com/newsroom/id/3185623
After all, and
there
is alittle
t intime
These
ciencies
cost
where
overall
cost and
of installavesselvehicle
offers and
a potenoperations.
It has
lowbene
mainteintractable,
policy
impacts
take
to be the
felt.ef
That
organisations
that include
food, motor
even
4
www.cisco.com/c/en/us/solutions/internetnew
applications
if
they
comprosavings
will
enable
plants
to
focus
costnot
of take
maintenance
and so ongroups
tial
solution.
nance
and requires
minimal
is
whyburden
representatives
of our
industry tion,
should
environmental
and, of course, the new adminof-things/iot-fog-computing.html
mise
the for
core
capabilities
of to
thebe on
the areas
where istration.
greater thought
is much
higher.
operator
intervention
normal
anything
granted.
Weinneed
as
assertive
and
With the 5United
States producing more of its
www.thinnect.com/mist-computing
automation
whetherown
thatoil
is and
accomAcknowledgements
operation. in system.
Combustion
also is
aggressive
our
publicThese
affairs chalefforts,
as required:
we have been
gas than
at any time in history, and prolenges
are
not
insurmountable,
modating
and integrating
legacy
authors thank
Costain
for support
Conclusion
destroys
anyyears,
hydrogen
sulphide
the
last eight
to
achieve
our policy
goals.
ducing it in
a clean The
and efficient
manner,
the reasons
for
in
producing
this
article,
which
was first
however,
and
IIoT
technologies
will
equipment
and
IP,
or
considering
There
arefor
a number
of technologies
present
in foremost,
the vapour,
avoiding
First and
we at
AFPM are
eager
our
RFS
policy are no longer valid.
presented
at the GPA
GCC Joint
McAdam
is Europe/GPA
Chief Architect
with
also
help
transition
carefully
where
in the
architecture
available
for VOC
recovery
from Rohan
complications
venting
gasesand
atabout
nation
to smooth
haveofa the
rational
debate
a sustainable
This
increased
production has made us keenly
Conference inProcess
Athens, September
Honeywell
Solutions 2016.
and has
implementation
of
new
capabilities
to
place
capabilities.
displaced
ships
thepast few years of our industrys great
levels that
may One
be harmful
to oper-the vapour
energy
policy.
that includes
realisation
that from
aware
overinthe
worked primarily on the development of
by
removing
mainte- marine
As said,loading
in manyneed
we are
of ways
crude
oil.
ators
workingburdens
locally.from
Combustion
to focus
on attracting
a new
generation
ofindustrial
skilled
References
machine
interfaces
for HPS
nance
and engineering
at theisstart
of this
journey.
The human
Selection
dependent
on the scale
does generate
nitrous teams.
oxides, but still
professionals.
1 ec.europa.eu/eurostat
(from
European
process
control products. He
now leads
the
Animpact
appliance
based
approach
to IIoT
undiscovered
of the remains
facility andanother
site
any
can be
assessed
relative
In
the specific
coming years,
we expect
have thein need
Environment
Agency)
development
of IIoT to
architecture
HPS.
new
Edgeconsiderable
capabilities, for
instance,
waiting ofto
be
factors. Many
available
to the
benefits
of country,
forthe
up explored.
to
two million
skilled
workers
degree
fromship
welders
2 Moylan
Shaw
T P, Minimise
loading
He
holds
aJ P,mathematics
and
a PhD
will
signi
cantly
cut
down
set-up
However,
the
lessons
from
decades
a crude
oil engineers
destroying VOCs. Assist fuel, processes incorporate
emissions,
Hydrocarbon
Aug1998.
in
computer
from
Charles
Sturt
to PhD
chemical
science
due toProcessing,
retirements
and
and
maintenance
plug-anddevelopment
of allowing
automation
absorption
column,industry
the We
supplied
as liquid,work:
is required
at of
University
a masters
degree inand
cognitive
growth.
knowand
that
the refining
petplay
boxes the
canVOC
connect
and DCSs
can recovered
serveindustries
us science
fromgreat
the University
New comSouth
VOC content
to rochemical
be
times when
contenttoof the
the systems
offer
careers,ofwith
Cloud
to
automatically
con
gure
well.
If
we
draw
on
them
wisely,
Wales.
conveniently into the
loadedsalaries
crude and
vapour is low.
Grantbenefits,
Johnson is Front
Solutions
Manager,
petitive
andEnd
are
focused
on
and
update,
relievingoffer
teams
of the we
of mapping
responsible
for aactivities
& gas, and
with
oil. can be condent
Other
technologies
a potenprogrammes
that reach
out to
varietyinofoilgroups
Paul
McLaughlin
isUK.Chief
Development
responsibility
set up a PC,
and out
a routecontrol
to bringatgenerations
usoffshore
safely toloadthe help
Costain,
Manchester,
He holds
a masters
Vapour
tial return on toinvestment,
through
that
us spread
the word.
Engineer
withchemical
Honeywell
Process Solutions
and
install,
con
gure
and
maintain
new
world.
degree
in
engineering
from will
the
facilities presents
greater
recoveryis of
liquids
or generation
energy
vital
for modern
day life,ofanding
the recognition
Lastly, the
big issue
on
everyones
mind Council
is how
leads
the
HPS
Chief
Engineering
and
the
University
of
Cambridge,
UK,
and
is
a
Member
software.
Virtual
engineering
platchallenges.
If it is technically
feasipower,
theirisadditional
that
thebut
world
going to complexuse petroleum
for decades
the regulatory
culture
change
under
a Trump
adminisGlobal
Architecture
working
on system
of the Institution
of Team,
Chemical
Engineers,
and a
forms,
meanwhile,
entire
to are
install
equipment
to transport
ity,come.
footprint
weight
count
to
We willand
win enable
this debate.
The ble
facts
on our
tration.
We believedesign,
he will
end
this
governments
war
product
line
architectures,
formal
Chartered
Engineer
in
the
UK.
automation
systems
to
be
engithe vapour
to fuels
shore,
significantly
them.
With
side.
But ouragainst
challenge
will
be the
to convert
this to by
the pipeline
on fossil
and roadmaps,
take affirmative
to reduce
the
References
investmentsteps
prioritisation,
strategic
neered
theof
Cloud.
Equipment
an onshore
plant headwinds
may planning,
relativelyinlow
value
of the
recovered
1recovery
www.tec
hradar.com/news/internet/
development
sound
policies
that balance
thein
need
regulatory
that
have
impeded
economic
intellectual
property
management,
Mike Dodd is a Design Safety & Environmental
such
as computers
and
networking
begrowing
most economical.
If this
is job
notcreation.
product
at the and
timereliable
of writing,
techand
cross-business
major
initiatives.
He
holds
for
affordable
fuels,
and asamsung-it-s-time-we-shut-up-about-theecongrowth
and
Wewith
certainly
do
not believe
we
Engineer
Costain,
Manchester,
UK.
He
hardware
only
need
to be procured
internet-of-things--1314979
practical, for example
if aloading
nologies
offer
a modest
payback
masters
degree
in in
engineering
from
omy
withthat
sound
environmental
policies.
will get
free pass;aholds
nor
we
want
one. But,
we the
do
awould
masters
degree
chemical
engineering
at
the2017,
end
of the
2from
www.businessinsider.com/iot-ecosystemof Pennsylvania,
and
a bachelor
SPMofbuoys
long way
fromget University
onshore
in we
addition
to their dramatienvironIn
willprocess,
continue
to fight for
repeal
the a believe
we will
reasonable
regulations
that
are
from
the University
of Manchester,
UK good
andof
is
internet-of-things-forecasts-and-businesscally
streamlining
the
automation
science
degree
in
mathematics
and
computer
a vapour combustion
system and
mental benefits
cannot
realise (RFS)
such a or shore,
anenvironment.
Associate Member of the IChemE.
Renewable
Fuel
Standard
for significant
for the economy
opportunities-2016-2?IR=T
from the University of Delaware.
system
process.
Email:ismike.dodd@costain.com
installed
a moored
barge I or
return engineering
offshore
applications
reform.
I in
am optimistic
that
we could see
changesonwith
One thing
will science
say
that 2017 is certainly stacking
this policy this year, due to growing support in the
up to be an interesting year.

In the coming years, we expect

to have the need for up to
two million skilled workers
from welders to PhD chemical
engineers

NEXT GENERATION

VAPOR PRESSURE TESTING

MINIVAP VP VISION

Premium Vapor Pressure Tester for Gasoline, Crude, LPG

Widest Pressure Range: 0 - 2000 kPa
Highest Precision and Accuracy
Certified for Robustness and Durability
Monitor Your Processes Through Cockpit Software
- 24/7 Access to Multiple Analyzers
- Enables Remote Diagnostics, Help and Troubleshooting
- Enables Global Results Download, Collaboration & Audit

ACCESS.
ANYWHERE.
ANYTIME.
Phone +43-1-282 16 27-0 | Fax +43-1-280 73 34 | info.grabner-instruments@ametek.at
http://www.grabner-instruments.com/Products/VaporPressure/VpVision.aspx

VP Vision half page.indd 1

www.eptq.com
www.eptq.com
www.eptq.com

outlook copy 8.indd 4

25.02.2016 11:29:54

PTQ Q1 2017 9
PTQ
PTQ Q3
Q4 2016
2016 47
65

14/12/2016 11:59

New 2nd generation HyBRIM NiMo catalyst

For significantly
better performance,

The answer is HyBRIM

The new 2nd generation TK-611 HyBRIM catalyst for high pressure
ultra-low sulfur diesel and hydrocracker pretreatment service.
25% higher HDS and HDN activity
Unmatched stability throughout the cycle
Superior volume swell for increased profitability
Now thats performance that pays.

Scan the code or go to

info.topsoe.com/tk611

topsoe.com

haldor.indd 1

14/12/2016 09:54

ptq&a
Q

We are interested in buying new crudes on the spot market.

How do we assess the potential side effects of processing them
in refinery units?

Andrea Fina, Technical Coordinator Development &

Marketing Division, Chimec, process@chimec.it

When a refinery introduces new or unknown crude oils

to its typical crude oil slate, possible side effects that
cannot be predicted with conventional tools used by the
planning department (crude assay, linear programming
model and so on) may arise.
Indeed, unlike distillate yields, finished products
specifications, fouling and emulsion issues induced by
the feed blend incompatibility are very hard to foresee.
If not predicted to be avoided or promptly solved,
once these side effects arise it might be too late for corrective actions: the consequences could be huge, resulting in economic losses that can eventually wear away
the profits linked from the purchasing of opportunity
crudes on the spot market.
Emulsion stabilisation and fouling deposition in a
crude distillation unit (CDU) can originate from many
sources, however the most common is asphaltenes
precipitation.
Asphaltenes are by definition the fraction soluble in
toluene and insoluble in heptane, characterised by a
polynuclear aromatic core with various heteroatoms in
the structure, and some aliphatic chains. Asphaltenes
are dispersed in the oil thanks to the formation of a
resins-asphaltenes complex; its solubility varies according to the relative contents of aromatics and paraffins. It
is clear that such a delicate balance can be easily upset,
leading to asphaltenes destabilisation and eventual
precipitation.
For example, mixing a paraffinic crude oil with one
rich in asphaltenes can lead to asphaltene precipitation.
The temperature plays an important role as well: while
the solubility of the asphaltenes increases with temperature, at the same time the consequent reduction of the
oil viscosity allows a higher collision factor of Brownian
flocculation, hence resulting in possible precipitation.
As already stated, asphaltenes precipitation promotes
undesired side effects as strong emulsion stability in the
desalting vessel and fouling build-up in the CDU preheat train.
Flocculated asphaltenes are able to stabilise the waterin-oil emulsion because they are adsorbed at the water
droplet interface, avoiding, or at least inhibiting, coalescence of droplets. Indeed, asphaltenes are surface active
molecules able to reduce the interfacial tension: this
leads, together with the creation of interfacial gradients,

the droplets to behave elastically to tangential stresses.

The formation of an elastic oil-water interface is the
main factor of emulsion stabilisation. As a result, a stable water-in-oil emulsion leads to:
High water content in the desalted crude oil with consequent increase in energy consumption
High salts content in the desalted crude oil, thus an
increased risk of corrosion in the main fractionator
overhead system, increased fouling deposition in the
crude distillation unit and in downstream processes
and increased catalyst poisoning of residue hydrocracking (if present)
High hydrocarbon content in the brine water that can
affect operations at the wastewater treatment plant.
On the other hand, fouling deposition in the heat
exchangers train reduces the overall heat transfer efficiency, and consequently the heat recovered from the
hot fluids in the preheat train. This translates into a
steeper furnace inlet temperature (FIT) decay: a higher
fuel consumption is then required to reach the distillation temperature. For instance, a CDU processing
10 Mt/y of crude oil, a nFIT reduction of 10C/y has an
energy cost of 2500 toe/y.
Finding a reliable tool to predict in advance the
asphaltenes potential instability, in any possible blend
the refiner can process, is a keystone to anticipate and
handle the related issues. For this reason, Chimec
built in new research on crude oils and their blends to
develop the Crude Oil Stability Model (COSMO), a tool
able to predict the stability of crude blends in terms of
asphaltenes precipitation risk, starting from the stability
analysis of each blends components.
The COSMO program is based on the Chimec
Stability Index (CSI), a value that directly indicates the
intrinsic stability of the asphaltenes in the oil matrix.
CSI is measured with ASTM D7157, by means of a
Rofa Fuel Stability Analyser, so it can be applied to any
hydrocarbon stream containing asphaltenes, crude oils,
gasoils, residues and fuel oils.
Thanks to the CSI it is possible to predict both the
fouling tendency of each crude oil and/or the tendency
in stabilising the emulsion. Long research efforts both in
Chimec laboratories and directly on the field led us to
develop a huge database of CSI for the most common
crudes processed.
Refineries seldom process a pure crude oil; indeed the
feed blend can usually consist of three to 10 crude oils.
The stability of the feed blend can be calculated starting from the stabilities of the individual blends constituents thanks to the rigorous COSMO model. Mixing
two or more crude oils can lead to a mixture that is less

Additional Q&A can be found at www.eptq.com/QandA

www.eptq.com

Q&A copy 24 (original).indd 1

PTQ Q1 2017 11

14/12/2016 12:05

more recent assays in order to have

more accurate results.
CSI Real
Prior to the utilisation of a LP
80
CSI Cosmo
model, crude properties should be
70
reviewed using the selected crude
assays. Properties such as cut yields,
60
sulphur content, density and TAN
50
(acidity) will allow a comparison of
a new crude oil with the usual crude
40
oils processed in the refinery. Then
30
distillation cuts can be calculated
20
using specific software, critical properties for process units feedstock such
10
as sulphur content, mercaptans in ker0
osene, cetane index in the gasoil cut,
0
10
20
30
40
50
60
70
80
90 100
metal content in vacuum gasoil can be
Isthmus, vol%
reviewed as they have an important
impact on product properties, metalFigure 1 Asphaltenes stability of a crude mixture
lurgy, catalyst life cycle and so on.
From this first phase of the crude
stable than all its components. This is due the non- evaluation the conclusion may be that the concerned
linear behaviour of asphaltene stability. For instance, crude cannot be processed individually but in a blend
mixtures of Isthmus and Western Desert have a lower with other crudes available in the refinery. In this case,
stability than the pure crudes (see Figure 1).
a LP model will be able to optimise the crude blend
It is important to highlight that the major part of suitable for the refinery while maximising refinery
crudes exhibit a behaviour on blending similar to the margin.
one showed by Isthmus and Western Desert.
Regarding process units, the LP model will take into
Thanks to the predictions of COSMO, it is possible to account capacity constraints or other hydraulic limitacalculate the stability of a blend whatever its composi- tions as well as operating modes. Ranges of variation
tion is. Once all the crude oils processed have been ana- on cut points for the specific crude units are also intelysed in terms of CSI, no other analysis is required until grated in the LP model.
a new crude oil is processed. When this occurs, CSI
The main LP model outputs are the optimal refinery
analysis of the new crude(s) is required only.
margin, routings and unit operating modes when proThe CSI calculated by means of COSMO for a cessing the new crude oil (individually or in a blend).
selected blend can be used to:
A Tim Olsen, Refining Consultant, Emerson Automation
Continuously evaluate the asphaltenes deposition
Solutions, Tim.Olsen@emerson.com, and Jake Davies, Global
tendencies of the CDU feed
Predict the behaviour of blends planned for the Marketing Director, Permasense, jake.davies@permasense.com
future, or even to define a proper one to minimise the First, be sure to understand whether the opportunity
crude you are interested in buying is from one source or
risk of asphaltene precipitation
Optimise the demulsifier and antifouling dosage a blend of two of more crude oils. More and more refinaccording to the real emulsion and fouling tendency of ers are finding they have purchased a blend of crude
oils, and many times this blend is considered incomthe crude oil blend.
patible (crude unit preheat exchanger fouling accelerA Miren Carro-Usle, Studies and Support Group Manager, ated with asphaltene precipitation). This was a topic of
Performance Programs Business Unit, Axens, Miren.CARRO- discussion at the 2016 Crude Oil Quality Association
USLE@axens.net
(COQA) meeting in Houston.
On the spot market crude oils are traded on short terms.
If the crude oil is from a single source (and not selfThe advantages are the attractiveness of the price as the incompatible), hopefully the crude oil properties match
spot market is influenced solely by supply and demand.
the design and capabilities of the refinery configuraAnother characteristic of such a system is the place- tion (sulphur content and major cut yields of naphtha,
ment of oil for disposal around the world no mat- distillate, gasoil, and resid). More than likely, a single
ter what country it comes from. Therefore, crude oils crude oil will not match the capabilities of the refinery
on the spot market can be unknown for a refinery and and thus need to be blended with other crude oil(s).
have to be evaluated in order to assess the refinery Crude oil blending introduces the potential for insolumargin that can be obtained and how to process a new ble asphaltenes (incompatible crude blend).
crude oil in the refinery.
With crude oil blending, a refiner must learn the maxThe planning LP model of the refinery or a less com- imum content of one crude oil that can be blended with
plex LP model adapted to crude evaluation is helpful another crude oil without having a significant impact
for such an assessment. The crude assay data will come on the risk of asphaltene precipitation. This includes the
from a crude assay library. It is important to select the order in which one crude is blended with the other if
Chimec stability index

90

12 PTQ Q1 2017

Q&A copy 24 (original).indd 2

www.eptq.com

14/12/2016 12:06

Better Technology and

Increased Profits

ISOMIX-e reactor internals design has been

setting industry standards based off of more than three decades

worth of experience, research, and hard work. It is todays most
innovative and environmentally-friendly design. It not only provides
safer and cheaper operations and easier maintenance, it also
maximizes catalyst utilization so that you can get the most out of the
catalyst youve paid for. Furthermore, these state-of-the-art internals
increase the run length and unit throughput so that you can be even
more productive than ever before.
Please contact us at: techma@chevron.com
clg.indd 1

14/12/2016 11:25

done in a single tank prior to being processed. If crude

oil A cannot be above 30% when blended with crude oil
B, then you do not want to add crude oil B into a tank
of crude oil A (rather, just the opposite adding crude
oil A up to 30% into a tank of crude oil B). Percentage
matters, and this is valuable information to determine
the maximum amount use of a discounted opportunity
crude that will not increase the risk of fouling and/or
corrosion.
Emerson has significant experience with automating crude unit preheat exchangers to detect when the
rate of fouling has increased. At the inception of the US
tight oil boom when refiners on the Gulf Coast started
using Eagle Ford, many refiners asked Emerson to assist
with monitoring beyond the traditional monthly Excel
spreadsheet manual checks. With WirelessHART (the
process industrys first international open wireless communication standard (IEC 62591, EN 6259)), refiners are
able to easily add instrumentation around each bundle to monitor the rate of fouling; fouling is not linear
across all bundles, therefore it is important to monitor
each heat exchanger bundle individually.
Tight oils tend to have more naphtha and less resid,
and the extra naphtha content can destabilise a crude
with high asphaltene content that may on its own have
a good asphaltene solvency (asphaltenes do not dissolve in crude oil but exist as a colloidal suspension).
Mixing stable crude blends and asphaltic and paraffinic
oil creates the potential for precipitating the unstable
asphaltenes by reducing the solvency capability of the
base crude. It is important to note that asphaltene content is not directly correlated with asphaltene risk. High
asphaltene content may be stable asphaltenes while low
asphaltene content may be unstable asphaltenes; it is
more of a factor of the solvency capability of the base
crude. Emerson provides the automation tools to monitor and enable a refiner to determine the crude oil blend
percentages that produce accelerated fouling.
Assessing the corrosion risks of opportunity crudes is
very difficult. Most are more corrosive and some idea
of the increased risk can be ascertained from the total
acid number (TAN). However, there are many corrosion risks that are carried in crudes that are not captured in total acid number. Often these corrosion risks
are only activated under certain operating conditions,
for example at elevated temperatures. Operators considering opportunity crudes should carry out a risk evaluation for the components of the process equipment, and
ensure that they have reliable corrosion monitoring in
the areas of increased risk to ensure the impact of that
increased risk (internal metal loss) to the asset integrity
is adequately controlled in service.

Michael Tinning Schmidt, Product Manager, Refinery

Business Unit, Haldor Topsoe, mtp@topsoe.dk

Buying spot market or opportunity crudes is a great

approach to increase refinery margins. However, processing the cheaper crudes will impact the refinery
hydroprocessing units due to lower feedstock quality in terms of, for instance, contaminants, density,
and TAN number (acidity). For the contaminants part,

14 PTQ Q1 2017

Q&A copy 24 (original).indd 3

these crude challenges will directly influence the performance of the installed hydroprocessing catalysts
and will, as a side effect, trigger that more additives are
needed for corrosion and fouling control. From a hydroprocessing catalyst point of view, the spot market crude
thus implies more exposure to Ni, V, Fe, and As from
the crude itself and exposure to more Si, P, Ca, and so
on from the increased additive injection.
The increased level of contaminants that has to be
handled by the hydroprocessing catalysts makes it even
more important to install the optimal guard and grading catalysts to protect the main high-active catalyst.
The key to handling high levels of heavy metal and
other catalyst poisons is best solved by using alumina
based catalyst where the porosity (pore volume, pore
size and surface area) is designed to allow diffusion and
storage of the unwanted contaminants.
Being the leading supplier of guard and grading catalysts, Haldor Topsoe has the specialty products and

Buying spot market or

opportunity crudes is a
great approach to increase
refinery margins
experience to solve contaminant problems. Dedicated
Haldor Topsoe catalysts for trapping phosphorous, silicon, nickel, and vanadium are recognised to be top tier.
Inorganic iron, frequently seen to increase from running
acidic crudes, is effectively stored in the inert TopTrap
product. Faced with the challenges of spot market
crudes makes it very important for the refiner to consult with the hydroprocessing catalyst supplier because
almost all catalyst solutions are unique and have to be
designed to match the specific situation.

Joe Ritchie, Integrated Project Solutions, Honeywell UOP,

Joe.Ritchie@Honeywell.com

New crudes are typically discounted when initially

marketed. This presents an opportunity to improve
margin, but measures should be taken to anticipate
crude behaviour to mitigate risk. Potential side effects
of a new crude depend on refinery configuration and
prevailing constraints. The further such a crude departs
from the refinery design basis, the greater the likelihood
there will be adverse side effects. This is why many
refiners run new crudes as part of a larger blend to mitigate risk. Key crude properties are API gravity, sulphur,
nitrogen, TAN, pour point, viscosity, metals, salts, and
mercury. Lower gravities impair desalting, potentially
increasing corrosion and heater fouling. Lower gravities or higher nitrogen levels decrease FCC and hydrocracker conversions. Higher sulphur levels increase
hydrotreater temperatures which shorten run length.
Higher TANs increase corrosion in gasoil circuits. Pour
point and viscosity will impact oil movements in the
tank field and to the crude unit. Higher metals levels
can shorten cat feed hydrotreater or hydrocracker run

www.eptq.com

14/12/2016 12:06

At the end of the day, you

want a technology supplier
who works with you.
Youre committed to progress and success. Were committed to you. And we demonstrate
our commitment through licensing world-class refining, gas, chemical technologies and
specialty catalysts that drive exceptional performance. You can count on our proven
technology and long-term collaboration to help you keep pace with the increasingly
complex challenges of todays evolving marketplace. From initial consultation through
plant startup and beyond, our global team offers you practical guidance based on years
of real-world operating experience. Our goal is your success.

2016 Exxon Mobil Corporation. All rights reserved. ExxonMobil, the ExxonMobil logo, the interlocking X device and all product names herein are trademarks of Exxon Mobil Corporation.

Learn more about how we can work together to grow your business.
www.catalysts-licensing.com

exxon.indd 1

14/12/2016 10:31

Preliminary assay assessment and evaluation of crudes

S.No Characteristics
Density and API, kg/m3 and API
1


2
Reid Vapour Pressure (RVP), kPa
3
Pour point and cloud point, C

4
Kinematic viscosity, cSt

5
Asphaltene, wt%
6
Carbon residue, wt%
7
Salt content, PTB
8
BS&W, vol%
9
TAN, mg KOH/gm
10 Sulphur, wt%
11 Filterable solids, PTB
12 ASTM distillation D-86, C
13 Metals (Ni, V, Fe, Ca, Cu, Si), ppm

Significance
Weight to volume and vice versa calculation, checking consistency of crude oil, control of refinery
operation and give a rough estimation of crude oil. API gravity of lighter crude oil may be of the
order of 45 whereas in heavier API is 10-12.
Volatility property of a liquid fuel.
For estimating the relative amount of wax present in the crude oil. Cloud point gives a rough idea
above which the oil can be safely handled.
Viscosity indicates the relative mobility of various crude oils. Temperature has a marked effect
on viscosity.
Indicates the presence of heavier hydrocarbons in the crude.
Measure of thermal coke forming property.
Potential corrosion related issues.
Potential tower operation issues and energy consumption.
Potential naphthenic acid corrosion in high temperature.
Impacts the finished product quality and environment.
Potential emulsion and fouling issues.
IBP, 5% vol recovered, 10% vol recovered..FBP.
Potential impact in downstream catalyst and furnace tube coking.

Table 1

lengths unless vacuum tower cut points are reduced.

Marketers can supply references for refiners who
have processed the crude and those refiners will often
share their experience. Integrating the crude assay
in the facility LP model will provide estimates of unit
throughputs and performance.

Parag Shah, VP, Technical Services, Dorf Ketal Chemicals,

paragshah@dorfketal.com

Crude assessment and evaluation is a crucial step

which can impact refinery operations and the quality of
finished products. Typically, evaluation can be classified
into different types:
Preliminary assay: reviewing key basic properties
and distillation data
Short evaluation: physicochemical properties, yield
and further characteristics of straight run products
Detailed evaluation: detailed studies of physicochemical properties, yields and side cuts.
For purchasing crudes on the spot market, detailed
information/data may or may not be available and hence
it is wise to review the preliminary assay (see Table 1).
As there are chances of low quality crudes being
pushed in spot markets at discounted prices,1 one
should typically look for the following bad actors apart
from preliminary assay review:
Organic chlorides: originating from chemicals used
upstream, these are known as non-water extractable
chlorides distilling into the column overhead and side
cuts, causing potential dew point and under-deposit
corrosion.
Tramp amines: find their source from amine based
H2S scavengers used during transportation of crude;
can impact the desalter performance and potential
under-deposit corrosion in column.
Hydrogen sulphide: one should typically look for
potential and existing H2S which contributes to health
hazards and potential high temperature sulphidic
corrosion.
Crude blend compatibility: understanding the blend

16 PTQ Q1 2017

Q&A copy 24 (original).indd 4

compatibility can overcome asphaltene precipitation,

which otherwise can cause potential sludge formation resulting in poor tank preparation, rag layer in the
desalter and fouling in preheat exchangers.
Dorf Ketal has vast experience in treating opportunity and problematic crudes. Various refiners around
the world have benefited from the crude management
programme which includes multifunctional emulsion
breaker chemistry, reactive adjunct to address the tramp
amine issues, and compatibility studies of over 200
crudes. Specific problems can be addressed separately
with respect to a particular crude purchased on the spot
market.
Reference
1. Asian refiners struggle to process light Mexican oil, Hydrocarbon
Processing, Article 3445593, 2015.

Randy Rechtien, Field Tech Services Manager, Randy.Rechtien@

bakerhughes.com, and Gerry Hoffman, Desalter subject matter
expert, Baker Hughes, Gerry.Hoffman@bakerhughes.com

The processing of a crude which has newly come to

market offers both opportunities and challenges. The
same can be said of any existing crude which has never
before been processed at a particular refinery. In these
cases, while there is no perfect method for determining
future processing risks, there are several key parameters which should be considered.
For existing crudes, the best method for identifying
potential problems is to review accessible documentation and databases which provide details of industry
experience. In addition, input from additive suppliers
should be solicited. Baker Hughes maintains a benchmarking database of numerous crudes, which includes
commonly faced processing challenges.
For new crudes where little to no experiences are
available, there are four aspects of crude processing that
should be examined to help determine risks:
Physical properties
Production location

www.eptq.com

14/12/2016 12:06

Tower Technical Bulletin

Random packing uniformity improves performance

Background
Random packing has been around for over 100 years. Over that
time period, its geometry has evolved to improve performance
significantly. Looking at the various stages of random packing
evolution, it is clear that structures with a more uniform surface
distribution within the packing volume have better performance.
Sulzers NeXRingTM random packing, designed with this in
mind, gives you an industry leading combination of capacity,
efficiency, and strength.
Why Geometry Matters
Column internals tend to flood from localized high velocity
regions. As such, designers want to make the restriction across
a bed as uniform as possible. So the apparent contradiction is
getting uniform flow through a random bed. The solution is to
make the individual packing structure as uniform as possible
so, regardless of its orientation within the bed, the fluids
flowing through the column experience approximately the same
resistance.

Column Size Reduction Using NeXRingTM Packing

Case Study: Amine Contactor
With a typical amine absorber, high performance NeXRing
packing can allow designers to substantially reduce costs in
revamps and grass roots applications.
Grass Roots: Using NeXRing instead of Pall Rings allows for a
column diameter reduction of 18% while maintaining the same
product quality. In high pressure applications, this can lower the
vessel wall thickness by 18% and the vessel cost by 35%.
Revamps: Changing from P-rings to NeXRing packing will
increase capacity by over 40% while maintaining the same
efficiency and product composition. Pressure drop will be
reduced by over 50%.
In either case, this advanced random packing will provide savings
due to improved operating performance and/or lower CAPEX.

Sulzer NeXRingTM Random Packing

NeXRing packings are designed with reinforced ribs that are

spaced evenly throughout the packing volume. This uniform
mechanical structure ensures uniform fluid flow through
the bed. There are other benefits to this design as well. The
end flanges combined with strengthened ribs make NeXRing
inherently strong. The open structure lowers the pressure drop
by 50% versus conventional packings.

Europe, Middle East and India

Sulzer Chemtech Ltd.
P.O. Box 65
8404 Winterthur, Switzerland
Phone: +41 52 262 50 28
chemtech@sulzer.com

www.sulzer.com
Please check for your local contact
The Sulzer Applications Group
Sulzer has over 150 years of in-house operating and
design experience in process applications. We understand
your process and your economic drivers. Sulzer has the
know-how and the technology to design internals with
reliable, high performance.

Asia Pacific
Sulzer Chemtech Co., Ltd.
10 Benoi Sector
Singapore 629845
Phone: +65 6515 5500
ctsg@sulzer.com

North and South America

Sulzer Chemtech USA, Inc.
8505 E. North Belt Drive
Humble, TX 77396, USA
Phone: +1 281 604-4100
ctus@sulzer.com

Legal Notice: The information contained in this publication is believed to be accurate and reliable, but is not to be construed as implying any warranty or guarantee of
performance. Sulzer Chemtech waives any liability and indemnity for effects resulting from its application.

sulzer.indd 1

14/12/2016 11:12

 Compatibility
Refinery equipment/process conditions.
A given crudes physical properties and assay data
can provide insight into the levels of contaminants that
may result in more serious processing issues. Crudes
with low gravity (API< 25) and/or high solids content
(> 50 ptb) place a greater burden on desalter operations
and result in poorer salt removal. High solids loading
along with high asphaltene levels can lead to preheat
train and furnace fouling problems. Crudes with high
sulphur content (>2 wt%) and/or high TAN levels (>
1.5 mg KOH/g) will generate more significant corrosion problems in the crude distillation units and units
downstream.
When new crudes are produced from geographic
areas that already have existing crude production, it
is reasonable to assume that these new crudes may be
comparable to those that have already been processed.
As such, known industry data can be used to predict
potential risks for the new crude.
The compatibility of new crudes with crudes already
being processed must be considered. In particular,
there is always the potential for asphaltene precipitation when asphaltic crudes are blended with paraffinic
crudes. In these cases, the new crude itself may not
represent processing concerns. However, the addition of the new crude to the overall blend can have a
significant impact on asphaltene precipitation. To this
end, Baker Hughes has developed ASIT (Asphaltene
Stability Index Test), which quantifies the extent of
asphaltene precipitation and allows refiners to make
proper blending decisions.
Crudes can be processed reasonably well under the
best operating conditions, but any crude can be problematic under less than optimal conditions. For example, some crudes should be avoided if desalter vessel
temperatures are too low to provide acceptable performance. Naphthenic acid corrosion risk may be too
severe if only lower grade metals are present in high
temperature circuits at the crude unit. As such, a complete understanding of a given refinerys limitations
is critical in evaluating whether a new crude will pose
processing issues.
ASIT is a mark of Baker Hughes Incorporated.

What are the chief causes of poor conversion in the FCC

unit?

Kate Hovey, Senior FCC Technical Service Engineer, Johnson

Matthey, Kate.Hovey@matthey.com

The FCC is one of the most complex, albeit interesting,

units in the refinery which can make troubleshooting
a conversion loss scenario highly difficult. However,
much like all refinery operations, the FCC feed quality has the biggest impact on conversion on the FCC.
The aromaticity of the feed, its hydrogen content, and
the amount of contaminant metals in the feed are a few
of the things that can affect the cracking chemistry in
the riser. Feed quality should be monitored on a daily
basis by the unit engineer in order to preempt how the

18 PTQ Q1 2017

Q&A copy 24 (original).indd 5

feed qualities will affect the unit and its performance.

Fortunately, the resilient nature of the FCC means that
it can process a wide range of different feedstocks; however it is vital that the unit engineer is able to make
adjustments to the operation in order to counteract
any negative impacts cause by changes in the feed. For
example, increasing the fresh catalyst make-up rate to
maintain constant E-cat metals during periods where
the feed metals are elevated, or reducing the preheat
temperature to maintain a constant catalyst to oil ratio
when a higher feed MCRT content results in a higher
delta coke. Many refiners will have analysis of FCC feed
density and metals on a daily basis and the distillation
data 2-3 times per week, however less commonly other
analysis such as refractive index or basic nitrogen are
also very useful parameters to monitor. Basic nitrogen
is an important feed quality when it comes to conversion and it is often overlooked. As it is a temporary poison, it can go undetected as it is not concentrated on the
E-cat and therefore will not show up on the E-cat analysis like the other contaminants.
If the feed qualities have not differed from typical,
then there are a few other potential root causes that
can be reviewed when trying to troubleshoot conversion loss. Linked with the feed quality is the E-cat qual-

Feed quality should be monitored

on a daily basis by the unit
engineer in order to preempt
how the feed qualities will affect
the unit and its performance
ity, however changes to the E-cat quality are not always
a result of a change in the feed. The E-cat microactivity and total surface area (TSA) reported on the E-cat
analysis should be maintained by continuous fresh catalyst make-up. If these are reducing with time then it
is possible that the fresh catalyst make-up is too low,
or there is a preferential loss of fresh catalyst from the
unit. It is possible that there is a mechanical problem
with the fresh catalyst autoloader, resulting in the target fresh catalyst make-up rate not being achieved,
and this can be identified by reviewing the full catalyst balance, accounting for fresh catalyst and additive
make- up, withdrawals, reactor side losses, regenerator side losses and the regenerator bed density and true
level. A reduction in MAT activity or TSA as a result of
deactivation of the catalyst can be caused by the use of
torch oil, direct air heater firing or excessive afterburn
in the regenerator. Similarly, poor stripper operation
can result in a high hydrogen in coke weight percent
and excessive temperature excursions. The enthalpy of
combustion of hydrogen in coke to form water vapour
releases more than three times more energy than the
combustion of carbon to form carbon dioxide. High
regenerator losses can be caused by things such as high
fines content in the fresh catalyst, maldistribution in the

www.eptq.com

14/12/2016 12:07

Shell Global Solutions

STRIVE FOR NEW HEIGHTS

Take advantage of our technology portfolio to help boost your
refinery margins.
Static or declining margins? Want to increase capacity or take advantage of
opportunity crudes while meeting stringent specifications? By working together,
we can help you to respond to todays demanding environment profitably and
responsibly while planning strategically for tomorrow. Ask us how our cutting-edge
licensed technologies can help grounded margins take off and reach new heights.
www.shell.com/globalsolutions

regenerator, or high bed levels to name a few examples.

Assuming the E-cat is looking healthy and the FCC
feed qualities are typical of normal operation, another
possible cause of conversion loss is due to the heat balance which directly influences the catalyst circulation.
Conversion is impacted by catalyst to oil ratio which
is a function of the riser outlet temperature, the feed
temperature, the regenerator temperature, FCC feed
rate and riser steam. Insufficient riser outlet temperature, high preheat temperatures and higher regenerator temperatures will slow the catalyst circulation. High
regenerator temperatures may be caused by insufficient
catalyst cooler duties.
On a similar note, the fluidisation of the catalyst can
play a role in the conversion. Good distribution of catalyst in the riser cross-section is important to ensure fast
and sufficient feed vaporisation and oil to catalyst contact. This minimises the thermal cracking which forms
mainly coke and dry gas. Additionally, if the catalyst
fluidisation in the standpipes is insufficient, the catalyst flow can become uneven and sluggish. The main
cause of uneven catalyst circulation, assuming the ABD
and Fprop of the E-cat are typical, is insufficient standpipe aeration. If the ABD and/or FProp have shifted
and the unit is experiencing standpipe fluidisation
problems, then it is also likely that there will be cyclone
dipleg bridging issues, which will increase the catalyst
losses. Standpipe fluidisation issues can be identified
by low slide valve differential pressures, or excessive
slide valve percent openings, and can be optimised by
adjustment of the standpipe aeration.
Of the above described causes of conversion loss, a
couple of the examples could be attributed to mechanical damage. For example, damage to the spent catalyst distributor can cause maldistribution and
afterburn in the regenerator. Damage to the stripper
internals can cause reduced stripping efficiency and
high hydrogen in coke content. Poor atomisation of
the feed and its dispersion in the riser cross-section
can be caused by erosion, coking or catastrophic failure of the feed distributors. Additionally, damage to
the lift steam distributor can impact the catalyst distribution and density at the feed injection zone. It is possible for the reactor bypass valve to pass, or the feed/
slurry exchanger to be leaking, both of which will
result in FCC feed being routed to the main column
bottoms. And a final example is coking in the riser or
excessive lift steam rates which will cause reduced
and potentially insufficient riser residence times. Lift
and feed dispersion steam should be periodically optimised by the unit engineer by carrying out step testing to identify the optimum rates for the current feed
quality and unit operation.
A few of the main causes of conversion loss have been
summarised here, some of which may be more easily
identified by the use of one of Tracercos diagnostics
studies, which can identify things like maldistribution in the riser, regenerator or stripper. Tracerco additionally has the ability to check for exchanger leaks as
well as identifying mechanical damage in the regenerator. Ultimately, however, FCC troubleshooting is

20 PTQ Q1 2017

Q&A copy 24 (original).indd 6

made a lot easier by having a clear understanding of

the normal operation, and this is facilitated by carrying out regular check runs and recording the unit operating parameters, the feed quality, the catalyst balance
and E-cat quality, and the full product yields and qualities. Familiarity is key when trying to identify change
on any refinery operation, however the complexity of
the FCC makes it even more vital to have a clear understanding of the normal operating parameters.

Alexis Shackleford, Technical Marketing Specialist, BASF

Refining Catalysts, alexis.shackleford@basf.com

Poor conversion can be the result of a number of things,

but chief causes include poor feed quality, catalyst activity and hardware malfunction/failures. The number one
most important factor affecting conversion is feed quality. A rule of thumb is a change in feed density of +0.01
S.G. (-1.5 API) causes a 2 wt% loss in conversion. Other
feed quality parameters to look at include Concarbon,
contaminants (nickel, vanadium, iron, sodium, calcium,
nitrogen), hydrogen content (UOP K factor), PIONA
analysis (aromatics are harder to crack), and distillation
curve (especially a high boiling point tail). Next cause of
poor conversion would be low catalyst activity. A rule
of thumb is one number decrease in FACT (fluidised
activity measured on the E-cat) results in 0.5-1 wt% loss
in conversion. There are a number of reasons catalyst
activity could be low from loader problems (not adding
enough catalysts), high regenerator temperature, feed
contaminants, poor stripping, and so on. Another cause
of poor conversion is a hardware malfunction or failure
(though uncommon). A malfunction could include poor
feed atomisation/vaporisation caused by heavier feed,
incorrect feed/steam ratio, too low feed temperature
causing high viscosity, uneven feed distribution through
nozzles, and so on. Other malfunctions/failures to look
for would be a refractory failure, steam ring failure, or
air distributor failure for example. If you suspect a hardware problem, after standard troubleshooting including
a pressure survey, consider more advance techniques
including a gamma scan.

Jeff Knight, Senior Chemical Engineering Manager,

Honeywell UOP, Jeff.Knight@Honeywell.com

Commercial FCC units operate over a broad range of

conversion values. At a high level, the conversion is a
function of the intended operating mode, the feed type
and quality, and the technology that is employed. Poor
conversion in the FCC unit can be viewed as a relative
term, as a comparison to its base or expected operation. Opportunities for conversion improvement and
optimisation occur when the FCC unit departs from its
intended operating objective or design envelope.
The chief causes of this poor conversion can be generally attributed to four primary categories: catalyst, hardware, feed quality, and operating conditions.
Because of the integrated character of the FCC process,
there can be overlap in the categorical causes. The FCC
operates at its best expected performance when it is
harmoniously optimised and this requires attention to
all four categories.

www.eptq.com

14/12/2016 12:07

Keep calm and optimize.

Get the most from your refining operation.

Linde delivers cost effective solutions to improve
efficiencies and recoveries. We supply and service
a full range of technologies, including PSA, steam
methane reforming, amine treatment, furnaces/
heaters, sulfur recovery, off-gas hydrogen
and liquids recovery, hydrotreating, emissions
abatement and more.

Some of the services we provide:

Efficiency improvements
Feasibility, debottlenecking and process studies
Furnace and heater re-lifes
Fitness for service inspections
Revamps, retrofits, upgrades and expansions
Engineering, procurement and fabrication
Overall project execution and commissioning
Spare parts
Training
Outage planning and optimization services

Linde Engineering North America Inc.

12140 Wickchester Lane, Suite 300, Houston, TX 77079 USA
Phone +1 281.717.9090 sales@leamericas.com www.leamericas.com

When it comes to choosing a compressor, a one-size-fits-all solution isnt always the best choice.
Thats why every Ariel compressor we produce is manufactured to order based on your needs
and requirements. ARIEL IS COMMITTED TO YOU.

Learn more about Ariel compressors

www.arielcorp.com/For-You/

ariel.indd 1

14/12/2016 11:07

Berthold Otzisk, Kurita Europe GmbH, berthold.otzisk@

kurita.eu

The observation of a poor conversion is usually the

result of feed quality changes, operating conditions,
mechanical influencing variables, catalyst properties
or combinations of them. Less paraffins and more aromatics or an increase of nitrogen, nickel, vanadium or
sodium may lead to a lower conversion. Metals in the
feed cause thermal deactivation in the regenerator,
resulting in a decrease in microactivity (MAT) and a
decrease in the surface area.
More carbon on regenerated catalyst (CRC) temporarily blocks some parts of the catalytic sites. E-cat and
fresh catalyst properties have a very large impact to
lower conversion. Changed operating conditions like
reduced reactor temperature or decrease in the addition rate of fresh catalyst significantly lower conversion.
Broken stripping steam distributors, plugged or damaged feed nozzles are known mechanical failures.
When you observe poor conversion you should check
if there are changes in feed properties, mechanical and
operating conditions, or catalyst properties. Review the
feed properties, check catalyst activity, fresh catalyst
availability and source and try to find out if catalyst to
oil ratio, dispersion steam rate or reactor temperature
conditions changed. Monitoring the feed nozzles pressure profile can help to observe mechanical failures.
Hydrogen in coke is also an indicator for damages.

We have found heavy deposits of catalyst fines in our FCC

units CO boiler stack. What could be causing this and what
solutions are available?

Jeff Knight, Senior Chemical Engineering Manager,

Honeywell UOP, Jeff.Knight@Honeywell.com

Catalyst fines in the boiler stack can result from a loss

in performance in either (or both) the regenerator secondary cyclones and tertiary collection equipment,
with the result a loss of fine particle size material to the
stack. This fine material will be the result of the natural attrition that takes place due to the circulation/deactivation/thermal cycling of the fresh catalyst. Atypical
attrition can result and will produce additional fine or
microfine material and could be a more likely cause
for this sort of stack deposition. A good loss troubleshooting protocol should be executed and this should
include a particle size count of the catalyst fines leaving
the regenerator, the distribution should be centred on a
mean and median size. A bimodal distribution almost
certainly indicates an atypical attrition process.

Alexis Shackleford, Technical Marketing Specialist, BASF

Refining Catalysts, alexis.shackleford@basf.com

Catalyst deposits form most commonly in expanders,

cyclones, and third stage separators which is where
the majority of the research has been. Over the course of
an FCC run length, there is always enough catalyst on
the regenerator side to form deposits. But what makes it
form or stick is the important part. So what causes catalyst fines to stick together? Often the glues are alkali

www.eptq.com

Q&A copy 24 (original).indd 7

Analysis of catalyst deposits

Sample
SiO2
Al2O3
Fe2O3
P2O5
TiO2
La2O3
Na2O
CaO
SO3
BaO
NiO
MgO
V2O5
K2O
PbO
ZnO
SrO
Cr2O3
ZrO2
MnO
CuO

Expander 3ry cyclone Refractory E-cat E-cat fines

52.8
49.9
25.1
51.6
53.7
36.2
36.9
30.9
41.8
37.5
2.45
5.00
6.18
0.87
0.96
2.92
2.10
0.14
2.53
3.41
1.34
1.32
1.08
1.46
1.80
0.98
1.14
0.02
0.94
1.35
0.93
0.92
0.25
0.27
0.46
0.75
0.54
14.1
0.08
0.11
0.46
0.76
20.4
0.00
0.28
0.19
0.14
0.00
0.00
0.00
0.22
0.21
0.02
0.07
0.05
0.22
0.00
0.68
0.00
0.00
0.15
0.05
0.04
0.09
0.06
0.10
0.09
0.76
0.04
0.08
0.08
0.03
0.00
0.04
0.05
0.04
0.02
0.00
0.02
0.02
0.03
0.01
0.02
0.00
0.00
0.04
0.02
0.02
0.00

0.52
0.02
0.17
0.01

0.07
0.06
0.11
0.01

0.01
0.02
0.00
0.01

0.02
0.04
0.00
0.01

Table 1

metals such as sodium, calcium, and magnesium, often

associated with high sulphur. Iron is another element
seen enriched in these samples. For stack deposits, low
temperature is a potential cause since the sulphuric acid
present can form sulphates deposits that are very sticky.
To help understand what is causing these deposits, you
should have the material analysed by your catalyst supplier including elemental analysis by XRF. Compare the
elemental analysis to samples of your E-cat, fines, and
perhaps even refractory to understand what elements
are enriched in the deposit and potential sources. In the
example here, this expander deposit is enriched with
iron, calcium, sulphur, barium and magnesium (a significant amount coming from the refractory). See Table 1.

We are feeding coker naphtha to our hydrotreater and

encountering heavy silicon build-up on the catalyst. Should
we be looking for a guard catalyst or is some other solution
available?

Michael Tinning Schmidt, Product Manager, Refinery

Business Unit, Haldor Topsoe, mtp@topsoe.dk

Most coker naphtha hydrotreaters are limited in cycle

length due to accumulating silicon deposits on the catalyst. Higher percentage of coker stock gives more Si
present in the feed blend and makes the problem worse.
It has been known for a long time that the silicone
oils added as de-foaming agents upstream the delayed
coker will thermally decompose and eventually end up
in the coker naphtha fraction.
These broken-down silicon species have a great affinity for reacting with the catalysts alumina surface, and
this mechanism is what causes severe deactivation of
the hydrotreating catalyst. To ease the problem, it is
absolutely essential that coker naphtha hydrotreaters

PTQ Q1 2017 23

14/12/2016 12:07

have catalyst systems installed with the highest possible surface area per reactor volume. This technical
approach will boost the silicon trapping capacity. Such
catalyst systems will include both low and high activity catalysts. Furthermore, they need to be tailor-made
to balance the required amount of silicon capacity and
deliver the needed activity for removing sulphur and
nitrogen from the coker naphtha stream.
Haldor Topsoe has conducted research in these types
of catalysts for more than 25 years and has recently
been able to bring a new type of catalyst to the market
with a significantly higher silicon capacity than seen up
to now. The new product line brand is SiliconTrap; it
will help the coker naphtha hydrotreater to reach longer and less troubled cycles than with conventional
catalysts.

Vivek Srinivasan, Assistant Manager, Technical services,

Dorf Ketal Chemicals, viveks@dorfketal.com

Silicon is known to be the most notorious in terms of

catalyst poisoning, as it restricts the catalyst pores,
eventually blocking the active sides permanently. A
subsequent consequence is reduced unit run length
ranging from three to six months.
Based on research papers and field experience, it is
seen that silicon can find its source from crude due to

Though various mechanical

solutions are in place to control
the poison, it is better to reduce
the contamination at source
use of silicone based antifoams in oil production wells,
which can land in the distillate/residual hydrotreater.
However, this is restricted to few oil fields depending
on the geographical location. The main source for silicon contamination in naphtha hydrotreaters is antifoams used in the coker unit to suppress coke drum
foaming. Based on the quantity of antifoam added,
one can expect degradation of silicone into low molecular weight silica fragments in coker gasoil and coker
naphtha.
Though various mechanical solutions are in place to
control the poison, it is better to reduce the contamination at source. Careful selection of silicone chemistry,
antifoam injection rate, injection philosophy and monitoring plays a major role. One can adopt the following measures to put a check on the silica content to less
than 2 ppm in coker naphtha as a global benchmark.
Silicone based antifoams are available in different
viscosities. Based on lab scale testing and field experience, it is seen that high molecular weight (high viscosity) products have a benefit of lower dosages and high
thermal stability to reduce the potential degradation in
coker naphtha.
Silicone antifoams are always recommended to be
dosed in the coke drum overhead with carrier media
like LCGO for better mixing and performance. Such

24 PTQ Q1 2017

Q&A copy 24 (original).indd 8

high temperature antifoams are not advised to be

added on continuous basis and one may have to review
the injection standard operating procedure based on
coke drum cycle and ullage.
The antifoam should be typically added with a proper
distribution system having appropriate quills installed
in the top of the drum and one should ensure that there
is minimum distance from the overhead vapour line to
avoid as such residual antifoam carryover to the main
fractionator.
Better selection of level indicators can minimise/optimise the addition of antifoams. Neutron backscatter
types have a cutting edge over gamma transmission/
nucleonic types in detecting true foam.

Gary Peterson, Senior Business Leader, Refining, Honeywell

UOP, Gary.Peterson@Honeywell.com

As the availability of heavier and more sour crudes

increases, many refiners find themselves feeding
coker naphtha containing high levels of silicon to
their reforming and isomerisation units. The traditional approach to removing silicon from these feeds
is to stack load silicon trap catalyst having a nominal
amount of desulphurisation/denitrification activity in
front of the main hydrotreating catalyst. A better option
is to use a dual purpose catalyst that exhibits both high
silicon capacity and high activity.

Bruce Wright, Senior Technical Engineer, Baker Hughes,

Bruce.Wright@bakerhughes.com

There are several changes that can be implemented

to reduce the amount of silicone that ends up in the
cracked products from a delayed coker, including
mechanical improvements, operational adjustments and
implementation of state of the art foam control products:
From a mechanical perspective, the defoamer injection system needs to be 180 degrees away from the overhead vapour lines the opposite side of the top of the
drum. This orientation ensures that entrainment of antifoam injected into the drum is not immediately carried
out with the high volume overhead gases. In addition,
state of the art foam measuring systems allow for critical assessment of the foam height in the drum. These systems allow operators to see how well and how fast foam
fronts are knocked down upon injection of antifoam.
Operationally, there are several items to consider.
Increasing the drum temperature reduces viscosity
slightly and foaming tendency decreases; however, the
benefits of this temperature increase can be offset by
increased furnace fouling. A carrier stream for the antifoam is a critical aspect to ensure foaming is controlled
with the minimum amount of product. The carrier
stream helps to carry the antifoam agent to the foam
front and should be injected at a rate of 100 parts carrier
to one part antifoam.
All antifoam injected into the drum will eventually
crack to fragments, distill in the coker main fractionator
and add silicon to the coker products. Therefore, minimisation of the amount of antifoam agent being injected
to the drum is the key to reducing silicon carryover into
the products. Controlled usage of antifoam agent is a

www.eptq.com

14/12/2016 12:07

Silicon in cracked products Plant 1

Naphtha
Si (ppm) using 60 000 cSt defoamer
34.0
Si (ppm) using 600 000 cSt defoamer 12.3
59%
Si reduction

LCGO
7.9
3.2
59%

HCGO
7.3
2.7
63%

Table 1
Silicon in cracked products Plant 2

Naphtha
Si (ppm) using 600 000 cSt defoamer 33.8
Si (ppm) using Foamstop LCI defoamer 8.2
Si reduction
75%

LCGO
28.7
3.2
88%

HCGO
1.8
0.9
50%

Table 2

critical aspect for minimising the amount that goes to

the cracked products; therefore, operator confidence in
the ability of the antifoam agent to rapidly knock down
foam and control further foaming can lead to minimising overall usage rates.

There are several antifoam agent options. Polydimethyl siloxanes (PDMS) have been used for many
years due to their ability to quickly control foaming.
These PDMS additives are available in several molecular weight ranges, classified in terms of the polymer
viscosity. PDMS viscosities being used in delayed coker
operations range from 60 000 to 600 000 centistokes.
If the refinery objective is rapid, consistent foam control, with minimal concern over cracked product contamination, then the lower viscosity PDMS additives
are the best choice. The industry has been switching to
higher viscosity PDMS for several years because these
larger molecules stay in the drum and control foam
longer before they are completely cracked into volatile
fragments, therefore less siloxane is required overall.
As a result, silicon poisoning of downstream HDS catalysts is reduced. Table 1 presents Si loading at a refinery where the defoamer was changed to the higher
viscosity program. Baker Hughes has developed an
antifoam additive containing a modified siloxane molecule, the Foamstop Low Catalyst Impact series of products. These products control foaming with less overall
silicone, therefore contamination of coker cracked products is further reduced as shown in Table 2.
FOAMSTOP is a mark of Baker Hughes Incorporated.

Is there a chemical treatment for heat exchanger fouling

that will avoid taking our units apart at every maintenance
break?

Tim Olsen, Refining Consultant, Emerson Automation

Solutions, Tim.Olsen@emerson.com

From Emersons automation perspective, having online

monitoring and analysis capabilities around the crude
unit preheat exchangers can identify when the rate of
fouling increases, thus inferring the crude oil blends are
incompatible. Percentages of crude blends matter; the
refiner can adjust the crude blend percentage to reduce
the rate of fouling once detected. The refiner can then
learn the maximum percentage of a crude oil that does
not result in accelerated fouling and share that information with production planners when purchasing quantities of crude oils.

Vivek Srinivasan, Assistant Manager, Technical Services,

Dorf Ketal Chemicals, viveks@dorfketal.com

Heat exchangers in any refinery unit serve to preheat

the feedstock or to cool down the product for effective
energy optimisation. Fouling in such crucial exchangers can lead to operational limitations and unexpected
downtime. Refiners across the world adopt different
cleaning frequencies and methodologies. Some have
exchanger bypass provision for online cleaning and
some clean the exchanger during annual turnaround or
opportunity shutdown. The question is, is it necessary
to clean the exchangers throughout the train during
a maintenance break which can take significant man
hours and can the same be avoided using a chemical
treatment programme? The answer can be explained
depending on the type of fouling, feedstock processed,
configuration and monitoring philosophy.
Basically, fouling can be classified into organic and
inorganic. Various factors like temperature, velocity and
feedstock properties influence the organic or inorganic
fouling. Table 3 shows an overview of fouling in key
exchanger circuits across a refinery.
Depending on the above mentioned location and type
of fouling, one can choose the Dorf Ketal antifoulant
chemistry to get key benefits like enhanced run length,
reduced cleaning and hence downtime. Adopting good
monitoring practices to assess the health of exchanger
like fouling factor calculation, heat duty and normalised pressure drop can help predict the intensity

Fouling in key exchanger circuits

S.No Unit
Fouling prone location
1
Crude unit
Preheat train and atmospheric

bottom exchangers

2
Vacuum unit
Vacuum column bottom exchangers
3
Hydrotreater unit Preheat train

4
Visbreaker unit
Preheat train and residue exchangers
5
FCC unit
Main column bottom train

Type of fouling
Dorf Ketal solution
Predominantly asphaltene (organic)
Asphaltene dispersant and metal passivator
and in some cases influenced by
inorganic (metals)
Asphaltene (organic)
Asphaltene dispersant
Organic (polymer formation)
Range of chemistries including dispersant
and inorganic (corrosion products)
and antioxidant
Organic (asphaltene) + inorganic (metals) Asphaltene dispersant and metal passivator
Organic (poly nuclear aromatic) +
Coke retardant and metal passivator
inorganic (catalyst fines)

Table 3

26 PTQ Q1 2017

Q&A copy 24 (original).indd 9

www.eptq.com

15/12/2016 16:27

Bringing Unity To
Hydroprocessing

Honeywell UOPs vast hydroprocessing

catalysts lineup is called Unity because
we bring a unified approach, delivering
complete catalyst, equipment, licensing
and technical support solutions for
hydrotreating, pretreat and hydrocracking.

2016 Honeywell International. All rights reserved.

honeywell.indd 1
SPM-UOP-138 UnityAd_A4.indd 1

Learn more at www.uop.com.

14/12/2016 11:11
12/7/16 4:19 PM

of fouling in a particular exchanger. Dorf Ketals heat

exchanger assessment tool not only determines the
exchanger performance but is also beneficial in strategising cleaning sequence. This way one can avoid the
humongous cleaning activity during a maintenance
break if online bypass provision is not available.

Andrea Fina, Technical Coordinator Development &

Marketing Division, Chimec, process@chimec.it

The loss of heat exchange efficiency is currently one of

the main issues for refineries, as it often happens long
before the date of the scheduled turnaround. After one
or two years, in some cases even less, the heat exchange
loss can be so important as to influence the performance
and the economics of the plant, forcing management
to plan dedicated stops necessary to restore the heat
exchange efficiency.
In recent years, there have been several attempts to
develop a method to clean the heat exchangers without
the extraction of the tube bundles. Many involved the
use of chemicals at a high dosage or a chemical cleaner,
to be applied through external circuits or using the systems pumps.
Anyhow, due to the poor results, the widespread
view is that extraction and hydrodynamic cleaning is
still necessary to recover proper efficiency in the time
frame that separates one turnaround from another.

The loss of heat exchange

efficiency is currently one of
the main issues for refineries
The refinery should act in advance to avoid the need
for this kind of cleaning, preventing fouling formation
through a continuous control of all the elements that
may affect the efficiency of the heat exchange train.
A preventive approach is the only technical solution
that can avoid or move over time the need for a treatment solution such as extraction and hydrodynamic
cleaning.
Chimecs approach is based on an advanced antifouling programme tailor-made for the unit where fouling
is observed. Chimec chemical and service programme
is based on continuous injection of a customised antifoulant and includes advanced tools for the prediction
and monitoring of the fouling trends (fouling factors, normalised duties and normalised furnace inlet
temperature).
Furthermore, the company has developed Cosmo, a
technology that enables prediction of the asphaltenes
stability of a planned crude slate and the related fouling potential. Based on the stability index calculated
by Cosmo, it is possible to optimise the refinery crude
slate and/or the antifoulant dosage for preventing and
reducing fouling issues in the unit.
If properly employed and implemented, such a
preventive approach allows one to minimise the loss
of heat exchange and reduce both the extra energy
costs and those related to the eventual extra plant

28 PTQ Q1 2017

Q&A copy 24 (original).indd 10

stop which is necessary to apply the cleaning.

In the event that the preventive actions indicated
have not been adopted and the fouling phenomena
are unusual Chimec is able to propose innovative solutions of online (Gain Back Technology) or offline (Deko
Efficiency Recovery) fouling removal.
Gain Back Technology has been developed for the
online removal of the fresh fouling deposit commonly
observed in a crude distillation unit. Its strong dispersion feature can efficiently recover the thermo-hydraulic
capacity of critical exchangers, maximising the throughput of crude oil and avoiding by-bass procedures.
Gain Back Technology is the solution when worsening of fouling trends is detected in real time, hence
when the fouling is still soft. The technology is applied
by shock on top of the usual antifouling and allows one
to fully recover the capacity of the unit in some hours of
treatment.
Deko Efficiency Recovery may be applied when the
refinery has already experienced the consequences
of a fouling deposition. The target of Deko Efficiency
Recovery is to inject innovative chemicals on the bypassable exchangers, removing the fouling up to a sustainable level. If the exchangers cannot be excluded, a
momentary stop of the plant must be taken into account.
Our solution allows one to eliminate in the shortest
possible time a significant amount of fouling, with a
rapid resumption of operation of the involved exchangers and the immediate recover of the heat exchange and
flow rate of the process. The efficiency of this solution
is the result of the innovative research made to develop
proper actives.
The characteristics of fouling are not equal in each
exchanger. The fouling of the heat exchangers before
the desalter is different from those after the desalter.
Several chemicals must be selected on the basis of
available information considering whether one or more
steps are needed. Chimec can apply products to solubilise organic deposits with a high level of unsaturation as well as dispersant actives on both organic and
inorganic deposits, using oil based carriers or aqueous
solutions.
The comparison between applying Deko Efficiency
Recovery and a shutdown of the plant to extract and
wash hydrodynamically the exchangers is very easy:
the washing procedure can take 7-10 days, depending on the procedures and the means put in place. It
involves removing of the content, blinding, extraction,
cleaning, reintegration and restart. All these steps force
the system to a stress and risks in each step.
Deko Efficiency Recovery only requires exclusion and
displacement of the fluids from the heat exchanger in
order to allow the start of operations. A period of 16-32
hours, depending on the type of fouling and of applied
phases, can enable recovery of heat exchange efficiency
between 50% and 70%.

Carl Weaver, Senior Scientist, Baker Hughes, CarlWeaver@

bakerhughes.com

A method of exchanger cleaning often referred to as

online cleaning is one means of removing unwanted

www.eptq.com

15/12/2016 13:23

YOUR BENEFIT:
LOWEST LIFE CYCLE COSTS

API 618
HIGHEST AVAILABILITY
COMBINED WITH BEST
MAINTAINABILITY; FAST
ACCESS TO ALL WEAR PARTS
Full range:
Rod load up to 1500 kN / 335000 lbs
Power up to 31000 kW / 42100 hp
Lubricated up to 1000 bara,
non-lubricated up to 300 bara
Advanced solutions for demanding
and sour gas applications

burckhardt.indd
1
02_API618_Ext_Range_210x297RA_RZ.indd
1

Over 120 years of experience in valve

design and manufacturing
Outstanding durability of Redura
sealing elements for longest mean
time between overhaul (MTBO)

Full range of services and top

performing components through
global organization and local service
centers
www.recip.com/api618

Worldwide Burckhardt Compression

quality, engineered in Switzerland
Your solution partner From bare
shaft compressors to turnkey
solutions

14/12/2016
09:57
13.06.16 14:21

fouling debris and restoring heat transfer without pulling bundles for mechanical removal and hydro-blasting
to remove unwanted fouling debris. Given that there
is redundant (two or more banks of heat exchangers)
equipment, the term online refers to keeping the unit
in operation at reduced throughput, leaving a bank of
exchangers in service while another bank of exchangers
is chemically cleaned.
The most common method of chemical cleaning the
out-of-service exchangers is typically a two-step process. The initial step, using an elevated temperature
aqueous circulation of a detergent or micro-emulsion
cleaner, first de-oils and removes organic binders liberating some debris. The next step, using a low pH aqueous
solution of an inhibited acid or organic chelant solution,
solubilises inorganic scales (for example, iron oxide,
iron sulphide and silicates) for removal of unwanted
inorganic debris while protecting the base metal of the
process equipment. In completion of this step, a passivation rinse flushes the process equipment, leaving internal surfaces in a passive state and ready to be put back
in service. The effectiveness of this approach is highly
dependent on both the composition of the unwanted
fouling debris (% organic material and how much of
that might be coke and % inorganic material) and being
able to circulate the appropriate chemical cleaning solution. If exchangers have fouled to the extent that there
are plugged tubes, then the chemical cleaning solution
will not have access to the area to be cleaned.
Continuous antifoulant treatment appropriate to process design and the hydrocarbon feed being processed
has shown significant benefit in reduced energy costs,
extended run length as well as a means of inhibiting
complete blockage of exchanger tube throughput which
would impede online cleaning efforts.

What is predictive data analytics? What benefits can

it bring to our unit control operations? And how do we
implement it?

Tim Olsen, Refining Consultant, Emerson Automation

Solutions, Tim.Olsen@emerson.com, and Jake Davies, Global
Marketing Director, Permasense, jake.davies@permasense.com

The traditional approach has been to collect and historicise process data, and then only use the data to look back
and evaluate after an incident. The new approach is to
utilise pervasive sensing and predictive data analytics
software to automatically analyse data and turn it into
information; this new approach looks forward and alerts
before abnormal operation or imminent failure, thus the
ability to take appropriate timely action to avoid asset
failure. It is important that a refiner should not forget the
staff training and behaviour changes that are required to
act properly on this new analysed information.
Predictive analytics utilises both process and asset
health measurements that involves extrapolation of historical trends (from previously acquired measurements)
into the future. In asset integrity terms, this can be
achieved with online wall thickness monitoring. A line
of best fit through the historical online wall thickness

30 PTQ Q1 2017

Q&A copy 24 (original).indd 11

measurements can be extrapolated to estimate when,

for example, the equipment will need to be replaced.
This information helps to better plan timing and scope
of turnarounds. See page 61.

Christophe Romatier, Senior Manager, Strategy &

Business Development, Honeywell UOP, Christophe.
Romatier@Honeywell.com

Data analytics represents a family of techniques that

can be applied to operating data to derive patterns.
These patterns can in turn be used to infer or predict
certain types of process unit behaviour. Honeywell
UOP recently launched a new business, Connected
Performance Services, which makes use of such techniques and is paired with our longstanding experience
designing these units, their catalysts, and well as assisting our customers in troubleshooting their operations.
With Connected Performance Services, customers have

Data analytics represents a

family of techniques that can
be applied to operating data to
derive patterns
access to a next generation process monitoring environment that can improve their unit reliability by detecting
issues before they become costly. This is an example of
data analytics in action and how it can lead to tangible
operational and financial benefits.

The feed flow rate to our sour water stripper is well below
design level when flooding starts to occur. Is the problem likely
to be foaming or tray damage and how do we investigate this
easily?

Nghia Pham Phu, Principal Engineer, Refinery Process

Department, Haldor Topsoe, npp@topsoe.dk

The flooding could be caused by different reasons, for

example:
Tray damage
Fouling
Trapping of hydrocarbons inside the column.
Tray damage or fouling could prohibit the liquid flow
downward and start flooding. The flooding will progress upward in the column.
The sour water feed could contain a small amount
of hydrocarbons due to entrainment. Hydrocarbons
in a certain boiling range could be trapped inside the
column and accumulate to a high level. These hydrocarbons will vaporise at high temperature close to the
bottom and condense at low temperature close to the
top, thus circulating inside the column. They accumulate gradually to a high level until foaming and flooding occurs.
Flooding, foaming, and tray damage can be detected
by gamma scan. A number of companies provide this
service.

www.eptq.com

15/12/2016 17:26

Meet our family...

Weve learned a lot in 70 years
And we know precious metals recovery and refining better than anyone!
Sabin Metal
(Canada), Ltd.

Sabin
International
Logistics
Corporation

Sabin Metal
Europe B.V.

Sabin Metal
Corporation
DMCC BRANCH

L CORPO
ETA
R
M

5-

94

SABI
N

TI N
PLA

RR-SabinCorpFamily.indd 1

ION
AT

ach of these organizations is

part of the global Sabin Metal
Group of Companies. Our combined
M
resources enable us to return highest
1
AN
NIV E R S A RY
possible value of recovered and refined
precious metals quickly, safely, and with assured peace-of-mind. We recover precious
metals from virtually all industries that use them in their processes or products. Were
proud to begin our eighth decade of leadership, and wed like to share our success
with you. Tell us what we can do for you today.

Sabin Metal West

Corporation

201

Sabin Metal
Corporation

2/10/15 2:01 PM

Variability
is predictable.
What if you could manage variability across
the refinery and safely add profit to your
bottom line? You can.
Our industry-leading integrated solutions
leverage decades of expertise, unrivaled
customer dedication, and innovative predictive
technologies to transform variability into a
competitive advantage.
Visit www.gewater.com/calculator to discover
your increased revenue potential today.

GE Water & Process Technologies

Predicting your success.

ge.indd 1

14/12/2016 10:35

Crude oil sourcing: price and opportunity

Sourcing crudes for the best refining margin needs to be supported by a detailed
procurement strategy
MISHA GANGADHARAN, S D POHANEKAR and M D PAWDE
Hindustan Petroleum Corporation Limited

he worlds crude oil market

includes not only spot markets featuring physical transactions but also highly developed
paper markets, notably futures
and forward delivery, thus forming a very composite market structure in which all these transactions
are interrelated and reciprocally
inuential. The variation in prices
between two grades can inuence
renery margins signicantly if
corrective action is not taken at the
appropriate time. Therefore identifying and procuring crude oil
that results in maximum margins
are the industrys top priorities. In
this article, various aspects relating
to crude oil procurement for optimisation of renery margins are
outlined.

How the oil market works

The price of oil is set in the global

marketplace. Oil is traded widely
all around the world and can
move from one market to another
easily by ship, pipeline or barge.
Therefore the market is worldwide
and the supply/demand balance
determines the price for crude oil
all around the world. If there is a
shortage of oil in one part of the
world, prices will rise in that market to attract supplies from other
markets until supply and demand
are in balance. If there is a surplus
in a region and the price drops,
buyers will soon be drawn to that
market. This explains why crude
oil prices are similar all around the
world. Prices vary only to reect
the cost of transporting crude oil
to that market and the quality differences between the various types
of oil. The global nature of the

www.eptq.com

q1 hpcl.indd 1

market also explains why events

anywhere in the world affect oil
prices in every market. Several
key factors inuence the oil price,
however the four major factors that
help in determining the price of oil
are supply, consumption, nancial
markets and government policies.
As per the basic principles of

The way oil is traded

on the financial
markets has a
massive influence
on its price
economics, prices will be low if
supply exceeds demand, and the
reverse applies: an increase in consumption over supply will lead
to an increase in prices. However,
crude oil pricing goes far beyond
just supply and demand. The way
oil is traded on the nancial markets has a massive inuence on its
price. Similar to the stock market,

people also trade in crude oil as a

commodity in nancial markets.
They purchase futures essentially
bets on how much oil will cost at a
later date and this in turn affects
how other people think oil should
be priced. It also affects how much
oil the petroleum companies will
release to the market. Oil trading in
nancial markets has been growing
bigger than ever in recent years. As
a result, speculation has come to
shape the price of crude oil more
than ever before. Crude oil trading
in nancial markets has a surprising
effect on crude oil prices speculators who buy large amounts of
futures can swing the price one way
or another. Here is an example: a
speculator who buys oil futures at
a price higher than the current market price can cause oil producers
to hoard their oil supply so they
can sell it later at the new, higher
future price. This cuts the current
supply of oil in the market and
drives up both present and future
prices.
Government regulation also has a
big impact on oil prices. Rules and

Dubai
Dubai & Brent
Brent
WTI
WTI & Brent

Figure 1 Marker crudes by area

PTQ Q1 2017 33

14/12/2016 12:18

Price differential, $/bbl

14

margin point of view can be less

attractive after a few months.

Brent / WTI differential

Brent / Dubai differential

12

Pricing and marker crude oil grades

10
8
6
4
2
0

Ja
n
2
M 014
ar
2
M 01
ay 4
20
Ju 14
l2
Se 01
p 4
2
N 01
ov 4
2
Ja 014
n
2
M 01
ar 5
2
M 01
ay 5
20
Ju 15
l2
Se 01
p 5
2
N 01
ov 5
20
15

Figure 2 Price differential movements between marker crudes

Characteristics of marker crude types
Key property
API
Sulphur, wt%
Pour, F
Acid number, mg KOH/gm

Brent
37.5
0.45
30
0.01

WTI
39
0.3
-0.4
0.09

Dubai
31.4
1.54
-11
0.05

Oman
30.4
1.45
-33
0.61

Table 1
Price determination methodologies for marker crude oil
Supplier
Marker
Marker price
Month of

crude
determination
announcing

OSP
Saudi Aramco
Oman Dubai avg.
B/L month average
M-1
Kuwait Petroleum Corporation (KPC) Oman Dubai avg. Nominated month average M-1
National Iranian Oil Company (NIOC) Oman Dubai avg. Nominated month average M-1
State Oil Marketing Organisation,
Iraq (SOMO)
Oman Dubai avg. Nominated month average M-1
Abu Dhabi National Oil Company
(ADNOC)
Outright Prices
Nominated month
M+1
Qatar General Petroleum Corporation
(QGPC)
Outright Price
B/L month
M+1
Yemen Oil (YOG)
Brent
5 days after B/L date
M-2
Egyptian General Petroleum Corporation
(EGPC)
Brent
B/L month
M
Brent
Five days avg. after B/L
Nigerian National Oil Company (NNPC)

Five days avg. before B/L M-1

Five days avg. after five days after B/L.
Sonangol, Angola
Brent
5 days around B/L
M-1
Sonatrach, Algeria
Brent
5 days after B/L date
M-1
Minas
B/L month avg.
M+1
BPMigas, Indonesia
Petronas, Malaysia
Brent
B/L month average
M
Brunei Shell Petroleum (BSP)
Outright price
Nominated month
M+1
Omans Oil Ministry (Oman)
DME Oman
B/L month average
M-2
National Oil Corporation, Libya
Brent
B/L month average
M-1

Table 2

regulation on the sulphur content

of fuel could raise demand for low
sulphur crude oil. The incentive
for fuel efficient cars, development
of more efficient alternate modes
of transport and so on will lead to

34 PTQ Q1 2017

q1 hpcl.indd 2

demand for oil prices to go down,

simply because the world will not
need it as much. Thus selection of
crude oil for processing at a refinery is always dynamic and crude
oil that was most desirable from a

The price payable for crude oil is

calculated based on the marker
crude oil price plus or minus a
price adjustment factor, which is set
by the seller, or is mutually agreed
between the buyer and the seller.
While the marker crude oil price
varies with events in international
markets, including speculations,
the adjustment factor is decided by
the seller, taking various elements
into account like quality difference
with the marker crude, transportation cost difference with alternate
grades, demand supply balance,
competition from other suppliers/
grades, and so on.
There are different crude oil
markers, each one representing
crude oil from a particular part of
the globe. The marker crude oil is
specific to the market and three
major markers used in pricing
of crude oil across the globe are
WTI (Western Texas Intermediate)
for the US market, Brent for the
European/West African Market
and Oman/Dubai for crude oil
grades in the Persian Gulf for the
Asian market. The applicability of
marker crudes across the globe is
shown in Figure 1. The details of
each of the markers are:
Dubai/Oman: this Middle Eastern
crude is a useful reference for oil of
a slightly lower grade than WTI or
Brent. It is heavier and has a higher
sulphur content, putting it in the
sour category. Dubai/Oman is the
main reference for Persian Gulf oil
delivered to the Asian market.
Brent Blend: about two-thirds of
all crude oil contracts around the
world are marked to Brent Blend,
making it the most widely used
marker. Brent actually refers to
oil from four different fields in the
North Sea: Brent, Forties, Oseberg
and Ekofisk. Crude from this
region is light and sweet, making it
ideal for the refining of diesel fuel,
gasoline and other high-demand
products. The supply is waterborne
and thus it is easy to transport to
distant locations.

www.eptq.com

14/12/2016 12:18

Table 2

www.eptq.com

q1 hpcl.indd 3

Price adjustment factor, $/bbl

2
0
2

Ja
n
2
M 014
ar
2
M 01
ay 4
20
Ju 14
l2
Se 01
p 4
2
N 01
ov 4
2
Ja 014
n
2
M 01
ar 5
2
M 01
ay 5
20
Ju 15
l2
Se 01
p 5
2
N 01
ov 5
20
15

Price differential factor, $/bbl

Arab Medium, API 30.6, Sulphur 2.6%

Kuwait, API 30.5, Sulphur 2.5%
Iran Heavy, API 30.2, Sulphur 1.8%

2
1
0

1
2
3
4

14
20
Ju 14
l2
Se 01
p 4
2
N 01
ov 4
2
Ja 014
n
2
M 01
ar 5
2
M 01
ay 5
20
Ju 15
l2
Se 01
p 5
2
N 01
ov 5
20
15

20

ay

ar
M

Ja

20

14

Figure 4 Price differential factor over Oman Dubai average

grades after the completion of the

month basis price is discovered
under spot market conditions.
Figures 3 and 4 show how the
price adjustment factors of grades
6
4

similar to each other but from different suppliers closely track each
other when the price adjustment
factor is declared before the beginning of the month. However, the

Arab Light, API 32.7, Sulphur 1.8%

Upper Zakum, API 33.0, Sulphur 1.8%

2
0
2
4

ar
2
M 01
ay 4
20
Ju 14
l2
Se 01
p 4
2
N 01
ov 4
2
Ja 014
n
2
M 01
ar 5
2
M 01
ay 5
20
Ju 15
l2
Se 01
p 5
2
N 01
ov 5
20
15

14

The price adjustment factor for

a crude oil grade is adjustment
over the marker crude oil price.
The adjustment factor is declared
before the beginning of the month
by many national oil companies
and is called the official selling
price (OSP), whereas if crude oil
is purchased on the spot market
the adjustment factor is as agreed
between buyer and seller.
The adjustment factor plays an
important role in crude selection
as the relative economics (grade
to be procured) changes based on
the adjustment factors declared by
the national oil company for their
various grades. To be competitive,
the national oil companies generally maintain the differentials of
their grades with other, similar
grades. Some national oil companies declare outright price for their

Price adjustment factor, official

selling price

Iran Light, API 33.1, Sulphur 1.5%

Arab Light, API 32.7, Sulphur 1.8%

Figure 3 Price adjustment factor over Oman Dubai average

Price adjustment factor, $/bbl

The determination of a marker

crude oil price varies between suppliers, and the various ways of
determining marker crude oil price
are: bill of lading (B/L) month
average, five days around B/L,
nominated month average, or as
agreed between buyer and seller.
Thus while finalising the sourcing
of marker crude oil, determination of its price must also be kept
in mind. Table 2 indicates marker
crude oil price determination methodologies followed by various
national oil companies for supply to Asia, as well as the time of
announcing the price adjustment
factor.

20

Determination of marker crude

oil price

Ja
n

West Texas Intermediate (WTI): WTI

refers to oil extracted from wells in
the US and transported via pipelines. The crude is very light and
very sweet, making it ideal for gasoline refining in particular. WTI continues to be the main benchmark for
oil consumed in the US. The characteristics of these marker crude oil
grades are shown in Table 1.
Figure 2 indicates how price differentials between marker crude oil
move for various reasons, swinging
sourcing decisions for a refinery.

Figure 5 Price adjustment factor over Oman Dubai average

PTQ Q1 2017 35

14/12/2016 12:19

Price adjustment factor, $/bbl

Murban, API 39.6, Sulphur 0.7%

Arab Extra Light, API 38.4, Sulphur 1.1%

4
2
0

Crude oil procurement strategy

The relative attractiveness of crude

oil for refinery margin maximisation takes into account the differentials in crude price, freight, and
product worth. If one scans the
data over a period of years, there is
no simple relationship on all three
parameters and thus decisions on
term contracting are required to
be taken basis various scenarios
on price variations and historical
performance. In addition, if product demand and specifications are
seasonal, the crude suitable for
one season may not be suitable for
another season. Gas oil rich crudes
will be most suitable in a high gas
oil demand/price season whereas
gasoline rich crudes will be suitable
when gasoline demand/prices are
high. Figure 7 shows monthly variation in gasoil and gasoline prices in
the Arabian Gulf market and indicates the suitability of gasoline rich
crudes since March 2015 over gas
oil rich crude oil grades.
The procurement strategy thus
requires detailed insight into market variations, supply security considerations and attractiveness of
grades traded on the spot market.
The most important decision is how
much to procure on term contracts,
which are generally for a period of
one year, to address concerns over
supply security if the grades contracted are not optimum for the
entire year.

Ja
n
2
M 014
ar
2
M 01
ay 4
20
Ju 14
l2
Se 01
p 4
2
N 01
ov 4
2
Ja 014
n
2
M 01
ar 5
2
M 01
ay 5
20
Ju 15
l2
Se 01
p 5
2
N 01
ov 5
20
15

Figure 6 Price adjustment factors over Oman Dubai average

20

Gasoline Dubai
Gasoil Dubai

Price differential, $/bbl

18
16
14
12
10
8
6

14
20
Ju 14
l2
Se 01
p 4
2
N 01
ov 4
2
Ja 014
n
2
M 01
ar 5
2
M 01
ay 5
20
Ju 15
l2
Se 01
p 5
2
N 01
ov 5
20
15
ay

20

ar

20

14

Ja

Thus, the economics of a grade

whose price is declared after the
end of the month vis vis a similar, alternative grade whose price
is declared before the beginning
of the month keep changing from
month to month as the differential
price between two grades varies
significantly.

Figure 7 Crude and product price differentials

Crude blend properties

Blend 1
API
Sulphur
% wt in blend

Arab Extra Light Soroosh

Blend
A
B
A+B
39.1
18.7
32.8
1.1
3.6
1.7
71
29

Blend 2
API
Sulphur
% wt in blend

Arab Super Light

A
50.6
0.04
10

Khafji
Blend
B
A+B
28.5
30.5
2.8
2.5
90

Blend 3
API
Sulphur
% wt in blend

Arab Super Light

A
50.6
0.04
30

Maya
B
21.8
3.3
38

Arab Light
32.7
1.8

Kuwait
30.5
2.5

Widening of the crude oil basket

Arab Heavy Blend
C
A+B+C
27.8
30.65
2.7
2.27
32

Arab Medium
30.6
2.6

Table 3

grades whose prices are declared

after the end of the month (outright
price) do not see any relationship

36 PTQ Q1 2017

q1 hpcl.indd 4

with similar grades whose prices

are declared before the beginning
of the month (see Figures 5 and 6).

The quality of crude oil varies

widely in terms of its properties:
API, sulphur, nitrogen, metals
and so on. The crude and vacuum
distillation unit in a refinery is
designed to operate within the limits of the hydraulics of streams separated from crude oil. However,
secondary units may be limited
by the quality and quantity of

www.eptq.com

15/12/2016 13:24

www.eptq.com

q1 hpcl.indd 5

Cost reduction, $/bbl

Blend 3
Blend 2
Blend 1

Ja
n
2
M 014
ar
2
M 01
ay 4
20
Ju 14
l2
Se 01
p 4
2
N 01
ov 4
2
Ja 014
n
2
M 01
ar 5
2
M 01
ay 5
20
Ju 15
l2
Se 01
p 5
2
N 01
ov 5
20
15

Figure 8 Reduction in crude cost through blending

Net gain, $/bbl

Blend 3
Blend 2
Blend 1

14
20
Ju 14
l2
Se 01
p 4
2
N 01
ov 4
2
Ja 014
n
2
M 01
ar 5
2
M 01
ay 5
20
Ju 15
l2
Se 01
p 5
2
N 01
ov 5
20
15

20

ay

ar

20

14

Ja

streams generated from crude oil.

The crude unit can handle crude oil
with certain API variations. If the
crude is very light (high API), the
column overhead section as well
as naphtha stabilisers will be limiting. Alternatively, if the crude is
heavy (low API), the bottom circuit
and vacuum unit may become limiting. Certain crude oil properties
like viscosity and metals content
can limit the preheat train/desalter
operation in a crude unit. Crude
oil with an acid number beyond an
acceptable limit cannot be accepted
in the absence of crude oil blending
facilities or a chemical programme
to mitigate corrosion.
The properties of product from
processed crude can limit the secondary processing units as well
as the nal product specication.
Sulphur recovery units may be limited for high sulphur crude, while
a hydrocracker may be limited due
to high nitrogen in vacuum gas oil,
and a diesel hydrotreater may be
limited for processing high diesel
yield crudes. Generally, linear programming (LP) models are used
for evaluation of crude oil grades.
However, there are limitations in
a LP model as processing in the
model is for a group of crudes
together and not for a single grade.
Thus widening of the crude oil
basket depends on the capability of
the renery to process crude oil on
a standalone basis in the absence
of crude oil blending facilities.
Alternatively, if blending facilities
are available, the widening of the
basket is not a challenge as difcult
to process crude oil grades can be
blended with other crudes so as to
have a composite crude oil suitable for renery conguration. The
challenge then shifts to scheduling and blending. As the economic
benet of widening of the crude
oil basket is expected to be high,
it is essential for every rener to
explore options that can give exibility to procure any crude oil
grade that will maximise margin.
Examples in Table 3 indicate how
grades that cannot be processed
neat, if blended and processed can
reduce the crude cost over procuring a single grade similar to the
blended crude oil grade.

Figure 9 Net gain through blended crude

Input

Crude oil grades

for evaluation
Product demand
Delivered-basis
crude oil prices
Product sales
realisation

Model representing
refinery units and
operation
LP output for various
pricing periods
LP output for
seasonal product
demand variations

Recommendations
of crude oil grades
under term and spot
contracts
Opportunity analysis
Most desirable grades
identifications
Availability checks
under term and spot
Quantity finalisation
under term and spot

Figure 10 Steps followed during crude oil evaluation

Thus widening of the crude oil

basket and blending is essential to
increase options and capture opportunities for maximisation of margins.
Figure 8 shows the price difference
between blended crude and neat
crude and indicates an opportunity
to reduce crude cost by widening the

crude oil basket with blending. The

net realisation of processing blended
crude over neat crude basis vecut (LPG, naphtha, kerosene, gasoil
and fuel oil) is shown in Figure 9.
The blended crude processing margins are directionally higher than
for neat crude of similar quality.

PTQ Q1 2017 37

14/12/2016 12:19

Formulation of crude oil

purchase plans

The salient features of procurement under term or spot

Procurement under term Procurement under spot
1 Available mainly from national oil companies or equity Available mainly from oil majors
producers. Some of the traders can also supply subject to
and traders.
them having a contract with oil major/equity producers.
2. The contract can be for a particular grade or for all
the grades available from the national oil company

The contract is for a specific grade of

crude oil.

3 The contract is finalised either through negotiation or

through participation in a tender enquiry by the national
oil company.

The contract is finalised either through

negotiation or through participation
by a supplier in a tender raised by
the buyer.

4 Period of contract is generally for 12 months starting

from April or from January.

The period of contract is specific to

the cargo procured.

5 Price as per official selling price (OSP) as declared by a

national oil company or at a differential to OSP as agreed
between buyer and seller. Grades that do not have an OSP
are priced as per the price declared by the seller according
to actual trading during the month.

Price at premium/discount to marker

crude or to OSP as agreed between
buyer and seller.

6 Terms and conditions as per general terms and

conditions of a national oil company or an oil major
holding equity. Few specific conditions of the contract like
payment terms and credit limit will be as agreed between
buyer and seller.

Terms and conditions as per general

terms and conditions of national oil
company or oil major holding the
equity. Few specific conditions of the
contract like payment terms and
credit limit will be as agreed between
buyer and seller.

7 The grade and quantity is finalised on a barrels per

day basis to be supplied during the year. Nominations
for loading month and cargo size are to be indicated
well in advance.

The grade, quantity and loading

month/date range for loading is
finalised at the time of contract.

8 The supply is mostly assured.

The most desired grade may not be

available if all the cargoes are tied up.

9 The grade offered by the supplier may not be suitable

for a particular pricing scenario and thus becomes
sub-optimum.

The contract is finalised only if

the grade appears suitable.

10 There are possibilities to surrender the cargo back to

the supplier with adequate notice, in the event of any
issues with refinery operations.

The cargo contracted is generally

required to be lifted unless agreed by
the seller. However, the option of
resale is available.

Table 4
Grades available only from national oil companies and only through term contracts
Country

Iran






Kuwait
Neutral Zone

Saudi Arabia

Crude oil grade

Forozan
Iranian Heavy
Iranian Light
Lavan
Nowruz
Sirri
Soroush
Kuwait
Hout
Khafji
Arab Extra Light
Arab Heavy
Arab Light
Arab Medium
Arab Super Light

Key properties
API
31.7
30.2
33.5
35.2
20.5
33.2
18.7
30.5
32.8
28.5
39.1
27.8
32.7
30.6
49.1

S%
2.1
1.8
1.5
1.7
3.5
2.0
3.6
2.5
2.0
2.8
1.1
2.7
1.8
2.6
0.1

Table 5

The crude oil grades considered for

blending need to be evaluated for
compatibility as well as evaluated

38 PTQ Q1 2017

q1 hpcl.indd 6

in detail by means of LP model runs

before a decision on procurement is
taken.

Sourcing of crude oil can be carried out under term or spot contracts with national oil companies,
oil majors and traders. Sourcing on
term conditions ensures security of
supply, whereas sourcing on the
spot market can capture opportunities for margin maximisation by
sourcing the most suitable grade
under the prevalent monthly market conditions. Most refiners use a
LP model (a true reflection of the
refinery configuration) for evaluation of crude oil. The product sales
plan, sales net realisation, refinery
processing units capacities/constraints, processing cost (fuel and
power consumption, chemical and
catalyst consumption), crude oil
availability, and crude oil landed
cost are input to the LP model for
evaluating the various options.
In developing a strategy for crude
oil procurement, the crude oil basket to be considered for evaluation
plays a very important role. Once
evaluation under different pricing scenarios has yielded appropriate crude oil grades, a strategy
for sourcing under term and spot
conditions needs to be developed.
Figure 10 shows the details of various steps followed during crude
oil evaluation. With various options
generated, a decision must be taken
on how much crude to term up and
how much to keep for procurement
under spot.
The sailent features of procurement under term or spot are shown
in Table 4.
If one reviews the features of
securing crude oil under term or
spot contract, the premium/discount applicable and date range for
loading a cargo secured under term
conditions is not known at the time
of confirming the contract, whereas
for crude oil procured under spot
conditions there is an agreement on
premium/discount to the marker
crude and also on a date range
for loading. Under term contract,
the supply of agreed quantity is
assured. As regards grade of crude
oil to be procured, there is no
issue if the contract is for a single
grade; however, if the contract is
for multiple grades the allotment of

www.eptq.com

15/12/2016 16:56

desired grades is subject to availability. The date of loading is subject

to acceptance by the seller.
Some national oil companies
market their crude oil grades only
through term contracts and also
with end user restriction. However,
the equity crude oil of some
national oil companies is available
under spot as well as under term
conditions without any end user
restriction. Table 5 details major
crude oil grades available only
through term contracts, whereas
Table 6 details availability of major
crude oil grades actively traded on
the spot market from countries of
interest to India.
As can be seen from Table 6,
there are many alternatives for
grades that are supplied only under
term contracts. If security of supply
is not a concern for a refiner, there
is very little reason why one should
seek procurement under term contract unless the grades available
under term conditions are robust
under all pricing scenarios.

Conclusion

The crude oil market is complex

and procurement of the right crude
can maximise refinery margins. The
procurement of a crude oil grade
under term conditions depends
on its attractiveness under various
pricing scenarios and any need
to address concerns over security
of supply. The opportunity exists
to source crude oil through spot
market and maximise margins, as
terming up of grades for a longer
period may not be attractive under
all pricing scenarios and may not
be required if there are no issues
on supply security. Widening of the
crude oil basket through state of the
art blending facilities can increase
the options for crude sourcing for a
refinery and will also contribute to
margin maximisation.

Misha Gangadharan is a Senior Engineer with

Hindustan Petroleum Corporation Limited India
(HPCL) with experience in crude oil evaluations,
production planning and optimisation. She
holds a BTech degree in chemical engineering
from Institute of Chemical Technology (ICT)
Mumbai, India.
Email: mishagangadharan@hpcl.in

www.eptq.com

q1 hpcl.indd 7

Grades actively traded on the spot market1

Country

Dubai
Iraq


Neutral Zone

Oman
Qatar


UAE



Yemen

Egypt


Nigeria
Angola
Algeria
Malaysia



Indonesia
Other WAF countries:
Cameroon, Chad, Congo,
Gabon, E. Guinea, Ghana,
Sudan
Mexico


Venezuela
Brazil
Colombia


Equador

Azerbaijan
Libya
Brunei

Russia



Kazakhstan

North Sea

Crude oil grade

Dubai
Basrah
Basrah Heavy
Kirkuk
Eocene
Ratawi
Oman
Al Shaheen
Qatar Land
Qatar Marine
Das Blend
Lower Zakum
Murban
Upper Zakum
Marib Light
Masila
Ras Gharib
Belyam
Suez Mix
Various Grades
Various Grades
Saharan Blend
Miri
Tapis
Labaun
Kikhe
Various grades
Various grades like
Kole, Doba, Djeno,
Nkossa, Rabi Lt, Ceiba,
Zafiro, Jubilee, Dar, Nile Blend
Isthmus
Maya
Olmeca
Various grades
Various grades
Cano Limon
Castilla
Cusiana
Napo
Oriente
Azeri Light,
Various grades
SLEB
Champion
ESPO
Sokol
Urals,
Vityaz
CPC Blend
Kumkol
Various grades

Key properties
API
S%
31.4
1.5
28.8
3.1
23.7
4.1
34.2
2.2
18.3
4.6
24.2
4.1
30.4
1.4
30
2.4
40.7
1.2
32.7
1.8
38.3
1.1
39.9
1.0
39.6
0.8
33.0
1.8
45.1
0.1
34.6
0.5
20.9
3.8
23.6
2.7
30.8
1.5
29-47
0.05-0.3
22-39
0.1-0.8
45.7
0.1
29.7
0.1
42.1
0.04
30.2
0.1
36.7
0.07
20-48
0.02-0.2

21-40
33.2
21.8
38.8
11 -39
17-29
30.2
18.2
43.2
18.6
24.7
36
37-42
38.5
33.1
36
36.1
32.7
34.4
45.4
41.4
30-47

0.10.4
1.2
3.3
0.9
0.1-5.2
0.3-0.8
0.5
1.5
0.1
2.0
1.4
0.1
0.1-0.4
0.1
0.1
0.6
0.3
1.0
0.2
0.6
0.1
0.2-0.8

Note 1: There are many more grades over and above those mentioned above

Table 6
S D Pohanekar is a Senior Engineer with
Hindustan Petroleum Corporation Limited
(HPCL) India with experience in refinery
production planning and optimisation. She
holds a BTech degree in chemical engineering
from Laxminarayan Institute of Technology,
Nagpur, India.
Email: sdpohanekar@hpcl.in

M D Pawde is Head of Economic Planning

and Optimisation with Hindustan Petroleum
Corporation Limited (HPCL) India with
over 30 years of experience covering
refinery operations, crude oil evaluation
and procurement, production planning and
optimisation. He holds a BTech degree in
chemical engineering from Nagpur University,
India. Email: mdpawde@hpcl.in.

PTQ Q1 2017 39

15/12/2016 16:32

OPTIMAL FILTRATION FOR

MAXIMUM PERFORMANCE
Optimize Your
Filtration Performance
Reenery Fuel Gas
Catalyst Protection
Amine Gas Treating
Compressor Protection
Final Product Protection
Turbo Machinery Protection

In the Reenery and Petrochemical Industry, an ecient ltration/separation

system is critical to the reliability and performance of the physical assets
that process petroleum products. Optimizing the reliability of these
ltration systems is essential in maximizing your operational performance.
Jonell has worked diligently to advance the economics of Petroleum
Filtration, delivering a variety of ltration/separation solutions to t your
speciic needs. Our goal is to ensure that our clients beneet from ltration
solutions that are constantly evolving and improving. Jonell not only makes
lters; we are constantly striving to make our lters better incorporating
the latest in ltration and separation technology.

www.jonellinc.com

Cubas oil: due for development

Cubas ambitions for energy self-sufficiency require major investment in the
nations oil reserves and refining industry
AMAURY PREZ SNCHEZ
University of Camagey

ubas internal power demand

is expected to rise in the
coming years in response to
growth in private business, expansion of foreign business investment
in the country, and progress and
expansion in important power consuming industries such as biotechnology, tourism and construction.
Today, Cuba generates about 95%
of its internal electricity from hydrocarbons and their byproducts.
Unin Cuba Petrleo (CUPET)
is Cubas largest oil company. It is
owned and operated by the Cuban
government and is involved in the
extraction of petroleum deposits,
rening and distribution of petroleum products (see Figure 1).
For the period to 2030, CUPET has
conceived several ambitious plans
and projects in order to enhance
its protability and productivity,
as well as to develop its oil rening industry. Among these can be
mentioned:
Expand the rening capacity of
the oil renery located in Cienfuegos
province to 150 000 b/d
Modernise and update equipment and accessories operating in
the reneries, mostly combustion
systems, boilers, vessels, tanks, heat
exchangers, wastewater treatment
structures, control valves and automation systems
Produce dielectric oils for 33kV
electrical transformers at Sergio
Soto renery
Install sweetening units for Jet A-1
(Merox) fuel in ico Lpez renery
Introduce liqueed petroleum gas
(LPG) into the national power mix
Elevate the quality of national
fuels to international standards
Increase storage capacity nation-

www.eptq.com

q1 reduc.indd 1

F Facilities
F

G GM G
F

F
M
F

M Maritime terminals
R Refineries

WF
F
F
M

W Oil wells

M
W

G Gas plants

F M

F
W WW W
W W W WW
W
W WW W
W W
W

F
F

F
F

RF

F M
F

Figure 1 CUPET infrastructure and facilities

wide, both for crude oil and petroleum products

Reduce/optimise the costs of
logistics operations
Intensify the exploitation of oil
deposits located inland and offshore
Expand the production level
of existing oil deposits by means
of enhanced oil recovery (EOR)
technologies.

Private investment

To achieve sustainability in energy

in the near future, the Cuban government has incorporated new perspectives and opportunities into
its oil industry offered by recently
approved foreign investment policies, in order to increase efciency,
reduce costs and boost production
capacity. Accordingly, the main
objectives of joint ventures and
associations with foreign rms and
companies are threefold: searching for unconventional oil; offshore
exploration; and EOR. There are
other collateral activities requiring
direct capital investment such as
technical services for oil extraction

operations, the supply of modern,

oil related technologies, equipment
and resources, as well as provision
of nancial, engineering and management services.
The rst business contract related
to oil extraction operations between
the Cuban government and a foreign
rm was signed in 1990, in order to
exploit the oil wells located at Block
III south of Varadero beach. In the
last three years, around 42 shared
production agreements (SPA) have
been agreed between CUPET and
foreign rms, in accordance with
the rst Cuban Foreign Investment
Law, approved in 1995, and 2014
legislation whose terms ease business transactions and increase the
scope of the proposed investment
objectives.
In Cuba, the National Ofce of
Mineral Resources (NOMR) is the
government institution in charge of
certifying and approving potential
investment projects and investors
in the oil industry. Each business
contract signed is protected under
Cuban government decree and

PTQ Q1 2017 41

15/12/2016 14:05

Foreign investment opportunities in the Cuban oil industry

Option 1

Title: Oil exploration risk contracts and shared production

agreements in production blocks
located in Cuban shoal waters
(eight blocks).
Description: Define the potential
oil and gas reserves existing in
determined areas of Cuban shoal
waters. If positive results are
obtained, exploit the energetic
resources with profitability.
Location: The eight blocks available are located predominantly
at the north of Pinar del Ro,
Matanzas, Villa Clara and Sancti
Spiritus provinces, as well as in
the south of the following provinces: Pinar del Ro, Artemisa,
Mayabeque, Matanzas, Ciego de
vila, Camagey and Granma.
Market: Cuban internal market in
first term, surplus production for
exportation.
Results: For an exploration block
presenting a 30-year exploitation
contract, considering an average
oil barrel cost of USD $122.8 and
applying 12% taxes, a net present
value (NPV) of $383.5 million, an
internal rate of return (IRR) of
39.5% and a payback period (PP)
of 2.8 years are estimated.
Option 2
Title: Oil exploration risk contracts and shared production
agreements in production blocks
located in the Cuban Exclusive

Economic Zone (EEZ) in the Gulf of

Mexico (52 available blocks).
Description: Define potential oil and
gas reserves in the Cuban EEZ. If
positive results are obtained, exploit
the resources profitably.
Location: The Cuban EEZ comprises
a 112 000 km2 zone in deep waters
in the Gulf of Mexico, just to the
north of the Pinar del Ro, Artemisa,
Mayabeque and Matanzas provinces, where there are 52 accessible
blocks for oil exploration and further exploitation (see Figure 3).
Market: Cuban internal market in
the first term, with surplus production for export.
Results: For an exploration block
with a 30-year exploitation contract,
considering an average oil cost of
$128.2/bbl and applying 10% taxes:
NPV: $1241.2 million;
IRR: 18.5%;
PP: 7.5 years.

interest to the Cuban government

are not considered.
Market: Cubas internal market in
first term, with surplus production for export.
Results: For a block with a 25-year
exploitation contract, considering
an average oil cost of $122.6/bbl
and applying 10% taxes:
NPV: $191.7 million;
IRR: 44%;
PP: 3.4 years.

Option 4

Title: Oil exploration risk contracts

and shared production agreements
in production blocks located in
Cuban national territory (25 blocks).
Description: Define potential oil
and gas reserves in the Cuban
national territory. If positive results
are obtained, exploit the resources
profitably.
Location: There are 25 available
blocks dispersed throughout Cuban
national territory. Urban zones,
protected zones, or areas of special

Title: Enhanced recovery of existing oil deposits.

Description: Increase the recovery coefficient of the oil reserves
contained in certain existing oil
deposits, which cannot be recovered by conventional extraction
methods.
Location: The oil deposits opened
for negotiation under this option
are Santa Cruz del Norte, in
Mayabeque province, and East
Varadero oil deposit in Matanzas
province. Other oil deposits will be
considered for future investments.
Market: Cuban internal market in
the first term, with surplus production for export.
Results: For a block with a 30-year
exploitation contract, using an
average oil cost of $95.1/bbl and
applying 12% taxes:
NPV: $142.4 million;
IRR: 46.3%;
PP: 2.1 years.

crude oil exists are well identified

and characterised.
The economic value of Cuban
crude oil in the international market
is about 60% of the reference crudes
WTI and Brent. It is estimated that
the production cost of a barrel of oil
from offshore platforms in Cuba is
about $20-35, while the production
cost of oil produced from inland oil
fields is $13-15/bbl.
Table 1 outlines foreign investment opportunities in the Cuban oil
industry.

If oil is found, it is estimated that

companies would have to invest in
developing production capacity for
at least three to five years before
production could begin. However,
production could be delayed due to,
mostly, availability of offshore oil
field development services. Once oil
production begins, it is expected to
grow slowly.
The main non-Cuban operator
in Cuba is Sherritt International, a
Toronto mining company which has
been active in Cuba for more than

Option 3

Table 1

could have a legal validity period

from 25 to 35 years. Taxes are not
paid for the first eight years, be they
municipal, regional, or those on the
repatriation of earnings or products. Bonuses are not paid upon
signing, and taxes are levied by the
National Tax Administration only
on net annual earnings. Nowadays,
foreign companies interested in
investing in the Cuban oil industry
prefer to participate mostly in EOR
operations since the exploratory risk
is smaller and the zones where the

42 PTQ Q1 2017

q1 reduc.indd 2

www.eptq.com

15/12/2016 14:05

10km

Canas
i
Escond Puerto
ido

Cruz

Santa

Via Bla
Boca d nca
e Jaruc
o

Block N37
3D seismic (HabanaMatanzas)
Leads
Leads too far from the coast (Block N37)
Prospects
ERW

S
O eb
e s or
te uc
o

Y
Le um
ja ur
no i

J
Le
j
Fr a n o
a
N il
or e
oe
Fr
st
ai
e
le

La Nin~a
Bacuranao
S Maria

La Pinta

Habana del
Este
Cojimar
Bacuranao

20 years. This rm operates Puerto

Escondido, Yumuri, and Varadero
West oil elds under two production-sharing contracts (PSC). In
May 2014 the company negotiated
a 10-year extension to the Puerto
Escondido-Yumuri PSC. It has
drilled eight wells, one more than
required by the extension terms,
and has ended the extension drilling
programme. Six of the wells produce oil, one is suspended, and one
has been abandoned.
Some studies have concluded that
Cuba could produce enough oil in
the future to become an oil exporter,
but there are some uncertainties
that still need to be considered rst
in order to support this conclusion.
First, there are tangible reservations
regarding when oil production will
start and at what rate it could be
obtained. Secondly, Cuba will need
to offset the roughly 130 000 b/d
of oil it currently imports to meet
existing demand prior to becoming a net oil exporter. Thirdly, oil
demand is expected to grow in the
near future, taking into account
growth in sectors such as construction, tourism and industry. In this
case, Cuba is still likely to trade
more oil especially as rening
capacity increases but its net trade
balance in oil may not necessarily
shift to a signicant oil export surplus. All will depend only on how
much oil is found and developed,
and what will happen with domestic Cuban demand. What is more
certain to take place is that an increment in oil production may reduce
Cubas dependence on oil imports
from Venezuela.

Yumuri
Seboru
co

o
er

ad

r
Va

1 B J Norte (Breccia)
2 B J Norte (Veloz)
3 Santa Cruz Este
4 Santa Cruz Norte
6km horizontal displacement

Figure 2 Mapped leads and prospects in Cubas Northern Oil Belt

was applied the oil production rate

increased markedly.
In the last 14 years, more than
245 million of oil barrels have been
extracted from the so-called Heavy
Crude North Fringe (HCNF) or
Northern Oil Belt (NOB), an oilrich area between Havana city and
Matanzas provinces (see Figure 2).
This area accounts for about 97% of
Cuban oil production (an oil deposit
located in that zone and qualied as
productive could generate about
2000 b/d), while there are other

AP

NG

R
TE

EA

small production sites located in

Ciego de vila and Sancti Spiritus
provinces, some of them with
exploitation periods of more than
60 years.
The most important oil deposit
located in the HCNF, and in Cuba,
is that of Varadero, which has 90
elds under exploitation and had
produced about 185 million barrels
of oil at the end of 2015. It is calculated that only about 6-7% of its
potential oil reserves (estimated at
1.3 billion barrels) have been recov-

GULF OF MEXICO
FLORIDA STRAIT

How oil is extracted in Cuba

FY
UC
N
ATA

q1 reduc.indd 3

IT O

www.eptq.com

A
STR

Cuba has the second largest

proven hydrocarbon reserves in
the Caribbean area, surpassed only
by those of Trinidad and Tobago.
Since 1996, and taking into account
that most oil deposits in Cuba are
located offshore, CUPET changed
its approach to extracting oil by
introducing the horizontal perforation method (HPM), which has led
to an increase in oil production levels since 2002. This change has had
a substantial economic impact on
the Cuban oil industry, considering that in most elds where HPM

CARIBBEAN SEA

Figure 3 Oil exploration blocks located in the Cuban Exclusive Economic Zone, Gulf
of Mexico

PTQ Q1 2017 43

15/12/2016 14:05

Varadero Field is the largest

in Cuba with over 11 billion
barrels of oil in place
(10.5-14.3 API)

Bolanos-1 (1991) recorded

a recovery of 22 API oil
from Shallow Sheet

Guadal-1 (1971) recovered

>30 barrels light oil (24.5 API)
on test from Shallow Sheet

Marti-5 (1984) flowed light oil

(24.0 API) from Deep Sheet at
unspecified rate
Surface geology / structural trend
Well oil shows
Well strong oil shows
Oil field
Oil and gas field
Lower sheet play leads

Motembo Field (1881) has

intermittent production of light
oil (50-64.5 API)

Does oil exist in Motembo?

km

50

Figure 4 Block 9 in Motembo region

ered to date because no secondary

or enhanced extraction methods
have been applied.
At present, annual oil production
capacity in Cuba is about 25 million
barrels (about 50 000 b/d). This is
used entirely for power generation
and meets about half of national
energy demand, while the rest
(about 90 000 b/d) is imported from
Venezuela under an exclusive payment agreement. Cuba accounts for
only 0.05% of the worlds total production of crude petroleum.
A new oil drilling campaign
started at the end of 2016, continuing throughout 2017, in the Cuban
Exclusive Economic Zone (EEZ) in
the Gulf of Mexico (a 112 000 km2
zone divided into 59 blocks, 30
of them assigned to foreign rms
under risk contract agreements, see
Figure 3), taking into account tangible evidence that there is about
15 000 million barrels of extractable
oil in this area. The rst exploration

ico Lpez
(Havana)
R

operations will be carried out by

the Venezuelan company Petrleos
de Venezuela (PdVSA) and the
Angolan rm Sonangol.
Up to 2020, there are concrete

At present, annual oil

production capacity
in Cuba is about 25
million barrels
(about 50 000 b/d)
plans to drill exploratory wells in
several land sites located between
La Habana and Santa Cruz del
Norte (mostly in the towns of Boca
de Jaruco, Tarar, Santa Mara and
Santa Cruz).
Since the end of 2015, oil deposits
in Boca de Jaruco are being treated
with steam injection technology in

Sergio Soto
(Sancti Spiritus)
R
R

Camilo Cienfuegos
(Cienfuegos)
R
R Refinery
Pipeline

Figure 5 Oil refineries in Cuba

44 PTQ Q1 2017

q1 reduc.indd 4

order to reduce the viscosity of the

extra heavy crude, in a joint venture project signed between CUPET
and the Russian state owned rm
Zarubezhneft.
Cuban oil is basically a sour,
heavy crude (8-16 API) and can be
used only for power generation and
to produce cement, lubricants and
asphalts, although there are some
oil deposits that supply substantial
amounts of light and medium oils
from time to time.

Hermanos Diaz
(Santiago de Cuba)

During July 2016 international

media publicised a report published by the Australian rm
MEO on the hypothetical discovery of high quality, light crude oil
reserves of about 8200 million barrels in a zone known as Motembo, a
2380 km2 area located in the central province of Villa Clara, which
is the rst region in Cuba where oil
was obtained. Logically, this notice
generated great expectations both
in Cuba and the rest of the world.
However, some CUPET authorities
declared in a press conference some
weeks later that there had been
erroneous misrepresentation of the
note published by the Australian
company, which did not use the
terms conrmed discovery or
nding, but only the identication of potential, important amounts
of oil which can be recovered under
the application of future development projects.
According to announcements by
MEO, conrmation and further validation that there are thousands of
millions of barrels of oil at this site
requires supplementary evaluation,
exploration and analysis operations
distributed in different phases: 2D
seismic operations, protability and
prospective studies, deep drilling
and oil quality testing.
MEO signed a shared production agreement with CUPET in
September 2015, and is carrying
out oil exploration studies at Block
9 in Motembo region (see Figure 4),
including review and reprocessing
of the existing 2D seismic data; geochemical samples acquisition and
evaluation; and completing new 2D
seismic operations in an additional
200 km zone.

www.eptq.com

15/12/2016 14:06

Fracking

Fracking is one of the most controversial oil extraction methods used

today. This method has never been
used for oil extraction in Cuba since
there are more efficient and profitable techniques available to extract
oil such as, for example, steam or
chemicals injection.
According to recent geological
studies made by CUPET, the oil reservoirs located in Cuba are in carbonate layers, so the best available
option to extract oil on those sites is
by means of acid or steam injection.
In that sense, fracking is not considered today a viable method for oil
extraction in Cuba.

Oil refineries in Cuba

There are four refineries currently

under operation in Cuba, all owned
and operated by CUPET (see Figure
5). They are:
ico Lpez, in Havana province
Camilo Cienfuegos, in Cienfuegos
province (see Figure 6)
Sergio Soto, in Sancti Spiritus
province
Hermanos Daz, in Santiago de
Cuba province.
Together, these refineries have
a total nominal crude oil distillation capacity of about 300 000 b/d,
although only 45% of this capacity
is currently exploited (135 000 b/d).
Table 2 shows the refining technologies applied by each Cuban
refinery as well as its refining
capacity.
The largest of the four refineries,
ico Lpez, accounts for nearly 40%
of total refining capacity (36 400
b/d). It operates a 12 500 b/d catalytic cracker, and is the only facility
with a catalytic conversion unit in
Cuba. There is a five-year plan to
modernise and increase its storage
capacity (mostly crude oil reception tanks and petroleum products
storage vessels), as well as its automation and steam generation/distribution systems. Similarly, there
are plans to install a Jet A-1 (Merox)
fuel sweetening plant.
The Camilo Cienfuegos refinery
was reopened in 2007 as the result
of an agreement between CUPET
and PdVSA, and is now operated
by Cuvenpetrol SA, a Cuban/
Venezuelan joint venture. This

www.eptq.com

q1 reduc.indd 5

Figure 6 Camilo Cienfuegos refinery in Cienfuegos province

refinery runs only on Venezuelan

crude, and has a nominal refining capacity of about 65 000 b/d.
More than 90% of the refinerys
output, including gasoline, diesel
fuel and fuel oils, is consumed by
the Cuban domestic market. Several
techno-economic design projects are
under way such as expansion of its
refining capacity to about 150 000
b/d; constructing a plant for olefin
and aromatics; increasing its storage
capacity from 8000 bbl to 139 000
bbl; building a liquefied natural gas
plant; and reactivating the pipeline
between Matanzas and Cienfuegos,
at a total preliminary cost of about
$6 billion. Several tests will be carried out to determine whether the
refinery can process up to 80 000
b/d. From July 2016, the refinery
reduced its actual processing capacity to 50 000 b/d due mostly to
reduced oil imports from Venezuela
and the implementation of exhaustive, complex maintenance operations in main equipment and related

accessories. In 2015, the facility

refined about 17.8 million bbl, while
it is projected that it will refine only
around 9.43 million bbl in 2016, 53%
of its planned output. Recently,
this refinery has been processing
and blending crude and petroleum
products for sale to neighbouring
countries.
Hermanos Daz refinery has a
nominal annual refining capacity
of about 1 500 000 cu m of crude
oil (30 000 b/d), although it is currently processing 22 000 b/d due
to reduced oil imports. Several oil
products are produced in this refinery such as liquefied petroleum gas
(LPG), naphtha, gasolines, vacuum
gas oil, fuel oil, diesel and asphaltic cement. It supplies the market of
Cubas eastern region and provides
raw material for the catalytic cracking unit in ico Lpez refinery.
Several actions have been accomplished throughout this year in
order to modernise some equipment
and accessories, such as the total

Refining technologies and capacities applied by Cuban oil refineries

Refining
Refinery
technology
ico Lpez
Camilo Cienfuegos
Sergio Soto
Hermanos Daz

(Medium conversion) (Hydro-skimming) (Hydro-skimming) (Hydro-skimming)
Refining capacity (tobd)
Atmospheric distillation
Vacuum distillation
Catalytic cracking
Reformer
Distillate hydro-refining

36.4
18.0
12.5
2.7
3.2

65.0
-
-
9.6
25.0

2.8
1.4
-
-
-

30.0
18.0
2.7
7.0

Table 2

PTQ Q1 2017 45

15/12/2016 14:06

Cuban production, exports, and imports of refined petroleum products,

2010-13 (000 barrels)
Product
production
Naphtha
LPG
Gasoline
Diesel fuel
Jet kerosene
Fuel oils
Total
Exports
Gasoline
Diesel fuel
Jet kerosene
Total
Imports
LPG
Gasoline
Diesel fuel
Jet kerosene
Fuel oils
Total

2010

2011

2012

2013

330
220
4163
8972
2412
17 856
33 953

601
183
3313
8943
2272
17 020
32 333

586
176
3225
8349
2214
16 720
21 270

513
147
2932
8063
2199
14 660
28 514

5336
528
5622
11 486

0
528
5417
5923

0
506
5087
5087

0
0
5116
5116

396
2,829
498
4193
4325
12 857

462
0
506
4200
3255
9031

469
0
147
4537
2404
8188

367
0
132
4552
2456
8114

Table 3

insulation of ovens and kilns, implementation of automation, more

efficient combustion systems, and
incremental storage capacity.
Finally, Sergio Soto refinery operates atmospheric distillation and
vacuum distillation columns, and
produces primarily asphaltic liquids
(AC-30 and RC-0 types), as well as
minor amounts of naphtha, diesel,
fuel oil, solvents, basic dielectric
oils, and chemicals to produce pesticides. It is the only Cuban refinery
dedicated exclusively to processing
and refining crude oil produced
in Cuba (about 2800 b/d). In 2015,
the amount of asphaltic liquid produced by this refinery was 15 400
t, the highest output in its history.
The vacuum distillation tower
was recently revamped in order to
increase production efficiency from
42% to about 58%. The production
of dielectric oils for 33-kV electric
transformers, and the introduction
of more efficient, high capacity boilers and cooling water systems is the
most important investment task for
the facility in the near future.
Cuban refineries are able to meet
current domestic demand for gasoline and diesel fuels, but the nation
still needs to import other refined
petroleum products such as kerosene jet fuels and fuel oils, mainly
from Algeria (which accounts for
80-85% of total imports of these

46 PTQ Q1 2017

q1 reduc.indd 6

products), Venezuela (about 8-10%),

the European Union (6%), Mexico
(2%) and Russia (2%). Such imports
account for about 25-30% of total
domestic demand for refined petroleum products (see Table 3).
From July 2016, PdVSA reduced
its crude oil exports to Cuba by
19.5% (about 83 130 b/d), alleging

The present day

represents a now or
never time for the
Cuban oil industry
regarding growth and
economic feasibility
declining crude prices in the international market. Cuba, in turn, has
replaced the shortfall with similar
volumes of crude and petroleum
products from the Caribbean transshipment terminals of Bonaire, St.
Eustatius, Aruba and Bullen Bay,
and Willemstad, Curaao.
Cuban refineries process small
amounts of Cuban oil. Most of the
crude oil obtained along the countrys northern coast is diluted with
naphtha and sent directly to power
plants for combustion.

Conclusion

Taking into account that about 95%

of Cubas potential oil reserves are
located either inland or in shallow
waters, they remain unextracted
because drilling operations require
the use of advanced recovery technology that Cuba does not possess.
The introduction of foreign capital
technologies in this field is a high
priority for the country in order to
achieve its intended self-reliance on
energy.
The proven existence of relevant oil reserves throughout the
country, the application of attractive tax related policies to foreign
firms interested in oil related operations in Cuba, the strong support
received by the Cuban government and the socio-political stability of the country, the availability
of skilled and experienced human
resources, as well as the reputation of CUPET regarding business
procedures and contracts are the
most significant advantages offered
by the country to develop its oil
industry by means of direct foreign
investment.
The Cuban energy sector needs
to update and modernise its technological systems and infrastructure to modern standards in order
to increase its profitability and
economic viability with respect to
the environment, to meet future
increases in internal power demand,
and to compete with the renewable sector. Thus, several strategic
programmes and projects are being
designed or evaluated to cope with
expectations and also to develop the
country to higher standards. In this
sense, the present day represents a
now or never time for the Cuban
oil industry regarding growth and
economic feasibility.

Amaury Prez Sanchez is Adjunct Professor

at the Department of Chemistry, University of
Camagey and a Research Engineer with CUPET
company. With over 10 years experience in
chemical/petrochemical processes, mostly
in mass/heat transfer operations, he holds a
masters degree in chemical process analysis
from the University of Camagey, Cuba.
Email: amaury.psanchez@reduc.edu.cu

www.eptq.com

15/12/2016 14:06

rezel.indd 1

14/12/2016 11:08

Safety Measures
Hard Hat
Goggles
High-Visibility Vest

Remote Mount
Capability Keeps
Workers Off Top of Vessel
for Switch Modification

Best-in-Class
Safe Failure Fraction
>91%

Insulated Gloves

Steel-Toed
Boots

Advanced
Self Diagnostics
Assures Reliable
Performance

Safety Harness

Dual-Point
Option for Two-Alarm
Safety Protocol

Protect your plant with Echotel

Ultrasonic Level Switches
ECHOTEL liquid level control technology measures up to the most rigorous safety standards,
with intelligent design that ensures outstanding quality and reliability.

echotel.magnetrol.com

magnetrol.com 800-624-8765 info@magnetrol.com

magnetrol 2.indd 1

2016 Magnetrol International, Incorporated

14/12/2016 10:44

Control system security

A major cyber attack on a refinery is a clear and present danger. What lessons
can be drawn from other sectors?
SINCLAIR KOELEMIJ
Honeywell Process Solutions

ince the 2010 attack on Iranian

uranium enrichment facilities,
the cyber security of industrial
control systems has attracted significant attention and investment as
companies have sought to improve
their protection against cyber
threats. Refining companies and
other industrial businesses have
revisited their approach to cyber
security, which was until then
focused on protecting systems
against malware and hackers. An
attack on the physical installation
had been inconceivable.
Even so, the 2010 attack required
considerable resources and skills to
succeed.1 Several factors contributed to its complexity:
It required stealth. Focused on
damaging the centrifuge equipment, it needed to be done quietly
to prevent early detection. To hide
what was happening from the
process operators, the feedback
from the control system also
required manipulation.
It was targeted, aimed at one
particular installation. It seems
unlikely it was intended to infect
other systems, since when it eventually did so resulted in the attack
being discovered and stopped.
It bridged an air gapped/isolated
system the control system for the
installation. A method had to be
developed
to
infect
mobile
computer equipment that was ultimately connected to the control
system network for engineering/
maintenance purposes.
It required knowledge of the

physical system. The attacker
needed to know the variations
in speed that would cause excessive wear of the centrifuge

www.eptq.com

q1 honeywell.indd 1

bearings, the exact structure of the

physical installation, and the exact
code running in the control equipment and feedback from the
sensors.
All this meant the attack required
a substantial effort. For many, it
therefore seemed such an attack
was unlikely to occur against
targets where no significant political motivation existed, since only
nation states would have the
resources, focus and motivation to
undertake it.
Damaging a physical plant installation or removing it from service
does not always require such a
complex attack, however. There is a
variety of ways to create damage
through thermal, mechanical or
hydraulic stress in a plant or to
cause physical systems to wear
more quickly than anticipated. The
operation window of the equipment is in large part determined by
the data in the industrial control
system. A determined attacker
successfully accessing the control
system will therefore not have
much difficulty causing a plant
shutdown.
With the recent attacks on the
steel industry (Germany, 2014), the
power industry (Ukraine, 2015),
and water treatment (US, 2016), an
expanding range of business leaders are becoming aware of the risks.
And while the refining community
has thus far been spared a major
attack, its leaders are looking across
industry for best practice security
approaches. The net result of this
heightened awareness is that chief
information security officers and
chief security officers are slowly
expanding their focus beyond

corporate IT systems to ask critical

questions
of
the
automation
systems, too.
This article explores the differences between information security
and control system security, and it
uses learnings drawn from the
manufacturing
and
processing
industries to inform refining decision makers.
While information security methodologies aim to secure the
integrity and availability of the
information in the control system,
they do not analyse the consequences for the underlying physical
system. This limits the ability of
information security to predict and
prevent attacks on the physical
system.
The main focus of control system
security, by contrast, is on identifying and analysing the different
types of failure scenarios and
designing security to withstand
attacks that could cause these.
Control system security develops
scenarios for such smart attacks
and identifies counter-measures
against them. The article provides
some examples and discusses ways
to protect the system.

Background

Industrial control systems typically

have three tiers:
The physical system (production
system)

The production management
system, including the distributed
control system (DCS), supervisory
control
and
data
acquisition
(SCADA) system and safety instrumented system (SIS)
The operations management system
(optimisation, quality management,

PTQ Q1 2017 49

16/12/2016 15:30

Business Management System (L4, L5)

(information management)

Operation Management System (L3)

(information management)
Industrial
Control
System

Production Management System (L2, L1)

(automation)

Information
System
Security

Control
System
Security

Sensors and Actuators (L0)

Refinery Installation
(production processes)

Crude

Petrol

Figure 1 Industrial control system

tanks and pipelines), including the

sensors and actuators.
Figure 1 shows the high level
structure, including the levels
assigned by the ISA 95 standard.2
The focus of this article is on levels
0, 1 and 2, but some applications at
level 3 can also impact the physical
system, so it is not always possible
to distinguish a clear boundary
between control system security and
information
system
security.
Depending on the application, there
may be an overlap between the two.

Knowledge

environmental management, planning, and maintenance).

Operations management systems
focus on the efciency, effectiveness
and
quality
of
the
manufacturing process. Production
management systems focus on the
manufacturing process itself the
automation of the production
process. Finally, the physical
system (production system) is
composed of the various process
units in the plant (such as the distillation column, boilers, furnaces,

Covert
attack
Disclosure
Zero dynamics
attack

Eavesdropping
DoS

Bias injection
attack

on

i
pt

ru

is

Figure 2 Cyber physical attack space model

50 PTQ Q1 2017

q1 honeywell.indd 2

Replay
attack

The security of the systems at the

business management layer differs
considerably from the control
system. A frequently mentioned
example is the difculty in updating the control system or installing
security patches for it. Stopping
production for this is difcult to
justify, and changes in general are
seen as a risk to the business continuity.3 This is why so many legacy
systems are still active despite no
longer having security patches
available for them, with the result
that they contain well-known
vulnerabilities that can be exploited
by attackers.
Another difference is that an
automation system such as a DCS
requires a stricter operational environment than the average computer
system.
Real-time availability ensures
every automation task in the
system has sufcient time to
complete its task. Real-time environments differ from a time sharing
environment where tasks are allocated a specic time slot to
complete their workload, over-runs
are not allowed, and ultimately
tasks may not be completed. In a
real-time environment, overloading
the system can result in resource
starvation. The environment is
therefore more vulnerable to denial
of service attacks than business
management systems, which can
cope with a very broad range of
delays.
Finally, risk is different in an
industrial control system. Consider
the attack model developed by
Andr Teixeira, Henrik Sandberg,
Daniel Prez, and Karl H.
Johansson.4 Unlike cyber security
for the business management
systems where cyber security is
focused on protecting data, they
noted that cyber attacks on control
systems have the potential to
change the physical production
process in different ways depending on the specic attack scenario.
The attack space for a control
system can be described by their
model (see Figure 2).
The attack space for cyber physical systems has three dimensions:
System knowledge: attacking the
cyber physical part of the system

www.eptq.com

14/12/2016 12:21

requires an understanding of the

production process, the various
parameters, measured values, and
actuators that can be used in the
attack.
Disclosure resources: an attack
can also require knowledge of the
real-time state information of the
system, process values, limits and
control parameters.
Disruption resources: an attack
requires the capability to disturb
the system, perhaps by modifying
an output setting, a control parameter, or disabling a system action.
Complex attacks, like the attack
on the Iranian nuclear facility, use
all three dimensions, but a simpler
attack such as replaying Modbus
trafc to freeze a process value
only needs two dimensions. An
attack like the HAVEX attack5 used
only one dimension a Trojan
horse for eavesdropping.
Risk for a cyber physical system
is a function of the likelihood and
impact of a scenario: risk =
(scenario, likelihood, and impact).
Scenario describes the system
under attack and what steps the
attacker needs to take to achieve
the goals.
Likelihood is a function of the
effort an attacker needs to invest to
execute the scenario. The more resilient a system is against a specic
scenario, the higher the investment
in effort required from the attacker
to achieve his goals, and the lower
the risk of him doing so.
Impact is the consequence of the
attack in principle the attackers
goal. Risk calculations within information security primarily use
scenarios to determine the business
impact, irrespective of the likelihood. The scenario is a dimension
not used in information security,
but essential in control system
security because of the direct
connection to the physical system.

Control system security

For control system security the

central question is what an attacker
can do to harm the physical system
(production system) or disturb the
production process.
There are three potential objectives for an ICS attack:
To damage the process equip-

www.eptq.com

q1 honeywell.indd 3

Failure occurs quickly

Critical limit high

Standard level high

Operating
window

Failure occurs with

sustained operations

Target range high

Stable
Reliable

Optimal

Target

Safe to
operate

Target range low

Standard level low
Critical limit low

Failure occurs with

sustained operations
Failure occurs quickly

Figure 3 Operating limits and boundaries

ment for example, to increase

wear or corrosion of the system,
rupture a pipeline, damage furnace
hosing or damage a distillation
column.
Disturb the production process
stop cooling systems, activate a
bypass or reduce product quality,
for example.
Compromise compliance such
as with health, safety, and environmental regulations
To accomplish any of these objectives, the attacker has three
potential targets:
Plant operability, which includes
the operating window, the safety
function, and monitoring and diagnostic functions. An attack on the
operating window, for example,
may increase it to cause damage.
The measured or calculated
values, with the attacker deceiving
operators or an automated function
by modifying the apparent realtime state of the system.
The control actions initiating
an unwanted action, blocking
execution of a required action or a
combination of the two.

Attacking the operating window

The operating window is dened

by parameters in the production
management system for example,
range values for ow, pressure,
temperature, rotation speed, level,
signal characterisation parameters,
rate limiting values, travel time,
output limits, and so on.
Figure 3 illustrates various types
of operating limits that create
boundaries for any specic operating window.

The middle zone (green), between

the standard limits (high and low),
is the zone designated for achieving
operational targets. Outside those
limits, operator intervention is
generally required to return the
process to this zone.
Depending on the process variable, some limit ranges may not
have an upper and lower boundary. For example, tube skin
temperatures generally have only
upper limits.
If the attacker breaches the integrity of the operating window, the
physical process moves into a range
where failure will eventually occur
if it remains there or, if too far
outside the operating window, may
even occur immediately.
During plant design, the integrity
operating window (IOW) is dened
and all conguration parameters are
set in such a way as to protect the
physical
system
from
being
damaged. As an additional safeguard
for
the
production
management system, the safety
instrumented system (SIS) is used to
force an emergency shutdown when
such a condition would occur and
safety would be at risk. Not every
operating window is guarded this
way, however, so in general, even
when the ICS includes a safety function, modifying the operating
window can cause damage.

Attacking the safety function

Safety functions are generally difcult to attack. They run proprietary

code and a physical key switch
usually controls downloads of the
code to the safety controllers. It is a

PTQ Q1 2017 51

14/12/2016 12:21

well-protected environment and

not easily breached. However, there
are other tactics that can have an
impact on the safety function, such
as attacking the safety transmitters.
Transmitters
are
frequently
centrally managed. If the management system is attacked it can be
used to modify the sensor configuration and indirectly influence the
safety systems trip point. This will
not always be possible, but where
the management system has a
direct connection to the IO multiplexer boards that connect to the
field instrumentation it is a risk.

Attacking the actuator function

Sometimes actuators are directly

connected to the control network,
are indirectly connected through
Ethernet to serial converter boxes,
or use Ethernet connected relays
operating a circuit breaker. Even
when actuators are connected with
controllers or PLCs there are ways
to (de-)activate them. They can be
operated by the attacker through
either spoofed message injections,
modified messages or message
replays. Situations where the
attacker could initiate an action and
obstruct the correction of this action
or obstruct the operation of an
actuator as a denial of service
attack should also be considered.
Finally, elements of the operating
window, such as the travel time of
a valve, are sometimes defined in
the actuator. In these cases this
information can be targeted, too.

Attacking the sensor function

Sensors can be attacked in several

ways, depending on sensor type
and the way they are connected
and managed. Attack scenarios
might involve replaying sensor
reading messages to freeze the
sensor reading, injecting sensor
messages to modify the readings,
or perhaps just blocking the sensor
messages. An alternative could be
to modify the sensor configuration,
for example changing thermocouple characterisation settings to
manipulate the sensor values.
Sensors and actuators that allow
over
the
network
firmware
upgrades are also vulnerable to
denial of service attacks in which

52 PTQ Q1 2017

q1 honeywell.indd 4

the attacker installs corrupted firmware to disable the sensors

function.

Analysing control system security

The attack surface of a control

system is much bigger than that of
open systems such as Microsoft
servers or network equipment. The
internal hardness of a control
system, achieved through the
segmentation of the network into
distinct
security
zones
with
restricted access, is vital. Making
the attackers life difficult once he
gains access to the control network
is just as important as keeping him
out to begin with. The more time
the attacker requires to reach his
objective, the more time you have
to spot the attack and respond to it.
Approaching control systems
through the eyes of information
security results in many vulnerabilities being missed. Control system
security therefore uses a different
approach, based upon cyber failure
scenarios. A cyber failure scenario
describes an action in the control
system and how this can be
executed by an attacker. A very
simple scenario may describe how
to open a circuit breaker in a power
grid and prevent remote action to
correct this.
Cyber failure scenarios can be
very process specific, but most are
generic and can be used for a wide
range of processes. The control
system security analysis process
will investigate which scenarios
apply for the specific system, how
the system is or can be protected
against them, and how it detects
the attack. This method uses the
scenarios (documented as attack
trees) and evaluates the likelihood
that each branch of the attack tree
is successfully executed. The critical
branch is the branch with the highest likelihood (and therefore
constituting the highest risk since
the root of the tree is common to all
branches). This can be used to rank
the various cyber failure scenarios.
Finally, the cyber attack space
model (Figure 2) shows us that
some (low effort) attacks require
little system knowledge and can be
relatively easily constructed in the
absence of specific security func-

tions. Other attacks require more

dimensions in the attack space
model to accomplish and are therefore more difficult to undertake.

Summary and conclusions

Control system security looks at

how the control system can be used
to attack the physical system and
damage it. This dimension is not
addressed by the information security approach, which, if relied on,
will result in gaps in the internal
hardness of an industrial control
system. Recent attacks such as that
on a steel factory in Germany6 and
on the power grid in the Ukraine7
have exploited these gaps. While
there has not yet been a major attack
on a refining company or infrastructure, there remains a high degree of
risk and, as we have seen, learnings
can be drawn from other sectors.
It is a common misunderstanding
that securing the open systems in a
control system will secure the
control system. During the design of
the control system, many decisions
are made that can result in potential
vulnerabilities in the control system
security. Ignoring these can lead to
serious incidents. The industrial
cyber security team from Honeywell
is one of the few teams in the world
to combine the disciplines of information security, control system
security, and process automation.
References
1 Zetter K, Countdown to Zero Day, Stuxnet
and the launch of the worlds first digital
weapon.
2 ANSI/ISA95.00.04,
Enterprise-Control
System Integration, Part 4: Objects and
attributes for manufacturing operations
management integration.
3 Krebs B, Cyber Incident Blamed for Nuclear
Power Plant Shutdown, Washington Post, Jun
2008.
4 Teixeira A, Sandberg H, Prez D, Johansson K
H, Attack Models and Scenarios for Networked
Control Systems, Royal Institute of Technology.
5 Langill J T, Defending against the Dragonfly
Cyber Security Attacks.
6 Bundesamt fr Sicherheit in der
Informationstechnik. Die Lage der IT-Sicherheit
in Deutschland 2014.
7 E-ISAC. Analysis of the Cyber Attack on the
Ukrainian Power Grid.
Sinclair Koelemij is Technical Team Leader,
Industrial Cyber Security EMEA with Honeywell Process Solutions.

www.eptq.com

14/12/2016 12:22

Neles NDX valve controller

delivers performance perfected
Thats how we make the big difference, the Metso Way.

The Neles NDX has been carefully designed and manufactured to make
the big difference to our customers, regardless of end use industry or
application. Savings are created through accurate, long-lasting and
maintenance-free performance and through time saved as a result of
extremely easy installation and use. The Neles NDX provides the reliability and robustness youd expect from a valve controller by Metso,
for all valve brands in standard applications.
Find out more about the savings, safety and reliability that each
Neles NDX valve controller offers at metso.com/ndx
#TheMetsoWay

Metso_NDX_PrintAd_A4_ENG.indd 1

13.4.2016 10.06

Intelligence on Duty

Get The Power of Big Data

A PROGNOST diagnostic system acquires and diagnoses 65 terabytes of sensor data
every day. Impressed?
Data itself means nothing
The challenge is to convert data into meaningful information. Our intelligent machine
and component diagnostics retrieve the knowledge operators need, to boost uptime
while reducing costs. PROGNOST keeps operators informed. Non-stop and in real-time.
Thats impressive.

reply@prognost.com www.prognost.com

prognost.indd 1

14/12/2016 09:59

Data operations transform fuels value

A refiner applied advanced analytics to develop techniques for processing
opportunity crudes which minimise negative effects on the plant
CRAIG HARCLERODE
OSIsoft

he petroleum industry is once

again in the midst of titanic
changes.1 Declining prices,
expanding sources of supply, rising regulatory requirements and,
perhaps most importantly of all, a
dramatic shift in markets like transportation are forcing companies
across the value chain to reconsider
long held assumptions about expansion, growth and customer demand.
Luckily, these new challenges are
coinciding with advances in big
data, Internet of Things (IoT) and
predictive analytics and the ability to leverage to process opportunity crudes and be more proactive
and predictive in decision making.
While the upstream oil industry has
been a somewhat enthusiastic adopter of digital technology, the midstream and downstream segments
have been conservative and slow
to adopt. That is changing with the
Industrial Internet of Things (IIoT),
advanced analytics and big data.
Collectively, we are inundated with
marketing messages that are adding confusion and false promises,
resulting in a good number of projects that go awry with limited or
no business value and, worse, lost
opportunity costs.
But we will also see implementations that will effectively serve as a
blueprint because they will demonstrate how digital technology
can reduce risks and costs while
improving asset utilisation, yields,
integrity and, most importantly,
protability.
In fact, we already have such an
example. MOL, based in Hungary,
has been on a journey to reinvent
its operations by better leveraging operational data already being

www.eptq.com

q1 OSI soft.indd 1

An integrated
downstream value chain
Integrated fuels value chain
4 refineries,
2 petrochem plants
Logistics including
2000 retail stations

POL

Bratislava
Refinery

PI System overview
4 HA collectives,
~400K tags
Elements
~300 smart templates
~21K elements
and growing

CZR

Duma
Refinery

AUT

SVK

TYK
ROU

ITA

SLO

Rijeka
Refinery

Notifications
~150K templates
~6K notifications

CRO

Sisak
Refinery

SRB

BIH

~61K event frames

including dynamic
PI Coresight is the primary visualisation tool

Figure 1 MOLs downstream operations and operational technology infrastructure

generated by its distributed control

systems (DCS) and other systems
as part of its operations. In 2012,
MOL leadership, in response to
European competition resulting in
low cracked spreads, embarked on
a business transformation enabled
by digital technologies.
The results? MOL has developed
techniques for processing opportunity crudes while minimising
the negative consequences such
as corrosion, operational issues
in areas such as the cokers, and
yields. Advanced corrosion analytics such as high temperature
hydrogen attach (HTHA) and other
forms of predictive corrosion have
been implemented across multiple sites. In all, MOL estimates it
increased earnings before interest,
tax, depreciation and amortisation

(EBITDA) by $1 billion over a ve

year period ending in 2016 through
more aggressive data modelling
and analytics.
Petroleum Economist named
MOL Downstream Company of the
Year in 20162 while the FieldComm
Group gave the company its Plant
of the Year Award for its Danube
facility.

Background

MOL is one of Central Europes

largest downstream companies. It
operates four reneries and two
petrochemical plants in eight countries along with 2000 lling stations
across 13 countries. To organise
data across its production facilities,
MOL has been using the PI System
from OSIsoft since 1998. The system, which has expanded steadily,

PTQ Q1 2017 55

15/12/2016 11:39

MOL downstream OT data model based applications

Safety (PSM) and asset integrity
Interlock governance/DCS role tracking
Operating envelopes
Integrity operating windows (IOWs)
Advanced alarm management

Yields
Crude blending control
Yield optimisation/reporting
Product quality
Analyser reliability

Energy
Energy monitoring management
Energy KPI breakdown (6 tiers)
Column energy efficiency dashboards
Hydrogen, utilities and energy balances
Flaring

Operational optimisation
Plan vs actual analytics with future data
NG and fuel demand gas forecasting
Peak electrical forecasting
Normal mode of control loops
APC monitoring
PI AF and Sigmafine (PI AF) used for
yield accounting and material
movement

CBM asset reliability

All critical rotating equipment
Hydrogen pressure swing absorbers

Figure 2 MOLs downstream OT data model based applications

is divided into four high availability collectives with a combined

total of approximately 400 000 tags
or data points. More importantly,
MOL utilises PI Asset Framework
with smart asset objects to provide a congurable, dynamic smart
operational technology (OT) infrastructure. Currently, MOL has over
300 smart asset object templates
300 templates, 21 000 elements,
and over 61 000 event frames for
signalling the occurrence of key
parameters or events (see Figure
1). Tibor Komroczki, who leads
the Information Integration and
Automation team at MOL, refers to
the PI System as the MOL common
language as it enables the abstraction and nomination of a diverse
tag and asset naming, units of measure, and time zones. MOL generates
over 80 billion data points per year.
The PI System served primarily
as an operations system of record
until 2010 when Komroczki led an
effort for digital transformation. As
a rst step, MOL adopted PI Asset
Framework to create a so-called
digital twin of different processes
and equipment sets in a facility.
With PI Asset Framework, all of
the relevant data streams, meta
data, calculations and analytics,
and alerts and notications from a
process step are combined into a
comprehensive, digital replica of
the plant. Additionally at this time,
it adopted PI Coresight, a visualisation tool for displaying and/
or analysing AF models. Taken

56 PTQ Q1 2017

q1 OSI soft.indd 2

together the smart OT infrastructure with PI Asset Framework,

and PI Coresight MOL had built
a self-serve analytics and business
intelligence environment where
operators and engineers who traditionally used Microsoft Excel can
congure their own smart asset
objects, combine them like Lego
blocks and create their own digital replica and experiment with
potential improvements, and then
execute changes across the MOL
enterprise with governance.
Challenge
critical availability problems

Hydrogen production plants

are critical units in the refinery
Pressure-swing absorbers (PSAs) are
critical equipment in unit operation
Cyclic operation: heavy load on valves
(9-10 open-close hourly)
$1.2M loss in three years due to PSA
valve failures
UPTIME program: 97% operational
availability

Figure 3 Advanced analytics can

predict the impact and ripple effects of
opportunity crudes

With the smart OT infrastructure

in place, MOL established a foundation for higher level efciencies
because it could connect its assets
relatively easily and track its performance backwards and forwards.
New applications can be added
rapidly. Komroczki asserts that
greater control over data has enabled MOL to move from managing in a reactive sense to predictive
management to management by
exception as indicated by the existence of over 61 000 event frames.
Some of the achievements include
improved asset integrity and safety,
asset health, improved energy efciency, increased yield, reduced
hydrocarbon loss, improved environmental reporting, and reduced
maintenance costs (see Figure 2).
Another plus: MOL reduced its IT
costs and reliance on outside vendors because employees were able
to quickly build their own functionality on top of their infrastructure
and then replicate it across foundries and, in doing so, simplifying
and standardising its application
and solutions portfolio. Different
data streams can also be analysed
in tandem so that MOL could
determine the full impact (nancial,
maintenance, energy consumption)
on changes to output.
MOL employed analytics to
reduce the risk of high temperature hydrogen attacks (HTHA). By
studying the relevant operational
data, the company was able to pinpoint the temperature and pressure
parameters that increased the risk
of HTHA. They developed a smart
asset HTHA application template
that was deployed in six units in
less than a week. Following the
successful test, it was rolled out
across MOLs plants in 2015 to over
50 pipe nodes.
Advanced analytics potentially
can be applied in a wide variety of
ways: energy modelling optimisation; the impact and ripple effects
of opportunity crudes in areas of
corrosion, fouling, and efciencies;
the economic gains to be achieved
through opportunity crude processing; better understanding of
advanced control; and preventative
and prescriptive maintenance (see
Figure 3).

www.eptq.com

15/12/2016 16:59

Together facing a
brighter tomorrow
At Yokogawa, we believe the skys the limit.
And to reach beyond todays horizons,
we work step-by-step with you to make
the unimagined a reality. Thats how we
move forward, through the synergy of
co-innovation partnership. Join hands with
us, and together we can sustain a brighter
future. Yokogawa: Building a better
tomorrow with you today.

Please visit www.yokogawa.com/eu

Facing-a-brighter210x297-EU.indd 1

08/09/16 11:14

The integrated OT Smart infrastructure

Yield accounting
P&S Unit models
Financial data

DCS

Varies by site

ERP

Financial data

EAM

IIoT/Edge
Laboratory data

LIMS

PI System

OT Object Model
OT data model /
infrastructure
PI integrator
for Azure

Opralog
e-logbook

NICE
Natural info
centre

Microsoft
Azure

Figure 4 MOL adopted Microsoft Azure machine learning for a production environment

Machine learning

Once MOL had the smart OT

infrastructure across its value
chain with associated IIoT analytics, focus was turned to machine
learning and big data analytics.
MOL has become one of the first, if
not the first, large refiner to adopt
Microsoft Azure machine learning
in a production environment (see
Figure 4). Microsoft Azure works
in conjunction with the PI System:
operational data is uploaded to the
cloud and then analysed across
Microsofts cloud infrastructure.
MOL has developed Azure
machine learning to predict the
impact of sulphur levels in variable feedstocks in their various
desulphurisation units. MOL had
been using offline models for analysing sulphur. Not only did using
offline models increase time, it
also increased the potential for
error. MOL estimated it was losing $600 000 per year across four
units because of its inability to
adjust unit parameters to optimise
sulphur content in the products.
MOL eliminated the losses thanks
to better forecasting and continues
to roll out the technology across
its infrastructure. As with its other
improvements, MOL was able to
leverage its previous technology
investments: the new applica-

58 PTQ Q1 2017

q1 OSI soft.indd 3

tion layered on top of what it had

already implemented.
Following these successes, MOL
turned to improving the performance of its delayed coking units.
By using opportunity crudes, MOL
estimated that it could gain $6 million for each 1% gain in DCU yield.
Gains in DCU yields with variable feed from opportunity crudes,
however, also increased the risk of
steam explosions during the hydrocutting step.
Azure analytics combined with
continual data feeds from the PI
System enabled MOL to thread the
needle. DCU yields were increased
by 2%, yielding an estimated gain
of $12 for each unit per year. At the
same time, steam explosions went
down by 75%. Machine learning
enabled the ability to achieve two
seemingly contradictory goals at
the same time. The company has
now positioned machine learning
for its four DCU units across its
enterprise to take full advantage of
opportunity crudes.

On-premise vs cloud?
No: on-premise plus cloud

While data gets transferred and

stored in the cloud with machine
learning analytics, cloud systems
typically will not replace on-premise storage systems. Real-time con-

trol and insight are required for

operational efficiency as well as
safety. Transferring data to the
cloud invariably increases latency:
data simply has to move far further before it can be used. It also
increases risk because a disruption
in the network can lead directly
to disruptions in operations costing millions in downtime. (Anyone
who has worked with offshore
upstream companies is likely familiar with the risks of satellite links.)
Instead, these systems complement each other. Companies are
opting to maintain on-premise systems and transfer data, or summaries of data, on an as-needed basis,
preferably during hours of low network traffic.
A substantial part of the success
revolved around the use of integrators that effectively automate the
translation process of bringing OT
data to IT-based analytics systems.
Using a CAST (clean, augment,
shape,
transmit)
methodology,
MOL was able to avoid the data
prep and data janitor problem that
can take up to 80% of the time of
projects. Other companies in similar
heavy industries have experienced
similar results: Cemex, the large
cement manufacturer, has reduced
the amount of time required for
preparing and gathering data
across 70 plants for its reports from
over 700 hours to less than one
through CAST automation.

The financial bottom line

One of the more compelling features of MOLs transformation, and

likely a phenomenon others will
experience, is that the changes are
additive. Once the foundation for
digital transformation is in place,
additional applications can be
added on top of the now existing
digital infrastructure. As a result,
incremental
improvements
can
accelerate savings, rather than result
in ever shrinking marginal gains.
Over the four year period running from 2011 and 2014, for
instance, MOL estimates that its
digital transformation programme
accounted for an additional $500
million in EBITDA. During the next
two years, however, MOL added an
additional $500 million to EBITDA,

www.eptq.com

15/12/2016 13:27

bringing the total over ve years to

$1 billion. An equivalent amount
of savings was achieved in roughly
half the time.
While the curve may change over
time, one can expect that savings
will compound. Each new improvement potentially will cost less than
the ones that went before it because
MOL can leverage all of the previous advances. Improvements made
in the rst year should also continue to grow as additional data is
continually fed back into the system to achieve Kaizen-style gains.

Next steps

MOL continues to mine for ways

to apply analytics to its business.
In 2017, it wants to increase white
product yield by 2.5% through
increased conversion and more efcient crude processing. To increase
its buffer against market swings, it
will additionally improve the diesel
to mogas ratio from 2.4 to 2.8. Flare
gas recovery and hydrocarbon loss
management initiatives, tracked
through continuous improvements
to monitoring and tracking systems, are under way.

Lessons learned

What did MOL and other leading

companies do differently in applying IIoT? There are ve lessons:
1. Do not forget that it is about
delivering business value and not
applying IIoT and advanced analytics for technologys sake. MOL has
seen technology as a means to an
end, not the other way around.
2. Start the journey by creating a
foundation for data. By creating
a digital infrastructure MOL gave
itself a scalable, coherent infrastructure. It created both a virtual
model of its plant through AF and
a means to implement and measure
those results in reality. The infrastructure approach also made it far
easier to develop new functionalities because the same basic foundation could be used for multiple
functionalities.
In the end, MOL created what
one could call an OT chart of
accounts where all OT data gets
aggregated across a portfolio similar to the nancial or IT chart
of accounts which structure has

www.eptq.com

q1 OSI soft.indd 4

been mandated by regulations.

3. Just do it. It is a journey of continuous improvement. Search for
improvements that can be implemented now and add others as
time goes on. Separating problems
can allow plant managers to resolve
individual problems more quickly
as well as document progress for
upper management.
4. Determine where analytics are
performed. Calculations such as
exchanger and pump efciencies,
energy utilisation, and yields or
advanced CBM can and should
be done in the OT infrastructure
closer to the assets. Performing OT
analytics in the OT data infrastructure will also enable the migration
of analytics to the edge over time.
Meanwhile, more extensive analytics that might require thousands of
servers and multiple data streams
are better suited for the cloud. One
way to think of the difference is
that analytics for individual plants
or processes are best conducted in
OT while enterprise-wide analytics
are most suited for the cloud.
5. Bridge OT and IT through
Automation. This can be accelerated
by the use of an integration layer
that CASTs operational data so that
it can be consumed in unstructured
IT systems. These data integrators
effectively automate data preparation and translation.
IIoT, advanced analytics and big
data are here and growing, make
no mistake. They will dramatically
transform our largest and oldest
industries. If you approach their
implementation and use strategically with the approach presented
above, you will increase the probability of value sustainable attainment from your IIoT, advanced
analytics, and big data initiatives.

Level Measurement
LNG & Low Density applications

Meets the design standard

of the LPG/LNG industry:

Includes NACE requirements
Reliable level indication without

external power with three

functions in one device:
Liquid level indication
Liquid level switch
Liquid level transmitter
Equipment with optional bi-stable

change over or on/off switches and

linear or non linear transmitters
(also as 2-wire systems with
420 mA output, optional with HART)
CE-conform, Ex-approval (ATEX & IECEx)
For operating pressures from vacuum up to 50 bar,

operating temperature from 196 C up to +400 C

Densities > 0.27 g/cm3
Closed, non-pressurized floats
Construction in stainless steel grade 316 & 316L

as standard.

References
1 www.iea.org/publications/freepublications/
publication/medium-term-oil-marketreport-2015.html
2 w w w. p e t r o l e u m - e c o n o m i s t . c o m /
articles/corporate/pe-award-winners/2016/
downstream-company-of-the-year-2016mol-group

Casing of indication rail in stainless steel,

float failure indication on request

sit us
Please vi
ebsite:
w
w
ne
r
at ou
ag.ch
aek
.w
w
ww

WEKA AG Switzerland
Schrlistrasse 8 CH-8344 Bretswil
Phone +41 43 833 43 43 Fax +41 43 833 43 49
info@weka-ag.ch www.weka-ag.ch

ARCA Flow Group worldwide: Competence in valves, pumps & cryogenics

Craig Harclerode is the Industry Principal, Oil

& Gas, with OSIsoft.

PTQ Q1 2017 59

15/12/2016 11:40

5th

REFINING
INDIA

PTQ's international conference for India

held in partnership with Industrial Development Services

Lalit Hotel, New Delhi

3-4
18-19 September 2017

ptq
PETROLEUM TECHNOLOGY QUARTERLY

refiningindia.com

ref indiaad.indd 1

15/12/2016 10:46

Profiting from plant data

Predictive analytics is an evolving technology with many potential applications
in and gains for the process industries
DOUGLAS WHITE
Emerson Automation Solutions

ypical plant problems include:

1. You are a process engineer
and suspect that the catalyst
in one of the sites reactors is deactivating faster than it has historically.
Is your suspicion correct, and if it
is, how much faster and what is the
cause? Will the catalyst last until the
next shutdown? If not, what operating changes does the plant need to
make to ensure it will last?
2. You are a reliability engineer trying to understand why some pumps
need frequent overhauls and some
do not. What are the characteristics
of those that repeatedly require this
service? Are there leading indicators that will allow you to identify
beforehand when a pump problem
is about to occur?
3. There have been several significant are releases over the past
few years with associated visible
plumes. The source and composition
of the releases are not obvious. You
have been asked to nd the cause
and recommend changes to avoid or
at least reduce the incidence of these
releases in the future.
What is the common factor in all
of these issues? Their resolution is
going to require deep analysis of
plant data including data beyond
standard process measurements.
Another commonality is that time
is an explicit variable and solutions should generate predictions of
future plant and equipment behaviour that supports decisions. How
can these questions be addressed
more efciently more quickly and
with less manpower requirements?
In this article, some of the new
methods that are available to assist
with this data analysis and development of solutions will be reviewed.

www.eptq.com

q1 emerson.indd 1

Previous

Today
Status information

Diagnostic information
Travel deviation
Cycle counter
Air supply pressure
signal
Output signal

Valve signature
Step response
Dynamic error drive
...etc.

Valve position

Figure 1 Valve feedback output data

In recent years, rapid progress has

been made in innovative algorithms
and approaches. How are the process industries taking advantage
of these new techniques today and
how might they apply them in coming years?
In a related trend, process plants
are producing more and more data.
Big data is a popular topic and the
process industries are no exception.
One rening company indicated it
produced 80 billion data items for
storage in one year from four sites.
Another process company referenced a corporate historian with 10
million tags across 15 sites and plans
to implement online access to the

most recent three years of historical

one minute sampled data for these
points.
Even more data can be expected
from process plants in the future.
Much of the new equipment and
devices purchased for process
plants is already equipped with
multiple sensors to monitor internal conditions and performance,
embedded computing to analyse the
data produced, and enhanced connectivity to transmit the results. This
continuing evolution in the number of sensors and their computing
and communication capabilities is
sometimes referenced as the
Industrial Internet of Things (IIoT).

PTQ Q1 2017 61

14/12/2016 12:40

What is
forecast to
happen
Predictive
analytics

What is
happening
Real-time
analytics

Future

What has
happened
Descriptive
analytics

Present

Past
Figure 2 Analytics time line

But how can plant staff extract value

from all of this data?
Figure 1 shows one example. In
years past, there might be one (or
even zero) feedback signal from a
field control valve. Today there can
be more than 100 including sophisticated diagnostic information and
generated graphical performance
representations.
Of course the increase in data is
not limited to the process industries.
Much larger increases are reported
in other industries. Walmart is
reported to collect and store 2.5
pentabytes of unstructured data
every hour from the more than
one million customers around the
world currently purchasing items.1
Amazon is estimated to have more
than 1.5 million servers in their data
centres.2
In this article the focus is on the
plant operational impact of these
developments and their potential
business impact. One characteristic
of most of these operational issues is
that they have a temporal/dynamic
character which leads to special data
requirements. There has also been
an equivalent and important impact
on other corporate activities such as
trading, customer and market analysis and financial services which
has been covered in many other
publications.

practice of capturing, organising,

and analysing data to determine
patterns, correlations, and conclusions. Engineers, of course, have
been doing analytics, in practice, for
as long as they have been employed.
However, the continuing increase
in computing capabilities has led to
development of practical algorithms
that can address significantly larger
data sets and more heterogeneous
data types.

Refinery

Action

Predictive analytics applications

Algorithms

Acquire
Dynamic event detection
and sensor data capture

Analyse
Extract information
patterns

Prediction
Models

End users
Present:
HMI
Distribution

Data analytics

Data analytics is the theory and

62 PTQ Q1 2017

q1 emerson.indd 2

McKinsey3 has ranked big data

analytics as one of the top potential
technologies that can increase productivity and GDP over the next
few years. Among business sectors,
manufacturing was the lead with
an estimate of overall GDP yearly
increase of $125-$250 billion.
Data analytics can be classified in
three categories based on the time
scale of interest as illustrated in
Figure 2.
Descriptive/historical
analytics
is analysis of what has happened,
when it happened and why it happened. This includes producing retrospective performance measures.
An example is calculation of actual
versus planned energy usage for
plant equipment as well as the overall site for the past day, week or
month.
Real-time analytics is the use of
the most recent and historical data
to interpret current plant conditions.
An example is comparison of plant
operating conditions against the
approved operating window and
issuing an alert if the operating window is being violated.
Predictive analytics is prediction
of what will happen and when it is
likely to happen based on an analysis of current and past data trends
and patterns. An example is estimating the likelihood that a pump will
need servicing before the next shutdown. Predictive analytics is the
focus of this article.

Figure 3 Predictive analytics implementation

Figure 3 shows a typical predictive

analytics application in the process
industries.
Relevant sensor and event data
is captured in real time. They are
automatically analysed for patterns
of significance and if one or more
patterns is detected a prediction is
generated from the model. If appropriate, an alert is generated or specific compensating action is initiated.
An example is the pump condition question at the beginning of this
article. The spectral properties of
the vibration measurement from the
pump and the statistical properties
of the outlet pressure measurement
might be the inputs of significance.
If both of these variables increase
significantly at the same time the

www.eptq.com

14/12/2016 12:40

digitalrefining.com is the most extensive

source of freely available information on
all aspects of the refining, gas and
petrochemical processing industries.
063_dig ref full page ad.indd 1

It provides a constantly growing

database of technical articles,
company literature, videos, industry
news and events.
15/12/2016 09:28

Common predictive model algorithms

Problems types

Plant application
examples

Types of algorithms

Types of models

General application
examples

Comments

Regression

Predict energy use in

plant based on planned
operation and models
Spectral data
interpretation
Video/photograph
characterisation

Generalised linear
regression

Linear continuous
variables output

Marketing campaigns

Still the most common

algorithm used

Neural nets

Convolutional neural
nets

Non-linear
continuous variables
Non-linear
continuous variables

Image recognition

Training requires lots of data

Anomaly detection Fault detection;

product quality tracking

Principal component
analysis (PCA)

Linear continuous
variables

Human gene to
disease correlation

Fingerprint identification;
photo sorting

Classification

Fault detection
Logistics regression
Decision trees
Sorting documents

random forest
Product quality prediction, Support vector
data preclassification
machine (SVM)

Binary outcome Yes/No Fraud detection

Discrete outcome
Recommend movies
A,B,C,D
to watch; music
Binary outcome
Image recognition
Yes/No

Clustering

Spectral data encoding;

characterisation

Nearest neighbour;
K-means,

Clusters similar data

Recommend movies
to watch; music

Natural language
processing

Text sorting

Text analytics

Textual analytics

Email spam
identification

Random forest often most

accurate but hard to interpret
Handles high dimensionality
and large data sets

Table 1

model might suggest a potential

future pump problem and suggest
action by operations and maintenance to address the problem.

Predictive model development

Central to the success of any predictive analytics project is the development of an appropriate model
which accurately forecasts the
occurrences under investigation.
Many of the popular algorithms for
model building have been known to
statisticians for many years. What
has changed is the ability to apply
these algorithms to extremely large
data sets and widespread availability of easy-to-use software for these
applications. This has resulted in
many recent applications in the consumer area including such familiar ones as fingerprint recognition
on smart phones, recommendations for music and movies from
previous selections, and driverless
cars. Table 1 lists some of the major
classes of algorithms and their existing and potential use in the process
plant area and in more general consumer applications. There is much
overlap in terms of potential applications and the approach used often
depends on the preferences and
experience of the user.
For more background on these
algorithms see references 4 and 5.

64 PTQ Q1 2017

q1 emerson.indd 3

Statistical methods for regression have been and remain the most
popular for plant applications with
a further subdivision into those
employing linear models versus
non-linear ones. Principal component analysis (PCA) is a subset of linear analysis that identifies a reduced
set of derived variables from the
data set that approximate the entire
set. These variables can be used to
identify the normal operating region
for a piece of equipment. Deviations
outside this region are often characteristic of faults. For non-linear
models, neural net algorithms have
been gaining importance because
of their ability to model quite complex data such as photo recognition. However, in contrast to linear
regression, the resulting models
have no easy physical interpretation. Classification algorithms are
often used when the results are discrete: binary such as yes/no, or multi-class such as A,B,C or D. Logistics
regression is regularly used for
binary cases with if-then splitting
(CART trees) often used for multi-output problems and large data
sets. Support vector machine algorithms are popular for binary problems with a large number of input
variables and large data sets and
non-linear boundaries. Clustering is
a technique to determine if there are

distinct groups within data sets with

similar characteristics. It can be used
to sub-divide large data sets into
more manageable and meaningful
sub-sets. Natural language processing is used to convert unstructured
text into structured indices that can
then be used for further analysis and
to segregate the source text into relevant sub-groups.
Most applications to date in the
process industries have been in the
equipment
availability/reliability
area with the system intended to
warn of impending faults or of deteriorating performance. These applications will often use a combination
of the algorithms listed above.
For example, clustering can be
used to identify groups of pumps
with similar maintenance histories and these groups can then
have individual predictive models
with resulting higher accuracy than
lumping all pumps together.
The methods discussed above are
primarily examples of supervised
model building. In this case there is
a data set where the output is generally known and the desire is to
determine if the other variables in
the data set can predict the outcome.
An area of current research is what
is termed unsupervised learning
where the system processes a large
amount of data to determine if there

www.eptq.com

14/12/2016 12:41

break through
the

NOISE

Better Signal-to-Noise Ratio Means

Better Level Control Performance

3X
Higher

4.40
Model 706
SNR

1.57

Competitor
SNR
The ECLIPSE Model 706 transmitter has a
signal-to-noise ratio nearly 3 times higher
than competitors.

While transmit pulse amplitude (signal size) has helped to make

guided wave radar technology the standard for accurate, reliable level
measurement, the fact is signal-to-noise ratio (SNR) represents a far
more critical indicator of level control performance. For superior SNR
in all process conditions, no other GWR device beats the Eclipse
Model 706 transmitter from Magnetrol.
To learn more about the breakthrough ECLIPSE Model 706
GWR transmitter visit radar.magnetrol.com or contact your
MAGNETROL representative today.

magnetrol.com 1-630-969-4000 info@magnetrol.com

magnetrol.indd 1

2016 Magnetrol International, Incorporated

14/12/2016 10:38

Unstructured text

Equipment
condition and
performance

Daily operating plans

Targets
Notes, etc.
Email

Relational data
Maintenance
Inspection
Work orders
...etc.

Model
building

Spectral data
Rotating machinery
Vibration
Thermography
Infrared/acoustic/
ultrasonic

Spatial

GIS/location

Operational real-time
and historical data
Process
Offsites
Utilities
Terminal data
Analysers (NIR)
Lab data

Video

Flares
Process equipment
...etc.

What are the expected benefits?

Figure 4 Typical plant predictive analytics data requirement

is any structure to the data or relationships without prior specification

of the desired answer.

Process plants and analytics

Much of the work on analytics has

taken place outside the process
plant environment. Why are plant
problems more difficult?
The first area of difficulty is the
character of the data required. Figure
4 shows typical data inputs for plant
predictive analytics problems. Note
the diverse data types.
For the catalyst example, operational real time and historical data
is required but also types of feedstocks and their analyses. To predict
whether the catalyst will last requires
data on the planned feedstocks, rates
and modes of operation. The pump
example requires time synchronisation of operational data with maintenance records and use of equipment
data sheets and history. As mentioned earlier, time is an explicit variable which means that all of these
records have to be accurately time
synchronised. For the flare example,
how can the historical video footage
of the flare be automatically searched
to identify when there was a visible
plume and coordinate the incidents
with operational data and equipment
work orders at preceding time periods? Much useful information may
be in unstructured text such as operational logs that need to be analysed.
To further complicate the issue each
of these data sources typically has
its own physical storage, data organisation, search programs and user
interface.
For operational real time and his-

66 PTQ Q1 2017

q1 emerson.indd 4

The other issue is that few of the

analytic algorithms were developed
to model dynamic systems, that is
systems that are time varying. As
emphasised here, that is exactly the
type of models that are needed for
most plant problems. Fortunately,
there has been a recognition of this
deficiency and suitable extensions
of the algorithms have been and are
being developed.

torical data, process plants often

have residence times that are significantly longer than typical data
sampling times. For the refinery
crude unit shown in Figure 5, a fluid
element entering the unit can take
one to two hours to exit. If product
quality is measured at the current
time and you want to correlate process conditions with the quality, you
have to use the conditions at the time
the fluid element was transiting in
that part of the plant. To accomplish
this requires selecting the measurements from that part of the plant at
past intervals corresponding to the
residence time between that point
and the sampling point. This is
called time shifting the data. If the
product was stabilised naphtha and
the time of sampling is t, you might
have to use data from the stabiliser
column at t-10 minutes, data from
the crude column overhead system
at t-20 minutes, data from the crude
heater at t-55 minutes, and crude
composition data at t-85 minutes to
properly analyse the unit relationships. Identifying the actual delay
times empirically from the data is
commonly one of the first steps that
has to occur in analysis.
In addition, process data is
natively of poor quality. It is often
corrupted by noise and has poor statistical qualities. It can be auto-correlated, cross-correlated, collinear, and
non-stationary. Data cleaning is the
term used to describe the necessary
and very important data preparation
step in analysis. It includes handling
missing data, removing outliers,
scaling, compensating for non-stationarity, and so on.

To evaluate the benefits from the

increased use of predictive analytics first consider the operating objectives of a typical large continuous
process plant. The goals can be summarised as The Four Zeros:
Safety: zero safety incidents
Sustainability: zero environmental incidents, excess energy use, and
waste
Availability and reliability: zero
unscheduled downtime
Financial optimisation: zero lost
profit opportunities.
How can predictive analytics contribute to meeting these objectives?
Although there are many possibilities, the most common areas are:

Improving safety performance

Avoiding incidents through early

detection of potential hazardous situations multivariable equipment
operating window

Reducing personnel exposure
to hazardous conditions through
enhanced detection of likely hazards.

Supporting sustainability

Comparing current utilisation of

resources, such as energy and other
utilities, to their expected usage
under current conditions with a
determination of possible causes of
variation
Early identification of potential
environmental emission events to
facilitate mitigation.

Increasing plant availability and

reducing unplanned downtime

Anomaly detection identifying

precursor events to unscheduled
equipment outages or problems,
allowing required maintenance
activities to occur before the event
Performance monitoring for condition based and preventative main-

www.eptq.com

14/12/2016 12:41

Stabiliser data at
time (t10) mins
Column overhead data
at time (t20) mins

Reflux
drum

Atmospheric
tower

Stabiliser outlet
sample at time t mins
Off gas

Naphtha
stabiliser

Light
naphtha

Pumparounds
T-6

Crude oil feed quality

at time (t85) mins

T-5

Heavy
naphtha

T-4

Jet fuel /
kerosene

T-3

Preheat

Stripper
columns

Crude

T-2

Diesel
Intermediate
gasoil

Desalter
Heavy
gasoil
Desalter outlet data
at time (t70) mins

T-1

Heater

Steam

Heater outlet data at

time (t55) mins

Atmospheric
residue

Figure 5 Time shifting data

tenance detecting loss of process/

equipment performance before it
impacts production capacity allowing for scheduled rather than
unscheduled maintenance
Reducing the impact of corrosion and fouling on equipment with
enhanced corrosion monitoring and
forecasting
Reducing turnaround duration
and frequency through better knowledge of actual equipment conditions.

Increasing margin and

productivity

Energy and utilities demand forecasting predicting potential contract penalties before occurrence
Detecting and dissecting complex
interacting constraints (sometimes
multi-unit) on production to help
overcome bottlenecks
Increasing the yield of most valuable products by identifying optimum operating conditions

www.eptq.com

q1 emerson.indd 5

What are the typical steps in executing a predictive analytics project?

and the estimated financial or

other benefits. While sometimes
difficult, this is a prerequisite to gaining approval and support from senior management which is important
for eventual programme success.

Problem definition

Inventory and acquire available data

Cost and benefits estimation

Data cleaning, consolidation and

transformation

Determining reasons for product

quality/yield issues.

Typical analytics project

The first step is to develop a clear

idea of the objective of the project.
What do you want to understand and
predict? How will results be used
to improve business performance in
the plant? What are the specific performance measures or metrics to be
used? For the flare example listed
at the start of the article this objective might be: identify the cause(s)
of recent flaring events in the plant
and possible mitigating strategies to
avoid them in the future.
Next, it is necessary to develop an
expected cost for the programme

Identifying, locating and gaining

access to available data is usually
the next hurdle. While process data
in a historian may be easily available, it is seldom all that is required.
For the example, flaring events
may be documented in environmental reports, plant operating instructions in daily instruction documents
and unusual operating situations in
log books. Maintenance issues on
equipment may be in work order
records.

It is often the case that gathering

PTQ Q1 2017 67

14/12/2016 12:42

and preparing the data for analysis

consumes from one-third to one-half
the total project hours. What data
should be collected? How should
it be stored, particularly non-numeric data? How should search
and retrieval of data of interest be
implemented?
Next is the data cleaning phase,
as discussed above. Data needs to be
brought to a consistent format with
a uniform timescale. Missing data
and outliers need to be excluded.
For numeric analysis, scaling is
often required. The better the training data, the better the model.
For the flare example, data time
synchronisation is likely to be
important and consistent data formatting an issue.

Model building and evaluation

As discussed, there are many algorithms that can be used for model
building. Before embarking on complex models and analysis it is often
worthwhile to start with visualisation and simple analysis with
linear models, to see if any relationships seem to exist. Confusion
between correlation and causation is
common and needs to be addressed.
At the end of the project, the model
has to make sense for it to be used
with confidence.
Previously, use of the various predictive analytics algorithms required
significant experience and expertise.
More and more easy-to-use software
is becoming available which allows
those interested to build analytical models without first becoming
an expert. These also often provide
an infrastructure for implementation, which will be a requirement for
eventual use.
Testing and validation of the model
are the next step. Ideally there is
enough data to segregate testing and
validation data sets from the training
data. After the model is developed
with the training data, it is tested
against the test data set. If parameter
changes are then made to the model
the improved model is checked
against the validation data set and
the accuracy reported. This testing and validation step is extremely
important in developing an estimate
of the accuracy of the model.
The flare example is a case where

68 PTQ Q1 2017

q1 emerson.indd 6

Plant 1
Plant 2

Surge
basin

Plant 3
Plant 4
Plant 5

Denitrification
basin

Project review

Plant 6
Plant 7
Plant 8

Clarification
basin

Plant 9
Plant 10

racy and increase uncertainty in the

model predictions. Accuracy should
be audited regularly and model
parameter updates performed when
appropriate.
Current data has to be cleaned and
transformed prior to use by the prediction model just as the training data
was. This procedure is automated in
the production environment.

Filtration

Plant 11

At the conclusion of the project the

actual return should be evaluated
against the initial objective. What
was the actual financial return versus the investment required? What
are the lessons learned and best
practices?

How to get started

Plant 12
Settling pond
Outfall

Figure 6 Wastewater treatment plant

there may be relatively few actual

events from which to build a model.
This typically requires special algorithm approaches.

Model deployment, monitoring and

updates

For the model to be useful it needs

to be moved into the daily decision making process. Who will take
action based on the results of the prediction? What are their needs for a
user interface? Standard questions
include: what infrastructure should
be used?; what tools should be used
for deployment?; should the data
be stored on premises or off-site (in
the cloud)? As with model building,
there has been recent development of
a number of new software offerings
in this area. In contrast to the past,
it is possible to deploy a solution
using these new infrastructures without requirement for an experienced
expert on the specific software. There
is a standardised documentation for
models Predictive Model Markup
Language (PMML) that can facilitate movement from a development
environment to deployed status.
There are often changes in the
plant over time that degrade accu-

If you are considering your first

application, how do you get started?
Generally a phased approach that
is targeted at a specific limited,
but important problem that can be
addressed in a few weeks or months
is better than a big bang type programme. A pilot project that generates some benefits can help build
support for more extensive future
programmes. Finding an executive
sponsor who is interested in the
programme and the results is much
better than conducting an isolated
study. Experienced consultants can
help.

Case studies
1. Reducing effluent emission
violations

A US chemical company installed

biological denitrification in their
wastewater treatment plant to
reduce ammonia in their water
discharge (see Figure 6). The process uses aerobic bacteria to convert ammonia to nitrates. However,
the company experienced occasional breakthroughs of ammonia
that caused the water discharge to
exceed specified effluent ammonia limits. To assist in identifying
issues, they decided to conduct a
data analytics study to see if the
breakthroughs were associated with
specific plant conditions. One of
the issues in the identification was
the large residence time in the system, from one to two weeks for the
total system depending on individ-

www.eptq.com

14/12/2016 12:42

ual plant discharge rates, although

there were intermediate measurements that did identify likely breakthrough. Another was that there
were multiple plants discharging to
the waste water system, each with
different product production schedules. More than 200 measured process variables were considered as
candidates and a thorough correlation study over ve years of data
was performed using PCA. The conclusion of the study was that a drop
in the pH in the efuent from one
of the plants preceded many of the
breakthrough indications by one to
two days. This drop in pH corresponded to a particular type of process operation in the plant.

2. Reducing maintenance, increasing

plant availability

A large Middle East chemical company commissioned a major new

site with over 6000 important
equipment assets requiring regular
maintenance. After initial commissioning of the plant, an integrated
maintenance
programme
was
instituted centred on reliability

centered maintenance and predictive diagnostics. The results were a

12% decrease in maintenance costs
while achieving a 2% increase in
overall plant availability.

Conclusion

In summary, predictive analytics is

an evolving technology with many
potential applications in the process
industries. However, implementation in these industries has distinctive issues due to the time series
character of the data and associated
complex data requirements. Current
applications are demonstrating
benets in process fault detection,
increasing equipment availability,
improving safety performance and
supporting nancial optimisation.
There will be more of these applications in the future.
References
1 McAfee A, Brynjolfsson E, Big Data: The
Management Revolution, Harvard Business
Review, Oct 2012.
2 Clark J, 5 Numbers That Illustrate the MindBending Size of Amazons Cloud, Bloomberg,
2014,
www.bloomberg.com/news/201411-14/5-numbers-that-illustrate-the-mind-

bending-size-of-amazon-s-cloud.html
3 Bloom A, McKinsey on Big Data Analytics:
The #1 Key to US Economic Growth, 2013,
blog.pivotol.io/pivotal/p-o-v/mckinseyon-big-data-analytics-the-1-key-to-useconomic-growth
4 Hastie T, Tibshirani R, Friedman J, Elements
of Statistical Learning Data Mining, Inference
and Prediction, 2nd Edition, Springer, 2008,
Free download at statweb.stanford.edu/~tibs/
ElemStatLearn/
5 Rudin C, Prediction: Machine Learning
and Statistics, Course (free) under
MITOpenCourseware, ocw.mit.edu/courses/
sloan-school-of-management/15-097prediction-machine-learning-and-statisticsspring-2012/index.htm
Doug White is a Principal Consultant with
the Process Systems and Solutions Group of
Emerson Process Management. Previously
he held senior management and technical
positions with MDC Technology, Profitpoint
Solutions, Aspen Technology, and Setpoint
where he was responsible for developing
and implementing advanced automation
and optimisation systems in process plants
around the world. He has published more than
50 articles on these subjects and holds a BChE
from the University of Florida, an MS from
California Institute of Technology, and an MA
and PhD from Princeton University.

EXPOEUROPE
28 - 30 MARCH 2017 - AHOY, ROTTERDAM

Register now qq www.stocexpo.com

Stocexpo2017_HalfpageVIS.indd 1

www.eptq.com

q1 emerson.indd 7

Organised by

30/11/2016 09:05:57

PTQ Q1 2017 69

14/12/2016 12:42

PTQ-FFC_Plus-FY17_300dpi.pdf 1 9/7/2016 12:04:39 PM

CM

MY

CY

CMY

Advanced FIBER FILM Contactor Technology

Reduce
Costs
Improve
Efficiencies

Hydrocarbon Treating Made Better

Improve Treating Rates
Reduce Carryover
Reduce Plot Space

Introducing the next generation of FIBER FILM

Contactor technology. FFC Plus is an advancement to
FIBER FILM technology that makes mercaptan extraction
more efficient than ever. The latest innovations to this
commercially proven technology allow for more than
double the treating capacity while reducing overall costs
and plot space.
FFC Plus can also be easily retrofitted into existing
FIBER FILM units as a drop-in solution that is simple to
install during a unit turnaround, requiring less time than a
standard cleaning. In addition, turndown can be
maintained at current levels. Add capacity for tomorrow
without losing operability today.

Optimal processing scheme for producing

pipeline quality gas
Exploring processing line-ups for optimal design of sour gas processing plants

SAEID MOKHATAB Gas Processing Consultant

GERRIT BLOEMENDAL Jacobs Comprimo Sulfur Solutions

Sulphur
recovery

Gas processing
plant

Raw
gas

Producing
pipeline quality
gas
Condensate

Inlet
separation

Raw
gas

Water

Sulphur

Gas processing
plant

Raw
gas

hu

Su

lp

at

ne

ns

de

ta

Lean gas
to
sales/LNG
plant

on

Bu

an

Producing
natural gas
liquids

Figure 1 Two configurations of gas

processing plants1

of the hydrocarbon liquid products can also be drivers for process complexity, which determines
Tail gas
treating
Sulphur
degassing

Acid gas
removal

Pipeline
quality
gas

Ethane

Pr
op

he gas processing plant must

be a fit-for-purpose design,
meeting project economics
and environmental requirements.
When determining the proper solution set for a gas processing project, it is not only important to
select the proper technology within
each processing unit but also to
adjust sequencing of the processing steps in order to optimise the
whole operation, thereby increasing operating flexibility within the
overall gas processing plant. This
article discusses the considerations
required to determine the optimal
processing line-up for sour gas processing plants producing pipeline
quality gas.
The choice of gas processing
plant configuration and its complexity depend upon the feed gas
compositions and the levels of
treating and processing required
in meeting product specifications
and emission limits. The values

Gas
dehydration

Incineration

the levels of natural gas liquid

(NGL) components to be recovered
(see Figure 1). While contaminants
and sulphur compounds must be
removed to meet emissions requirements, the extent of processing to
recover NGL components is project
specific.1
A typical scheme for a gas processing plant producing pipeline
quality gas is shown in Figure 2. As
can be seen, different gas processing steps are required to meet the
product specifications defined by
the gas processing plants owner.
The number of combinations of
possible processing steps presents
a challenge to determine the best
processing scheme to meet technological and economic targets
while providing flexible and reliable operability. In configuring the
optimal processing line-up, the
plant designer must understand
the technology options available,
their integration opportunities and
Flue gas

Liquid sulphur to
export/solidification
Mercaptans
removal
(if required)

Hydrocarbon
dewpoint
controlling

Gas
compression
(if required)

Pipeline
quality
gas

Off-gas

Condensate
stabilisation

Condensate to
storage and export

Figure 2 A typical gas processing plant producing pipeline quality gas

www.eptq.com

jacobs.indd 1

PTQ Q1 2017 71

15/12/2016 13:33

Sulphur

Sulphur
recovery

R-SH

Regen-gas
treating
(physical solvent)
H2O

H2S, CO2

Feed
gas

Gas dehydration and

mercaptans removal
(molecular sieve technology)

Acid gas removal

(promoted MDEA)

Hydrocarbon dewpoint controlling

(propane refrigeration system)

Regeneration gas

Pipeline
quality
gas

C4+

Figure 3 Processing line-up A

their limitations. Technical risk,

licensor experience, degree of commercialisation, safety, health and
environmental aspects all need to
be weighed along with the process
and economic performance of the
technologies concerned.

Consider interactions between

different gas processing units
Adjust the sequence of gas processing units.
This section proposes three integration schemes of the gas processing steps for a sour feed gas

Proposed gas processing line-ups

Changes in the
product market
continue to drive
improvements in
gas processing
technology

Natural gas processing is often considered a mature industry with little opportunity for improvements
or innovations. However, changes
in the product market continue to
drive improvements in gas processing technology. Experience has
proven that operating costs and
investment constraints are becoming more important when selecting
the proper gas processing solution
for developing more unconventional and stranded gas reserves.
In general, the following steps shall
be followed for determining the
proper technology line-up for a gas
processing project:2
Select the appropriate technology
for each gas processing step

containing mercaptans. Typical

schemes of each line-up have
been illustrated in Figures 3 to 5.
Note that for the selection of the
acid gas removal technology to
remove hydrogen sulphide (H2S)
and carbon dioxide (CO2), differences are only marginal, where

more often the technology selection is driven by the requirement

to remove trace components such
as mercaptans (R-SH) and carbonyl sulphide (COS). Whereas
deep removal of H2S and CO2 is
now mastered, mercaptan removal
from a sour gas is still considered a
challenge depending on the concentrations
involved.
Different
options for the combined removal
of acid gas components and mercaptans have been used in various
gas processing plants. The optimum solution in many cases is
the distribution of the mercaptans
removal capabilities over the acid
gas removal unit (AGRU), utilising
a mixed chemical and physical solvent or a promoted methyl diethanolamine (MDEA) solvent, as well as
the molecular sieve unit (MSU).3
In line-up A, shown in Figure
3, the sour feed gas is sent to the
AGRU utilising promoted MDEA
solvents in which all acid gas components (H2S and CO2) and some
parts of the light mercaptans are

Off-gas to tail gas

treating unit

Sulphur

Sulphur
recovery

R-SH

Regen-gas
treating
(physical solvent)

H2S, CO2

Feed
gas

Acid gas removal

(promoted MDEA)

Gas deydration and

hydrocarbon dewpoint controlling
(propane refrigeration system)
MEG
injection

H2O

C4+

Mercaptans removal
(molecular sieve technology)

Pipeline
quality
gas

Regeneration gas

Figure 4 Processing line-up B

72 PTQ Q1 2017

jacobs.indd 2

www.eptq.com

15/12/2016 13:33

Off-gas to tail gas

treating unit

Sulphur

Sulphur
recovery

R-SH

H2O

H2S, CO2

Feed
gas

Regen-gas
treating
(physical solvent)

C4+
Gas dehydration and
hydrocarbon dewpoint controlling
(silica gel technology)

Acid gas removal

(promoted MDEA)

Mercaptans removal
(molecular sieve technology)

Pipeline
quality
gas

Regeneration gas

Regeneration gas

Figure 5 Processing line-up C

removed from the feed gas. The

sweet gas is then routed to the
gas dehydration and mercaptans
removal unit utilising molecular
sieve technology, and finally to the
hydrocarbon dew point controlling
unit (DPCU) utilising a propane refrigeration system, which
requires greater operator attention
and maintenance than a silica gel
system, but offers greater flexibility.
In this case a side (slip) stream of
cold propane can be used to keep
the feed gas temperature of the
MSU at 40C or below (as required
by most molecular sieve vendors).
Silica gel technology may be a
feasible and competitive alternative for some feed compositions
(not rich in C3-C4 components) and
operating conditions (temperatures
below 45C and pressures above
28.5 bara). Since total pressure drop
across the silica gel bed is very low
(about 0.65 to 0.85 bar), this process
does not require product gas compression. However, this process
may require a pre-cooling system
to achieve a maximum inlet gas
temperature of 45C.4

The Joule-Thomson (JT) expansion method would not be attractive unless the required outlet
pressure is much lower than the
inlet pressure or the inlet pressure
is too high, having the operating
range in the critical region. In the
case of using a turboexpander for
the hydrocarbon DPCU, the outlet gas usually needs supplemental compression to fulfill product
gas pressure at the delivery point,
requiring high power demand,
though some of the power can
be recovered from the expansion
process.
In line-up B, shown in Figure 4,
the sweet gas from the AGRU is
firstly routed to the gas dehydration and hydrocarbon DPCU using
a propane refrigeration system in
which monoethylene glycol (MEG)
is injected to prevent hydrate formation. Although methanol is a more
effective hydrate inhibitor than
MEG at low temperatures, MEG is
typically chosen since it is adequate
for dewpointing temperatures, safer
to handle, and easier to regenerate than methanol. In this line-up,

mercaptans are removed utilising

molecular sieve technology.
In line-up C, shown in Figure 5,
silica gel technology allows the single step removal of both water and
heavy hydrocarbons from natural
gas to meet the required pipeline
quality gas specifications. In this
line-up, mercaptan removal would
be done with a molecular sieve
system.
In proposed processing lineups A to C, tail (off) gas from a
Claus sulphur recovery unit (SRU),
which invariably contains small
quantities of sulphur compounds,
shall be sent to a tail gas treating
unit (TGTU) to remove the residual
sulphur species in order to meet
emissions regulations. In the past,
the most common approach for a
99.8% plus sulphur recovery was
a SRU followed by an amine-based
TGTU.5 In this method, all sulphur
compounds in the tail gas are converted to H2S by hydrogenation followed by H2S scrubbing by one of
the selective amine-type processes
so that H2S-rich gas can be recycled to the inlet of the Claus unit.
To
incinerator

CO2/mercaptans

Lean acid gas

from AGRU

AGEU

H2S-rich gas

Claus
SRU
Sulphur

TGTU
hydrogenating
section

TGTU
absorber

Rich amine solvent

Lean amine solvent

Figure 6 Typical integration of tail gas treating and acid gas enrichment units

www.eptq.com

jacobs.indd 3

PTQ Q1 2017 73

15/12/2016 13:33

SuperClaus process
Steam

Combustion
chamber

Stack

Waste heat
boiler

Incinerator
QC

Caustic
scrubber

ABC

FC

Caustic
Acid Air
gas

Air

Figure 7 Jacobs Superclaus process followed by a caustic scrubber

Therefore, the only emission is

from the CO2-rich vent gas. In this
case, the TGTU can also be integrated with an upstream AGRU or
acid gas enrichment unit (AGEU)
to reduce emissions, simplify operations, and reduce capital cost.1
Figure 6 shows an integrated tail
gas treating and acid gas enrichment scheme in which the AGEU
absorber can be used to selectively
absorb H2S from the lean acid gas
feed, producing an H2S-rich solvent
and a CO2 overhead gas. Since the
AGEU uses an aqueous solution
of a selective amine, the CO2 reject
gas still contains some traces of
H2S and most of the mercaptans as
aqueous solvents have little affinity for these species. Upon passing
the reject gas over the hydrogenating section, most of the mercaptans
(more than 80%) will be converted
to H2S, which could be subsequently absorbed in an absorber
using the same solvent as in the
AGEU.5
In view of cost, complexity and
reliability, it may be worthwhile
to explore alternatives to the wellknown Claus sulphur recovery
plus amine-based tail gas treating technology. One of the interesting options is the combination
of two well established processes;
Jacobs Superclaus process followed by a caustic scrubber (see

74 PTQ Q1 2017

jacobs.indd 4

Figure 7). A number of advantages

can be achieved by this combination: in the forefront are the lower
installed cost and higher reliability.
Investment costs are reduced, the
process is less expensive to operate
and to maintain, requires a smaller
footprint and greatly simplifies
overall operation. By combining
the Superclaus and caustic scrubber technologies, the overall system
can achieve greater than 99.9% sul-

ide content of less than 20 ppm in

the flue gas is achievable without
excessive cost.

The best processing

line-up is dependent
on the initial feed gas
conditions, the treated
gas specifications
and environmental
requirements

Determining the best gas processing line-up is very much dependent

on the initial feed gas conditions,
the treated gas specifications and
environmental requirements. As
an example, if drying and mercaptan removal is necessary, drying
options other than molecular sieve
technology (such as glycol) likely
require an additional processing
unit for the removal of mercaptans.
When the molecular sieve beds
are regenerated, the water and mercaptans desorb into the regeneration gas, which can then be treated
using a physical solvent process to
remove the mercaptans and provide a concentrated mercaptan
stream that can be sent to the SRU.
Removing mercaptans in this way
reduces the quantity of sulphur species sent to the hydrocarbon recovery unit and thus into the liquid
products. This can result in a sig-

phur removal at compelling capital and operating costs. At least

99.0-99.4% of the H2S is captured
and recovered as elemental sulphur by the Superclaus process and
the remaining sulphur is scrubbed
and converted to sodium sulphate
by the caustic scrubber. Because of
the nature of the caustic scrubber
process, the residual sulphur diox-

Line-up selection

In selecting the best gas processing

line-up that satisfies the business
objectives, factors such as plant reliability, operational flexibility, process guarantee, and life cycle costs
must be considered.

Design flexibilities and issues

www.eptq.com

15/12/2016 13:34

EGYPS

BOOTH 34 ABU DHABI

12-16 FEBRUARY 2017

STAND 1E36, HALL1 CAIRO

14-16 FEBRUARY 2017

R ION 017 E 17
HU AT 2 EEC ur20
LP IC ING S, GR ulph
SU DIF DL EN m/s
LI AN ATH url.co
SO H IL iny
D APR s at t
AN & 28 etail

MIDDLE EAST SULPHUR

d
27 nar
i
m
Se

SEE US AT:

THINK SULPHUR THINK SANDVIK

We design, manufacture and
commission equipment for every
aspect of sulphur processing, from
upstream handling and a range of
granulation options to downstream
conveying, storage and bulk loading.
Engineering & consulting
Ecient sulphur degassing
Premium Rotoform pastillation
High capacity granulation
High performance remelting
Handling, storage and loading
Global service and spare parts

A WORLD LEADER IN SULPHUR PROCESSING AND HANDLING

Sandvik Process Systems,

Division of Sandvik Materials Technology Deutschland GmbH, Salierstr. 35, 70736 Fellbach, Germany
Tel: +49 711 5105-0 Fax: +49 711 5105-152 info.spsde@sandvik.com www.processsystems.sandvik.com

sandvik.indd
1
SANDVIK_Sulphur_Ad_viz_2017.indd
1

14/12/2016
02/12/2016 11:09
16:16

nificant reduction in hydrocarbon

liquid treating equipment size, and
can create new technology opportunities in this area (for instance, use
molecular sieves instead of caustic
to clean up the residual mercaptan).
Considering this, process line-up A,
which utilises molecular sieve technology for simultaneous removal
of water and mercaptans from the
feed gas, would result in an appropriate design and would provide
operating flexibility over the feed
gas range.
The volume of mercaptan-rich
molecular sieve regeneration gas
and therefore the physical solvent
cost in line-up A is greater than
that in line-ups B and C. However,
line-up B has some challenges:6
The mixture of MEG and condensed water/hydrocarbon has
high viscosity and requires heating to ensure good separation.
Even then, uncertainty in composition and the extent of hydrocarbon
absorption in the MEG may cause
operational difficulties.
Glycol has some affinity for
hydrocarbons and co-adsorbs some
BTEX (benzene, toluene, ethylbenzene, and xylene) compounds that
will end up in the regeneration
vapour stream.
Operation is completely dependent on the propane refrigeration.
When the propane system is unavailable, there is no dehydration.
In fact, the purpose of the injected
MEG is not to dehydrate the
gas but to prevent formation of
hydrates.

Hydrocarbon
dewpoint
controlling

Feed
gas

MSU regengas treating

absorber

AGRU
absorber

Dehydration /
mercaptans removal

Pipeline
quality
gas

Figure 8 Feed gas pretreatment using silica gel technology7

There is considerable loss of
glycol caused by its solubility in
hydrocarbon condensate.
Rich MEG sent to the MEG
regeneration system will add to the
system cost.
In line-up C, a slip stream of the
gas containing mercaptans (from
downstream of the silica gel unit)
is used for regeneration of the silica
gels that would reduce the lifetime
and performance of the system.
Also, the silica gel adsorption times
are usually in the range of dozens of minutes up to 2-3 hours,
which on the other side reduces
the expected life time of the adsorbent. The other disadvantage is the
high-pressure adsorbent vessels,
which can be expensive.
As mentioned above, one of the
main parameters affecting selection of the processing line-up is
initial feed gas composition. For
example, if the feed gas contains
a highly aromatic corrosion inhibitor or elemental sulphur solvents,
these can cause heavy hydrocarbon condensation and foaming
problems in the AGRU that prevent the unit from performing
at its design capacity. This may
impact the entire availability and
capacity of the plant. In these
cases, it is necessary to pretreat the

feed gas to remove heavy hydrocarbons. The most commonly used

option for such a purpose is using
silica gel technology,7 where heavy
hydrocarbons are usually removed
to such an extent that the requirement for downstream hydrocarbon
dew point controlling is eliminated
(see Figure 8). It should be stressed
that the treated gas hydrocarbon
dew point increase will become
more severe for sour feed gases
with a high acid gas content,
which also affects the silica gel
performance.7

Licensors preference

Leading technology licensors commonly take an integrated treatment scheme presented in line-up
A for optimal design of gas processing plants with a sour feed
gas containing mercaptans.8 For
example, Figure 9 illustrates the
overall arrangement of a typical integrated treatment package offered by Shell in which a
mixed physical/chemical process
(Sulfinol) is used for the AGRU.
This integrated treatment line-up
can deliver enhanced flexibility
and operability of the whole gas
processing plant. However, the
best choice for hydrocarbon dew
point control needs to be selected

Treated
gas

MSU

Sour
feed gas

Acid gas
removal

AGRU
regenerator
Mixed solvent

AGEU
absorber

AGEU
regenerator
Aqueous amine
solvent

Offgas

TGTU

Incinerator

SRU

Liquid
sulphur

Flue
gas

Figure 9 Overall arrangement of the Shell optimised solution with respect to maximisation of treating train operability and flexibility9

76 PTQ Q1 2017

jacobs.indd 5

www.eptq.com

15/12/2016 13:34

Process importance

Indicated power, %

obvious or intuitive rst order

for-fast-moving-innovators-a-threat-to-slowimprove
yield
profiles
and
and
tighter
several
into
process H,that
typically
effects
onby
thecontrol
process.
Journal,
29
JanJ,a1979.
for
each
project
inisof
order
to operatobtain
Burlington,
MA,
2015.
9 Klinkenbijl
Grootjans
Rajani
J B, Best
increased
delta
coke
at a given
stripping
steam
rate
or
contents.
Blending
also
used
to reduce
the naphthenic
Keystages
variable
modifications
to
the
plant
topreserve
be
able
reused
re-routing
the
sample
of
theUSA,
seven
rules
ofBoth
sedimentation
Parameters
affecting
the
developand
microbiological
aspects
of Summary
biofilm-metal
interface.
aeromoving-dinosaurs
No.
1
100
References
system
integrity.
As
this
field
grows
ing
parameters
to
reduce
variability
3
Respini,
Jones,
Spanu,
Sesselego,
Avoiding
foulof
seal
drum
off-gas
contains
10-40%

Multiply
the
volume
of
air
by
its
result
in
better
handling
of
fouling,
which
provided
refinery
personnel
2
Basham
S,
Luinstra
E,
The
effects
runs
at
70-90%
load
of
the
rated
2
Liebert
R,anaerobic
Flexible
Processing,
LNGhydrogen
Industry,
Practice
for
Deep Treating
Sour
Natural
Gases
Ultimately
this
gives
the
operator
the
most
cost-effective
and
fit-fora
higher
in
coke
measurement
through
heat
acid
content
of the
feed
and
reduce
the
corrosion
rate.
to
maintain
operation.
The
modifilines.
ment
of
biofilms
include:
microbial
corrosion.
bic
and
organisms
play
Step control
4
Industrial
Internet
Insights
Report
for
2015,
1
Troubleshooting
Vacuum
Systems,
Wiley
play,
Hydrocarbon
Engineering,
Nov
2006.
methanol
on
the
Claus
process,
Brimstone
in
industry,
and
specifically
in
refinimproved
fractional
salting
from
S.
The
fatal
concentration
ofto
H
S capacity.
H
density,
to
obtain
pounds
air
and
the
same
time
with
the
right
information
be
39-43,
May/Jun
2013.
(toThis
LNG
and
GTL),
presented
at
the
GasTech
When
considering
pneu90
apurpose
much
greater
understanding
ofmayan
gas
processing
line-up.
balance
calculations.
delta
coke
is especially
detriBlending
two
different
feedstocks
lead
tocould
incomcations
bemethods
as
simple
as
only
One atof
the
advantages
ofincrease
being
2 important
2

Temperature
of
the
system
or
Microbial
activities
act
asflow
aofdrivrole
in
the
initiation,
Key
variable
Reverse
control
Accenture,
2014,
www.accenture.com/us-en/_
Rule
Parameter
Higher
more
Mitigation
Energy
savings
Publications.
Sulphur
Symposium,
Vail,
Colorado,
Sept
2000.
3
Mokhatab
S,catalyst
Hawes
P,mental
Optimal
Mercaptans
2005 Conference
and
Exhibition,
Bilbao,
Spain,
eries,
opportunities
for
data
over
40%
to80
6.5%.
is
0.1%.
Fresh
air
equipment
flow
per
second
HVGO
directly
(less
gasoil
recycled
able
tonew
constantly
keep
the
at
matic
suction
valve
unloaders
and
No.
2unit
the
process
and
how
to
maximise
due
to
the
high
hydrogen
content
that
results
patibility
of
the
heavy
crude
constituents
(asphaltene)
in
changing
to
requiring
the
able
to
reuse
both
the
feed
preparaambient
temperature
ing
force
for
biocorrosion.
propagation,
and
inhibition
of
2
Process
Equipment
Malfunctions,
McGrawacnmedia/Accenture/next-gen/reassemblingsediment
Operation
Catalyst
Design
Ideal
control system
Matteo
Virzi
is Proudfoot
Seniorcontrol,
Technology
3
Ismailian
F,
G, Ng aK,Manager
Pushing
Removal
Solution
in conversion
Gas
Processing
Plants,
14-17
Mar
2005.
scientists
and
are
emerg3.
aparaffinic
set
of
defined
constraints
asAs
wash
oil).
Figure
17
shows
the
the reason,
best
process
severity
for
any
content
has
increased
when
obtaining
samples
is spillback

theTemperature
N
valve
fourprocess
performance.
Effective
use
incomplete
increased
steam
production
in past
the the
regenerator.
the
more
light
crude
oil;
for this
there
Conclusion
1Given
Yes
Lower
High
asphaltene
Mix
temperature,
plenum,
cup
installation
ofanalytics
athese
new
tion
section
incinerator
is
2 and the
Hill.
industry/pdf/Accenture-Industrial-Internet
Water
flow
rate
surface
Microbiologically
induced
corrocorrosion
in
different
ways
by
their
70
with
ISAB
Priolo
refinery
in
an
the
Claus
envelope.
10%
H2SItaly.
acidHe
gasistwo
with
presented
at
the
GPA
Europe
Spring
Conference,
ing,
with
the
potential
to
change
the
around
any
or
several
operating
trend
of
wash
oil
rate.
processed
feed.
by
20%
(25-5%),
you
may
assume
advisable.
2
Incompatible
Yes
Minimise
None
HPS
cyclones,
aromatic
for
quench
cylinder
compressor
with
of
data
analytics
platform
can
Changing-Competitive-Landscape-Industries.
causes include
or
damaged
steam
are
many
models
claimed
predict
3
Lines inplugged
J distillation
R, Understanding
ejector
systems
Comparison
of to
the
proposed
altersmaller SRU next
toPossible
the oldderiving
existing
thethis
possibility
design
atorelatively

Nutrient
availability
sion
causing
organisms
areincompatibility.
different
energy
expert
and
thermal
conversion
60
significant
aromatic,
LRGCC
proceeds,
Feb
Hamburg,
Germany,
22-24
Apr
2015.
Saeid
Mokhatab
is
an
internationally
3 Asphaltene
conversion
No optimised
Maximise
High
asphaltene
conversion and
Mix temperature,
plenum, cup per compression stage
The
decrease
in
wash
oil
rate with
face
of
refinery
operations
usher
parameters,
calculate
an
that
the
air
flow
above
cylinders
pdf
necessary to troubleshoot vacuum distillation,
help
mitigate/minimise
potential
lines/nozzles,
fallen
refractory
restricting
theyears
steam
High
asphaltene
content,
together
incompatibilnatives
for
most
sour
prolarge
unit.
The
reuse
ofMixexisting
compact
SRU
thatcalculated
can
begas
skidded.
Surface
of the
substratum
sulphate
bacteria
processes.
more
than
20
of
2003.
ppreducing
No Limit
ofThe
~1800pathways.
psia
None
temperature,
plenum,
cupWith
4
H50
4
Zoccante
M
L, of
Di Campli
Y, Adsorption
recognised
gas processing
consultant
whonow,
has
2 can
was
continuously
optimised.
in
a
new
era
achieving
operacase
that
achieve
multiple
5
The
25
best
jobs
in America
right
Conclusions
is
about
20%
of
the
non-condensiGraham
Corp.
Technology
Bulletin,
Batavia,
would
be
required
tooperations
enable
at
future
that
impact
on
flow,
orHPS
damage
to
the
stripper
internals.
Damaged
ity,
lead
sludge
precipitation
storage
tanks.
equipment
will
largely
depend
on
Jacobs
prepared
a leads
preliminary
cessing
requirements
toLow
the

pH
of
water
in
the
system
(Desulphovibrio,
experience
in
technology,
and
Fungi
and
algae
may
also
be
5 willproblems
Coketo Desulphotomaculum,
Yes
tempin
High
asphaltene
conversion
cyclones,
mix
temperature,
plenum,
cup,
aromatic
for
quench
4
Paskall
H
G,
Reaction
furnace
chemistry
and
Alternative
for
Lean
Gases
Dewpoint
Control,
been
actively
involved
in
several
large-scale
according
employees,
Jacquelyn
Smith,
40 asanalysed
HVGO
was
on
areducdaily
NY, US.
tional
calm
in
theconversion
face
of are
increasing
goals,
such
reducing
fractional
The
examples
cited
here
acyclones,
tiny
ble
flow.
Therefore,
multiply
25%to
load
increments
to
run,
steam
lines
maybe
identified
by
measuring
the
flow
ITW
can
help
in
solving
the
above
by
margins,
such
as
shutdown
or problem
automation,
he
holds
aprojects,
degree
infor
chemical
6 gas
Yes
None
High
asphaltene
HPSReid
temperature,
plenum,
cup,
aromatic
quench
the known
turndown
limitations
of mixleast
layout
ofSolubility
the
plant
with
common
recommendation
of
the
integrated

Effectiveness
of
biofouling
remeand
Desulphomonas
sp.),
iron
involved
inusing
metal
In
operational
modes,
Sulphur
Recovery
textbook,
presented
at
the
59thdeterioration.
Laurance
Gas
gas
field development
concentrating
Business
Insider,
20
Jan 2016,
www.
4
Putman
R
E,
Steam
Surface
Condensers,
basis
with
respect
to
the
level
of
variability
and
tighter
regulations.
salting
below
5%
while
maximising
segment
of
the
vast
number
of
the
pounds
of
air
per
second
7
Aromatics
No
Maximise
None
HPS
cyclones,
mix
temperature,
plenum,
cup,
aromatic
for
quench
relatively
efficiently,
at
loads
30
engineering
from
the
University
of
Palermo.
through
line. dial
Damage
topre-commissioning
stripping
may
patented
asphaltene
stabilisers
in the
crude
blend.
deferment.
equipment,
which
are
back each
burning
shells
for
the
converters
and
treatment
scheme
presented
Western
Research,
1988. internals
measures.
ing
bacteria
(Gallionellea
ferrugine
fuel
and oil
storage
tanks,
fungal
Conditioning
Conference,
Norman,
OK,
USA,
on design,
and start-up
businessinsider.com/best-jobs-in-americaASME Press, 2001.
contaminants,
using
the
VCI
techthe
amount
distillate
problems
that
steam
ejector catalyst
calculated
above
by
approximately
Email:
mvirzi@isab.com
between
70%
and
However,
cause
a restricted
flow
resulting
the
need
to
Additionally,
when
accumulates
into
the
tanks,
for
burners,
channelling
for
catalyst
condensers.
line-up
A due
to
itssludge
practical
and species
20 of
formation
isinfor
the
result
and
Ferrobacillus
sp.),
acidproduced.
producsuch
asbeset
aspergillus,
penicil22-25
Feb
2009.
ofBiofilm
processing
plants.
He 75%.
has
presented
onof
gas
right-now-2016-1
5
Applied
design
chemical
&
Marco
van process
Son is the
Technical
Manager
of
Acknowledgements
nique
that
measures
the
coke
This
allows
advisory
control
for
the
Table
1
and
surface
condenser
perforfour
between
100%
and
76%
the
open
up
the
spentan
catalyst
slide
valve
forOilload,
theand
same
patented
ITW
Cleaning
can belium
used
to
The
data bacteria
challenge
beds
andremove
sulphur
fogging
in That
the
financial
benefits.
In fact,
the treat5
Bloemendal
G, There Are
More
Roads
accumulation
process

not
processing
technologies
worldwide
and
has
ing
(Pseudomonas,
and
fusarium
may
grow
on
Marco
Respini
is
a
Senior
Technology
Expert
10 Online
6
Low
Oil
Prices
Prime
Gas
petrochemical
plants,
Ejector
&
Vacuum
Comprimo
Sulfur
Solutions
with
Jacobs
Canada
3,4,5
The
author
would
like
to
recognise
the
particles
entrained
within
the mance.
crude
unit
overhead,
to scheme
enable
There
is
always
an

If data
the
molecular
weight
of
the
Temperature
Pressure
Flow
dispensable
process
gas
needs
flow.
If this
occurs,
there
is
a nearly
high
likelihood
precipitated
sludge,
without
the
need
to
open
the
tank
Lead
toof
Rome,
Which Ishistorical
the
Bestproduce
Route
for
Conclusion
authored
or
co-authored
250
technical
case
condensers.
However,
even
As
analytics
transforms
the
with
Baker
Hughes
Downstream
Chemicals,
necessarily
uniform
in still
time
or
ment
processes
usedBacillus),
inNPRA
this
Aerobacter,
and
and
fuel
components
and
Industry
for
IoT
Adoption,
John
Systems, 218-222.
Selection
and
Compatibility,
Ann.wash
Mtg.
Inc, Calgary,
Canada.
has close
to
20 Ward,
years
Correlations
from a Commercial
Unit, 1988
Robert
Ohmes
is sensors
aHe
Principal
Consultant
sensors
0 isnot
significant contributions
ofsensors
Dr Collin
Cross,
HVGO
(thus
removed
by
meeting
operational
goals
while
understandable
temptation
to
cracked
gas
different
from
the
to
be
recycled
through
the
spillYou?,
presented
at
the
2013
Sour
Oil
&
Gas
specialising
in
refinery
and
petrochemical
publications.
He
has
held
technical
advisory
that
there
is
increased
flow
on
one
side
of
the
stripper
and
require
manpower
to
manually
clean
it.
This
will
Forbes
Business,
1
Feb
2016,
/www.forbes.
As aAM
result
theleading
continued
with
turndowns
close
to
10:1
it
may
industry
at of
increasing
pace,
space

that
starts
immediately
sulphur
oxidising
bacteria
acids
which
corrode
are
based
on
0an
10
20 technol30deple40 carboxylic
50
60
70
80
90
100
2011
11-46.
experience
with
sulphur,
sour
water
and
NPRA
MtgHenry
3/20-22/88,
AM-88-59.
DavidAnn
Mason,
Grichuk,
and Steve Bakas with KBC Advanced Technologies, Houston,
oil). of
avoiding
impact
to
reliaconsult
the
equipment
vendor
for
molecular
weight
of
air, Commercial
multiply
back
valve.
process
inworst
Advanced
Technology,
Abu
Dhabi,
UAE,
24-28
positions
leading
journals,
camine
om
smaximising
i timprovements
efor
s / s aIn
pThe
/ 2this
0 1mass
/fraction
0metal
2fouling
/flux
0 1case,
/ l in
ocontrol.
wseveral
-the
oup
iland
less
on
side.
rate
in
oral,negative
dramatically
downtime
and
tion
existing
sour
wells
in
stilltheir
be
possible
to
utilise
aflow
lot
of another
Figure
3 Relative
importance
process
variables
to/ His
MGO
many
companies
struggle
toknowtake
after
immersion
of6professional
(Thiobacillus
sp.). proprietary
iron.
ogy
licensors
2avoid
Sayles
et
Texaco
H-Oil
Unitreduce
Capacity,
%
technologies
and
has
published
Texas.
primary
responsibilities
are
profit
for
help
in the
providing
dataofand
feedback
Also,
the
rate
of
antifoulant
on
2 engineer
bility
due
to
salt
formation.
He
has
15
years
of
refining
experience
and
advice
and
guidance,
when
trying
the
pounds
of
flow
by
the
ratio
of
prices-prime-oil-and-gas-industry-for-iotMar
2013.
societies,
and
conferences
in
the
field
of
gas
to
24%
of
the
rated
compressor
Norman
Lieberman
is
a
chemical
.
If
the
conventional
strippers
is
about
1650
lb/min/ft
concerns
related
to
conventional
tank
cleaning
Alberta
cost
thethis
existing
equipment.
the end, aqueous
advantage
of the
new
opportunities
Operation,
Japan
Pet.current
Inst.ubiquitous
Pet.
Ref.prohiConf.
environment.
TheMSc
growth
Microorganisms
are
in Scott
how
thatand
take
the
feed
stream
and
papers on these
topics. He
holds
degrees
Sayles
a PrincipalIn Consultant
improvement
initiatives,
hydroprocessing
unit
for
work. was
the
wash
oil
was
optimised
and
isadoption/#469b81a75ce2
currently
involved
in are
developing
newof
S,
An
Optimum
Line-up
for
Sour
6
Mokhatab
processing,
and
has
received
a
number
to
troubleshoot
steam
ejector
the
molecular
weight
of
the
who
specialises
in
troubleshooting
refinery
power
is
wasted.
Additionally,
stripper
nozzles
are
damaged,
there
temporary
techniques.
(10/19-20/88).
bition
to
develop
new
sour
gas Role
the KBC
combination
of acidconnecting
gas
flow, of
in
chemical
engineering
fromwith
thetoTechnical
productivity
Hub
andand
its
af
liation
Robert
it
affords.
Data
science
professionals
with
Advanced and
Technologies,
Houston,
of
biofilms
consulting,
consulting
biofilm
isorganisational
considered
be for
a
nature
and
grow
at very
rapid
rates
deliver
the
required
end
product
7international
Key Performance
Management
work
A
technologies
for
improving
refinery conversion
increased
when
needed.
Gas
Processing,
PTQ,
Gas,35
19-24,
2010.
awards
in
recognition
of
his
non-catalytic
process
equipment
and
in
performance
problems.
Typically,
non-condensible
gas,
divided
by
Figure
6sludge
Comparison
of
thethat
compressor
power
consumption
between
step
control
and
installed
fixed
clearance
pockets
References
3Conclusion
Kunnas
etrequired
Mitigate
fouling
in
ebullatednozzles
that
help,
opening
up
the
reactor/stripper
The
isand
transformed
a fully
reusable
prodUniversity
Eindhoven,
The
Netherlands,
and
in
Texas,
with
more
than
years
of
domestic
and
international
clients.
He
holds
fields,
itwater
isal,
expected
more
and
acid
gas
and
the
presexperts.
provides
resources
and
Gordon
University
is
bringing
with
skills
are
ininto
great
The
termItcomposition
biofilm
refers
torefinery
the but
result
of
complex
processes
involvin
soil,
air.
They
show
in
anthe
optimal
manner.
However,
Comprehensive
Look
At
Performance,
Sam
processes.
With
extensive
experience
in
7 Zhang
I,
Graham
C,
Nielsen
D,
Schulte
in
the
natural
gas
industry.
retrofit
process
design.
He
has
taught
his
reverse
flow
control
as
aProc.,
function
of evaluthe
compressor
capacity
1
Underutilization
of
refinery
Connecting
data
with
a sludge
common
this
does
help.
Only
field
the
molecular
weight
of10/10.
air
(29).
would
improve
the
situation
bed
hydrocrackers,
Hydro.
refining,
engineering
and
gas
processing
from
and
petrochemical
experience,
ranging
from
degrees
from
Kansassolution.
State
University
(BSChE)
and performing
istransport
the
best
uct
and
the
recovered
can
beGonzalez
reprocessed
more
gastolerance
plants
will
have
to
ence
ofR,
willrepairs
ofcontaminants
microbial
commuing
of 2012
organic
and
extreme
to
varying
envifunding
tonot
kick-start
projects,
together
industry
and
universities
demand
and
short
supply
and
there
may
be in
some
cases
where
a development
Lordo,
Nalco
Champion,
Q&A
and
asphaltene
related
problems
inAFPM
oil
production
D,
Debest
M,
Impact
oftoMagazine,
Sulphur
Solvent
or slightly,
Gerrit
Bloemendal
is(MBA)
a technology
specialist
Troubleshooting
Seminar
since
1983
to
Conclusions
automation
systems,
Fuel
Dec
2015,
knowledge
management
platform
4
Sayles,
Romero,
Understand
the
differences
investigation,
experience,
careful
See
Table
1
for
an
actual
set
of
test
ENSPM,
Rueil-Malmaison,
France.
but
the
control
logic
refinery
plant
manager
research
engineer.
and
Tulane
University
and
is
a
licensed
without
any
concern.
ate
their
SRUs
for
future
turndown
determine
what
may
or
may
not
be
nities
on
submerged
surfaces
in
inorganic
molecules
and
microbial
ronmental
conditions
such
as
acidic
deliver
skills
through
education
with
world-class
research
in
informuch
of the
expertise
in advanced
techno-economic
analysis
of
alterTechnology
Forum.
and
refining,
he
is
an
inventor
of
five
US
patents
Corrosion
Inhibitor
Solvent
Injection
on
with
Jacobs
Comprimo
Sulfur
Solutions
over
18
000
engineers
and
experienced
VisTec
is
a
mark
of
Baker
Hughes
Incorporated.
39.
This
article
describes
an
example
of
between
thermal
and
hydrocracking,
Hydro.
A
Shashank
Gujale in
isLouisiana.
a complicated
SeniorEngineer,
Process
allows
improved
accelerHis
technicalAs
areas
ofexperimentation
expertise
include
professional
engineer
measurements
and
results
from
low
sulphur
refinery
become
more
automation
allows
for
precise
of
the
compressor,
the
number
of of
Charles
Radcliffe,
Senior
Technical
Service
operation.
InapH,
order
toprove
evaluate
the
possible.
well,
the
presence
of would
aqueous
environments.
Biofilm
cells
toApublished
the
surface,
adsorption
ofin
and
alkaline
low
and
higher
programmes,
and
help
develop
the
matics
and
computer
science.
(The
analytics
resides
invisibility,
academic
instinative
schemes
may
to
be
and
has
10
technical
papersProject
and
Treated
Gas
Hydrocarbon
Dewpoints
andApril
Gas
in Leiden,
the
specialising
plant
operators. The
first
of
his
eight
books,
Jeffrey
Zurlo
isPrincipal
Senior
Strategic
2 Big Success
withand
Bigebullated
Data,
Accenture,
high9/11.
conversion
visbreaking
with
Proc.,
Technologist/Specialist
with
Comprimo
operation
and
design,
bed
residual
Rick
Manner
istoaNetherlands,
Consultant
with
ates
response,
and
enables
deeper
appear
to
offer
any
hope
of
a
vacuum
tower.
and
access
the
compressor
forin
plant
control
fine-tuning
of
cylinders
per
stage
can
be
reduced.
Europe,
Johnson
Matthey
Process
Technologies,
Charles.
options
for
these
units,
it
is
imperaacid
gas
enrichment
in
a
plant
or
the
physic-chemical
molecules
to
the
surface
and
initial
temperatures
as
shown
in all
Table
1. influences
seven
conference
papers
on
visbreakers
and
local
ecosystem
by
building
a
coheData
Lab
has
other
hubs
tutions
and
research
laboratories.
superior
after
considering
techTreating
Unit
Performance,
presented
at
the
amine
treating,
sulphur
recovery
and
tail
gas
Troubleshooting
Process
Operations,
has
Manager
with
GE
Water
&
Process
2014, www.accenture.com/us-en/_acnmedia/
increased
runa calibration
length
when
Sulfur
Solutions,
Jacobs
Canada
Inc., Calgary,
5analysis
Sayles,
Bailor,
Catalyst
addition
in ebullated
hydrocracking,
hydrotreating,
FCC,
and maintenance
KBC
Advanced
Technologies,
Parsippany,
New
Q
toair
improve
refinery
operasuccessful
outcome.
The
ejectors
If
using
as
gas
is process
impeded.
However,
parameters,
such
as and
gas
Hence,
the
capital
costs
and
footRadcliffe@matthey.com
We
have
had
a fall-off
inwhat
steam
stripping
in
our
heavy
fuel technology.
oil With
stability
problems.
Aexperience
graduate
tive
to
understand
the
anefficiency
existing
tail
gas
unit
could
save
interactions
between
metal
attachment
of
microbial
cells
Microorganisms
and
their
metasive
data
science
community.
Edinburgh
and
Glasgow.)
Scottish
innovation
centre
The
nical
and
economic
factors.
87th
Annual
GPA
Convention,
Grapevine,
TX,
treatment
He
has
published
articles
been
continuously
in
print
for
35
years.
He
Technologies.
over
25
years
Accenture/Conversion-Assets/DotCom/
Canada,
where
has
beenstage
involved
inover
various
bed
units,
PTQ, to
Q2argon
2005.
practical understanding
of mostratio
processes.
Jersey.
He
has he
been
with
KBC
for
27
compared
typical
visbreaker
run pressure,
tions,
reliability,
and
profit
potential.
should
and
will
perform
on
the cause
unacceptable,
or
helium
may
two
cylinders
per
still
hydrogen/feed
and
print
of
compressor
are
of
University
with
a degree
inare
industrial
References
FCC
unit
with
carryover
of
hydrocarbons
into
the
The most
likely
for
aby
decline
in
stripping
effilimitations
ofresulting
the
existing
equipthe
day
when
it processing
comes
to enhances
the
turnenvironment,
frequently
followed
their
irreversible
adhebolic
products,
including
previously worked for Amoco Oil in Whiting,
USA,
2-5
Mar
2008.
onMilan
low
temperature
TGTU
catalyst
and
has
Hydrocarbon
compaData
Lab
isthe
focused
onenzymes,
helping
in
process
engineering,
refinery
treatment
Documents/Global/PDF/Industries_14/
sulphur
recovery
unit, amine
recovery
unit
and
6Bringing
Sayles,
Ebullation
Factor,
Hydro.
Eng.,
3/06.
He
is
a
member
of
the
American
Fuel
and
years,
primarily
specialising
in
hydrocracking,
lengths.
Professor
John
McCall
is
Director
of
the
Smart
dark
data
to
light
reveals
manufacturers
performance
be
employed,
but
is
a fewer
time
chemistry,
he
has
been
a through
Research
Fellow
in
required
asservices,
compared
with 1980,
stepless
flow
rate.
minimised.
Furthermore,
1 Petralito
G,and
Respini
M,
optimal
regenerator.
Causes
andthis
solutions,
please.
ciency
cancurves.
be
divided
into
two
areas:
are Characterization
and
what
the
regulatory
down
situation.
Indiana
and
Texas
City,
and
8 Mokhatab
S,a
Meyer
P,Achieving
Selecting
Best
given
presentations
on
international
seminars.
corrosion,
leads
tofootprint
deteriorasion
facilitated
by
production
of
sales
and
hydrocarbon
product
exopolysaccharides,
organic
and gas
Accenture-Big-Data-POV.pdf
nies
have
strong
in
industry
to
capitalise
on
data
sour
water
stripper
debottlenecking
studies,
7ment
Wisdom,
of
Asphaltenes
Petrochemical
Manufacturers,
holds
a
BS
hydrotreating,
hydrogen
processes,
hydrogen
This successful
result
was
made
Datafield
Technologies
Centre catalysts
at Robertand
Gordon
the
of organometallic
is the
a
trends
that
would
otherwise
go
severity,
PTQ,
Q1,
2010.
The
main
ingredient
needed
to control.
consuming
option.
But
take
care
to
The
following
example
explains
cylinders
mean
mainteReferences
1.
Process
related:
until 1983 was Refinery Manager for the Good
Technology
Line-up
for
Designing
Gas
He graduated
from
Utrecht
University,
applications
and
technical
support,
he
requirements
arereduced
inMogas
the
case
of tion
3visbreaking
Industrial
IoT:
An
Opportunity
for
Fastof
the
metal.
Biofilm
induces
extracellular
polymeric
substances
inorganic
acids,
and
volatile
Scotland
and
are
already
science.
Itby
aims
to
revolutionise
and
inreforming,
theand
development
of
process
design
During
Hydro
Conversion,
3rd
Eb Bed
degree
in chemical
engineering
fromworking
Michigan
plants,
isomerisation
and
alternate
possible
coupling
Baker
Hughes
University
incumbent
Chair
of
The
Data
registered
professional
chemist
in
Italy.
He
is
2
A
Phase
Separation
Kinetic
Model
for
Coke
A
unnoticed,
drives
visibility
to
the
achieve
this
result
is
the
dedication
select
a
screwed
fitting
size
that
The
same
application
achemical
stepthe
impact
on
the
compressor
nance
costs,
shorter
1
Mokhatab
S, July
Poe
W
Makammonia
J shutdown
Handbook
Hope refinery in Louisiana. He graduated from
provides
marketing
and
technical
to
Processing
Units,
atan
the
GPA
Europe
Netherlands,
degree
inwith
chemistry,
and
SUPERCLAUS
isData
apresented
mark
of
Jacobs.
CliffConf,
Avery,
Global
FCC
Process
Albemarle,
a.
An
increase
in
will
have
a bachelor
significant
effect
Moving
Innovators;
ain
Threat
to
Slow-Moving
lower
capacities.
In
most
cases,
many
changes
the
type
and
(EPS).
Once
attached,
the
organisms
compounds,
such
or Advisor,
packages.
He with
holds
aHe
ofsupport
Hvy
Oils
2007.
Technological
University
and
MS
degree
in pressure
fuels
technology.
He aholds
BS
MS
degrees
in
with
The
Lab
and
Scottish
Scottish
industry
inA,as
how
itY,develops
Lab aEducation
Board.
isNACE.
aand
founding
director
VisTec
anticoke/antifoulant
treat- arrangement
also
memberTechnical
of
ACS
and
Formation,
Preprints
ACS,Conference,
Div.of
Petthe
Chem,
38,
current
of
system
performance,
of
Natural
Gas
Transmission
&
Processing,
and
the
determination
field
will
notisstate
greatly
impact
the
vacuum
Cooper Union in New York City in 1964.
less
full-range
flow
control
system
based
on
the
chosen
periods
for
overhauls,
and
fewer
the
refining
and
gas
processing
industries.
Sour
Gas
Processing
Sitges,
from
Twente
University
with
a
degree
Dinosaurs,
Industrial
IOT/Industrie
4.0
cliff.avery@albemarle.com
on
stripping
efficiency.
Possible
causes
are
overhead
there
likely
an
optimum
with
engineering
degree
from
the
Institute
of
8
McDaniel
et
al,
AMOCOs
LC-Fining
chemical
engineering
from
Lamar
University,
chemical
engineering
from
the
Massachusetts
concentration
of
ions,
pH
values,
begin
to
produce
material
termed
hydrogen
sulphide,
can
alter
elecReferences
to explore the opportu- Email:
and
applies
cutting-edgetechnology,
analytics universities
of Celerum
Ltd., which provides consultancy
marco.respini@bakerhughes.com
428-433, 1993.
ments
withGulf
monitoring
3rd
Edition,
Professional
Publishing,
Email:
norm@lieberman-eng.com
and
clearly
identifies
efficient
paths
He
holds
aTechnology.
BE in
chemical
engineering
from
troubleshooter.
tower
operating
pressure.
Also,
the capacity
13-15
May
2009.
Barcelona,
Spain,
in chemical
engineering.
would
require
only
one
cylinder
control
system.
A
reciprospare
parts
to
be
kept
in
stock.
The
Viewpoints,
6Loiselle
bJan
2016,
http://industrial-iot.
Poorer
stripping
efficiency
is
generally
measured
as
line
coking,
fractionator
flooding,
and
wet
gas
Chemical
Technology
(formerly
UDCT,
Mumbai
Resid
Hydrocracking
Yield
and
Performance
Texas.
Institute
of
respect
to
capital
and
operating
and
oxidation
reduction
potential.
extracellular
biopolymer,
or
slime
trochemical
processes
at
the
1
Royan
T
S,
C
E,
High
sulphur
recovery
and data science, capturing new nities that data analytics offer. For and optimisation services, and PlanSea Ltd.,
Stevens
Institute
of Technology,
New Jersey.
com/2016/01/industrial-iot-an-opportunitytoward
optimised
systems
to cating
compression
stage (two
cylinhigher
efficiency
reduces
electric
University),
Mumbai,
India.
achievedcompressor
from
acid shall
gas, Thecompress
Oil and Gas per
expenditure
to determine
the
best
which focuses
on optimising
marine logistics.
example
Thelean
Data
Labs
Aberdeen
market
opportunities,
boosting
power costs, and the high degree process gas in two compression ders in total for the compressor)

A controlled decrease
in wash oil can result
in better handling of
fouling, and at the
same time increase
HVGO directly

alves
Best V 67
8
since 1

www.eptq.com
www.eptq.com
38 www.eptq.com
PTQ
Q2 2014
2014
www.eptq.com
www.eptq.com
www.eptq.com
98
PTQ
Q3
www.eptq.com
www.eptq.com

hoerbiger.indd
5
baker hughes.indd
8
bpcl6.indd
2
jacobs.indd
liberman.indd6 5

PTQQ2
Q42015
2016 97
95
PTQ
Q3
2016
57
PTQ
www.eptq.com
PTQ
Q1
2015
27
PTQQ1
Q42014
2015 131
95
PTQ
www.eptq.com
PTQ
Q1
2017
77
PTQ Q4 2014 129

10/03/2014 13:40
13/12/2013
10/06/2014 11:38
14:26
14/12/201616:38
12:48
10/09/2014

Support grid
Scallops

Outer basket

Centerpipe

Outlet Collector

From design to manufacturing and field installation, Johnson Screens provides an

extensive offering of products and services for the hydrocarbon processing industry.
Johnson Screens offers a complete solution to your hydrocarbon processing industrial needs, including:
Down Flow Reactor Internals
Radial Flow Reactor Internals
Installation of Internals

Hydrocarbon Refining Reactor Internals

Petrochemical Processing Reactor Internals
Gas Processing Column Internals

Please visit our website to see the full line of Johnson Screens products, or find a reactor internals specialist
engineer at one of our global manufacturing plants to assist you in finding the proper solution for your
application.
We look forward to providing the quality engineered products you have come to expect when specifying
Johnson Screens.
North & South America
Phone +1 651 636 3900
Fax +1 651 638 3281
usa.water@aqseptence.com

Asia Pacific
Phone +61 7 3867 5555
Fax +61 7 3265 2768
asiapacific.water@aqseptence.com

Europe, Middle East & Africa

Phone +33 5 4902 1600
Fax +33 5 4902 1616
france.water@aqseptence.com

www.tiny.cc/johnsonscreenshp
johnson screens.indd 1

14/12/2016 11:06

Correcting vacuum column design flaws

Conventional design practices have been proven inadequate when maximising
diesel recovery and gas oil cut point
GARY MARTIN
Recon Management Services, Inc.

o achieve a good return on

investment (ROI) from a
project requires developing
minimum cost designs regardless
of whether it is a grassroots or
revamp project. However, low cost
designs do not necessarily guarantee a good ROI. In recent years, a
number of new fuels vacuum units
have been designed to produce
diesel off the top of the vacuum
column (see Figure 1). The vacuum
units are commonly designed for
low pressure damp or wet operation to enable maximising HVGO
cutpoint.1 These design conditions,
in addition to higher LVGO pumparound duties associated with
maximising diesel production, lead
to large diameter LVGO pumparound sections. The operating
conditions associated with the
design make the possibility of
excessive overhead slop oil production much more likely. Design
practices of the past are inadequate
and must be re-evaluated to obtain
an acceptable design.
This article is based on a revamp
design to fix the operation of a large
scale Gulf Coast vacuum column
with excessive liquid entrainment
into the overhead system. The
LVGO pumparound section of this
vacuum column is 38ft (11.58m) in
diameter and typically has a diesel
product yield of 17 500 b/d (115.9
m3/hr). The vacuum column is
designed for damp operation, meaning that it has a resid stripping
section using stripping steam and
the charge heaters utilise coil steam.
Spray header distributors produce a
range of droplet sizes and essentially all LVGO pumparounds
designed with these distributors

www.eptq.com

q1 recon.indd 1

Motive
steam

Ejector
sytem

Noncondensibles
Sour water
Slop oil

Vacuum
column

LVGO pumparound

Diesel
Reflux

Figure 1 Upper section of vacuum column

have some overhead entrainment.

This unit typically operates at about
1700 b/d (11.26 m3/hr) of slop oil
yield, with approximately 40% of
this from equilibrium oil and the
remainder from entrainment. While

Low cost designs

do not necessarily
guarantee a
good return on
investment
this may seem high, in actuality it is
worse. Due to problems with the
overhead ejector system design, the
column operating pressure is
normally higher than its original
design value. In addition, the operators limit the LVGO pumparound
circulation rate to limit overhead
entrainment. During operation with

increased LVGO pumparound circulation or during the winter when the

conditions enable lower pressure
operation, the vacuum column overhead slop oil make has been in the
range of 2750 b/d (18.22 m3/hr) and
at times higher. Due to economic
yield incentives, the refiner would
prefer to operate at the lower
column pressure and higher LVGO
pumparound rate if these problems
did not exist. At the lower column
pressure and higher pumparound
rate, the liquid entrainment into the
overhead is in excess of 2000 b/d.
Slop oil make is detrimental to the
plants bottom line. At minimum, it
adds to the operating costs to reprocess the oil and in the worst case it
consumes plant capacity to process
additional crude. To reduce the
excessive slop oil make at lower
pressure operation, the operators
must adjust the heat balance on the
column. It is modified to lower the
LVGO pumparound heat duty. This

PTQ Q1 2017 79

14/12/2016 13:03

ments or, as addressed in this

article, that ignore the effects of
liquid entrainment.

Equilibrium and non-equilibrium oil

in overheads

Figure 2 Spray header above a packed bed

reduces the required LVGO pumparound rate and consequently

reduces the entrainment rate to a
somewhat more acceptable overhead slop oil yield. This however
significantly reduces the diesel yield.

LVGO pumparound

The typical design of a vacuum

column LVGO pumparound section
includes a product/pumparound
collector tray, a packed bed, and a
spray header for pumparound
return liquid distribution. A spray
header has been the industry
distributor of choice because it is
relatively cheap and it provides for
good heat transfer. All spray headers produce a range of liquid
droplet sizes. Normally, the majority of droplets are large enough
and the vapour velocity low
enough that the entrainment is
minimal. When using spray headers, a damp vacuum column design
always makes the entrainment
problem worse because of the
corresponding higher top of the
column vapour velocity. The problem becomes more of an issue with
higher coil and stripping steam
rates. Dry columns have very little
vapour compared to damp columns
when operating at the same column
overhead temperature and pressure. Also, the more spray nozzles
the higher the potential entrainment. This LVGO pumparound
distributor has 61 nozzles while
most vacuum units have 7 to 19.

80 PTQ Q1 2017

q1 recon.indd 2

Spray header design

Spray header distributors, as shown

in Figure 2, are good distributors
for packed bed heat transfer
service. This is due to the high
surface area contact between liquid
and vapour due to atomisation of
the liquid. Heat transfer between
the liquid and the vapour phase is
occurring prior to entering the
packed bed. However, in general,
the liquid distribution to the bed is
only roughly uniform and every
nozzle has different distribution
characteristics.
While spray headers are relatively simple distributors, many
engineered designs that have been
produced are very poor. Common
design mistakes include: layouts
that produce excessive liquid overspray onto the vessel wall; too
small an nozzle orifice size that is
not practical and leads to fouling;
and designs that do not take into
consideration turndown requireVacuum
column
overhead
slop oil slop oil
Vacuum
column
overhead
Volume %
1
5
10
30
50
70
90
95
99

Table 1

D2887F
93
219
268
378
483
558
617
638
687

High vacuum column overhead

slop make can occur from operating problems in the vacuum
column or from light material
carried over from the atmospheric
crude column. A case study of the
latter is provided in the literature.2
Table 1 lists the vacuum column
overhead slop oil make distillation
curve determined by the plant laboratory during one set of operating
conditions. The column top pressure and temperature measured
were 38 mmHg and 125F (51.7C)
respectively. Proper evaluation and
fixing of the problem requires good
test run data.3 Using a full set of
test run data to model the vacuum
unit yields a much lighter overhead
slop oil composition based on equilibrium
calculations
at
the
measured operating conditions. The
measured overhead slop oil make
on this day of operation was 1726
b/d (11.43 m3/hr) with a 38.1 API
gravity. To match the model results
with actual operation, 978 b/d (6.48
m3/hr) of entrained LVGO pumparound liquid had to be added into
the column overheads to match the
material balance and slop oil
composition.
Equilibrium oil in the column
overhead vapour is a function of
the column overhead operating
pressure
and
temperature.
Non-equilibrium oil is heavier oil
that should not be in the vapour
phase at these operating conditions.
In this case, the heavier oil is from
the smaller atomised oil droplets
formed from the pumparound
distributor spray nozzles that are
entrained by the rising vapours.
The vacuum column also has
temperature indicators measuring
the vapour temperature in the top
head of the vacuum column and in
the column overhead vapour line.
There was a temperature drop of
21F (11.7C), which is further indication of entrainment. As the rising
vapours from the LVGO pumparound bed exit the column, they are
mixed with the cold LVGO

www.eptq.com

14/12/2016 13:03

pumparound
liquid
that
is
entrained into the overheads,
resulting in a reduction of the
measured vapour temperature in
the overhead vapour line.

Mist eliminators

In some applications, mist eliminators located at the top of the

column can be used to remove or
minimise the entrained liquid.
However, years of bad experiences
by the industry have proved that
mesh pads should not be used for
this application. Experience has
proved that they will foul, leading
to high pressure drop and a corresponding decrease in gas oil yields.

Phase Doppler interferometry

To properly design a spray header

requires knowing the estimated
entrainment rate for the selected
nozzles. The tendency for the atomised drops to be entrained is a
function of the drop size distribution, the vapour velocity of the
up-flowing vapours, and the
vapour and liquid physical properties. There are three forces that act
on the liquid droplets from the
spray nozzles that determine if
entrainment of liquid will occur.
The forces are gravitation, the
buoyant force which acts in the
opposite direction from the gravitational force, and a drag force due to
the relative motion between the
particles and the rising vapours.
The drag force acts to oppose the
motion of the liquid droplets in
the opposite direction of their
movement.
In a revamp case such as this,
operating data to determine the
overhead vapour rate and composition is available from plant meters
and lab data. However, if inadequate meters are available or if this
is a new design then air leakage,
cracked gas, and so on must be
estimated. Guidelines regarding
estimates for these values are available in the literature.4
The quantity of entrainment can
be determined by the particle
terminal velocity, obtained using
Stokes Law, and knowing the drop
size distribution for the selected
nozzle. Stokes Law can be used at
low Reynolds numbers and can be

www.eptq.com

q1 recon.indd 3

found throughout the literature.5

Terminal velocity can be calculated
using the equation:
Terminal velocity =
2 ! [! ! ! ]

where
g = gravitational acceleration
p = particle density
= actual vapour density
m = mass of particle
Ap = projected area of particle
Cd = drag coefficient
It is understood that liquid distribution within a spray cone varies
for a given nozzle and further variation is found from one nozzle type
to the next. It is also known that as
the nozzle pressure drop changes
from one flow rate to the next the
spray cone diameter varies, as well
as the liquid particle sizes.
However, reliably determining
drop size distribution was not
available
until
testing
with
phase
Doppler
interferometry.
Fractionation Research Inc. in
Stillwater, Oklahoma has recently
tested this method.6 Their test
results also showed that an increase
in liquid rate results in a decrease
of the entrained droplet Sauter
mean diameter and a corresponding increase in total liquid
entrained. The tests also showed
that increasing the gas rate
decreases the entrained droplet
Sauter mean diameter and has a
corresponding increase in the total
liquid entrained.
Using phase Doppler interferometry to determine the drop size
distribution along with determining
the vapour velocity of the up-flowing vapours and the vapour and
liquid physical properties, the
designer can evaluate the entrainment rate expected for a selected
nozzle. Different nozzles provide
varied liquid distribution and drop
size distribution. In fact some
nozzles are designed to produce
very small droplets that if used in
this application would be even
more detrimental to operation. This
method can be used to predict the
entrainment rate and to effectively
select optimum nozzles.

Conclusion

Maximising production of diesel

from the top of a vacuum column
combined with a damp vacuum
column operation to maximise gas
oil recovery makes old design practices for the column internals
inadequate.
While
properly
designed trough distributors can fix
this problem, they do not have as
good heat transfer efficiency as
spray headers. Vessel height limitations to increase the packed bed
height may not enable achieving
the required pumparound heat
removal and thus not enable using
a trough distributor. Using spray
nozzle phase Doppler interferometry test data can be used to
appropriately select the optimum
LVGO pumparound spray nozzle
for the specific operating conditions
to minimise overhead oil entrainment. This approach can also be
used for other column designs.
References
1 Martin G R, Vacuum Unit Design Effect on
Operating Variables, PTQ, Summer 2002, 8591.
2 Martin G R, Lines J R, Golden S W, Vacuum
System, Fundamentals, Encyclopedia of
Chemical Processing and Design, Vol. 61,
McKetta J J, Weismantel G E, editors, Marcel
Dekker, Inc., New York, NY, 1997, 5-18.
3 Martin G R, Cheatham B E, Keeping Down
the Cost of Revamp Investment, PTQ, Summer
1999, 99-107.
4 Martin G R, Nigg J M, Vacuum Pressure
Control: Impact on Profitability, PTQ, Summer
2001, 73-81.
5 McCabe W L, Smith J C, Unit Operations of
Chemical Engineering, McGraw-Hill Book Co.,
New York, 1976.
6 King B, et al, Measuring Entrainment from a
Spray for Industrial Applications, AIChE Annual
Meeting, Salt Lake City, 11 Nov 2015, Paper
155B.

Gary R Martin is a Senior Process Specialist

with ReCon Management Services, Inc. He has
been involved in revamps, including conceptual
process design and process design packages
for large capital revamps, optimisation, and
troubleshooting services to the refining
industry worldwide. He previously was a
Principal of Process Consulting Services Inc.
and also worked for Total Petroleum, El Paso
Refining, and Glitsch Inc and holds a BS degree
in chemical engineering from Oklahoma State
University.
Email: gmartin@recon-group.com

PTQ Q1 2017 81

14/12/2016 13:03

ITW

Innovative
Technologies
Worldwide

CHANGE YOUR MIND

ON FOULING REMOVAL

ARE THESE IMAGES FAMILIAR TO YOU ? FORGET ABOUT THEM !

Patented ITW Online Cleaning can remove any type of fouling from equipment, including the polymers, without the need to
open or enter hazardous process equipment.
An entire Process Unit can be cleaned by utilizing ITW Online Cleaning in as little as 24 hours on an oil-to-oil basis.
Online Cleaning can be applied at any time during the run of the Unit, in order to solve the problems when they start
appearing, rather than when they are no longer sustainable.
This will in turn avoid throughput reduction, giveaway and energy loss, associated with fouling.
The application of ITW Online Cleaning will be therefore driven by performance recovery and Opex improvement, rather than
the economics for placing a turnaround.

ITW Online Cleaning can be applied to all Refinery/Petrochemical/Gas Field/Oil Field production Units, as
well as storage tanks, and is the only technology which has an immediate Return On Investment (ROI).
Regular application of ITW Online Cleaning will target an increased run length under clean conditions, with a far higher ROI.
For Turnaround applications, ITW Online Cleaning can eliminate/dramatically reduce the need for mechanical cleaning,
thereby reducing downtime and improve operational HS&E.
In a Turnaround application, an additional ROI increase can be targeted by applying ITW Improved Degassing/
Decontamination to achieve quick and effective safe entry conditions.
Our patented chemistry does not create any emulsion and fluids can be easily handled by Waste Water Treatment Plant.

EVALUATE ITW ONLINE CLEANING AND ITW IMPROVED DEGASSING/DECONTAMINATION TODAY TO

IMPROVE YOUR PLANTS PROFITABILITY !
For more informations contact :

ITW S.r.l.- C.da S.Cusumano - 96011 Augusta - Italy

Tel. +39 (0931) 766011
E-mail: info@itwtechnologies.com
www.itwtechnologies.com

Now hiring
Professionals
Worldwide

Join ITW Team worldwide and send your Curriculum Vitae to : jobs@itwtechnologies.com

itw.indd 1

05/06/2014 19:38

Analysing FCC hot spots

Finite element analysis can be employed to improve the safety and quality of old
or new designs and troubleshoot problems in the field
KENNETH FEWEL
Technip Stone & Webster Process Technology

ot spots on FCC units are a

common occurrence but relatively little is known about
the implications for long term operation. Standard procedures in a FCC
plant are to monitor on a scheduled basis the thermal scans of their
equipment and use steam to cool if
hot spots exceed a certain temperature threshold. Although this has
served operators in a practical manner for years, there are larger safety
issues that should not be ignored.
ASME has developed standards for
analysing thermal stresses that are
suitable for design, maintenance
and operation. Operators should be
aware of these methods and take
advantage of them for their hot spot
maintenance and procedures.
Finite element analysis (FEA) is a
method that enables accurate stateof-the-art analysis of stress and
strain in all types of solid materials
for all types of loadings, including
thermal. Thermal stresses are usually calculated for a design that is
working perfectly to insulate the
steel shell from high temperatures
inside the FCC vessel. FEA has been
used to validate traditional calculation and more accurately define
the state of temperature, stress and
strain that will occur under normal
operation in FCC units made from
composite, refractory-lined steel,
materials.
Recently, as a result of hot spots
appearing on FCC equipment,
there was a need to analyse and
determine the severity of stresses
and strains under hot spot operation. This has provided an excellent
opportunity to use FEA to explore
these situations using quantitative
analysis and make recommenda-

www.eptq.com

q1 technip.indd 1

tions to operators about the safety

of their particular situation.
Before further explaining FEA,
it is important to understand why
hot spots occur and how stress and
strain occurs.

Hot spot causes

Hot spots are caused by several

mechanisms of operation that come
together like a perfect storm. They
do not appear randomly as there are
reasons why hot spots form in FCC
equipment.

Hot spots are caused

by several mechanisms
of operation that
come together like a
perfect storm
1. Erosion by catalyst

Catalyst is, by nature, irregular in

shape and very hard compared
to carbon or stainless steels. The
refractory used to line steel vessels or piping is designed to provide insulation and strength for
stresses and thermal strains; it is
not designed purely for erosion
resistance. Therefore, the material
is susceptible to erosion by catalyst. Fortunately, this insulating
refractory is thick, usually 4-5in in
piping and vessels. However, catalyst can quickly erode through this
thickness with the help of the next
mechanism.

2. Thermal stress and strain

a. Wall stress: thermal wall stress

can easily reach levels that will

crack the refractory. In many

designs some cracking cannot be
avoided. Longitudinal cracks in the
refractory outer diameter of lining
will form as soon as temperatures
reach operational levels. This is
due to thermal wall stresses in the
refractory. These stresses have been
determined from research experiments by Wygant and Crowley1
among others.2 The inner diameter
of the lining is very hot, particularly
in a regenerator, typically 1300F
(700C). The inner refractory will
go into compression when internal temperatures start to rise. The
outer surface of the lining may not
expand as much as the steel. Thus
the lining will go into tension stress
and the refractory may not bear
upon the inner diameter of the steel
piping. If so, thermal wall stress can
cause the lining in the outer edges
to crack.
b. Thermal bending stress: thermal expansion forces in the piping
and vessel can crack the refractory
in bending, often transverse to the
longitudinal direction. In addition,
compression bending stress can
open gaps between the steel and lining, leading to large gaps through
which vapour and catalyst can easily circulate. These are the most
damaging stresses to the refractory
that can occur. The resulting hot
spots will exacerbate the problem,
making the thermal stress and gap
worse.

3. Worm holes

As a direct result of wall stress

cracks in the refractory lining and
gaps opened by thermal bending of
the refractory on the outside of the
lining, there is a strong possibility

PTQ Q1 2017 83

14/12/2016 13:10

Change in hot spot equivalent

stress, psi

50000

y = 73.789x

45000

R2 = 0.6434

40000
35000
30000
25000
20000
15000
10000
5000
0

50

100

150

200

250

300

350

400

450

500

Change in temperature of the hot spot, F

Figure 1 Rise in equivalent stress as a function of rise in hot spot temperature

that the hot vapours and catalyst

will find alternate paths through the
piping. The wall stress cracks are
probably not large enough for catalyst to circulate through, and they
may fill with catalyst instead. A
larger problem is the thermal bending strain in the refractory. Thermal
bending stress and strain will open
existing cracks or form new cracks
large enough for catalyst to circulate through. Worm holes can then
form, leading to catalyst circulating past the steel, overheating, and
rapid erosion of the steel skin from
the inside out.
The most apparent result of a
combination of these mechanisms is
a hot spot. The steel will often reach
temperatures in excess of piping
or vessel design values, typically
around 650F (340C). When the
skin reaches temperatures higher
than design, the situation requires
cooling steam to keep the temperatures under control. This is standard maintenance procedure, but
what are the resulting safety and
long term implications? When is a
shutdown necessary? If a shutdown
is planned, what needs to be done?
Is the steel and/or lining material ruined or can it be reused after
shutdown?
Usually, the answers are not so
clear to the operator as a potentially
dangerous situation arises. What if
the steam pressure is lost temporarily? What if the temperature rises?
Continuous monitoring of the steam
cooling system is rarely done. The
thermal scans are not continuously

84 PTQ Q1 2017

q1 technip.indd 2

reviewed either, but done only

periodically. How can an operator
know that his plant is safe for the
next day or week, or until the next
scheduled shutdown?
These questions should not be
answered by rough estimates or
guesses. Hot spots, once formed,
cause additional thermal stress.
When a hot spot appears, the situation will worsen and thermal stress
and strain can quickly become
unstable.

When a hot spot

appears, the situation
will worsen and
thermal stress and
strain can quickly
become unstable
How hot spots cause thermal stress

Take, for example, a thin plate held

rigidly at the edges and heated with
a torch at the centre of the plate.
The material in the centre will begin
to expand and thermal stress will
occur. The edges, if truly rigid, will
prevent expansion from occurring
in the plane of the plate. Once the
stresses exceed the proportional
limit, the plate will begin to warp
out of plane. Imagine that this happens on the surface of a large catalyst transfer pipe. The steel must
bulge outwardly but will be resisted

by the geometrical constraints of the

adjacent cooler pipe and, to a lesser
extent, refractory anchors. But the
gap between the steel and refractory
is larger now and the circulation of
catalyst is naturally increased. The
hot spot temperature is increased
and the thermal stress becomes
higher. This instability will not stop
until one of the following mechanisms occurs:
1. The catalyst streaming in builds
up to fill the void left by the steel
strain and the gas bypass flow is
reduced or eliminated.
2. The strain hardening in the steel,
already in the plastic region, reaches
a point of equilibrium before the
rupture point is reached. This is
usually assisted by a geometric constraint and/or steaming.
3. The steel reaches rupture point.
Catalyst and vapours begin to spew
from the opening.
Obviously every operator hopes
for number 1 or 2 to occur. Despite
denials, number 3 has been known
to happen.
Based on the Goodier3 equation for
thin circular plates unconstrained at
the edges, the following relationship
can roughly estimate the thermal
stress rise due to additional temperature rise in the elastic region:
d r

d t

d rt

T
0.5 E dT

where:
sr = radial stress in the hot spot
st = tangential stress in the hot spot
rt = shear at the edge of the hot spot
dT = temperature change
E = elastic modulus of the material
a = coefficient of thermal expansion
Of course this must be added to
the normal operating weight, pressure and thermal stress state which
existed before the hot spot formed.
The sample field data employed in
Figure 1 show that geometric configuration and pre-stress matter greatly
because there is much scatter in the
equivalent stress data. Where prestress is largely compressive, hot
spot stress will simply add more
compressive stress to the negative
components of pre-stress. Where
pre-stress is mostly tensile, the hot
spot will lower or reverse the tensile components and raise the compressive components. In addition,

www.eptq.com

15/12/2016 13:36

www.cat-tech.com

MAKING YOUR
PLANT WORK...
TOWER FIELD SERVICES
Cat Tech Tower Field Services offers bespoke
mechanical services to oil refinery and protochemical

SMARTER
We thrive on exceeding our clients expectations and understand the need for
upfront communication in relation to timescales and costs.
No-one likes surprises in this industry and we ensure that everything is done, from
pre-job planning, on site activities to post job reporting in a mannerwhich allows our
customers to concentrate on their core business.

plants all around the world.

We work hard to maintain an industry reputation for safety, speed and dependability.

Our team has vast experience in overseeing and

Our in house training allows us to staff a project with the relevant skilled personnel
to produced customer satifaction.

undertaking the installation of all types of Tower

internals.
Our tower services devision has deep roots, with

We would value the opportunity to demonstrate the benefits of utilising Cat

Tech for all of your tower services.

several of the senior team at Cat Tech being

The benefits of engaging Cat Tech are:

involved in the founding of the Cana-Tex

organisation.
Since 1971, Cana-Tex had been the premier
global Tower speciality contractor; performing
Tower maintenance and plant turnaround
projects around the world. These origins, with
the addition of new team members, make Cat
Tech Tower Field services a global operation
which is both cost effective and technically
experienced.

Our well established sales team have

a wealth of experience. We take pride
on ensuring all enquires are dealt with
quickly and efficiently.

Experienced Tower specialists.

Proven track record.
Ability to supply spares and replacements.
Spade to Spade service.
Revamps or modifications to Towers / Internals.
Inspection of Tower internals.
Supervision of external/client labour force.
Detailed inspection reports.
Cleaning of tower internals.
Reactor tray field services.
Turnkey reactor tray field services.

UK: Cat Tech (Europe) Ltd.,

1 South Park Road,
South Park Industrial Estate,
Scunthorpe, North Lincolnshire,
DN17 2BY UK
Phone: +44 (0) 1724 871747
Fax: + 44 (0) 1724 861928
Bulgaria: Cat Tech Services (South
Eastern Europe) Ltd.,
Devnia, Varna District,
Povelyanovo, 20 Droujba Str,
R Bulgaria, 9160
Phone: +(35) 9519 92052
Fax: +(35) 9527 46770

China: Cat Tech Services (Shanghai) Ltd.,

Room 1230 No.1, 58 Nong Huachi Road,
Putuo District,
Shanghai 20061 CN
Phone: +(86) 2152 9175 08
Fax: +(86) 2152 9175 05
Singapore: Cat Tech Asia Pacific Pte.
Ltd., 17 Fan Yoong Road,
Singapore 629794
Phone: +(65) 6264 6261
Fax: +(65) 6264 6401
South America: Please contact the UK
office on +44 1724 871 747

Thailand: Cat Tech Services (Thailand)

Ltd., 2/1 Soi Banbon Nuenpayom
Road, Maptaphut Sub District, Muang
Rayong District, Rayong Province
21150 TH
Phone: +(66) 894 064 100
Fax: +(66) 638 692 382

Lining

Temperature, F
1423.10 (max)
1294.20
1165.40
1036.50
907.68
778.84
650.00
500.00
450.00
400.00
352.97
305.94
258.91
211.88 (min)

200C skin
Steel pipe

0 2.5 5.0 7.5 10.0

1382.00

750C

Typical hot spot

788.00
576.60
716.64
467.83
435.06

0 2.5 5.0 7.5 10.0

inches

Figure 2 Thermal gradients in typical lined FCC pipe modelled with heat transfer through
both materials using FEA

geometric conditions such as crevices, bending stress and other stress

risers will play an important part.
Finally, vibration will also increase
the general state of stress with
reversing components that will complicate the situation. Vibration stress
is almost always present in an FCC
and is largely disregarded, despite
the fact that it can be signicant.
Combining all of these factors
makes a traditional analysis very
challenging. It can be done, but
the results will be highly uncertain, could be inaccurate, and likely
misleading.
Methods are available to determine the complete state of stress
and to mitigate the causes. This is
where the application of FEA modelling can help.

The FEA modelling approach to lined

pipe hot spot analysis
An FEA model can analyse hot
spots in detail. With the assistance

of thermal scans, the hot spot can

be replicated in a model. By doing
so, the state of stress caused by the
hot spot can be determined with a
high degree of accuracy. These thermal stresses can be compared with
ASME allowable to determine if
continued operation is safe. Figure 2
illustrates a FCC lined pipe hot spot
model and the associated thermal
state.
ASME Section VIII, Division 2,
has developed an analysis method
for determining the safety of steel
thermal stress.4 The ASME piping
code B31.35 can be used also, but
this will result in more conservatism
as the piping code is based on traditional calculations. Both have been
used for most studies, as a check
against one another. The results
were similar although stresses
checked against B31.3 had a slightly
lower margin of safety.
Refractory stress can be checked
also. Although allowable material

Figure 3 FEA sub-model of steel plate,

refractory lining and stainless anchors

limits are not available in ASME,

the manufacturers strength properties can be used with a safety factor
applied. Compared to steel, refractory strength is very low. Although
compressive strength has been
stated as high as 5000 psi, tensile
or rupture stress strength, determined from bending specimens, is
generally less than 1500 psi, even
though stainless needles are used in
many formulations to improve tensile strength. Fortunately, the elastic
modulus is generally less than steel
and therefore exibility is greater.
In addition, the exibility of the
anchors is fairly high and this isolates some of the steel strain from
the refractory strain and vice versa
(see Figures 3 and 4).
Despite the use of anchors
and needles, refractory is very

662.00
482.00

755.60
Temperature, F
1423.1 (max)
1036.5
650.00
500.00
450.00
400.00
352.97
305.94
258.91
211.88 (min)

788.00
inches
0.51241103 (max)
0.45547103
0.39854103
0.34160103
0.28463103
0.22774103
0.17080103
0.11387103
5.6934e5
0 (min)
0

2.5

5.0

7.5

698.00
377.12

Max
899.60
791.60

Min
10.0

inches

Figure 4 ANSYS FEA sub-model of steel and lining pulling apart. Colours show
deformation with highest shown in red and lowest in blue. Refractory has been cut away
to show the anchors clearly

86 PTQ Q1 2017

q1 technip.indd 3

225

450

675

900

inches

Figure 5 Field case situation: hot spot

temperatures found from thermal scans,
steam on

www.eptq.com

14/12/2016 13:10

susceptible to cracking under operating conditions due to thermal stress.

If the design is not analysed carefully
with the use of FEA, areas of critical
bending stress in the refractory are
usually missed. This is because most
piping is analysed with a one-dimensional piping stress program.
These programs do not provide for
placing discrete refractory material
in the piping model. Although the
added strength of the refractory can
be factored in, stress in the refractory material often cannot be determined. If the refractory reaches the
bending rupture point, the contribution to overall strength is lost. Yet the
one-dimensional approach would
not take this into account. Therefore,
this approach could be non-conservative if the refractory strength is factored in and the refractory is actually
breaking.
FEA is the most advanced means
to analyse the refractory-lined pipe
in FCC systems. No other method
can supply the detailed stress
state or the design quality for both
steel and refractory lining in one
model.

Case history: FCC hot spot FEA

A major European operator was

experiencing an increasing number

Temperature, F
1423.10 (max)
1036.50
650.00
500.00
450.00
400.00
352.97
305.94
258.91
211.88 (min)
50

100

150

200

Figure 7 Standpipe to riser wye piece; hot

spots appear in high bending stress areas

q1 technip.indd 4

20

30

40

inches

Temperature, F
1425.5 (max)
1036.3
650.00
500.00
450.00
400.00
352.97
305.94
258.90
211.87 (min)

662

Figure 6 FEA replicated hot spot on regenerated catalyst withdrawal well nozzle, top
side. High stress existed here because of normal operating thermal tension load in the
regenerated catalyst standpipe

of hot spots on transfer lines and

nozzle connections. The plant had
vessels bulging, pipes rupturing, and
hot spots occurring across the piping systems slide valve, cyclone dipleg nozzle and other transfer lines.
The long standpipe and riser transfer piping had no expansion joint so
thermal stress and strain were suspected. A transfer line model was set
up to determine the complete state of
stress in the affected piping. Figure 5
shows hot spot tagged by temperature with existing steam cooling.
In addition to a model of normal

Stress, psi
55852.00
49653.00
43454.00
37255.00
31057.00
24858.00
18659.00
12460.00
6261.20
100.47 (min)

791.6

inches

www.eptq.com

10

operation under weight, pressure

and thermal loads, a second model
was set up and the hot spots were
added. Figures 6 to 8 show the temperature contours of model setup
and resulting stresses and deformations for typical parts of a catalyst
transfer system.
The results of the study were very
revealing. Figure 9 shows computed
displacements later veried by the
operator. Figures 10 to 11 illustrate
the stresses in the reactor cyclone

47439
42300

899.6

Deformation, in
4.9886 (max)
4.4343
3.8800
3.3257
2.7714
2.2172
1.6629
1.1086
0.55429
0 (min)
Max

53780
27937

25 50 75 100

225

450

675

900

inches

inches

Figure 8 Stresses found with FEA in the

wye, probably due to bending failure in the
refractory

Figure 9 As a benefit of an accurate

thermal strain state, FEA will yield very
accurate thermal displacements. The
operator measured a 4in lateral movement
with hot spots, matching the FEA closely

PTQ Q1 2017 87

15/12/2016 17:10

Maximum
principal stress,
psi
1.0363e5 (max)
7000.0
6222.2
5444.4
4666.7
3888.9
3111.1
2333.3
1555.6
777.78
0
500
1000
50397 (min)

Stress, psi
1.2519e5 (max)
55000
48127
41255
24382
27509
20636
13764
6891
18.237 (min)
54654

10

15

20

inches
500

Use of FEA to analyse FCC designs

The detailed design process for FCC

equipment should rely on an FEA
stress model to determine the complete thermal stress state before an
FCC is built. FEA will reveal where
thermal stress and strain is high,
allowing custom changes to the

88 PTQ Q1 2017

q1 technip.indd 5

ature, size, pre-stress, position,

geometric factors and material
properties. It is difcult to know
with certainty if a hot spot is a
safety issue without a detailed FEA
analysis.
The analysis can take all of these
factors into consideration: anchors,
refractory
strengths,
refractory
expansion coefcients, piping material, weight, pressure and thermal
stress. Even vibration and fatigue
can create damaging refractory
stress and strain to occur, leading

100

DPT normal FOS

DPT ASME FOS
DPT B3.13 FOS
10

pl

eg

se
ne

di

Ri

e
ju

nc
tio

va
lv

lo
yc

A6

A4

A7

e
id
Sl

ye

45
p

ip
e
45
p
D
W
ip
Y
e
ju
nc
tio
n

0.1

of

In almost every case, high levels of

operating stress and strain coincide
with hot spot locations. The hot spot
temperature can be correlated by
the operating stress and strain at the
hot spot. Figure 13 illustrates this
with a graph of hot spot temperature vs steel operating stress.
Although the data are scattered,
the relationship between pre-stress
and hot spots is clear. Faulty thermal design causes high operating
stress and this leads to cracking of
the refractory, perhaps during the
rst start-up. From there, it may be
only a matter of time before the hot
spots appear.

Factor of Safety

Pre-stress and hot spots: causal

relationships

dipleg skin and lining. Figure 12

summarises safety factors. All of
the thermal hot spot stresses could
pass ASME Section VIII, Div 2,
part 5 rules. The operator decided
to continue running the plant to
the next scheduled shutdown with
condence.

design and possibly preventing hot

spots. This premise is based on experience with hot spots in the eld.
Most, if not all, cases are associated
with thermal stress and strain, the
key causal elements in hot spot formation. If these are eliminated, erosion is the likely reason for a hot spot
to form and that could take many
years to happen.
Once a hot spot has appeared, an
FEA analysis can determine if the
stresses are high enough to cause
alarm. Variables include temper-

Figure 10 Skin stress on an external FCC

reactor with hot spot around the nozzle
connection

Figure 11 Dipleg refractory lining stress contours at the hot spot indicating ultimate
failure of refractory throughout thickness

inches

of
W

375

tto

250

To
p

125

Bo

Figure 12 Factor of safety (FOS) comparison for dead weight, pressure and thermal
(DPT) case at new condition vs hot spot operation. A factor of safety of 1 or greater is
a passing grade, indicating that computed stress is lower than allowable stress. Normal
FOS indicates as designed. For hot spot cases, ASME indicates Sec 8, Div.2 evaluation and
B31.3 is a piping evaluation

www.eptq.com

14/12/2016 13:11

to hot spots and the possibility of

rupture. This can be analysed with
a dynamic stress FEA model of the
piping or vessel.

Hot spot temperature, F

1000

Conclusion

FEA is a maturing technology used

more frequently as the tools become
more widespread, user friendly and
real world effective. With proper
setup, verification of computations and experience to evaluate the
results, this engineering tool can
be a means to improve the quality of old or new designs and troubleshoot problems in the field with
the vision like no other method. Do
not discount FEA as unnecessary,
too complicated or expensive as
the investment returns can greatly
advance engineering and product
quality.
References
1 Wygant J F, Crowley, American Ceramic
Society Bulletin, Vol. 43, No. 3, 1964.
2 Buyukozturk, Oral, Tseng, Journal of the
American Ceramic Society, Vol. 65, No. 6, Jun
1982.

900
800
700

y = 0.0128x + 670.72

600

R2 = 0.7968

500

400

Hot spot temperature

Linear hot spot temperature
0

5000

10000

15000

20000

Operating thermal stress as designed, psi

Figure 13 Correlation of design stress levels and eventual hot spot temperatures
experienced in those stress locations in an FCC catalyst transfer line. This is one
indication that higher operating stress leads to hot spots
3 Goodier J N, Thermal Stress, ASME J of Applied
Mechanics, Vol 4, No 1, 1937.
4 ASME Section VIII, Division 2, Part 5: Design
by Analysis, 2007.
5 ASME Process Piping, B31.3, 2006.

Kenneth Fewel is a Senior Supervising

Engineer with Technip Stone & Webster

Process Technology. With 40 years of industry

engineering experience, he holds a bachelors
degree in mechanical engineering from
Southern Methodist University and a masters
from the University of Texas. PE licensed, he
holds seven US patents, received numerous
professional awards, and is a Fellow of the
American Society of Mechanical Engineers.
Email: kfewel@technip.com

A New Way
bolted bulk material silos
for all kinds of bulk solids
simple and quick installation
short delivery times
comes with complete finish
resistant to explosion and
pressure shock
SILOBAU THORWESTEN GmbH, D-59269 Beckum (Germany), www.thorwesten.com

TW_Silobau_Anz_A_new_way_PTQ_1_3_17_125x178.indd 1

www.eptq.com

q1 technip.indd 6

08.12.16 12:45

PTQ Q1 2017 89

14/12/2016 13:11

OPTIMIZE
INJECTOR

PERFORMANCE

Torch Oil Injector

SUPERIOR SPRAY. SERIOUS RESULTS.

Whether you need to cool gas, dissolve salts in an overhead line or inject chemicals to prevent corrosion,
we can help optimize injector performance. Here's how:
Assistance with nozzle selection, spray direction and injector placement. There are dozens
of factors to consider before choosing a spray nozzle, determining whether to spray co- or counter-current
and identifying the proper placement of an injector in a vessel. We can help you evaluate your process
conditions and then design an injector to provide optimal performance
Design validation using Computational Fluid Dynamics (CFD) and Fluid Structure Interaction (FSI).
We use powerful modeling tools to simulate your environment, confirm the injector will provide the
expected spray performance and withstand process conditions such as thermal stresses, heat transfer,
vortex shedding and more
Proven track record. Companies like Technip, Mustang Engineering, Bechtel, Shell and many others
rely on us to manufacture B31.1 and B31.3 code-compliant injectors and conduct radiographic,
hydrostatic, ferrite tests and more

Learn More. Call 1.800.95.SPRAY or visit spray.com/injectors

CFD MODEL ILLUSTRATES SPRAY
CHARACTERISTICS BASED ON INJECTOR
PLACEMENT IN PIPES & VESSELS

NEW STEAMMAX NOZZLES

USE PLANT STEAM TO
ATOMIZE LIQUID INSTEAD
OF COMPRESSED AIR

SLURRY RECYCLE INJECTOR.

DOZENS OF OTHER TYPES
ALSO AVAILABLE

U NMATCH ED G L O B A L EN GINEER ING, MANUFAC TUR ING & TEC HNIC AL S UP P OR T

N O ZZL ES | CO NTRO L SYSTEMS | HEADER S & INJEC TOR S | R ES EAR C H & TES TI N G

Balanced distillation equipment design

Fouling resistance and efficiency requirements for distillation equipment are
balanced and optimised for reliable unit performance
SOUN HO LEE
GTC Technology

ouling tendency is a critical

issue in crude distillation units
and should not be overlooked
when designing crude distillation
columns. Corrosion tendency can
influence fouling issues as well.
Since fouling resistance has an
inverse relationship to efficiency in
distillation equipment design, optimising equipment design between
fouling resistance and efficiency
requirements must be precise. Poor

application know-how as well as

poor equipment design often
downgrade column performance
and reduce unit run length.
This article will discuss common
fouling issues associated with crude
distillation column design. Actual
retrofits for crude atmospheric
columns are demonstrated through
two case studies. These studies
examine how fouling resistance and
efficiency requirements for distilla-

tion equipment are balanced and

optimised through careful evaluation and design methodologies.

Case study 1: crude distillation unit

description and background

The configuration of the crude

distillation unit in this case is illustrated in Figure 1. Fractionated light
and heavy kerosene streams
through the crude atmospheric
column and the side strippers are

Off gas

Crude
atmospheric
column

Unstabilised
naphtha

Steam
Light kerosene

Heavy
kerosene
pumparound

Kerosene

Steam
Heavy kerosene

Light diesel
pumparound

Steam

Preflash
drum
Desalted
crude

Light diesel

Heavy diesel
pumparound

Diesel
Steam
Steam

Heavy diesel

Atmospheric
residue

Figure 1 Case study 1: crude distillation unit configuration

www.eptq.com

Q1 gtc.indd 1

PTQ Q1 2017 91

14/12/2016 13:16

fractionation sections. Kerosene or

diesel intermediate product yield
limitation was also experienced
when kerosene or diesel boiling
range material composition was
increased in the charged crude
slate. Charge crude compositions
were frequently varied during
operation.

Case study 1: root cause

identification

Figure 2 Case study 1: valve/perforation hole wearing and corrosion

Figure 3 Case study 1: underside view of fouled trays

combined and rundown as a single

kerosene
intermediate
product
stream. A diesel intermediate product stream is also formed from a
combination of light diesel and
heavy diesel streams. These crude
atmospheric columns and side
strippers were originally designed
with conventional movable valve
trays, traditionally selected in the
past. The exception was the wash
section which was arranged with
structured packing. Three pumparound circuits are arranged at the
heavy kerosene, light diesel and
heavy diesel range material locations.
The
naphtha/kerosene
fractionation section is positioned
as the crude atmospheric column
top section. This column was
designed without a top pumparound circuit in order to maximise
fractionation between unstabilised
naphtha and kerosene at a given
column height.1

92 PTQ Q1 2017

Q1 gtc.indd 2

This crude distillation unit faced

two problems: fouling and corrosion of the distillation equipment in
the unit were found during a turnaround inspection. Valve perforation
hole wearing and corrosion were
found in the trays for naphtha/

Fouling tendency
is a critical issue in
crude distillation
units and should not
be overlooked when
designing crude
distillation columns
kerosene fractionation. Tray fouling
was also identified in the trays for
the light kerosene/heavy kerosene
and heavy kerosene/light diesel

Figure 2 shows that the naphtha/

kerosene
fractionation
trays
suffered from valve/perforation
hole wearing and corrosion. Some
movable valve units dislodged
from the tray deck. Perforation hole
sizes on the tray deck were
increased by wearing and corrosion
actions.2 Low column top temperature required for target operation
could accelerate hydrochloric acid
corrosion and valve/perforation
hole wearing. Significant fractionation
efficiency
loss
between
naphtha and kerosene was not
recognised during the operation.
The bulky fractionation nature of
crude distillation service might
result in fractionation efficiency
being insensitive to tray weeping.
However, if this valve/perforation
hole wearing progresses, significant
fractionation efficiency loss will be
noticed
through
substantial
weeping.
Fouled trays located for the light
kerosene/heavy
kerosene
and
heavy kerosene/light diesel fractionation sections are shown in
Figure 3. A tar-like substance was
discovered around the periphery of
the valve legs. Phosphates used for
crude
oil
production
were
suspected as the root cause. Boiled
phosphates may react with kerosene boiling range material and
make fouling deposits.
A dedicated process evaluation
for kerosene or diesel yield limitation was conducted. The original
column and tray drawing revealed
that intermediate side product and
pumparound streams were withdrawn from fractionating trays
directly. The originally designed
side draw configuration is illustrated in Figure 4. Flow from the
crude atmospheric column to the
side stripper relies on gravity flow.

www.eptq.com

14/12/2016 13:17

If the liquid head formed on the

collector tray is not high enough to
overcome total friction losses from
the crude atmospheric column to
the side stripper, the flow rate can
be limited. Moreover, frothy liquid
withdrawn from the fractionating
trays active area can contain
vapour. The presence of vapour can
limit this gravity flow. Rigorous
pipe line hydraulic evaluation
revealed that the gravity line
hydraulics could be limited at a
maximum target draw rate.3

www.eptq.com

Q1 gtc.indd 3

LC

Side
stripper

Pumparound
LT

Steam
Rundown
product

Figure 4 Case study 1: original design side draw configuration

0.7

Freezing-point reduction, F

Based on the aforementioned

process evaluation and root cause
analysis, the original movable valve
trays were replaced by fixed valve
trays. This tray type conversion
improved equipment resistance
against fouling and valve/perforation hole wearing.
The original fractionating trays at
draw locations were converted to
chimney trays to increase the liquid
head for gravity flow. This chimney
tray conversion also eliminates the
chance of yield loss and start-up
trouble through fixed valve tray
implementation and increases draw
liquid residence time for vapour
disengagement
from
liquid.
However, this conversion resulted
in losing one tray for each fractionation section: light kerosene/heavy
kerosene, heavy kerosene/light
diesel and light diesel/heavy diesel
fractionation. GT-Optim high performance
trays
with
various
performance-enhancing features and
fixed valves were implemented for
the rest of the fractionating trays.
As described earlier, light and
heavy
kerosene
streams
are
combined and rundown as a single
kerosene
intermediate
product
stream. Therefore, fractionation
performance between light kerosene and heavy kerosene streams is
not critical. The same rundown
configuration of light and heavy
diesel streams does not necessitate
sharp fractionation between the
two streams. However, fractionation performance between heavy
kerosene and light diesel streams
affects rundown kerosene and
diesel intermediate product quali-

FT

0.6
0.1F

0.5
0.4

Pre-retrofit
test run

0.3

Retrofit design
prediction

0.2
0.1
0

Number of theoretical stages gain for

heavy kerosene light diesel fractionation section

Figure 5 Case study 1: pre-retrofit sensitivity analysis

1.4

Freezing point reduction, F

Case study 1: equipment

modification

Crude
atmospheric
column

FT

1.2
1.0
0.1F

0.8
3.0%

0.6

Pre-retrofit
test run

0.4

Retrofit design
prediction

0.2
0

10

20

30

40

50

60

70

80

90

Top simulated reflux L/V for

heavy kerosene light diesel fractionation section, %
Figure 6 Case study 1: pre-retrofit sensitivity analysis

ties including the kerosene freezing

point, one of the key specifications
for kerosene rundown. To predict
kerosene freezing point change,
dedicated sensitivity analysis was
conducted.

The aforementioned high performance tray implementation could

improve individual tray efficiency.
Nevertheless, extra individual tray
efficiency improvement was not
counted to predict the retrofit

PTQ Q1 2017 93

15/12/2016 13:41

Case study 1: performance summary

Case
Pre-retrofit
Post-retrofit
Parameter
Test run
Test run
Yield balance
Crude charge, BPD
Base
+12
Unstabilised naphtha, LV%
Base
+6.0
Kerosene, LV%
Base
+19.5
Diesel, LV%
Base
+12.1
Fractionation performance
Base
+1.0
Light kerosene 5% - naphtha 95%,1 F
Light diesel 5% - heavy kerosene 95%,1 F
Base
-1.0
Base
Kerosene flash point, F
0
Base
+0.2
Kerosene freezing point, F
Heavy kerosene/light diesel internal reflux L/V,2,3 weight basis %
Base
-15
Heat balance
Heat removal - overhead condenser,2 % of total BTU/hr
72
62
4
6
Heat removal - heavy kerosene pumparound,2 % of total BTU/hr
Heat removal - light diesel pumparound,2 % of total BTU/hr
7
10
Heat removal - heavy diesel pumparound,2 % of total BTU/hr
16
21
1. ASTM D86 (LV%)
2. Simulated value
3. At the top tray of the section

Table 1

heavy kerosene/light diesel fractionation performance.

0.1F by using a chimney tray

conversion scenario. Figure 6 also
shows another kerosene freezing
point sensitivity curve per heavy
kerosene/light diesel fractionation
section internal reflux L/V (liquid/
vapour) ratio. A freezing point
increment of 0.1F was predicted at
a 3% lower heavy kerosene/light
diesel fractionation section internal
reflux L/V ratio. Undetected kerosene freezing point changes were
anticipated through the sensitivity
analysis.

Case study 1: sensitivity analysis

For reliable sensitivity analysis,

simulation modelling was first validated with pertinent unit test run
conditions. Simulated kerosene
freezing point value was reasonably matched to actual value. The
tray efficiency and internal vapour/
liquid traffic profile for each fractionation section were quantified
through model validation. A
constructed kerosene freezing point
sensitivity curve per varied theoretical stages is plotted in Figure 5.
This curve predicted that the freezing point could be increased by

Case study 1: performance summary

The pre- and post-retrofit performances are summarised and

compared in Table 1. Both retrofit

Freezing point reduction, F

1.4
1.2

Post-retrofit
test run

1.0

15%

0.8

0.2F

0.6

8%

0.4

Retrofit design
prediction

0.2
0

10

20

30

40

Pre-retrofit
test run

50

60

70

80

Top simulated reflux L/V for

heavy kerosene light diesel fractionation section, %
Figure 7 Case study 1: pre- and post-retrofit sensitivity analysis

94 PTQ Q1 2017

Q1 gtc.indd 4

90

test run conditions were obtained

through the same operating mode,
SOR (start of run) for fair comparison. Since the internal vapour/
liquid traffic and heat balances of
the crude atmospheric column were
not measurable operating parameters, these values were quantified
through simulation modelling. Like
the simulation model for the
pre-retrofit case, the simulation
model for the post-retrofit case was
also
validated
with
selected
post-retrofit test run conditions.
The crude charge rate was
increased during the post-retrofit
test run. The product yield balance
reveals that the post-retrofit test
run crude slate contained more
kerosene boiling range materials
compared to pre-retrofit crude
slate. Kerosene and diesel yield
limitation experienced in the past
was eliminated.
The simulated crude atmospheric
column heat balance of the postretrofit case was also shifted from
that of the pre-retrofit case due to a
change in the crude slate composition. Laboratory test results showed
that the post-retrofit kerosene freezing point was relaxed by 0.2F.
Meanwhile, simulation modelling
showed that the post-retrofit internal reflux L/V ratio for the heavy
kerosene/light diesel fractionation
was reduced compared to the
pre-retrofit value.
The heavy kerosene/light diesel
fractionation section performance
through the post-retrofit test run
was evaluated and compared to the
pre-retrofit section performance.
The post-retrofit kerosene freezing point was plotted and
compared to the pre-retrofit sensitivity curve in Figure 7. Relaxing
the kerosene freezing point by 0.2F
predicted an internal reflux L/V
ratio reduction of 8% on the
pre-retrofit sensitivity curve. A
lower internal reflux ratio of 15%
was simulated with the same 0.2F
freezing point relaxation at the
post-retrofit test run conditions.
Results indicated that the actual
post-retrofit heavy kerosene/light
diesel fractionation section efficiency was more satisfactory than
the predicted retrofit efficiency
value.

www.eptq.com

15/12/2016 13:41

es
ot
N
Process
Crude units can be designed for reliability.

Maximize Reliability in
Grassroots Crude Units

DESALTING

Crude unit operators are far too familiar with a long list
of crude unit reliability problems including fouling in
heat exchangers and fired heaters, poor desalting, corrosion of piping and equipment, and coking in the vacuum column wash zone. Many millions of dollars have
been spent fighting these problems, yet they continue
to force unplanned shutdowns with depressing regularity.
Revamps must address reliability issues, but project
scope is hindered by the limitations of existing equipment. Grassroots design of crude and vacuum units
presents an opportunity to get everything right the first
time. Here are a few tips for designing a reliable and
profitable crude/vacuum unit.

HEAT EXCHANGER AND HEATER FOULING

High velocities in heat exchanger tubes produce high
shear at the walls, preventing foulants from accumulating. High shell-side velocities, coupled with exchanger designs that minimize dead zones in the flow,
eliminate shell-side fouling. In fired heaters, high mass
fluxes maximize wall shear, shorten residence time,
and lower wall film temperatures, all of which reduce
coking. Furthermore, reliable heaters must have correctly sized burners with proper burner-to-burner and
burner-to-tube spacing.

Desalter size is highly dependent on crude blend due

to dramatic variation in required centerline velocity.
A unit must be designed with the flexibility to carefully control desalter temperature, which can range
from 110C to 150C, by shifting heat from upstream to
downstream of the desalters. Vendors are often judged
on cost alone, which results in minimum sizing for the
design crudes and rates. Carefully consider whether
long-term crude trends will soon render these desalters inadequate.

CORROSION
In grassroots design, be realistic about metallurgy. Because modern refineries do not run a steady diet of the
same crude, consider the sulfur and TAN numbers of
potential crudes outside the units design blend. Chronic corrosion issues, or the inability to process high-margin opportunity crudes, will quickly overshadow the initial savings from choosing too low of a metallurgy.

VACUUM COLUMN WASH ZONE COKING

Wash zones are not for fractionation, they are for
de-entrainment! Pursuing fractionation efficiency by
specifying a deep bed with small crimp packing is a
recipe for rapid coking. The correct choice of packing
combined with the right wash rate and good distribution will properly de-entrain while preventing coke formation.

3400 Bissonnet St.

Suite 130
Houston, TX 77005, USA

+1 (713) 665-7046
info@revamps.com
www.revamps.com

Case study 1: fractionation efficiency comparison

Heavy kerosene/light diesel
Fractionation section
Kerosene freezing point, F
Internal reflux L/V,1,2 weight basis %
Section efficiency,1,3 %

Pre-retrofit
Test run
Base
Base
55

Retrofit
Prediction
+0.1
- 3
50

Post-retrofit
Test run
+0.2
- 15
65

1. Simulated value
2. At the top tray of the section
3. Overall efficiency (number of theoretical stages/number of fractionating trays)

Table 2

Table 2 shows simulated heavy

kerosene/light diesel fractionation
section efficiencies. As one fractionating tray was converted to a
chimney tray, a lower number of
theoretical stages were counted for
the retrofit performance prediction.
Nevertheless, the simulated theoretical stage count through the
post-retrofit test run condition data
was maintained at the same level as
the pre-retrofit theoretical stage
count, resulting in improved fractionation efficiency between heavy

kerosene and light diesel. The

aforementioned GT-Optim high
performance tray implementation
contributed to the fractionation efficiency improvement.

Case study 2: crude distillation unit

description and background

The second case study also examines a crude distillation unit. Figure
8 illustrates the schematic of the
unit. The crude atmospheric
column in this case was designed
as a fully structured packed frac-

tionator excluding the bottom

stripping section.
Unstabilised naphtha, kerosene
and atmospheric gas oil intermediate products were distillated
through the crude atmospheric
column and the side strippers. The
unstabilised naphtha stream was
further separated into LPG (liquefied petroleum gas), light naphtha,
and heavy naphtha through the
naphtha stabiliser and naphtha
splitter. The heavy naphtha stream
was directed to the reforming unit
for aromatic component production. The unit utilised two
pumparound circuits as well as an
overhead condenser for heat
removal. One circuit was positioned as the top pumparound. The
other circuit was located between
kerosene and atmospheric gas oil
boiling range materials. A second
packed bed at the top of the crude
atmospheric column had the function of fractionating naphtha and
kerosene boiling range materials.

Off gas

Crude
atmospheric
column

LPG

Light
naphtha

Top
pumparound

Naphtha/kerosene
fractionation section
Heavy
naphtha
Steam

Bottom
pumparound

Kerosene

Preflash
drum
Steam

Desalted
crude

Atmospheric
gas oil
Steam
Atmospheric
residue

Figure 8 Case study 2: crude distillation unit configuration

96 PTQ Q1 2017

Q1 gtc.indd 5

www.eptq.com

14/12/2016 13:17

Register your attendance now at artc.wraconferences.com. Save 15% on your booking by using the code MP15
Register your attendance now at artc.wraconferences.com. Save 15% on your booking by using the code MP15

20th Annual ARTC Meeting

20th
Annual
ARTC
Meeting
29
- 30 March
2017, Jakarta,
Indonesia
29 - 30 March 2017, Jakarta, Indonesia

Asias most respected event for core refining technology

Asias most respected event for core refining technology
ARTC 2017
speakers
include:
ARTC 2017
speakers include:

Asit Das
Head of Refining
Asit Das
R&D
Centre
Head of Refining
Reliance
Industries
R&D Centre
Reliance Industries

Dempsy Robby Kambey

GM Asset Management
Dempsy
Robby
Kambey
PT
Kreasindo
Resources
GM
Asset
Management
Indonesia
PT Kreasindo Resources
Indonesia

Karambir Anand
Partner
Karambir Anand
EY
Partner
EY

Matthew George
Chief Manager,
Matthew George
Petrochemicals
Marketing,
Chief Manager,
Indian
Oil Corporation
Petrochemicals Marketing,
Indian Oil Corporation

Paul Kennedy
Vice President, Asia
Paul Kennedy
Solomon
Associates
Vice President, Asia
Solomon Associates

Michael Costello,
Technical Manager, JVs
Michael Costello,
Chevron
Technical Manager, JVs
Chevron

Clive Gibson
Partner
Clive Gibson
Nexant
Partner
Nexant

LTG Nugroho
Widyotomo
LTG Nugroho
Secretariat
General
Widyotomo
National
Resilience
Secretariat
General
Council
Indonesia
National Resilience
Council Indonesia

Are you a refiner or petrochemical producer?

The
Association
has reserved 50 complimentary
AreWorld
youRefining
a refiner
or petrochemical
producer? passes

for refiners and petrochemicals producers at attend the ARTC 20th

The World
Refining
Association
Annual
meeting
as our
guests. has reserved 50 complimentary passes
for
refiners
and
petrochemicals
at attend Producers
the ARTC 20th
Places are reserved for Refinersproducers
and Petrochemical
such as
Annual
meeting
as
our
guests.
Pertamina, Petronas, Petrovietnam, S-Oil, Thai Oil, Reliance Industries,
Places are Refining
reservedCompany
for Refiners
Singapore
andand
willPetrochemical
be given on a Producers
first come such
first as
Pertamina,
Thai Oil, Reliance Industries,
serve
basis. Petronas,
Apply for Petrovietnam,
your free pass S-Oil,
at artc.wraconferences.com
Singapore Refining Company and will be given on a first come first
serve basis. Apply for your free pass at artc.wraconferences.com

Contact:
Kelly
Tea :
Contact
kelly.tea@clarionevents.com
Kelly207
Tea384 7807
+44
kelly.tea@clarionevents.com
artc.wraconferences.com
+44 207 384 7807
artc.wraconferences.com

Case study 2: distributor modification summary

Naphtha/kerosene fractionation
Distributor drip point density, drip point/ft2
Distributor operating range
Drip hole elevation,1 inch
Distributor drip hole diameter, inch

Original
16
2.5:1
1.5
3/16

Modification
4.6
1.7:1
3
7/16

1. From the bottom of the trough

Lv

H =

k N HA

Table 3

A particular unit limitation the

rener faced was that fractionation
performance between naphtha and
kerosene was substantially downgraded after a four-month operation.
Substantial amounts of kerosene
boiling range materials were downgraded to the naphtha stream. This
downgrading not only limited the
kerosene yield but also inuenced
the downstream reforming unit
performance. The high rear end
distillation point of the heavy naphtha stream adversely affected the
reforming reactor catalyst activation.

Case Study 2: root-cause

identification

Inspection during unit turnaround

showed that the trough-type liquid
distributor for the naphtha/kerosene fractionation section was
fouled. Several root causes of the
fouling were identied through
rigorous evaluation.
Amine-based corrosion inhibitor
was injected into the crude distillation unit. Chloride present in the
column overhead may react with
the inhibitor and form ammonium
salt, which can foul the distributor.
Formed ammonium salt particles
could reside in the top pumparound
bed and also migrate to the naphtha/kerosene fractionation bed.

Findings
showed
that
the
columns inside wall cladding
using Monel metallurgy was only
applied to the portion where the
top pumparound section was positioned. The column inside wall
portion of the naphtha/kerosene
fractionation section remained as
carbon steel. Rusted wall pieces
from corrosion could accelerate
fouling.

A minimum liquid
head needs to be
maintained to ensure
uniform liquid
distribution
A review of the original distillation equipment drawing revealed
that the gravity ow trough-type
liquid distributor for the naphtha/
kerosene fractionation section was
designed with high drip point
density and small drip hole size.
The naphtha/kerosene fractionation packed bed was equipped with
structured packing with a 1in crimp
size and a 45 inclination angle.
This packing size at the given bed
height was suitable to achieve

Temperature, F

330

Post-retrofit
Pre-retrofit

320
310
300
290
280
270

50

100

150

200

250

Unit run length, days

Figure 9 Case study 2: performance trend naphtha end point

98 PTQ Q1 2017

Q1 gtc.indd 6

target
fractionation
efciency.
However, the liquid distributor
drip point density originally
selected was excessive for the 1in
crimp size.
The basic equation used to size
gravity liquid distributors is:

300

350

400

H = Liquid height (Head) above

round shaped hole
Lv = Liquid volumetric ow
N = Number of drip holes
HA = Hole area
K = Orice coefcient
A minimum liquid head needs to
be maintained to ensure uniform
liquid distribution. A certain
number of drip holes, which indicates drip point density, is required
for the desired distribution quality.
However, unnecessarily high drip
point density reduces distributor
drip hole size and increases a chance
of fouling. Distributor operating
range affects hole size because the
liquid head should be maintained at
the minimum rate for uniform
liquid distribution. The original
liquid distributor (as designed) was
not properly optimised between
fouling resistance and liquid distribution quality.

Case study 2: distributor

modification

The liquid distributor modications

for the naphtha/kerosene fractionation section are summarised in Table
3. In order to enlarge liquid distributor drip hole size, the drip point
density was reduced in a new
design. The new density was carefully selected by considering the
commercially proven drip point
density in the given size packing
and application. Distributor operating range was also adjusted further
to increase distributor drip hole size.
The minimum end of the distributor
operating range was increased. This
adjusted distributor operating range
does not reduce the unit operating
range. The minimum rate of the
liquid distributor does not have to
be matched to the minimum unit
charge rate.4 Heat balance shifting
through pumparound adjustments
or increasing furnace coil outlet

www.eptq.com

15/12/2016 16:35

114
112

Temperature, F

temperature can maintain the

required minimum distributor rate
during lower unit charge rate operation. This strategy can increase
energy consumption during minimum charge rate operation. But it
can assure efcient unit operation in
the entire charge range and more
efcient overall unit economics can
be achieved. Distributor drip hole
elevation from the bottom of the
trough was increased to slow down
distributor fouling.
The measured naphtha end point
and kerosene ash point trends are
plotted in Figure 9 and Figure 10
respectively. Plots in red indicate
values gathered during pre-modication operating periods while
plots in blue represent values
achieved after the modication.
Stable naphtha end points and
kerosene ash points were maintained for more than eight months
of operation.
These case studies show how
fouling resistance and efciency
requirements for distillation equipment are balanced and optimised

110
108
106
104
Post-retrofit
Pre-retrofit

102
100

50

100

150

200

250

300

350

400

Unit run length, days

Figure 10 Case study 2: performance trend kerosene flash point

for reliable crude distillation unit

performance.
This article is an updated version of a
presentation given at AIChEs Spring Meeting
Distillation Symposium, 11-14 Apr 2016, in
Houston, TX.
GT-OPTIM is a mark of GTC Technology US
LLC.
References
1 Lee S H, et al, Optimising crude unit design,
PTQ, Q2 2009.

Middle East Sulphur 2017

2 Kister H Z, Distillation Operation, McGrawHill Company, 1990.

3 Libermann N P, Process Design for Reliable
Operations, Gulf Publishing Company, 2nd
Edition.
4 Bonilla J A, Dont neglect liquid distributors,
Chemical Engineering Progress, Mar 1993.
Soun Ho Lee is Manager of Refining
Application for GTC Technology US, LLC, in
Euless, Texas, specialising in process design,
simulation modelling, energy saving design
and troubleshooting for refining and aromatic
applications. Email: Sounho@gtctech.com

Platinum Sponsor:

12-16 February 2017 Jumeirah Hotel At Etihad Towers, Abu Dhabi, UAE

Exploring the sulphur value chain

Organisers of the 33rd Sulphur Conference

CRU is delighted to announce a new conference devoted to exploring the entire sulphur value chain and the growing role of
the Middle East as the worlds largest producer of Sulphur.

The commercial programme will be devoted to:

The technical sessions will focus on:

A thorough exploration of economics underpinning the oil

and gas markets and the subsequent impact on the global
sulphur market
The role of the Middle East as the worlds largest sulphur
producing region
Demand updates from key consumers
Maintaining the quality of sulphur and identifying alternative
uses for sulphur and new areas for sulphur demand

Promoting best practice on the technical and operational aspects

of sulphur production with a particular focus on establishing
operational best practice via the sharing of operational
experience through operator-led papers incorporating:
The latest technologies and case studies for sulphur
production
SRU performance and reliability
Energy efficiency
Emissions compliance
Sulphur forming & handling

Event Partners

Official Publications

Media Partners

CRU
VIEW
Global update
on the sulphur
market

Technical
workshops &
training

Exhibition &
sponsorship
Enquiries contact:
Michelle Fisk
+44(0)20 7903 2159

To book your delegate place go to www.middleeastsulphur.com and register TODAY.

www.eptq.com

Q1 gtc.indd 7

PTQ Q1 2017 99

14/12/2016 13:18

WWW.ZWICK-ARMATUREN.DE

TRI-SHARK
100 % CONTROL VALVE
100 % TIGHT

Firing high sulphur fuel

An evaluation of schemes to improve firing efficiency and ensure the reliability
of combustion equipment
ADIL REHMAN, C STEVEN LANCASTER, SANDEEPAN GHOSH, OM PRAKASH SAHU and PAWAN KUMAR SHARMA
KBR Technology

ressure to reduce the carbon

footprint of processing facilities
means a reduction in fuel consumption by the energy consuming
plant. At the same time, the ever
increasing price of fuel means that
refiners are compelled to utilise
cheaper fuels like fuel oil, refinery
gas, or sub-grade fuels with a high
content of sulphur and other impurities. On the credit side, there is the
possibility of increased profit margins as a result of increasing energy
demand worldwide.
Thus refiners and petrochemical companies aim for process
revamps in order to improve their
profit margins as well as decrease
their carbon footprint. Furnaces are
major consumers of fuel in a typical
refinery or a petrochemical plant,
and various process schemes are
explored here in order to process
dirty fuels without severe corrosion
problems or compromising the life
of the equipment and avoiding any
unwanted shutdowns.

Basis of study

For the sake of the study, hypothetical base case and revamp case
models have been developed, and
simulations for a hydrotreater fractionator feed furnace have been
carried out using typical industrial
data and studied using commercial
software, FRNC-5PC Version 4.18
Mod 7.6.

Definition of cases

The various process cases considered for the study are as follows:
Base Case A is the existing base
case of a natural draft heater with
an absorbed duty of 15 Gcal/h.
Revamp Case B1 [Option I] is

www.eptq.com

q1 kbr.indd 1

Exhaust to
safe location

Fuel

Combustion
system
Ambient air
Air circuit
Flue gas circuit

Figure 1 Existing process scheme (Base

Case A)

the revamp case of a balance draft

heater with an absorbed duty of
15 Gcal/h with on-board heat
recovery in the form of a conventional air preheater with multiple
carbon steel tube bundles.
Revamp Case B2 [Option II] is
the revamp case of a balance draft
heater with an absorbed duty of
15 Gcal/h with heat recovery in the
form of an on-board glass coated
tube type air preheater.
Revamp Case B3 [Option III] is
the revamp case of a balance draft
heater with an absorbed duty of
15 Gcal/h with heat recovery in
the form of an off-board ground
mounted cast and glass type air
preheater.
Revamp Case B4 [Option IV] is
the revamp case of a balance draft
heater with an absorbed duty of
15 Gcal/h with heat recovery in
the form of an off-board ground
mounted cast type air preheater
with an upstream steam air preheater required to maintain the lowest tube metal temperature above
acid dew point.
Revamp Case B5 [Option IV] is

the revamp case of a balance draft

heater with an absorbed duty of
15 Gcal/h with heat recovery in the
form of an off-board glass coated
tube type air preheater.

Comparing the base case scheme

with revamp schemes

Basic schematics of the existing base

case along with various other process revamp schemes are shown in
Figures 1, 3, 4, 5, 6, and 7.

Methodology for study

Typical base case and revamp case

process flow rate, duty of the furnace, process inlet and outlet temperature, process inlet and outlet
pressure have been considered and
a feed property grid has been generated using commercially available
process simulator PRO II. Results of
the various process schemes were
compared to evaluate their relative
advantages and disadvantages
The study was carried out with
the following assumptions and data
for modelling: the base case is considered to be a process scheme with
a 15 Gcal/h natural draft heater
without any heat recovery, radiant
heat loss of 1.5%, excess air of 20%,
ambient air datum temperature
of 15.6C and relative humidity of
60%. The revamp case is considered
to be a process scheme with a 15
Gcal/h balanced draft heater with
heat recovery, radiant heat loss of
2.5%, excess air of 15%, ambient air
datum temperature of 15.6C and
relative humidity of 60%.
The typical refinery fuel gas composition in mole% is as follows:
hydrogen, 44; methane, 13; ethane, 15;
propane, 22; isobutane, 5; hydrogen
sulphide, 0.6; and isopentane: 0.4.

PTQ Q1 2017 101

14/12/2016 13:44

Parameters and simulation results for base case and revamp cases
Step 1

Building the base case simulation

model, and running the simulation
base case.
Step 2

Building the revamp case

simulation models, and running
the simulation revamp cases.
Step 3

Evaluating various process

schemes available for efficiency
improvement for the revamp cases.
Step 4

Comparison of various revamp

options with the base case.
Step 5

Comparison of various revamp

options with each other.
Step 6

Evaluating merits/demerits of
various revamp options.

Heater system under study Base

Revamp
Case A Case B1

Process flow, kg/h
259 416 259 416
Process in/out temp., C
317/378 317/378
Absorbed duty, Gcal/h
15
15
Fired duty, Gcal / h
18.47
16.72
1551
1411
Fuel flow, kg/h
% excess air
20
15
Efficiency,%
81.2
89.7
Acid dew point,C
140
140
Lowest tube metal temp., C
293
155
Max tube metal temp., C
415
428
Avg. radiant heat flux, Kcal/h.m2 18 869
19 617
Bridge wall temp., C
787
797
APH duty, Gcal/h
N/A
1
APH air in/out temp., C
N/A
851/233
105
95
Total CO2 emitted MTPD
NOx2 emission, ppm
46
81

Revamp Revamp Revamp

Case B2 Case B3 Case B4
259 416 259 416 259 416
317/378 317/378 31/378
15
15
15
16.25
16.25
16.72
1372
1372
1411
15
15
15
92.3
92.3
89.7
140
140
140
N/A3
1555
155
429
429
428
19 758
19 758 19 617
799
799
797
1.8
1.8
1
15.64/295 15.66/295 856/233
93
93
95
95
95
81

Revamp
Case B5
259 416
317/378
15
16.25
1372
15
92.3
140
N/A3
429
19 758
799
1.8
15.64/295
93
95

1 During low ambient conditions such as in winter, a steam air preheater or electric heater may be required
to increase air inlet temperature for avoiding any acid corrosion issues.
2 Considering low NOx burner design.
3 Lowest tube metal temperature is not applicable here as the metal tubes are coated with glass. Thus
more appropriate terminology may be the lowest glass temperature. Acid condensation would not cause any
corrosion problem as metal is coated with glass in the cold zone.
4 During low ambient conditions, the air inlet temperature would be lower but would not have any acid corrosion
issues due to glass tubes construction in the cold zone.
5 Lowest tube metal temperature shall be maintained by vendor in the cast section.
6 During low ambient conditions, the air inlet temperature would be required to be increased using a steam air
preheater and/or air may be partly bypassed for avoiding any acid corrosion issues.
*Note the figures mentioned in the tables are indicative in nature and meant for study purposes only.

Table 1
Step 7

Conclusion/inference.

Exhaust to
safe location

Figure 2 Algorithm followed for the study

A typical algorithm followed for

the purposes of the study is shown
in Figure 2.
Important parameters and simulation results for the base case and
revamp cases are shown in Table 1.

Results and discussions

Revamp Option I

This scheme is identical to the base

case scheme except that conventional
on-board metallic tube bundles have
been used for heat recovery, thereby
targeting improvement in fuel efficiency. Overall structural weight
would be increased due to the addition of an on-board air preheater,
thereby increasing load on the foundation as compared with the existing
base case configuration.
Although extra footprint is not
envisaged for an on-board air preheater, ducting and footprint would
be required for a flue gas fan and
combustion air fan to take care of
extra pressure drop across the flue

102 PTQ Q1 2017

q1 kbr.indd 2

On-board
conventional
APH
Fuel

Combustion
system
Preheated
air
Air circuit
Flue gas circuit

Figure 3 Process scheme (Option I:

Revamp Case B10

gas side and air side. Thus, where

there is a plot limitation, for example in a revamp, such a scheme is a
feasible option for improving fuel
efficiency, provided space is available for fans and ducting.
In the event of acid condensation, the carbon steel tubes of an
on-board air preheater are expected
to corrode at an estimated rate of
approximately 2 mm/y. Thus for
low ambient or winter conditions, a

steam air preheater is envisaged for

avoiding any acid corrosion issues,
which will add to the hardware
requirement of the Option I configuration. This situation is not envisaged with the base case as there is
no heat recovery and no chance of
acid corrosion.
Efficiency
improvement
for
Option I is around 9% compared
with the base case and thus would
bring down fuel consumption substantially. CO2 emissions would fall
from 105 to 95 t/d, thereby saving 3300 t/y of CO2 being emitted
to the environment, considering
330 on-stream operation days.
However, quantities of other emissions like NOx will go up for the
revamp case compared with the
base case configuration.
Because this scheme introduces
fans, electrical power consumption
would be required compared with
the base case, which essentially does
not require any electric consumption.
Soot deposits on the outside of the
tubes, especially in the case of fuel
oil firing, is a matter of concern as it
reduces the efficiency of the system.
In the case of on-board air preheater

www.eptq.com

14/12/2016 13:45

"

Comparison with the literature method



A simple example is illustrated here, which

designs, water washing in general is not recommended
uses the above relation in Equation 14 to
for cleaning of soot since water may spill below Tap
and
calculate the total aeration rate for the standnumber
may cause unnecessary damage to coils, refractory
Inlet
pipe in a FCC unit. If, in a standpipe, density
and so on. Thus
steam cleaning can be recommended
1
of 575 kg/m3 (mf) is to be maintained for
by installation of soot blowers in the on-board air 2presmooth catalyst circulation, with catalyst skele3
heater region. This would 3require installation of access
4
tal density of 2500 kg/m (sk), circulation rate
platforms for the operation and maintenance of these
5
of 100 T/min (MC) and top pressure of 3.8 kg/
6
soot 2blowers.
cm (PT) for a standpipe with height of 16 m
7
Carbon steel tube bundles are easy to manufacture
Total
(H), then from Equation 14:
and can be fabricated by most construction vendors
3
Qa =1789 mwith
/hr (29.81
m3/min)
compared
a cast
and glass type air preheater or a
Table 1
glass coated tube air preheater, which are specialised
If one calculates the gas requirement from
designs available from only a few vendors. This option
Equation 4 directly considering inlet pressure and outlet
in general is mechanically more robust compared with
pressure in the standpipe, one would get a total aeraa glass type air preheater
or glass coated tube air pretion rate of 1604 m3/h or 26.73 m3/min. But the sum of
heater, which are thought of as fragile in nature. Since a
aeration rate (total aeration rate) calculated from
carbon steel tube bundle is relatively robust compared
Equation 4 at different aeration points along the height
with a glass type air preheater or glass coated tube air
of the standpipe (see Table 1) comes to a value of 29.34
preheater,
there are no transportability issues with this
m3/min (1760 m3/h). This would be still clearer from a
design. Modularisation is also possible with this type of
calculation for determining aeration rate at different
design.
aeration taps by both correlations.
If designed and operated properly always maintaining the lowest tube metal temperature above acid
Location of standpipe aeration
dew point this type of design would require minimal
Once the desired amount of aeration medium flow is
maintenance. However, if acid corrosion is encounestablished, it is necessary to add it at appropriate
tered, then carbon steel tubes would corrode in a few
locations so as to get uniform density across the length
years and unwanted maintenance shutdowns/turnaof the standpipe. For FCC particles, it is necessary to
rounds may be required.
add aeration medium uniformly along the standpipe to
Delivery at site would be comparatively quicker
get uniform void fraction in the entire standpipe. If
than for a cast and glass type air preheater or glass
aeration is added at only one location, a large bubble
coated tube air preheater, which are relatively fragmay form at the aeration point. The large bubble may
ile. Construction in terms of installation and erection
obstruct the path of solid catalyst particles. Therefore
would be relatively easy and quick, and can be done
the aeration medium is added at several locations to
by most construction contractors compared with cast
avoid formation of a larger bubble and provide suffiand glass type or glass coated tube designs, which may
cient area for the aeration medium to dissipate through
require supplier supervision and services during instalthe very fine particles at a rate almost the same as the
lation and erection.
rate of addition at that time.
In a commercial FCC unit, the amount of aeration is
Revamp Option II
divided at several parts via aeration taps located
This scheme is identical to Option I except that conapproximately two to three metres apart. If the
ventional on-board metallic tube bundles have been
distance between several aeration taps is known, one
replaced by glass coated metallic tube bundles in the
can easily determine the aeration flow rate at every
cold zone region for heat recovery, thereby targeting
aeration tap, an example of which is shown in Table 1.
improvement in fuel efficiency. Overall structural
It is clear from Table 1 that the total aeration rates
weight would be increased due to the addition of
calculated from Equations 14 and 4 determined from
an on-board air preheater, thereby increasing load
the aeration rate for individual aeration taps are close
on foundations compared with the base case
enough compared with the total aeration rate calculated
configuration.
directly considering inlet and outlet pressure for the
Although extra footprint is not envisaged for an
standpipe. Integration over the length of the standpipe
on-board air preheater, ducting and footprint would
enables one to calculate the total aeration rate required
be required for a flue gas fan and combustion air fan to
for the standpipe directly from Equation 14 as against
take care of extra pressure drop across the flue gas side
Equation 4.
and air side. Thus where there is a plot limitation, for
example in a revamp, such a scheme is a feasible option
Conclusion
for improving fuel efficiency, provided that space is
From this discussion, it is apparent that the smooth
available for fans and ducting. Since metallic tubes are
operation of a standpipe requires appropriate aeration
coated with glass, acid corrosion issues are not envisto maintain constant void fraction throughout the
aged in this scheme.
length of the standpipe. A simple expression has been
Maximum efficiency improvement can be achieved
derived for determining the aeration rate at different
with Option II of around 11% compared with the base
case and thus would bring down fuel consumption

!

!

Aeration rate at aeration taps

substantially. CO2 emissions would fall from 105 to

Tap location
from saving
H,
Pressure,
required,emitted
m3/min
Acid gas
93t/d,
thereby
3960
t/y ofAeration
CO2 being
top of standpipe metres
kg/cm2
of aeration medium to SRU
to the environment, considering 330 on-stream opera3.8
Eq. 14
Eq. 4
tion days.
emissions3.00
like NOx will
Cooler
1.5 However,
1.5 other3.89
2.97 go
2.5case compared
4.03
4.86 the base
4.77case
up for 4.0
the revamp
with
Effluent
6.5
2.5
4.69
4.61
Stripper
configuration.
Cooler 4.17
pump
9.0
4.32
4.53
4.45
Effluent to 2.5
As this
scheme introduces
power
11.5
2.5
4.46fans, electrical
4.38
4.31conbiotreater
14.0 would be
2.5 required
4.60 compared
4.24 with the
4.18base
sumption
Pump
16.5
2.5
4.11
Feed/effluent
Flash
case. Soot
deposits
on gas
the4.75
outside
of
the tubes,4.05
espe16.5
29.81
29.34
Flash
cially in the
casedrum
of fuel oil firing, isexchanger
a matter of concern
HydroReboiler
as it reduces the efficiency of the
system. In the case of
carbons
Sour
to
slops
on-board air preheater designs, water washing in genprocess
eralwater
is not recommended for cleaning soot since water
aeration
taps.
It can be used to evaluate the theoretical
Sour
water
may spill.
Thus
steam cleaning can be recommended by
aeration rate
at
pump each aeration tap in the standpipe.
the installation
of soot blowers in the on-board air preheater
region.
Buffer tank
Reference
Sour waterdesign availThis
type
of
air
preheater
a specialised
1 Mott R W, Troubleshooting
FCCis
standpipe
flow problems, Catalagram
pump
able
106,from
11-20, only
2009. a few vendors. In general, it would be
mechanically less robust compared with an on-board
Figure 3 Pumparound sour water stripper scheme
metal
preheater
design
but may
slightly
betRahultube
C Patilair
is Senior
Manager
of the Refining
R&Dbe
group
with Reliance
in Navi
Mumbai,
India. He holdsmay
a masters
chemical
terIndustries
than aLtd
glass
type.
Transportability
be ainmatter
condensable
hydrocarbons
by allowing
engineering
fromliquid
IIT Kanpur
andmetal
a degree
in chemical
engineeringthe
from water in the ov
of
concern compared
with
designs.
water
to
de-pressure
and
settle
for
at
least
five
minutes.
ICT
Mumbai.
Modularisation is possible with this type of design. being sent to th
2.
degassed
water
flows
to a Refining
holding,
or at
settling
Ajay
Gupta
is Assistant
Vice
President,
Reliance 6. The stripped
TheThe
only
concern
is that
this
option
may R&D
offer
standtank.
In
this
tank
there
may
degassing
Industries
Ltd, Jamnagar,
Gujarat,
India.be
Hefurther
has bachelors,
mastersas
and exchanger then
ard module dimensions governed by glass coated tube
doctoral
degrees
in chemicalof
engineering
from IIT, Delhi, India.
well
as
the
separation
liquid
hydrocarbons
and
oil,
lengths and other fabrication aspects. However, tai- processing (cru
Asit K float
Das isto
Vice
President
the Refining
R&D
at Reliance and so on).
which
the
surfaceand
ofheads
thebewater.
If the
residence
lor-made
modularisation
may
offered
by suppliers
Industries
Ltd,
Jamnagar,
Gujarat,
India.
He
holds
a
degree
in
chemical
time request.
is high This
enough,
composition
of the
water
upon
type the
of
design
will require
minimal
engineering
from
Jadhavpur
University,
West
Bengal,
a
masters
stabilises and allows for a consistent flow andin Effects of hydro
maintenance.
chemical engineering from IIT, Kanpur, and a doctoral degree in chemical
Preventing hyd
engineering from Ghent University, Belgium.
stripper will in
from entering t
it is advantage
SRU feed, the m
Difficulties m
Decreased ca
Lower efficie
Potential iss
sulphur quality
Sulphur plan
reaction furnace
the feed gas to
stream catalytic
The modified C
its success is h
of H2S with o
water. Having
will limit the e
stages and imp
ciency of the SR
As previously
tion furnace b
However, since
stream reaction
furnace is cont
H2S is combust
furnace that op
makes the air
parameter.
In order to co
use a set air-to

www.eptq.com
www.eptq.com

q1 kbr.indd 3

PTQ Q3 2016 79
48 PTQ Q4 2016

PTQ Q1 2017 103

14/12/2016 13:45

Exhaust to
safe location
On-board
glass coated
APH
Fuel

Combustion
system
Preheated
air
Air circuit
Flue gas circuit

Figure 4 Process scheme (Option II:

Revamp Case B2)

Delivery at site may be in general

comparatively slower than for the
other options since manufacture
of glass coated components may
be limited to few manufacturers.
However, delivery time can be
improved and is a matter of further
optimisation during project execution. Construction and erection may
be relatively slow as an on-board
glass coated air preheater may not fit
exactly over the existing fired heater
convection section. This type of
design may require supplier supervision and services during installation.

Revamp Option III

This scheme is identical to Option II

except that an on-board glass coated
metallic tubular air preheater is
replaced by a cast and glass type offboard air preheater mounted on the
ground for heat recovery, thereby
targeting improvement in fuel efficiency. Overall structural weight
would be increased due to the addition of an off-board air preheater,
but load on the heater foundation
would be less compared with the
existing base case configuration and
on-board air preheater designs since
the load would not be on the furnace
but shifted to the ground.
Extra plot area is envisaged for
an off-board air preheater design
scheme. Ducting and footprint
would be required for a flue gas
fan and combustion air fan to take
care of extra pressure drop across
the flue gas side and air side. Thus
where there is a plot limitation

104 PTQ Q1 2017

q1 kbr.indd 4

such a scheme does not seem to be

a suitable option for improving fuel
efficiency.
Since this design comprises a
glass module in the cold zone, acid
corrosion issues are not envisaged.
Maximum efficiency improvement
can be achieved with Option III and
should be around 11% compared
with the base case, which would
bring down fuel consumption substantially. CO2 emissions would fall
from 105 to 93t/d, or 3960 t/y, considering 330 on-stream operation
days. Other emissions like NOx will
rise for the revamp case compared
with the base case.
As this scheme introduces fans,
electrical
power
consumption
would be required. In the case of
off-board air preheater designs,
water washing is generally available for soot cleaning. This would
require installation of access platforms for the operation and maintenance of soot blowers.
This type of air preheater is a
specialised design available from a
few vendors. It would in general be
mechanically less robust compared
with an on-board metal tube design.
Transportability may be a matter
of concern compared with metal
designs due to their more fragile
construction. The cast components also contribute to weight.
Modularisation is possible with this
type of design. The only concern is
that this option may offer standard
module dimensions governed by
glass tube lengths and other fabrication aspects. However, tailor-made
modularisation may be offered by
suppliers upon request.

Exhaust to
safe location
Off-board
ground mounted
cast APH

Steam APH

Fuel
Combustion
system
Preheated
air
Air circuit
Flue gas circuit

Figure 6 Process Scheme (Option IV:

Revamp Case B40

Exhaust to
safe location
Off-board
ground mounted
cast and glass
APH

Fuel
Combustion
system
Preheated
air
Air circuit
Flue gas circuit

Figure 5 Process Scheme (Option III:

Revamp Case B3)

This type of design may call for

maintenance along with service by
the supplier as and when required.
Breakage of glass tubes may also
occur during operation if excessive
vibration is induced due to factors
like fan motor and blade induced
vibration, and air and flue gas
induced turbulence. There are other
maintenance issues like damage of
polymeric sealing material in the
tube sheets of the glass module at
flue gas temperatures greater than
280C. However, these types of
issues may be addressed during the
project execution stage.
Delivery at site may be comparatively slow since the manufacture
of glass components may be limited
to a few manufacturers, whereas
installation and erection are comparable with other options. This
type of air preheater may require
supplier supervision/services during installation/erection.

Revamp Option IV

This scheme is identical to Option

III except that an off-board cast and
glass type air preheater has been
replaced by a cast only type offboard air preheater coupled with a
steam air preheater upstream of the
air circuit mounted on the ground
for heat recovery. Overall structural
weight would be increased due
to the addition of an off-board air
preheater and steam air preheater,
but load on the heater foundation would be less compared with
the base case configuration and
on-board air preheater designs since
the load of the air preheater would
not be on the furnace.

www.eptq.com

15/12/2016 13:42

Extra plot area is envisaged for an

off-board air preheater and steam
air preheater design. Ducting and
footprint would be required for a
flue gas fan and combustion air fan
to take care of extra pressure drop
across the flue gas side and air side.
Thus where there is a plot limitation
such a scheme does not seem to be
a suitable option for improving fuel
efficiency.
Since this design comprises a
steam air preheater to maintain the
lowest tube metal temperature above
acid dew point, acid corrosion issues
are not envisaged. However, there is
an additional requirement for steam
in the steam air preheater, so the utility requirement would be increased,
which might be a limitation for some
plants.
Efficiency improvement for Option
IV would be around 9% or more
compared with the base case whilst
the impact on CO2 and other emissions would be comparable with the
revamp options already described.
The scheme introduces fans,
thus electrical power consumption
would be required compared with
the base case. This type of air preheater is a specialised design compared with an on-board metal tube
bundle type air preheater and may
be limited to a few suppliers, but it
would be mechanically more robust
compared with an air preheater
having glass components.

Revamp Option V

This scheme is identical to Option

II except that an on-board glass
coated type air preheater is replaced
by an off-board glass coated type
configuration for heat recovery.
Overall structural weight would be
about the same as for an on-board
glass coated air preheater, with the
load shifted to the ground. Extra
plot area is envisaged and footprint would be required for a flue
gas fan and combustion air fan.
Where there is a plot limitation,
such a scheme does not seem to be
a suitable option for improving fuel
efficiency.
The design includes a glass
coated tubular construction in the
cold zone, so acid corrosion issues
are not envisaged in this scheme.
Maximum efficiency improvement

www.eptq.com

q1 kbr.indd 5

Fuel
Combustion
system
Preheated
air

Exhaust to
safe location
Off-board
ground mounted
cast APH

Air circuit
Flue gas circuit

Figure 7 Process Scheme [Option V:

Revamp Case B5]

with Option V should be around

11% more than with the base case,
whilst the impact on CO2 and other
emissions would be comparable
with the revamp options already
described.
This scheme introduces fans,
so electrical power consumption
would be required. Water washing
is generally available for cleaning
soot from off-board air preheater
designs. This would require instal-

Various process
schemes are available
for efficiency
improvement in high
sulphur fuel scenarios
lation of access platforms for the
operation and maintenance of
soot blowers.
A specialised design, this type of
air preheater may only be manufactured by a few vendors. It is less
robust compared with an on-board
metal tubes design but may be
slightly better than a cast and glass
type air preheater and would need
minimal maintenance and service
by the supplier.
Compared with the base case,
which is a natural draft system, all
revamp options would necessitate
replacement of natural draft burners with forced draft types because
of the additional air side resistance
of the air preheater. However,
this is subject to review since the
original equipment supplier may
recommend some modifications to

existing burners and total replacement might not be required.

Conclusion

Various process schemes are available and currently in operation in

plants worldwide for efficiency
improvement in high sulphur fuel
scenarios. Each of these schemes
has its merits and demerits. The
aim of this study is only to provide
basic guidelines for comparison of
the available options. Contractors
and suppliers will typically provide
tailor-made engineering solutions
and provide project-specific optimisation. Selection of a particular
scheme is frequently driven by
project economics and client and
operating companies previous
experiences and preferences.
Acknowledgement
Special thanks to Mr Sandeepan Ghosh for his
contribution in carrying out detailed analysis
of the study work.
Adil Rehman is Technical Advisor Process
with KBR Technology, Gurgaon, India. He has
over 14 years of experience in process design
in various companies and holds a masters
degree in petrochemical engineering from
AMU Central University, India.
Email: Adil.Rehman@kbr.com
C Steven Lancaster was formerly with KBR
Technology. He has over 31 years of experience
in furnace and fired heater design and
specialised heat transfer applications and holds
a bachelors degree in chemical engineering
from Vanderbilt
University, Nashville,
Tennessee.
Sandeepan Ghosh is Principal Technical
Professional Process with KBR Technology,
Gurgaon, India. He has over 12 years of
experience in process design in various
companies and holds a bachelors degree in
chemical engineering from Delhi College of
Engineering, India.
Email: Sandeepan.Ghosh@kbr.com
Om Prakash Sahu is Technical Professional
Leader Mechanical with KBR Technology,
Gurgaon, India. He has over 16 years of
experience in mechanical design with various
companies and holds a bachelors degree in
mechanical engineering from Barkatullah
University, India.
Email: Omprakash.Sahu@kbr.com
Pawan Kumar Sharma is Chief Technical
Advisor Mechanical with KBR Technology,
Gurgaon, India. He has over 25 years of
experience in mechanical design in various
companies and holds a bachelors degree in
mechanical engineering from Kurukshetra
University, India.
Email: Pawan.Sharma@kbr.com

PTQ Q1 2017 105

14/12/2016 13:45

crystaphase ptq revamps supplement-stability.pdf

2016-09-07

10:55 AM

Whats eluding your reactor?

Stability.
C

CM

MY

CY

CMY

For all the engineering that goes into

stabilizing your reactors performance, the
most important factor is the most difficult
to control: your feedstock.
With Crystaphase reactor optimization, you
can take control of the particles, poisons,
and maldistribution that otherwise wreck
your best-laid plans.
The result: predictable performance, year
after year, cycle after cycle.
Put your reactor back on schedule. Email
stability@crystaphase.com to find out how.

crystaphase.com
+1 281-874-2110
sales@crystaphase.com

crystaphase.indd 1

14/12/2016 10:33

Combating reactor pressure drop

A refiner investigates the causes of pressure drop due to fouling in a fixed bed
reactor and considers an appropriate mitigation strategy
ANKIT A JAIN and AJAY GUPTA
Reliance Industries Ltd

www.eptq.com

Q1 RIL.indd 1

this fouling problem have been

probed in detail. These scenarios
are: deposition of nes in the voids
between the catalyst particles; and
irregular swelling of catalyst particles in the voids due to chemical
reactions (see Figures 3b-c). The
pressure drop across the catalyst
bed of a single phase packed bed
reactor is estimated by the Ergun
equation. In this article, the correlation is used to quantify with time
the pressure drop due to fouling
across the bed. Similarly, the
approach discussed here may be
extended to multiphase packed bed
reactors by making appropriate
changes in the governing pressure
drop correlations.1,2

Case study: fouling of a packed

bed reactor

The two cases of fouling observed

in the catalyst bed are shown schematically in Figures 3a-c. Case 1
represents a situation in which
fouling arises due to the accumulation of ne particles in the voids

Frictional losses
1. Inlet diffusers
2. Distributors
3. Trays, etc.

Particulates
1. Iron sulphide
2. Phosphate products
3. Carbon particles
4. NaCl

Organic species

Pressure
drop
(reactor)

1. Olefins/diolefins

2. Metal naphthanetes
3. Asphaltenes

Coke formation
1. Insufficient hydrogen
2. Maldistribution

3. High bed temperature

Figure 1 Possible causes of pressure

drop across a hydrotreater (black depicts
pressure drop due to inherent sources;
blue depicts fouling sources)

Fouling
Normal operation

Pressure drop

he dominant factor that limits

the run length of xed bed
units in a renery is the
build-up of pressure in the reactor.
The possible causes of pressure
drop in the reactor are shown in
Figure 1. Pressure drop due to
fouling requires scientic understanding and root cause analysis in
order to develop an appropriate
mitigation strategy. The typical
pressure prole seen in one of our
commercial xed bed reactors is
shown in Figure 2. A rapid increase
in pressure drop across the catalyst
bed was seen in both the single and
multiphase xed bed reactors operating in our renery, leading to a
premature shutdown of the reactor.
On probing the different layers of
catalyst bed, it was found that the
layers of the bed were either
plugged with nes (iron sulphide,
coke
and
other
inorganic
compounds) or in some cases there
was irregular change in the shape
of the catalyst particles. In a
scenario wherein each manufacturers aim is to increase the intrinsic
activity of the catalyst and, at the
same time, reners aim to pack the
maximum amount of catalyst into a
given volume, it is imperative that
the run length of the reactor should
be governed by catalyst activity
rather than by a rapid increase in
pressure drop due to fouling or
other factors.
This article describes an approach
to quantifying pressure drop due to
fouling in a reactor. The unit in
question is a single phase packed
bed reactor, which was experiencing a severe fouling problem and
subsequent pressure drop problems. Two phenomena that led to

Time
Figure 2 Schematic representation of a pressure drop profile in a commercial reactor

PTQ Q1 2017 107

14/12/2016 14:02

Fluid flow

Fluid flow

the concentration of ne particles in

the feed remains constant and
uniform over a period of time. As
the virgin xed bed is fed with
nes-laden feed, the void fraction
(B,t) will change with time according to the relationship seen in
Equation 1.
!,! = 1

Volume of catalyst in the fouled fraction of bed + Volume of fines

Total volume of fouled fraction of bed

!,! = 1 !,!

Fluid flow

(iron sulphide, carbon particles,

phosphate particles, and so on)
while Case 2 represents a scenario
wherein the catalyst particles
surface expands due to chemical
reactions, leading to a decrease in
the void space.
Figure 3a is a simplied
representation of a portion of
packed bed loaded with spherical
particles where we have assumed a
void fraction of 0.40.
The two cases represented in
Figures 3b and 3c have been
analysed separately in this work.
The work can be easily extended to
a case representing a combination
of both phenomena.
The volume of nes (Vf,t) entering
the reactor and the diameter of
particles of nes (dp,f) is assumed.
For both cases, we assume the fraction (length) of the bed (Lf) being
fouled out of the total length of the
bed (Lt).
For simplicity, it is assumed that

108 PTQ Q1 2017

Q1 RIL.indd 2

[1]

R! L! (1 !,!!!! ) + V!,!
= 1
R! L!

It should to be noted that the

volume of catalyst and total volume
of bed referred to are in the localised fouled region of the bed only.
For the rest of the bed (L =Lt Lf)
the voidage remains the same.

Case 1: fouling due to accumulation

of fine particles

Figure 3a Clean reactor catalyst bed

3b Catalyst bed with fines depositing in
the voids 3c Catalyst bed with deposition
on the catalyst particles

For calculating the equivalent

diameter, we have assumed that
the surface being formed due to the
deposition of particles is hemi-ellipsoidal in each part (assuming that
the catalyst particles are divided
into four parts, see Figure 3c) and
appropriate formulae have been

Assuming that the nes entering

the bed are spherical and are of
uniform size (rp,f), we can easily
quantify the number of particles
depositing in the bed with respect
to time.
The number of ne particles
deposited in the bed in time t is:

used to calculate the surface area of

the equivalent particle (Se,t).
For Case 2:
Equivalent diameter d!,! = 6 x
d!,!,! = 6

Volume of catalyst + Volume of fines

Surface area of the new shape of particles

!
r!,!
L 1 !,! + V!,!

4n!,! S!,!

[4]

The pressure drop in the two

cases has been calculated using the
Ergun equation (see Equation 5):
P! =

150L 1 !,!
!
!
d!,!,!
!,!

v+

1.75L 1 !,! !
v
!
d!,!,! !,!

[5]

Pt is the pressure drop across the

bed at time t; L is the length of the
bed (for pressure drop across the
(np,f,t)= Volume of fines deposited in time t
fouled fraction of bed: L = Lf and
Volume of single particle
for the rest L = Lt - Lf); dp,e,t is the
V!,!
n!,!,! =
[2] equivalent diameter of packing; is
4 !
r
3 !,!
the density of uid; is the
Various correlations are proposed dynamic viscosity of the uid; v is
in the literature to calculate the the supercial velocity; and B,t is
equivalent diameter of particles in the void fraction of the bed. The
a bed with multiple dimensions. In above calculation is carried out for
our studies, we have calculated the each iteration of time step t.
Various time dependent functions
Sauter mean diameter to calculate
the equivalent diameter of the (rate of ne deposition, rate of
particle size in the bed (see swelling, and so on) may be introduced in the above procedure to
Equation 3).
take into account
Volume of catalyst + Volume of fines
the more complex
Equivalent diameter d!,! = 6 x

Surface area of catalyst + Surface area of fines
dynamic behaviour
of the system.
!
!
d!,! = 2

n!,! r!,! + n!,!,! r!,!

!
! + n
n!,! r!,!
!,!,! r!,!

[3]
For Case 1:
np,c is the number of particles of
main catalyst with radius rp,c.

Case 2: accumulation of fouling

products on the catalyst surface

The number of particles in this case

remains the same as the initial
number of particles.

Results and analysis

The pressure drop in the bed for

the two examples of fouling
phenomena
is
calculated
as
discussed earlier. Sensitivity analysis with respect to parameters
including the fouled fraction of bed
length and the diameter of the
catalyst particles and the ne particles is carried out to understand

www.eptq.com

14/12/2016 14:02

45

40
2
Pressure
drop ratio

H in dry gas, vol%

ProgREss-26
Case 1
Supplier

Case 2

357
306
255
204

15

2
10
1400
1
0

1600

0.1

1800

2000

2200

2400

2600

0.2 Ecat
0.3 Ni0.4
0.5 0.6ppm0.7
equivalent,

2800

0.8

3000

0.9

1.0

Normalised time

45

H2 in dry gas, vol%

40
# TRACKS
Figure
Comparison
of pressure
drop for the
two cases of fouling
APC4KPI
THE OVERALL
PERFORMANCE
OF A DMC
APPLICATION
FOR
THE
MONTH.
IT IS EQUAL
TO THE
TWO PREVIOUS
ProgREss-26
APC
KPI tracks
the
overall
performance
of
a
DMC
application
for the month.
35
(SERVICE
FACTOR
AND
TECHNICAL
KPI, SHOWN
BELOW)
Supplier
ItKPIS
is equal
to the two
previous
KPIs
(Service Factor
and Technical
KPI)
A
10
MULTIPLIED
ANDso
SCALED
SO THAT
TARGET
multiplied
together
and scaled
that target
performance
is PERFORMANCE
100%
30 TOGETHER
L/Lt = 0.1 (Case 1)
IS 100%.

nes
ticles

L/Lt = 0.3 (Case 1)

25

Figure 4 KPI diagram from a monthly summary reportL/Lt = 1.0 (Case 1)

Pressure drop ratio

8
20
applications 7to
15
targets
6

achieve operating

Make APC
10 utilisation an expecta5
7.2
7.4 7.6 7.8 8.0
tion across the
organisation.

L/Lt = 0.1 (Case 2)

engineer
all aspects
of APC
0.3 (Case
2)
L/Lt = in
L/Lt = 1.0 (Case
2)
development
and usage.
8.2

8.4

8.6

8.8

9.0

9.2

C3=, wt%
References
Things to consider
3
Diaz
D, as
Conley
R C, etofal,(a)
Maintaining APC
7 ProgREss-26
reduced hydrogen
in dry1
gas,
here
a function
Ecat Ni and
Figure
Match
managements
APC
(b)
propylene
yield
(commercial
data)
program
applications
effectiveness,
presented
2 number and quality
ambition to the
at the 2014 AFPM Q&A and Technology Forum,
of people they
1 are willing to dediDenver CO, Oct 2014.
0 9 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
cate to the effort
10.5
2 Coker R, APC2: Aligning people with control,
ProgREss-26 with
Normalised
time
Choose a group
supervisor
10.0
PTQ, Q3 2012, 57-60.
Supplier
9.5
3 Friedman Y Z, Avoid advanced control project
B
10
9.0
mistakes, Hydro Proc, Oct 1992, 119.
L/Lt = 0.1 m
9 8.5
4 Conley R C, Diaz
It m
takes a village:
L/Lt D,
= 0.3
8.0
maintaining
APC
effectiveness,
presented
8 7.5
at
the
2015
ISA
Process
Control
and
Safety
7.0
7
Symposium, Houston, TX, Nov 2015.
6.5
5 Friedman Y Z, APC application ownership,
6 6.0
Hydro
Proc, Sept 2010, 13.
5.5
5 5.0
6 Ayral T E, Conley R C, Gaining operator
acceptance of advanced controls, Hydro Proc,
4.5
4
4.5
4.7
4.9
5.1 Jun 1987, 42-43.
5.3
5.5
5.7
5.9
6.1
Coke,
wt%
7 Friedman Y
Z, Why coker APC applications
3
are tough, Hydro Proc, Dec 2005, 98.
2 0.86
ProgREss-26
8 Ayral T E, Conley R C, Advanced control
Supplier
documentation for operators, Hydro Proc, Sept
1 0.83
0.10
0.15
0.20
0.25
0.30
0.35
0.40
1988, 103-4.
9 Friedman Y Z, Advanced process control: it
broad plant 0.80
experience: control,Voidage
operations and plant economics. A takes effort to make it work, Hydroc Proc, Feb
0.77 of pressure drop over the
1997, 17.
Figure 5a Comparison
entire length of the bed vs certain
part-time
supervisor
is acceptable
4

Bottoms, wt%

Coke, wt%

Pressure drop ratio

Involve console
operators, process
supervisors and
process engineer in
all aspects of APC
development and
usage

bed 5b Comparison
of pressure drop over the entire length of the bed vs
regions
Deal of the
with
DCS/instrument
0.74
certain regions
of the
bed
problems
before
beginning
APC
Partner with
0.71 a proven APC
Randy Conley is DCS/ SIS/ APC
its effect on the pressure drop the two cases (Case 1: particle
contractor
Implementation Supervisor for TOTALs Port
0.68
deposition and Case 2: particle
profile.
Focus APC engineering efforts on Arthur
Texas (PAR)
refinery.
His previous
200000
210000
220000
230000
240000
250000
swelling)include
has Profimatics
been compared
in
implementing and maintaining
and CITGOs
Feed employers
rate, kg/h
Comparison
of pressure drop due to Lake
Figure
4.
For
the
same
rate
of
foulsuccessful
applications
Charles, Louisiana refinery. He holds BS
two fouling
phenomena
ing (in terms of volume), the same
the
Involve
console
operators, and MS degrees in chemical engineering from
Figure
8 ProgREss-26
yield
improved
coke (length
relative toof
the bed)
The pressure
droplowered
dueand
tobottoms
fouling
in and
extent
of delta
fouling
Lamar University, Texas.
process
supervisors
process
alternate supplier

t / day), m3/d
www.eptq.com
www.eptq.com
www.eptq.com

Q1 RIL.indd 3

blocked-in, tha
the process chil
of time, the a
quench into the
gas will simpl
refrigerant inven
and build up in
a closed circui
load of the comp
possibly exceed
driver, causing a

CHANGE
ON FOUL

HIRING
EXPERIENCED
PROFESSIONALS
WORLDWIDE

Method 2 - Spar

Recently, proces
to utilise the f
continuous supp
erant in the LP
suction drums.
introduce the h
the lower liquid
of the suction d
sparging system
recycle gas to b
liquid invento
increased conve
inventory into
and, since the op
the drum does n
ably,
the
p
ITW is expanding and looking
temperature
sen
for experienced professionals
inlet remains un
worldwide. The candidates
The sparger n
should have a minimum five
to efficiently allo
years experience in at least one
of these fields:
the maximum a
ARE THESE IMAGES
cle gasFAMILI
as sma
refining/petrochemicals
liquid inventor
Patented
ITW Online
Cleaning can remove any type o
process
specialty
chemicals
the back press
open or enter
hazardous process equipment.
sales
surge
valve ITW
to b
An
entire
Process
Unit
can
be
cleaned
by utilizing
refining/petrochemicals
Online Cleaning can be applied at sizing
any time
during
t
the
anti-s
process technology, operations,
appearing, rather than when they aredesign
no longer
sustain
different
maintenance, turnaround
This will in turn avoid throughput reduction,
a
it, onegiveaway
must con
The application of ITW Online Cleaning will be therefor
pressure of the
and
be willing
to travel
theshould
economics
for placing
a turnaround.
nationwide and worldwide, along
plus the static p
with
being
prone
to hard work
by the
ITW
Online
Cleaning
can be applied
to liquid
all Refil
sales.tanks, and is the
maintained
as
well as and
storage
only technolo
the drum. This
Having
a technical
degree
and
Regular
application
of ITW
Online Cleaning
will target
the need
for q
good market knowledge is also
associated
contr
required.

For Turnaround applications, ITW Online Cleaning c

thereby reducing downtime and improve operational H
The available
positionsapplication,
will cover: an Calculating
suctiin
In a Turnaround
additional ROI
technical
sales, implementation
multi-stage
refrie
Decontamination
to achieve quick and
effective safe
of ITW
in thedoes
field,not create
Ourtechnologies
patented chemistry
any emulsion
In many
cases
sales and operations
provide
dedi
management.
EVALUATE
ITW ONLINE CLEANING AND ITW

controllers for
IMPROVE
YOUR P
refrigeration
Interesting compensation plans
means that the
willFor
be given
along with serious
more informations
contact :
ITW S.r.l.- C.da S.C
each
m
career possibilities.
Tel.
+39 stage
(0931) 766
directly
or inferr
E-mail:
info@itwtec
The problem
www.itwtechnologie
Please contact:
traditional desi
ITW S.r.l., C.da S.Cusumano,
Join ITW Team worldwide and send
yourmachines
Curriculu
stage
96011 Augusta ITALY

Email: info@itwtechnologies.com
www.itwtechnologies.com
itw.indd
1

14/12/2016 14:02

dp,f /dp,c = 50
dp,f /dp,c = 10
dp,f /dp,c = 5.0
dp,f /dp,c = 2.5

10

Pressure-drop ratio

9
8

Feed filter system

Scale trap baskets
Graded material

Sacrificial beds

High porosity
graded material

Antifouling
chemical

4
3
2

Figure 8 Pressure drop mitigation

measures in hydrotreaters
0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Normalised time
Figure 6 Comparison of pressure drop over the entire length of the bed vs certain regions
of the bed

10

dp,f /dp,c = 0.10 (Case 1)

dp,f /dp,c = 0.17 (Case 1)
dp,f /dp,c = 0.25 (Case 1)
dp,c = 2 mm (Case 2)
dp,c = 3 mm (Case 2)
dp,c = 5 mm (Case 2)

Pressure-drop ratio

9
8
7
6
5
4
3

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Normalised time
Figure 7 Effect of diameter of catalyst on the pressure drop profile

and same diameter of the particle

bed, the pressure drop is higher in
the case wherein nes are concentrated in the voids between the
catalyst particles compared to the
deposition of nes on the surface of
catalyst particles. At the end of the
same cycle, the pressure drop in
the rst case is roughly twice that
in the second case.

Effect of fraction of bed fouled

We varied the fraction of the region

in which fouling takes place
(usually the topmost layer of the
bed). It can be noted from Figure 5a
that the increase in pressure drop
is most sensitive to the increase in
the fraction of fouled bed; this is

Q1 RIL.indd 4

tance of strategies that ensure the

distribution of nes (fouling particles) over the entire length of the
bed. Skimming of the top layer of
the bed from time to time is
applied in various units. The length
of the bed to be skimmed can be
estimated from these calculations.

Effect of fines diameter

110 PTQ Q1 2017

Pressure drop
mitigation in
hydrotreaters

evident from the slope of the

increase in pressure drop for the L
= 0.1 m to L = 0.3 m case. It can be
seen from Figure 5b that at 0.27 of
voidage an exponential increase in
the pressure drop across the reactor begins. We dene this as the
critical voidage. Hence a reactor
showing an exponential increase in
pressure drop may have reached
the critical voidage of 0.27 at a
certain layer of the bed. It can also
be inferred from Figure 5a that the
distribution of fouling over the
entire length of the bed, instead of
being concentrated in a smaller
region of the bed, is crucial to
increasing the run length of the
reactor. This indicates the impor-

Simulations to study the effect of

nes diameter were done by keeping the volume of nes constant.
These simulations kept the diameter of the catalyst particles constant.
It can be seen from Figure 6 that as
the particle size decreases, this
leads to a higher rate of pressure
drop. Hence strategies to trap the
ne particles are important in order
to ensure an increased run length
for the bed.

Effect of catalyst particle diameter

Simulations wherein the diameter

of the catalyst particles was
increased (keeping the constant
ratio of dp,c : dp,f in Case 1) showed
that as the catalyst particle size
increases there is a decrease in the
rate of pressure drop, whereas in
Case 2 the change in the pressure
drop prole is negligible.

Mitigation strategies

Various technologies have been

employed for mitigating pressure
drop problems in a reactor. As
discussed, localised fouling causes
a rapid rise in pressure drop;
normally this would occur in the
topmost layer of the bed. One of
the strategies could be to remove
this fouled layer. For practical

www.eptq.com

15/12/2016 13:44

1 Al-Dahhan M H, Larachi F, Dudukovic M P, Laurent A, High-pressure

trickle-bed
reactors: a isreview,
IndustrialNorth
& Engineering
Chemistry
Justin Weatherford
Sales Manager
America with
ITW.
Research,
36,years
No. 8,experience
1997, 3292-3314.
With 15
in the refining/petrochemical business
2 Ranade
V V, Chaudhari
R, Gunjal
R, Trickle
Bed and
Reactors:
Reactor
he previously
worked for
a majorP oil
company
engineering
Engineering
Applications,
2011.
company&and
holds a BSElsevier,
in chemical
engineering from Kansas State
University.
Ankit
A Jain
is a Research Scientist in the Refining R&D department of
Email:
jweatherford@itwtechnologies.com
Reliance
Industries
He holds
a bachelors
and PhD in
Marcello
FerraraLimited.
is Chairman
of ITW.
With 30 degree
years experience
chemical
engineering
from
NIT
Surat
and
IIT
Bombay
respectively.
in the petroleum business, he holds international patents for new
Ajayprocesses
Gupta isand
Assistant
Vice
President and
heads control
the fixed
bed
additive
compositions
forcurrently
environmental
and
process
development
group of the Refining R&D
department
at Reliance
for improving
petroleum/petrochemical
processes,
and holds
a PhD
Industries
Ltd., Jamnagar,
in industrial
chemistry. Gujarat, India. He holds bachelors, masters
andEmail:
doctoral
degrees in chemical engineering from IIT, Delhi, India.
mferrara@itwtechnologies.com

www.eptq.com
www.eptq.com

Q1 RIL.indd 5

Quality in XXL

reasons,
thisfollowing
is possible
if weCleaning.
have two The
separate
applied
Online
vesselbeds:
is
a main
bed clean
and a tosacrificial/guard
bed. For
nearlycatalyst
bare-metal
begin with so there
is
lownopressure
systems,
we may
also
have of
two
guard
requirement
to degas
a huge
volume
sludge,
beds
swing
while
for high
pressure
andinthere
aremode
no HSE
concerns
as you
wouldsystems
have
onewith
needs
be careful
in employing such a strategy.
theto
typical
approach.
There are other commercial technologies such as catalysts
with macropores
at the top layer of the catalyst
Combination
of technologies
bedByto capture
injecting
antifouling
chemicals,
applyingfines,ITW
Improved
Degassing/
andDecontamination
so on. A short overview
of Cleaning,
the various
technoloafter Online
safe
entry
giesconditions
is shown
schematically
Figuretime,
8. and
The
will be
achieved in ainshorter
calculations
in thisconsumption
article provewill
the be
importance
steam-outshown
time/steam
further
of having
reduced.appropriate barriers to ensure that fines do
not enter
therecent
catalyst
In one
job bed.
in a 100 000 b/d crude unit, the
following was achieved:
Summary
and conclusions
The first- and second-stage desalter vessels were
Maximum
availability ofwith
the the
catalyst
bed, avoiding
cleaned simultaneously
remainder
of the
premature
shutdown
the reactor
due train,
to factors
such
unit (including
theof entire
preheat
atmosas pheric
fouling,and
plays
a defining
role
in to
thea profitability
vacuum
towers)
prior
turnaround. of
ITW Online Cleaning lasted 16 hours. Some 98%
therefinery.
of the original
sludge volume
was increase
dissolved
Various
factors leading
to the rapid
of and
presremoved
vessels
(seebeen
Figure
2).
sure
drop infrom
the the
reactor
have
discussed
in this
The washing fluids were sent to a crude tank
article.
and fully
(seea Figure
Two
casesreprocessed
of fouling in
single 3).
phase packed bed
The
subsequent
degassing/
of
reactor
have
been described.
Case decontamination
1 assumes that fouldesalter
wasofachieved
0% (fines)
LEL
ingthe
is due
to thevessels
deposition
spherical with
particles
within bed,
six hours
of Case
initiating
0 ppm
H2S catalyst
in and
the voids
in the
while in
2 the
steam-outisoperations.
assumption
that fouling occurs due to ellipsoidal
Only
inchesinoftheinorganic
remained
swelling
of 2-3
particles
voids of material
the catalyst
bed due
in the vessels
and
mechanical reactions.
cleaning Itwas
to deposition
of fines
or no
polymerisation
was
required.
observed
that the pressure drop in Case 1 is higher than
ITW
company
solving
all
in Case
2 helped
becausethis
of refining
the higher
surfacein
area
the fluid
the
conventional
problems
associated
with
cleanflow encounters.
ingapproach
their desalter
vessels. Because
the desalters
canto
to calculating
the pressure
drop due
An
be
cleaned
so
quickly
and
efficiently
utilising
ITW
fouling in the reactor is described. This approach can
following
options
arewith
available
for
be technology,
extended to the
other
fixed bed
reactors
an approfuture
consideration:
priate change in the pressure drop equations.
The job analysis
can be with
donerespect
duringtothe
turnaround
various
parame Sensitivity
simultaneously
while
online
cleaning
the
ters has been described. It is seen that the remainpressure
derisofhighly
the unit.
drop
sensitive to the length of layer in which

The
job place.
can beThe
done
during the
turnaround
fouling takes
calculations
prove
the imporfollowing
the
cleaning
of
the
reminder
of
the unit. in
tance of strategies to ensure that fines are deposited

The
job
can
be
pushed
outside
of
the
turnaround
the entire length of the bed rather than accumulated in
altogether with
minimal
consequences
a concentrated
region
of the operational
bed.
because
the
time
frame
is
dramatically
than
The approach discussed in this article willless
be useful
the
conventional
approach
(24
hours
oil-out/oil-in
in quantifying the pressure drop due to fouling in a
only).
fixed
bed reactor and in designing an appropriate mitiAllstrategy.
of the above is a powerful tool to improve
gation
current operations and to effectively overcome the
current challenges of desalter cleaning.
References

ARCA
Flow Group
worldwide:

ECOTROL control valve

ARCA Anti-Surge Control Valves
Flow- and noise optimized valve
housing and trim
Stable control even under severe
service conditions due to fully
pressure balancing of valve plug
Fast and accurate response
with negligible overshoot allows
efficient compressor operation
close to the surge limit
Rigid pneumatic actuators with
proven pneumatic instrumentation
Up to 28 (DN 700)
For further
product
information

ARCA Regler GmbH, D-47913 Tnisvorst

Phone +49-2156-7709-0, Fax -55, sale@arca-valve.com

www.arca-valve.com
Competence in valves, pumps & cryogenics
Subsidiaries and partners in Switzerland, the Netherlands,
India, P.R. China, South Korea, Mexico and USA!

PTQ Q3 2016 113

PTQ Q1 2017 111

14/12/2016 14:03

THE SINGLE
SOURCE FOR
EMISSIONS
CONTROL.
Weve been honing our expertise for more
than 85 years. Investing in experts and assets.
Innovating through research and development.
And earning unrivaled combustion and vapor
control experience at installations worldwide.
Let us put our expertise to work for you.

johnzinkhamworthy.com | +1.918.234.1800

j zink.indd 1

2017 John Zink Company LLC. johnzinkhamworthy.com/legal-notices

14/12/2016 09:56

The design temperature of flare systems

A key objective when setting flare system design conditions is to maintain the
integrity of the system during fire relief
PAUL DAVID
Paul David Process Ltd

great deal has been written

about the minimum design
temperature of flare and
blow-down systems because cold
temperatures can cause brittle failure or over-stressing of flare piping.
Less has been written about setting
the maximum design temperature
of flare system piping.
American Petroleum Institute
(API) Standard 521,1 which is used
throughout the industry by process
engineers, gives only general guidance on how to approach the task of
setting the (mechanical) design temperature. The standard advises that
the extremes of temperature of the
fluids entering the header should
be considered, heat transfer analysis
may be performed, it is common to
exclude the fire case, and careful
analysis is required. There is considerable scope for interpretation
by individuals and companies of
how this is to be implemented. The
authors experience is that company
standards differ in their approach
and in the level of detail they offer,
sometimes giving rise to more questions than answers. The refinery
process engineer, faced with tables
of relief case data, has to place a
number in the box on the line list
that says Design Temperature. It
is a number for which he or she will
be held accountable.
The design temperature of process equipment is usually the higher
of: the maximum normal operating
temperature plus a margin (typically of the order of 25C); or the
highest temperature expected during start-up, shutdown or upset.
Upset conditions include the operation of the pressure relief valve
and for equipment containing a

www.eptq.com

pdp.indd 1

saturated liquid the temperature is

higher at relief pressure than at normal pressure. The selection of relief
valve body and flange rating are
normally based on the equipment
design conditions.
The fire contingency is not normally considered when setting the
design temperature of the equipment
or its associated pressure relief valve
(PRV). The fire relief temperature for
a heavy hydrocarbon mixture, with
a wide boiling range, can be very
much higher than the equipment
design temperature. This can lead to
the somewhat uncomfortable result
that the PRV fire case operating
temperature is higher than its design
temperature.
If during the fire the vessel PRV
opens to relieve vapour, the temperature of the fluid entering the flare
system may be much higher than
the vessel design temperature. Such
a possibility gives rise to the question: should the design temperature
of the flare system be higher than
the design temperature of the vessel
it serves?
A fire on a hydrocarbon processing unit usually means that a
loss of containment has already
occurred. When a vessel containing hydrocarbon is subject to heat
input from a major fire, further
failure (for instance at the vessel
flanges) should not be unexpected.
Although this results in an escalation of the fire, the incident is still
contained in the same plant area.
If hot vapour relief to the flare
system causes sufficient stress at
any point in the flare header (by
thermal expansion) then significant damage can occur. This may
cause loss of containment at a

location remote from the original

fire. The result would be a major
escalation of the incident. It is therefore not illogical that parts of the
flare system should have a higher
design temperature than that of
the equipment from which the fire
relief stream originates. However,
the header piping in on-plot flare
systems is highly constrained (by
many piping branches) and likely
to be large in diameter. Setting an
unreasonably high design temperature will result in major mechanical
design problems.
There is no industry wide practice for setting design temperature
based on fire case data. This scenario
requires careful consideration to
avoid either an inadequate specification or an infeasible specification.
API Standard 521 states that it is
common practice to exclude the fire
relief scenario when specifying the
design temperature of the flare headers. It may be a common practice
but, in the authors experience, it is
certainly not ubiquitous. To consider
fire case for the majority of refinery
PRVs but to ignore it when specifying the disposal system to which
they discharge might be considered
questionable. After all, a fire on an
oil refinery is a reasonably foreseeable contingency.
The following sections assume
that the fire case will not be ignored
and consider some of the issues that
will therefore arise.

What is the fire case relieving

temperature?
Wetted wall vessels

Many refinery vessels containing

wide boiling range hydrocarbons
will have very high calculated fire

PTQ Q1 2017 113

14/12/2016 14:10

relief temperatures. The heavier

the hydrocarbon and the higher the
PRV set pressure, the higher will be
the relief temperature. If we take the
average boiling point as indicative
of the temperature a vessel might
be expected to reach, then a distillate stream with an average boiling
point of 300C might be over 400C
at a relieving pressure of 5 barg. The
design of the flare sub-header piping
would be very challenging at this
sort of temperature.
Actually, the relieving temperature is unlikely to reach 400C
and it is therefore unlikely that the
flare piping would ever reach this
temperature. Hydrocarbons tend
to start cracking if the temperature
exceeds a value of around 350C;
if the liquid in a vessel is boiling at
350C it is likely that some cracking
is occurring at the vessel walls. The
lower molecular weight materials
produced by cracking will tend to
reduce the effective vapour pressure
of the liquid and make it unlikely
that the temperature will continue
to rise in the way predicted from
the feed stream boiling range.
Depending on hydrocarbon type
and molecular weight, it may well
be that the hydrocarbon in the
vessel will become supercritical at
the relieving pressure and will no
longer boil. Prediction of what happens inside the vessel is now even
more difficult since cracking will
still occur (or increase since the wall
temperature is likely to increase).
The formation of light hydrocarbons
will tend to increase the critical
pressure of the mixture and also
cool the vessel contents since thermal cracking is endothermic.
If the calculated fire relief temperature is higher than 350C it should
be treated with extreme caution.

Vessels containing vapour

When vessels containing only gas or

vapour are subject to fire heat input,
very high initial relieving temperatures can be calculated depending
on the ratio of normal operating to
relieving pressure. In some cases the
calculated relief temperature will
be infeasibly high and failure of the
vessel would have occurred before
relief. In any case the mass relief
rate is likely to be low and the relief

114 PTQ Q1 2017

pdp.indd 2

relatively short lived. Failure of the

vessel is likely unless the vessel is
effectively cooled by fire-water. In
either case the relief will cease. For
these reasons, relief from a vessel
containing vapour only is unlikely
to heat up a significant portion of
the flare piping and is unlikely to
determine the design temperature
of the flare system.

What is the temperature

downstream of the pressure
relief valve?

Overall, the flow through a relief

valve is considered to be approximately isenthalpic the nozzle
flow is isentropic, but this does
not continue throughout the valve.
Therefore it is usual to expect a
temperature drop across the relief

If the calculated fire

relief temperature
is higher than 350C
it should be treated
with extreme caution
valve due to the pressure falling at
constant enthalpy. This is typically
of the order of 1.5C per bar drop
across the PRV for a hydrocarbon
stream.
The flow velocity at the relief
valve outlet is almost always
higher than at the relief valve inlet.
Rigorous flow simulators will indicate a further temperature drop
below the stagnation temperature
which would be calculated by a
process simulator at the valve outlet. (Stagnation temperature is the
temperature that would occur if the
flow was brought to rest isentropically.) This additional temperature
drop will generally not amount to
more than a few degrees and should
not be accounted for in simple heat
transfer calculations to estimate
the wall temperature. Due to the
velocity profile, the velocity of the
fluid next to the wall is zero and
the fluid temperature at the wall
would be nearer the stagnation
temperature than the bulk flowing
temperature for adiabatic flow.2 The
flow is not actually adiabatic, since

heat transfer through the pipe wall

occurs, but the principle that the
velocity related temperature drop
is not fully experienced at the wall
remains.
As vapour flows through pipe of
constant diameter, its pressure will
fall and its velocity will increase
this results in a reduction in temperature. This effect may briefly
be reversed when a flow from a
small pipe enters a larger header,
causing the velocity to reduce. Any
fall in temperature due to velocity
increase in the flare system will normally be less than 10C for heavy
hydrocarbons at typical back pressures. Again, care should be taken
in accounting for this temperature
drop the fluid temperature close
to the wall would be higher than
the bulk temperature for adiabatic flow. Note: API Standard 521
recommends the use of isothermal
flow calculations for estimating
PRV back pressures from the flare
piping. While this is reasonable
in that it results in a conservative
back pressure estimate (for gases
at above ambient temperature) isothermal flow calculations are not
generally suitable for estimating
the flare temperature profile.
Isothermal flow is equivalent to
adiabatic flow with the exponent in
the equation PV= constant set to
a value of 1. For heavy hydrocarbons, the ideal gas value of is not
much above 1, which is why the
calculated temperature drop (in the
absence of heat transfer) is usually
small.
While all these effects are directionally helpful for the flare
system, their combined effect is relatively small.
When hot gases flow through a
PRV into the flare system, the heat
transfer from the gas to the cold
pipe will initially result in relatively
rapid cooling of the hot gas. The
heat transfer coefficient between
the flowing gas stream and metal
is likely to be limited by the fouling
factor (refinery flare lines are typically heavily scaled). Nevertheless
the inside coefficient will be well
over an order of magnitude greater
than the outside coefficient and
this results in the pipe wall heating
quickly. Since the outside coefficient

www.eptq.com

14/12/2016 14:11

Advancing the Worlds

Gas and LNG Value Chain

Full Conference
Programme
Now Available

Advancing the Worlds

Gas & LNG Value Chain
Conference & Exhibition Programme

Official Knowledge Partner:

Organised by:

For the latest programme and speaker line-up, visit www.gastechevent.com

INSIGHTFUL PRESENTATIONS FROM OUTSTANDING INDUSTRY SPEAKERS

Patrick Pouyann
Chairman & Chief
Executive Officer
Total

Ryan M. Lance
Chairman & Chief
Executive Officer
ConocoPhillips

25,000+ 2,500+
Exhibition
Attendees

Conference
Delegates

Alexey Miller
Deputy Chairman
of the Board of
Directors, Chairman
of the Management
Committee
Gazprom

Sultan Ahmed
Al Jaber
Minister of State
United Arab
Emirates Chief
Executive Officer
Abu Dhabi National
Oil Company

Maarten Wetselaar
Integrated Gas
& New Energies
Director
Royal Dutch Shell

600+

70+

130+

International
Exhibitors

Countries
Represented

Conference
Papers

SUPPORTED BY THE JAPAN GASTECH CONSORTIUM

Download Gastech conference programme

on www.gastechevent.com/ptq1

DMG21.
Gastech
gastech.indd
1 Delprom A4_PTQs Q1.indd 1

02/12/2016
14/12/2016 10:19
10:34

is very low, the steady-state pipe

wall temperature is quite close to
the relief stream temperature. Near
the PRV, the steady-state temperature will be reached within minutes.
Once the steady-state temperature
is reached in the pipe segment just
downstream of the PRV, the gas
temperature into the next segment
is not much lower than at the PRV
outlet heat transfer, limited by
the outside co-efficient, is slow. The
wall temperature in the first segment is also not much lower than
the PRV vapour outlet temperature.
As we move downstream through
the flare system, this process keeps
repeating itself. The gas and wall
temperatures rise later (due to the
larger upstream mass of metal to be
heated) and the steady-state temperature is lower (due to the greater
upstream heat transfer area). The
rate of both these effects is dependent on the relief case and the system
geometry. However, on a typical
process unit the time to steady state
is still likely to be measured in minutes and the steady-state temperature is likely to be well within 50C
of the PRV outlet temperature.
Off-plot flare headers are often of
considerable length, are designed
with considerable flexibility, and
have far fewer connections. They
are downstream of the sub-header,
often downstream of a large knockout drum, and the header itself has
considerable mass and relatively
large heat transfer area. All the
considerations for the sub-header
apply but, for the off-plot header as
a whole, the rate of temperature rise
will be significantly slower and the
steady-state temperature will also
be lower.

What is the hazard?

Process engineers are accustomed

to think of the flanges as the weak
points in piping systems and most
of us are quite capable of assessing
their pressure-temperature rating.
For flare sub-headers with multiple
branches, the thermal expansion
of the header can cause stresses for
which the flange pressure-temperature rating is not directly relevant.
It is the temperature of the pipe wall
(causing thermal growth) that is
important, not necessarily the tem-

116 PTQ Q1 2017

pdp.indd 3

perature of the flanges. Resultant

over-stresses, which might occur
due to fire relief, can be in the subheader or one of the branches and
can only be predicted by a full stress
analysis. This stress analysis will
reflect the fact that only part of the
sub-header may be at high temperature, depending on the location of
the fire. Where the failure occurs is
not necessarily intuitive. It may well
be that the failure is at a flange, but
the flange that fails may not be one
that experiences the highest temperature. The stress may be caused by
the growth of piping elsewhere.
Failure in the flare system during
a localised fire is a highly undesirable event. The flare is an important
safety system. As the fire scenario
unfolds, operators may vent gas

Failure in the flare

system during a
localised fire is a
highly undesirable
event
from numerous locations into the
flare. Failure will result in highly
flammable or toxic gas venting
to atmosphere at an unexpected
location.

Setting the flare system design

temperature

The flare system does not usually

have a design temperature lower
than the flowing temperature for
non-fire contingencies.
Setting the design temperature
of the PRV tail-pipe to a value of
300C-350C, based on the fire case,
might not be considered unreasonable for equipment containing
heavy or wide boiling hydrocarbon
streams.
However, 350C would be a very
high design temperature for the
process unit flare sub-header or
downstream piping and refinery
headers generally have design temperatures lower than this. There
are several reasons for this, including those that have been discussed
above:
There will be some heat loss/tem-

perature drop as the vapour flows

through the PRV and flare piping
If the relief valve opens intermittently then flare piping may not
reach its equilibrium temperature
The relief stream from the vessel
in question may mix with other
material as it enters the sub-header
It takes time for the vapour
to heat the metal of the flare
sub-header.
The heat-up time may also allow
the fire brigade to start cooling
sprays on the area surrounding the
fire. Although the flare header is
unlikely to be their priority, cooling
of the process equipment in the fire
area should reduce the relief rate.
There are two methods by which
the design temperature of the flare
sub-header is usually set:
1. Large companies sometimes have
a not to be exceeded value based
on experience. This value is derived
from many years of operating multiple process units and is known to
be practical for design from the piping stress viewpoint.
2. The second method is based on
rigorous fluid flow with heat transfer calculations to give an expected
metal temperature for the flare
sub-header during the worst case
scenario. It is necessary to be careful
with this approach the fluid flow
and heat transfer in the flare system
can be modelled with rigour but the
process engineer should question
whether the model of the process
vessel, during the fire, is equally
rigorous. The calculated fire case
piping metal temperature is used
by the stress engineer to ensure the
integrity of the flare system and a
suitable mechanical design temperature is back calculated. All shortterm overstress allowances in the
relevant piping design code should
be taken into account for the fire
case.
For units where the fire relief
temperature is assessed at 350C, a
sub-header design temperature in
the range 250C to 300C may not
be unexpected.
The off-plot flare system design
temperature is found by similar
considerations and methods to
those applying to the sub-headers.
Generally, the design temperature
is lower than for the on-plot piping

www.eptq.com

14/12/2016 14:11

since there is more opportunity for

heat loss in the upstream piping.
It is not normally considered that
the PRV tail-pipe or flare headers
will experience fire engulfment.
On-plot flare sub-headers will normally be at an elevation above the
7.6m (25ft) normally considered for
fire relief calculations1 and PRV tailpipes should be situated above their
respective headers. Direct radiation
is therefore not normally taken into
account when determining piping
design temperatures. Where there
are particular concerns, it may be
advisable to provide fire-proof insulation although this is unusual on
flare piping. The provision of any
insulation, of course, affects the heat
loss calculations discussed above.

Conclusion

It is important to know the objective

when setting flare system design
conditions. For large hydrocarbon
processing plant, one key objective
is to maintain the integrity of the
flare system during fire relief. The
consequences of the fire may cause

upsets or shutdowns on other process units and the flare system is a

key utility under these conditions.
A flare system failure may result
in a large release of flammable
material at an unexpected location.
The author is aware of an incident
where a huge fireball was caused by
a process unit venting into a damaged flare system. Although in this
case the damage was not caused by
thermal expansion it emphasised
the importance of maintaining flare
system integrity, particularly during
upset conditions.
The use of relieving temperatures,
and hence flare design temperatures, of higher than around 350C
are unlikely to be warranted for typical refinery hydrocarbon streams.
For other materials, the behaviour
of the material at elevated temperatures needs to be reviewed.
Design temperature of downstream sub-headers and headers is
based on the maximum expected
metal temperature. This is often
set by experience since the exact
behaviour of the relieving material

when heated by fire and the actual

contingency is not known until it
happens. In the absence of extensive, relevant experience, the large
simulation effort required should be
tempered by good judgement. The
data generated during the fire case
study should be documented and
reviewed with the engineer responsible for the flare piping stress
analysis.
References
1 American Petroleum Institute Standard 521,
Pressure relieving and Depressuring systems.
Sixth Edition, Jan 2014.
2 Shackelford A, Temperature Effects for Highvelocity Gas Flow, Chemical Engineering, Jan
2015.

Paul David is Director and Process Engineer

with Paul David Process, providing process
engineering and overpressure protection
services. He holds a bachelors degree in
chemical engineering from the University
of Bath and has over 30 years experience in
the industrial gases, chemical and oil refining
industries. For more than 20 years he worked
at a major UK oil refinery.

www.sogat.org
SPONSORS

2630 March 2017,

Abu Dhabi

SOGAT Workshops

March 26-27

Workshops will focus on: Amine Treatment; Sour Gas Process Optimisation and
Simulation; Novel Methodologies in Mercaptan Removal, and are separately bookable.

13th International SOGAT Conference

March 28-30

SOGAT Exhibition

March 28-30

The Conference Programme will feature such technical topics as:

 Energy recovery in CO2 removal processes
 Case study of SRU and AGE facilities installed in a newly discovered gas eld in Egypt
 Successful implementation of are gas recovery systems
 Performance improvements in AGR from ultra sour wells
 Dealing with CO2 cycling due to CO2-EOR
 Operational process safety experiences at the Shah eld

Exhibitors include Energy Recovery, Huntsman, OHL Gutermuth Industrial Valves GmbH,
Al Hosn Gas, DOW, Worley Parsons, Sulphur Experts, John Zink Hamworthy and many
more. Please visit www.sogat.org/exhibition to review the shell scheme, oor plan and
availability.

Media Partners

www.eptq.com

pdp.indd 4

For further information on all aspects of SOGAT 2017 and to reserve your delegate
places please refer to www.sogat.org or contact Nerie Mojica at:
Dome Exhibitions, PO Box 52641, Abu Dhabi, UAE
E: nerie@domeexhibitions.com T: +971 2 674 4040

PTQ Q1 2017 117

14/12/2016 14:12

Whats missing in this picture?

YOU!

JOIN THE FACES OF THE INDUSTRY.

AFPM Annual Meeting
March 19 - 21
Marriott Rivercenter
San Antonio, TX

afpm.indd 1
AFPM_2017_AM_Meeting_Ad_PTQ_v1.indd
1

afpm.org/Conferences
#AM17

14/12/20162:20
09:58
12/12/16
PM

Laser scanning with dimensional control

Integrating traditional methodology with new scanning technologies to achieve
higher order accuracies for critical interfaces and tie-in points
PETER FIELD
Warner Surveys

or oil and gas companies operating in a volatile climate, careful planning and tight control
over construction and maintenance
programmes have never been so
vital. The sheer immensity of most
projects and the scale of investment
involved places huge pressures
on all parties from stakeholders
and engineers to engineering procurement, construction & design
(EPCD) contractors, with projects
often scheduled for completion
within extremely short timescales,
working around the clock.
Errors or misinterpretation during design and fabrication can
easily occur, leading to the need
to rework not just minor parts
but substantial plant items. A key
contributor to project over-runs,
reworking can account for up to
15% of total budget costs on new
build developments and a significant proportion of the cost for
scheduled maintenance and engineering works. These are costs
that could be totally removed by
anticipating issues ahead of time
through closer collaboration and
employing survey techniques up
front.
A huge leap forward for surveying, laser scanning enabled the
remote capture and dissemination
of large amounts of data, where
previously each data point had
to be measured individually and,
frequently, physically. The technology is now used extensively
for the accurate mapping of large
scale developments and complex
sites, including heavily congested
or inaccessible areas with complicated assets in the hostile environments such as those to be found in

www.eptq.com

q1 warners copy.indd 1

Figure 1 Laser scanning provides a reliable method for surveying congested and often
inaccessible areas of plant. However, hostile site conditions may affect accuracy, and
dimensional control techniques may be required for critical tie-in points

refineries and industrial plants (see

Figure 1).

Collaborative working

Being able to call on a comprehensive database offers benefits for collaborative working, where many
different influencers require input
into the design process. The interoperability of modern 3D models
created using a variety of software
packages, central to the building
information modelling (BIM) workflow approach just now emerging
in construction, has been employed
extensively by the oil and gas sector over many years. Providing a
visual representation of plant structural and operational aspects, the
digitisation of data in this way creates a central information resource
that saves time and greatly reduces

the risk of errors and need for

reworking to ensure projects are
delivered on time, on programme
and on budget. Having a shared
source of accurate data allows project engineers and contractors to
maintain control over all aspects
of the engineering and fabrication phases of oil construction projects, delivering improved returns
through an integrated approach.

Early intervention pays dividends

It is often the case that survey services are first deployed on large
projects at the point where fabricated elements built to original designs are brought to site to
check. Reworking may be required
where misalignment occurs, causing delays that could be avoided by
surveying earlier in the process to

PTQ Q1 2017 119

14/12/2016 14:15

could be up to 15 years old in

some cases, increasing the chances
of misalignment or poor fit where
plant elements are fabricated working to original designs. For this
reason, it is advisable to consider
commissioning additional laser
scanning surveys to help plan the
shutdown properly. Lack of detail
could otherwise leave asset owners
and contractors having to deal with
considerable variations in scope
during the execution of projects.
Completion of design work in the
point cloud or model environment
in advance provides the necessary
assurance that the process will be
clash free.

Unplanned shutdowns
Figure 2 Basic dimensional control techniques are used to ensure a single weld hook up
(SWHU) for fabrication and installation of new modules

approve the designs ahead of fabrication, allowing build issues to be

detected and rectified before on-site
construction commences.
Design errors or omissions can
lead to inadequate project specification, with estimates of cost accuracy being underestimated in some
cases. The early involvement of surveyors at cold eyes review stage
can help with assessing design concepts for constructability, providing
access to site-experienced expert
knowledge of the issues that can
occur and potential workarounds
to guide engineers and designers
with less experience to this level
of project complexity. Early screening and availability to survey data
plus a fresh pair of eyes can offer
new solutions to deliver cost savings and prevent project overruns,
helping to secure an earlier return
on investment.

Shutdowns and turnarounds

minimising your downtime

Shutdowns are an inevitable feature of refining and can represent

a significant proportion of a plants
yearly budget. Typically, these will
be scheduled every 3-5 years, taking plants or part-sections offstream
to undertake inspections required
to comply with regulations and
to carry out necessary repairs and
maintenance. A percentage of these
events will involve major plant

120 PTQ Q1 2017

q1 warners copy.indd 2

replacement, with planned downtime periods offering the opportunity for scheduling in revamps and
regeneration.
Whilst shutdowns are usually
planned when production is at its
lowest, and such maintenance is
important in maintaining productive capacity, extensive planning
and control are required, often
several years beforehand as well
as during execution. Turnaround
time is dependent on the extent of
the project and any problems that
occur. Any derailment of timetables for example where reworking
is required, with much of the work
difficult to scope in advance could
result in the loss of millions of dollars for every day of lost production
and incur additional direct costs for
labour and heavy equipment usage.
Key to the planning process is
the availability of accurate legacy
data for buildings and structures.
The availability of a comprehensive database generated through
laser scanning and the conversion
of point cloud data into computer
aided design (CAD) models provides a tool for asset management
and a valuable guide for extensions and alterations, particularly
in clash prevention, where pipes
and new elements of plant must be
integrated within already congested
sites or structures.
However, existing legacy data

Unplanned shutdowns can have a

significant effect on facilities and
their operators. In addition to the
production losses incurred during
shutdown, unplanned events place
huge pressure on resources when
operations are restored, with capacity needing to be restored quickly.
Outages of this kind can also affect
the wider economy, with the lack
of availability of fuel supplies creating a knock-on effect on industry
and deliveries.
Managing shutdowns and turnarounds successfully involves planning ahead to extend the period
between shutdowns and eliminate
unexpected downtime through preventative maintenance. Although
survey
companies
including
Warners will generally offer a rapid
response service to cover emergencies, regular surveys to assess plant
condition can avoid unexpected
maintenance costs and business
disruption, both of which may seriously impact on the bottom line.

Laser scanning or dimensional

control?

The choice of survey type can

sometimes be confusing for project teams, with laser scan surveys
often requested when the required
tolerances in fact necessitate dimensional control techniques with more
traditional instrumentation.
So how do you decide what you
need? These definitions may help:
Laser scanning is a rapid and reliable
method for surveying often inacces-

www.eptq.com

14/12/2016 14:15

sible, complex or congested areas.

Survey control is the essential, traditional survey activity providing
the auditable accuracy to so many
survey operations including laser
scanning.
Dimensional control is the name
given to high accuracy survey techniques used to achieve a good fit up
between new, basic pieces of plant.
Critical interface surveying raises
the bar for dimensional control
and relates to high accuracy techniques and instrumentation used to
achieve first time fit-ups between
new and old complex (often dimensionally corrupted) pieces of plant
or structures.
SWHU (single weld hook up):
although basically with the same
objective as critical interface surveying, this has become the term
generally applied to very large
projects with multiple modules
requiring a first time fit with no
additional spool connections (see
Figure 2).

Data accuracy

Whilst fit for purpose for many

design and construction applications, laser scanning is subject
to inherent limits on accuracy,
with terminal accuracy subject to
and dependent on a number of variables. Good quality survey control,
range, reflectivity of the surface,
type of scanner, temperature variations, stability of the survey subject
and even the symmetry of a survey
subject can all have an impact on
results.
In the real world, most survey
subjects are exposed to a significant
temperature change, which alone
can account for several millimetres
difference in the data registered.
Additionally, the noise produced
within a point cloud from a flat
surface can range from as little as
2mm to 10mm or more depending
on the quality of scanning instrumentation and the surface being
scanned. Interpolation of this
noise can lead to further distortion
in the data. Where the objects being
scanned are vibrating or subject
to other influences that can cause
movement, this will also affect
the results.
For this reason, claims that

www.eptq.com

q1 warners copy.indd 3

laser scanning can provide accuracy to within +/-2mm should be

approached with caution. Using the
highest possible specification terrestrial laser scanner in a stable, controlled environment at relatively
short range (10-20m) and with high
order survey control (+/-1mm), it
may be possible to produce data
to that kind of auditable accuracy.
If the data is not auditable, accuracy claims should be dramatically
downgraded, with +/-5mm being
more realistic.
Moreover, the concept that modelled surfaces are more accurate
than the basic cloud data is only
true in certain circumstances.
Surfaces are often irregular, the
geometry of old basic primitive
shapes becomes corrupted, and
even the lack of dimensional integrity of basic primitives can result
in some software squaring off or
smoothing out such irregularities. Although they result in a very
good looking model, they do not
accurately represent what actually
exists on site, so on this occasion
the accuracy of the model can actually be dictated by the quality of
the structure being surveyed.
While a number of software companies are actively involved in
investigating how these limitations
can be overcome, a commercially
viable solution is still some way
off. Another option available to
data users is to bypass the automated software corrections and
more complex as-built geometries
by interrogating the underlying
databases directly.
Frequently overlooked in the
delivery of 3D models, the point
cloud database captured by the
laser scanning process offers a valuable resource for designers and
contractors. Containing high quality, auditable data that can be interpreted factually, these databases
offer the flexibility to interrogate
and view point clouds to produce
2D plans and elevations directly.
This allows the freedom to request
additional deliverables from the
original survey team in the form of
sections or drill down into specific
areas at a minimal extra cost.
In addition, opening up wider
access to raw data within project

teams can facilitate greater interoperability, although this may be

hampered by the availability of the
necessary software in some cases.
Often however there is no substitute for a full CAD model.

Dimensional control raising the bar

For surveys of existing plant, where

the key aim is to get a record
of what is present, accuracies of
10-15mm, laser scanning and 3D
modelling will be sufficient to
prove the design is clash free.
A more inclusive and wide ranging approach, critical interface surveying calls for a different order of
complexity and accuracy in dimensional control. Essential in achieving a first time fit, this requires
working to tolerances of 1mm or
less (see Figure 3).
Applications for this higher order
accuracy include ensuring complex
new fit-ups such as cone to cone
connections between new and old
plant are achieved on a first time
lift and fit or high value reactor/
regenerator heads and internal
components are replaced in a single lift. Consistently achieving a
weldable 3mm root gap actually
requires surveying to a 1mm accuracy. Dimensional control is the key
component of major plant replacement projects involving any major
item of process plant whether cutting heads off reactors and regenerators or dimensionally controlling
the fabrication and the installation
of new PRT lines.

Off-site fabrication

In the quest for greater consistency

and faster construction, prefabrication where assemblies are manufactured under factory conditions then
transported to site for incorporation
into project civil engineering works
is increasingly common in revamp
and new build projects.
Offering greater programme
certainty and lower labour costs,
this approach also results in less
waste. However, the in-situ work
abutting
prefabricated
assemblies requires absolute accuracy
to avoid civils interface problems
and achieve single weld hook-up
(SWHU) objectives.
Warner Surveys applies high

PTQ Q1 2017 121

14/12/2016 14:15

Figure 3 Checking cone shapes for potential fit-up accuracy

order total station survey control,

combining laser scanning with
proven methodologies such as conventional survey control, dimensional control and critical interface
fit-ups and is active in the further
development of traditional survey
techniques for inspection and visualisation to include drone-based/
unmanned aerial vehicle (UAV) and
vehicle-mounted scanners.
It is the combination of laser
scanning, conventional survey control, dimensional control and critical interface surveying that can
provide the overall database for
design, BIM and clash prevention applications along with 1mm
accuracy dimensional control techniques for specific corridors of
interest or tie-in points.

Other issues where surveys can help

Tower installations

This task requires tried and tested

standard operating procedures
including at least one visit to the
fabrication environment, on-site
surveys to the civils interface,
pre-shimming and checks against
design elevations to ensure not
only a vertical installation but an
installation that successfully interfaces with adjacent plant.

Reverse engineering

Continually vibrating and often

hot existing plant can become
dimensionally
corrupted
over
time. Achieving a successful interface between new designs and old

122 PTQ Q1 2017

q1 warners copy.indd 4

plant or pipework connections will

therefore be problematic, especially
where designs are based on old
data. Planning ahead by testing the
designs against actual plant condition prior to manufacturing will
avoid delays on site and potential
reworking. Ambient temperature
correction enables matching of connections between the new plant
and existing structures subject to
temperature variations in operation. The availability of improved
intelligence removes the risk of
clashes and poor fit-ups.

Visualisations

Using survey data to build 3D

animations can help avoid problems where large built-in cranes
are operating in confined spaces.
Providing a visual walk-through,
these animations can track the trajectory of crane movements to
ensure there will be sufficient clearance with no part of the crane or
the load likely to hit surrounding
plant, something that is very useful
during constructability studies.

Case study

Warner Surveys was awarded

an
engineering
project
for
Tengizchevroil
Future
Growth
Project-Wellhead Pressure Management Project with survey teams
based in Kazakhstan, Korea and
Italy over a four-year period. The
company is part of the team providing front-end engineering design
(FEED) and engineering, procure-

ment and construction management

activities. The FEED stage is now
complete and detailed engineering
under way.
Warner
Surveys
has
been
appointed to oversee all technical
aspects throughout with the aim
of ensuring first time fit on site
(SWHU).
While laser scanning models are
being used to ensure clash prevention, process plant design and
asset management, the technology
will only be deployed during engineering and construction to provide final clearance checks prior to
shipping through dimensionally
restricted inland waterway systems.
The extent of the project and
the off-site fabrication of modules
require the application of high-end
dimensional control techniques
throughout to ensure final correct
positioning as per design, working
within very tight tolerances of less
than +/-2mm. Warner Surveys will
be involved in the full life cycle of
every module from fabrication and
transportation to on-site positioning
and the final hook up.
Working closely with the local
fabrication teams from its local
office in Kazakhstan, the company
is responsible for the establishment
of standard operating procedures
and validating the accuracy of
the fabrication process through to
sign-off. Constant monitoring and
reporting allows any occurrences
outside of accepted tolerances
to be checked back with designs
and the necessary amendments
made before shipping to site. As
part of this brief, Warner Surveys
is involved with marking the cut
lines for pipe connections. On-site,
the company will continue to manage the civils interface by setting
out demarcation lines and match
marks to allow cranes and other
module transportation systems to
manoeuvre and lower modules into
place on the concrete foundations
to achieve the delicate fit required.
The result is a perfect SWHU, minimising construction timescales and
avoiding costly on-site reworking.

Pushing the boundaries

Continued investment in research

and development focuses on the

www.eptq.com

14/12/2016 14:16

FULL PAGE: 175 x 250 mm

Providing 3D movement monitoring and change reporting, the automated system issues emails and/
350
or text messages
when pre-agreed
key trigger values are exceeded,
300
enabling timely action to be taken
to prevent250further deterioration,
which might lead to major failures.

and improved integration, with the

reassurance of independent results.

Built-in protection

H2 yield, SCFB

With the price of oil expected to

remain low, oil and gas companies
are increasingly concerned with
benchmark
howGlobal
theyindustry
can reduce
costs on their
200
capital projects and operations.
Structured 150
support
There will always be uncertainty at
Increasingly, suppliers are work- this level of complexity, but taking
100
ing with project
teams and organi- a few basic steps can eliminate the
sations on major long-term projects risk of additional costs and mini50
across a range of areas from con- mise delays.
CEPSA Gibraltar-San Roque
struction site0 engineering surveys
Land surveyors have a duty of
0 and
1000large
2000 site
3000top4000 5000
7000 the
8000services
9000 10000
to multiple site
care 6000
to ensure
provided
Ni equivalent,
wt ppm
ographic surveys as well as internaare fit
for purpose and delivered
V
tional oil, gas and energy projects.
to / the
required accuracy. Equally
Ni +
(0.4
0.33)Sb
4
Working within a framework however, designers and construcagreement
or2 yield
on ofaProgREss-26
call-off basis
tion
need(Ecat
to be
Figure
11 The H
is amongst
the professionals
lowest in the industry
ACEclear
testing
industry
benchmark) of support about what the project demands.
ensures
the availability
and expertise at every interface to Ultimately, good survey control
deliver a seamless
progression and techniques must lie behind and
3
offers an economic solution, fixing support the laser scan, however
costs through the establishment of high its quality, providing auditaindustrythat
benchmark
an agreed schedule of rates. With Global
ble proof
there is no degradathe necessary
certification, stake- tion in the laser scan due to other
2
holders also benefit from the added outside influences.
assurance of compliance with safety
A holistic approach to surveystandards.
ing, which combines laser scan1 Warner Surveys has
In addition,
ning and modelling for project
established a Quick Quote mech- stages where lower accuracies are
anism to simplify process
man- sufficient,
along with the higher
CEPSA Gibraltar-San
Roque
agement
and
communications order accuracy of dimensional con0
by providing0 a1000
fast2000
turn-round
trol 6000
for critical
interfaces
and tie-in
3000 4000 5000
7000 8000
9000 10000
for urgent and additional Ni
survey
points,
will always deliver the best
equivalent,
wt ppm
requirements. It was also one ofV results.
Ni +
(0.4 / 0.33)Sb
the first companies to set up an4
autonomous CAD department, a Peter Field is Managing Director of Warner
Figure
The coke selectivity
ProgREss-26 isSurveys
amongstLtd
theand
besta infellow
the industry
move 12designed
to giveofdesigners
of the Chartered
(lowest Ecat coke factor by ACE testing benchmark)
and contractors increased flexibility Institute of Civil Engineering Surveyors.
Coke factor

assessment of different technologies

and new instrumentation to find
for the
FCCway
in EMEA,
and holds aaccurate
masters
the
best
to generate
degree in chemical engineering and a PhD in
data that is fit for purpose across a
heterogeneous catalysis from the University
range of diverse applications. Often
of Barcelona.
the subjects to be surveyed can be in
aggressive
hostile
environments,
Christophe or
Chau
is Global
Marketing
subject
of temperature
Manager to
for extremes
Refining Technologies
at W.R
and
be hundreds
of offeet
away:
Grace.may
He has
over 20 years
experience
in refining
catalysts,
newthat
catalyst
new
solutions
are including
emerging
can
development,these
catalyst
evaluation and scaleovercome
problems.
up,
technical
service and
training
in EMEA/
For
example,
many
chimneys
CIS and AP. He holds a chemical engineering
and industrial plants are in disdegree and a PhD in zeolite catalysis from the
tress. The use of UAVs for inspecUniversity of Montpellier/TOTAL.
tion removes any risks and can
generate
inspection
Javier Llanohighly
Nogalesaccurate
is Operations
Manager
reports
onunit
movements
or decay
as
of the FCC
at Cepsas Gibraltar
- San
well
dimensional
Roqueas
refinery,
responsiblechange
as well where
for the
optimisationby
group
for FCCspecialists.
technology within
deployed
survey
Cepsa.
He was
previously
engineer
Warner
Surveys
is start-up
also currently
for
three
new
process
units,
including
steam
evaluating new products3 to ameasillion b/d, MMSCFD (cubic
feet /and
day),a m
/d
reformer
meta-xylene extraction
unit,
ure movement quickly and effithen moved to the CDU unit as a Process
ciently, including Leica GeoSystems
Engineer. He holds a masters degree in
Kumonos
system, which
to
chemical engineering
from appears
the Escuela
provide
a
reliable
method
to
map
Superior de Ingenieros (ESI) in Seville, and a
and
report
to cracks
on
masters
degreechanges
in petrochemistry
from the
structures
remotely,
thereby
removInstituto Superior
de la Energa
(ISE) in
Repsol.
ing access issues and improving
Bertahealth
Aramburu
Lopez-Aranguren
the
and safety
aspects of isa
currently
responsible
for
FCC,
visbreaking and
project.
the new Bottom of the Barrel Department at
Used largely by construction
Cepsas Research Centre. She holds a degree
companies to date, 24/7 automated
in chemical engineering from Complutense
monitoring
and reporting also
University, Madrid.
offers potential benefits for oil and
gas
companies.
Installed
set up
Rafael
Domingo Larraz
Mora and
was Director
in
just aR&D
fewuntil
days,
sysof Cepsa
2015.the
He remote
holds a PhD
in
chemical
engineering
from updates
the University
tem
provides
real-time
via
of La Laguna,
an Executive
MBA from
dedicated
websites
on how
plantIESE
is
Business
School
and
a
masters
degree
moving during operational cycles
in environmental engineering from the
and
whether it is performing safely
University of Las Palmas.
and as per design expectations.

digitalrefining.co
m: keep up to dat
e with
your industry and
make better decis
ions

digitalrefining.com is the most

extensive source of freely available
information on all aspects of the
refining, gas and petrochemical

www.eptq.com
www.eptq.com

q1 warners copy.indd 5

processing industries. It provides a

constantly growing database of
technical articles, company literature,
videos, industry news and events.

PTQ Q1 2017 123

PTQ Q4 2016 115

14/12/2016 14:26

Identifying contributors to flaring

Applying an integrated flare and fuel gas monitoring system to identify and
eliminate major contributors to refinery flaring
P SRIDHAR
Indian Oil Corporation

eople are often confused when

it comes to flare monitoring.
In a refinery, where hundreds
of control valves, dump valves and
pressure safety valves make their
way into the list of causes, it can be
difficult to monitor and minimise
flaring.
It is a very complex task for a
person to identify difficult and frequent occurrences of flaring, or to
pinpoint contributors to flaring.
Operators can come into contact
with a huge list of split ranges and
dead bands when they look into this
matter. In addition, those responsible for flare monitoring may not
have an in-depth knowledge of
every process involved, thus making the issue more confusing.

Common reasons for flaring

A few common reasons for flaring

(apart from acid gas flaring) in petrochemical plants and petroleum refineries are outlined below:
A. Imbalance in the fuel gas system
B. Process
requirement
where
column overheads are operated
at very low pressure (about 1 kg/cm2)
C. Dumping from the hydrogen
header
D. Process abnormalities leading to
flare
E. Process equipment abnormalities.
Point B above is a system requirement to reject lighter gas; it cannot be
controlled.
Point E can be solved during a
turnaround (if fouled) or by revamping (if under-designed). Points A, C
and D can be addressed by effective
monitoring of the system.

Proposed methodology

Trivial flaring points might none-

124 PTQ Q1 2017

q2 indian oil.indd 1

A typical monitoring table

Attribute\flare source
Point A
Control valve size in inches, D
1
% Opening
30
Upstream pressure in kg/cm2 g, P
0.5
No
Split range
No
Dead band applied
Actual control valve opening, %op 30
P*D3/100
0.3

Point B
Point C
10
6
7
60
2
30
No
Yes (>50% flaring)
No
Yes (+/-5%)
7
10 (including dead band ((60-55)*2)
140
648

Table 1

theless be identified or targeted as

major contributors to flaring depending on the percent opening of control
valves. To avoid any confusion, it is
proposed to sort flaring points based
on volumetric flow rate using the following formula:
P*D3/100*% op

This can be used to set the order

and magnitude of flaring. The formula cannot predict the exact
amount of flaring, but it can arrange
flaring points in order of priority (see
Table 1).
For all contributors to flaring, percent opening can be normalised on
the basis of actual percent opening to flare, considering dead band
and split range. This method can be
adapted wherever a flaring measurement is missing or not reliable.
For the above example, trivial Point
A can be ignored and Points B and
Point C can be focused upon to
reduce the level of flaring.
The basis of the above equation to
infer flaring is as follows:
At any point, the number of moles
being flared through a control valve
is .
For the number of moles present in
a control valve port, n:

n = P * V/ (R * T)
n = P * D3 (6 * R * T)
where P is the upstream pressure,
T is the temperature and V is the
volume of gas being flared. D is the
diameter of the control valve port.
Because the temperature from
the surge drum, from where most
of the flaring takes place, is almost
constant:
n

=K*P*D3 (K is constant)
/K=P*D3
dividing by 100 on both sides
n
/K/100=P*D3/100
n
/K/100 represents relative moles
n

Integrated flare and fuel gas

monitoring sheet

Data for an online system can be

imported directly from a real time
database in Excel format. Flaring
data should be retrieved on a five
minutes average basis to make the
system dynamic.
Flaring control valves can be
arranged in order of maximum flaring by using the expression:
Pressure in kg/cm2 g *(c/v diameter in inch)
3*% opening

www.eptq.com

15/12/2016 17:18

INTEGRATED FLARE AND FUEL GAS MONITORING SHEET

(IFMS)
TIME

1/8/2015 9:21

FLARE SYSTEM

TAG

RELATIVE MOLES

DHDS-HGU-2

GJA.1011PC1706.OP

160

HGU-3 PSA2

GJA.2041PIC3013.OP

94

JR current Flare factor

medium aring

FUEL GAS SYSTEM STATUS

Column1

no action
required
Column2

JR FUEL GAS GENERATION

61854

JR FG CONSUMPTION

61722

Kg/hr

FGRS DISCHARGE TO GRE HEADER

2954

Kg/hr

Column3
Kg/hr

GENERATION-CONSUMPTION

3086

Kg/hr

MAKE UP NG+LPG to FG SYSTEM

1190

Kg/hr

ERROR IN fuel gas system calculation

1322

Kg/hr

% ERROR
Crude processed

50800

MTPD

FG consumption excluding inerts

60978

Kg/hr

% Fuel gas consumption on Crude processed

Figure 2 Reduction in fuel gas to flare control valve (202 days trend)

Natural gas cost

21

USD/MMBTU

LPG cost

19

USD/MMBTU

IFO cost

11

USD/MMBTU

HSD for Gt

20

USD/MMBTU

Advantage in using LPG in place of NG

USD/MMBTU

Advantage in using IFO in place of NG

10

USD/MMBTU

Advantage in using HSD in place of NG

USD/MMBTU

Figure 1 Typical integrated flare and fuel gas

monitoring sheet consisting of total flaring
from each unit and highest contributor

q2 indian oil.indd 2

200

Time, days
Figure 3 Reduction in hydrogen to fuel gas header (200 days trend)

Reduction

Fuel gas hydrogen generation and

consumption data can be retrieved
on a last one average basis to make
it more reliable. These data can be
sorted on the basis of their deviation from the last one week average. Based on this, deviation in the
system can be identied to x the
actual problem.
All of these data can be integrated
onto one sheet, an integrated are
and fuel gas monitoring sheet (see
Figure 1). Continual updating of
these data facilitates easy, one-click
understanding of a fuel gas and are
system sheet dashboard.
Action to be taken can be incorporated into the sheet, based on the
upper and lower limits of individual
consumption points.
Predetermined
disturbances
should also be considered in order
to determine a strategy to mitigate aring. For instance, delayed
coker unit drum change-over and
FCC unit catalyst make-up lead
to extra gas generation which will
go to are if not absorbed into the
system.
Provision should also be made to
highlight recovery of poor data as
the system depends completely on
the quality of data recovered.

www.eptq.com

202

Time, days

Reduction

FLARE STATUS

365

Reduction

UNIT

Reduction

Current Highest aring points

Figure 4 Reduction in make-up gas to fuel gas system

Results

The real time trends shown in

Figures 2 to 4 demonstrate the
effectiveness of a close monitoring
system.

Conclusion

At present, there is no reliable and

effective method for determining
aring from process units. The integrated system for are and fuel gas
monitoring described here will facilitate monitoring of a fuel gas and
are system at a glance.
It will also enhance operational
competitiveness among the various
operating units for achieving max-

imum reduction in aring and fuel

gas consumption. Hidden problems
will be identied and resolved in an
appropriate manner. For instance, a
persistent problem may be resolved
by the addition of a condenser or by
more timely cleaning of a condenser.
Acknowledgment
The author would like to thank all colleagues
and associates without whose efforts this
article would not have been possible.
P Sridhar is the Deputy Manager Production
with Gujarat refinery, Indian Oil Corporation. He
has nine years operational experience in various
refinery areas including FCC, atmospheric and
vacuum distillation, and cryogenic, extractive
and reactive distillation units.

PTQ Q1 2017 125

15/12/2016 17:18

Technology in Action
Shaping up hydrotreating performance
Trilobe (TL) shape

A ne balance of activity and pressure drop has long created a challenge when considering the maximisation of
performance for hydroprocessing and hydrocracking
units. It is especially a critical balance for high prole
units in hydrocracking service that receive large margins for product upgrade and also have high incentive
for incremental processing capacity. Recent margins have
placed a great deal of pressure on reners to maximise
hydrocracking unit throughput up to hydraulic limitations, which in many cases is a limit set by reactor pressure drop. Limitations in reactor pressure drop can be
mitigated by many means, but ultimately catalyst selection is the most critical factor in hydrocracker optimisation. Criterion developed the ATX (Advanced Trilobe
eXtra) catalyst shape to allow hydrocracking units to
reduce pressure drop and improve activity simultaneously. There are several signicant advantages of the
ATX shape (see Figure 1).
Pressure drop in packed beds is commonly modelled by the Ergun equation which is applicable for single phase ow, but to model two phase ow through
packed beds there are modied versions of the Ergun
equation. Using the modied Reynolds number the
Ergun equation can be simplied as:

150
=
+ 1.75

!
1

!

=
=

This equation illustrates that pressure drop is

inversely proportional to both catalyst bed void fraction and the effective particle diameter. Void fraction
itself is a function of loading method (dense vs sock)
and particle shape. Differing diameter particles of the

Limitations in reactor pressure

drop can be mitigated by many
means, but ultimately catalyst
selection is the most critical factor
in hydrocracker optimisation
same extruded shape load at the same void fraction
and density since packing efciency is determined by
shape. However, larger particle diameter extrudates
of the same shape will result in lower pressure drop.

126 PTQ Q1 2017

tia.indd 1

Conventional HC catalyst shape

used in all HC catalysts offered
after 1994

Advanced Trilobe eXtra (ATX) shape

Proprietary technology offered for

majority of HC catalyst
Lower CBD lower fill cost
(~10% lower $)
Better liquid yield: shorter diffusion path
reduced over-cracking
Larger void fraction
higher particulate uptake
delayed onset of rapid P build
Lower SOR P (15 to 20% lower
commercial performance
demonstrated vs conventional TL)

Figure 1 The Trilobe design allows hydrocracking units to reduce

pressure drop and improve activity

This is due to the effect of liquid hold-up and relative

velocity in the bed voids. In addition to particle size,
particle shape has a signicant impact on pressure
drop.
The following two commercial examples demonstrate
the effect of the ATX shape applied in hydrocracking
units to unlock additional capacity, creating increases in
protability.

Case 1: Canadian refiner

Figure 2 shows the normalised pressure drop from
a hydrocracking unit at a Canadian renery where
the previous cycle applied Criterions earlier generation Z-3723 TL catalyst and the current cycle is operating with Z-FX20 ATX. The current catalyst load was
also dense loaded to maximise reactor performance
and yields while still achieving reduced bed pressure
drop. The reduction in pressure drop across the reactor has permitted an increase in unit capacity of 20%,

www.eptq.com

15/12/2016 16:25

P, kPa

Previous cycle TL shape

Current cycle ATX shape
0

100

200

300

400

500

600

700

800

900

Days on stream
Figure 2 Improved pressure drop in a hydrocracking unit

Normalised reactor P, psi

Previous cycle
TL shape

Current cycle
ATX shape

0
27
0
42
0

12

30

0
45
0
60
0
75
0
90
0
11
50
12
00

~15% lower P

15

Case 2: US Gulf Coast refiner

One of the highest capacity hydrocracking units in the world leverages the advantage of ATX shaped
catalysts to maximise unit capacity.
Over recent operating cycles the
following unit has transitioned to
Criterions ATX catalysts to reduce
unit pressure drop. While only
~30% of the overall catalyst load
is now ATX shaped cracking catalyst, the unit has been capable of
increasing throughput by 35% (see
Figure 3).
Figure 4 shows the relative difference between the various shaped
extrudates used in hydroprocessing applications in terms of
pressure drop per bed height contribution ratio with the typical trilobe setting the scale at 100%. The
chart shows that a quadlobe particle with the same outer dimensions as a trilobe particle will result
in higher pressure drop across
the same bed depth. This is well
explained when considering the
Ergun equation because the effective diameter of a quadlobe particle is smaller than that of a trilobe
particle. Further, the ATX shape
offers a signicant reduction in
pressure drop when compared to
standard TL particles.
While the pressure drop advantages of ATX are remarkable in
comparison to conventional particle shapes, the greatest advantage afforded by the ATX shape
is exceptional activity. Improved
reactor pressure drop proles with
improved activity enables performance gains for these hydrocracking units.

15% lower P

Days on stream

Figure 3 Pressure drop performance at a Gulf Coast refinery

120
110
100

P/ L, %

resulting in increased prots, better than $10 million annually. In

addition to increases in unit protability, the catalyst load itself saw a
reduced ll cost as a result of lower
compacted bulk density.

90
80
70
60

www.eptq.com

tia.indd 2

Sp
he
re

Pe
lle
t

de
r

be

AT
X

in
yl
C

Criterion Catalysts & Technologies

For more information:
Teresa.Brod@CRI-Criterion.com

Tr
ilo

lo

be

50

ua
d

James Esteban is a Senior Technical

Service Engineer with Criterion Catalysts &
Technologies.

Figure 4 Standard particle shapes used in hydroprocessing of equal outer dimensions and
loading method

PTQ Q1 2017 127

15/12/2016 16:26

Alphabetical list of advertisers

AFPM Annual Meeting

118

Albemarle Catalysts Company

ARCA Valves

111

Jonell Filtration Group

40

Linde Engineering North America

21

Magnetrol International

48 & 65

Ariel Corporation

22

Merichem Company

70

ARTC Annual Meeting

97

Metso Flow Control

53

Axens

OBC

MERTC 2017

IBC

BASF Corporation, Catalysts Division IFC

Middle East Sulphur 2017

99

Burckhardt Compression

29

OHL Gutermuth Industrial Valves

77

Cat Tech

85

Process Consulting Services

95

Prognost Systems

54

Refining India 2017

60

CB&I
Chevron Lummus Global

13

Criterion Catalyst & Technologies

Crystaphase Products

4
106

Rembe Safety + Control

103

Rezel Catalytic Technologies

47

63 & 123

Sabin Metal Corporation

31

DuPont Clean Technologies

25

Sandvik Process Systems

75

ExxonMobil

15

Shell Global Solutions

19

Silobau Thorwesten

89

DigitalRefining.com

Gastech

115

GE Water & Process Technologies

Grabner Instruments

32
9

SOGAT 2017

117

Spraying Systems

90

Haldor Topse

10

StocExpo 2017

69

Honeywell UOP

27

Sulzer Chemtech

17

WEKA

59

John Zink Hamworthy Combustion 112

Yokogawa Europe

57

78

Zwick Armaturen

100

ITW Technologies
Johnson Screens

82 & 109

For more information on these advertisers, go to eptq.com/advertisers.aspx

128 PTQ Q1 2017

www.eptq.com

Inaugural

MERTC
The New Flagship Refining and
Petrochemicals Event for the Middle East
23-24 January 2017 | Manama, Bahrain
Under the Patronage of
H.E. Shaikh
Mohammed bin
Khalifa bin
Ahmed Al Khalifa
Minister of Oil
Kingdom of Bahrain

Esteemed advisors include:

Eng. Awadh
M. Al-Maker
Executive Vice
President, Technology
& Innovation
SABIC

Mohammad Ghazi
Al-Mutairi
CEO
Kuwait National
Petroleum Company
(KNPC)

Ahmad A Al-Ohali
Chief Executive Officer
Saudi International
Petrochemical
Company (Sipchem)

To register go to
mertc.wraconferences.com or contact Rosie on r.brewster@theenergyexchange.co.uk
Save 10% when booking by quoting PTQ10 at the online checkout

Brought to you by the organisers of

mertc.indd 1

and

17th annual

14/12/2016 11:20