PTQ Q2 2014 PDF

SPECIAL FEATURES
SEPARATION
THERMAL TECHNOLOGY
REFINING
GAS PROCESSING
PETROCHEMICALS
PETROLEUM TECHNOLOGY QUARTERLY
ptq
Q2 2014
cover and spine copy 19.indd 1 10/03/2014 14:59
1
2
1
5
_
e
www.engineering-solutions.airliquide.com
Because our genes lead us to an irrepressible passion for new solutions.
The only thing longer than our company history
is the list of our innovations.
When it comes to providing innovative
plant and system solutions, Air Liquide
Global E&C Solutions is your partner of
choice worldwide. No surprise, since
we innovate relentlessly in order to bring
sustainable and cost-eective solutions
to society, levera ging partnerships with
customers, suppliers, academics and
communities. This is not only demonstrated
by thousands of patents, but by pioneering
reference projects around the globe as
well. Yours could be the next one.
AIR2362022_AZ_Techno-DNA_7-8-2013_RZ.indd 1 21.05.13 11:26 air liquide.indd 1 04/03/2014 10:54
2014. 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.
3 Tier 3: ready or not
Chris Cunningham

5 ptq&a

15 Reformer monitoring via in-tube temperature measurement
Oliver J Smith IV Air Products and Chemicals, Inc
Bill Cotton Johnson Matthey

25 Revamping the hydrogen generation unit
K V Patwardhan, V D Dhumal and P Balaramakrishna
R&D Centre, L&T Hydrocarbon Engineering, India

33 Enhance processing with capacity control of reciprocating compressors
Klaus Stachel and Markus Wenisch Hoerbiger Compression Technology
43 Removing salts and contaminants in glycol regeneration
Barry A Perlmutter BHS-Sonthofen Inc
Christian Gassen BHS-Sonthofen GmbH
51 Accurate prediction of tower relief
Harry Z Ha, Abdulla Harji and Jonathan Webber Fluor Canada Ltd
59 Myth of high cutpoint in dry vacuum units
Scott Golden, Tony Barletta and Steve White Process Consulting Services
67 Coordinating a complex automation project
Igor epic and Boris eelj INA Rijeka renery, Croatia
71 Dynamics of operation for are systems
Yahya Akta and zge zarik Tpras Izmir renery
77 Air pre-heater improves energy efciency
Bart Van Den Berg HeatMatrix
81 Maximising the use of process energy
Sunil Kumar, Shrikant Nanoti and M O Garg Csir-Indian Institute of Petroleum
93 Real Time Online Energy Management
Mohammad Ershaid KNPC Mina Al-Ahmadi Renery
Carlos Ruiz, Diego Ruiz and Nicols Visuara Soteica Visual MESA
Dhanasekar Periyasami EBS
103 Advanced solutions for efcient crude blending
Gregory Shahnovsky, Tal Cohen and Ronny McMurray Modcon-Systems Ltd
113 Performance of waxy crudes as FCC feeds
Erick Gamas, Cliff Avery, Gerbrand Mesu and David De Villiers Albemarle Corporation
Alan Yahev Neweld Exploration Company
117 Predicting severity, viscosity and yields in fuel oil visbreaking
Sepehr Sadighi and Reza Seif Mohaddecy Research Institute of Petroleum Industry
122 Recovering tungsten from spent hydrocracking catalysts
Jean-Pierre Dufour and Sophie Comte Valdi
125 Technology in Action
INA Rijeka renerys hydrocracker unit, where Emersons DeltaV automation system and DeltaV SIS system are installed
(see p67). Photo: Radovan Perii, HCU Complex Project Manager, INA Rijeka renery
Q2 (Apr, May, Jun) 2014
www.eptq.com
p
t
q
Y L R E T R A U Q Y G O L O N H C E T M U E L O R T E P
ed com copy 4.indd 1 11/03/2014 16:35
C
M
Y
CM
MY
CY
CMY
K
KBC Adv - PTQ Q1 2014.pdf 1 12/10/2013 9:29:32 AM
KBC.indd 1 10/12/2013 16:06
T
he European Union has arguably
been the global leader in biodiesel
production and use, with overall
biodiesel production increasing from 1.9
million tonnes in 2004 to nearly 10.3 million
tonnes in 2007. Biodiesel production in the
US has also increased dramatically in the
past few years from 2 million gallons in
2000 to approximately 450 million gallons
in 2007. According to the National Biodiesel
Board, 171 companies own biodiesel
manufacturing plants and are actively
marketing biodiesel.1. The global biodiesel
market is estimated to reach 37 billion
gallons by 2016, with an average annual
growth rate of 42%. Europe will continue to
be the major biodiesel market for the next
decade, followed closely by the US market.
Although high energy prices,
increasing global demand, drought
and other factors are the primary
drivers for higher food prices, food
competitive feedstocks have long
been and will continue to be a major
concern for the development of biofu-
els. To compete, the industry has
responded by developing methods to
increase process effciency, utilise or
upgrade by-products and operate
with lower quality lipids as
feedstocks.
Feedstocks
Biodiesel refers to a diesel-equivalent
fuel consisting of short-chain alkyl
(methyl or ethyl) esters, made by the
transesterifcation of triglycerides,
commonly known as vegetable oils or
animal fats. The most common form
uses methanol, the cheapest alcohol
available, to produce methyl esters.
The molecules in biodiesel are pri-
marily fatty acid methyl esters
(FAME), usually created by trans-
esterifcation between fats and metha-
nol. Currently, biodiesel is produced
from various vegetable and plant oils.
First-generation food-based feedstocks
are straight vegetable oils such as
soybean oil and animal fats such as
tallow, lard, yellow grease, chicken fat
and the by-products of the production
of Omega-3 fatty acids from fsh oil.
Soybean oil and rapeseeds oil are the
common source for biodiesel produc-
tion in the US and Europe in quanti-
ties that can produce enough biodie-
sel to be used in a commercial market
with currently applicable
PTQ Q2 2014 3
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
Bob Aldridge
sales@petroleumtechnology.com
Advertising Sales Ofce
tel +44 844 5888 771
fax +44 844 5888 662
Publisher
Nic Allen
publisher@petroleumtechnology.com
Circulation
Jacki Watts
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.
Vol 19 No 3
Q2 (Apr, May, Jun) 2014
Tier 3:
ready or not
T
he big question for refners is not whether gasoline containing 10 ppm
sulphur is on its way into American fuel tanks, but how soon. While for
environmental groups and vehicle manufacturers, reduction of gasoline
sulphur to a third its current average level could not come soon enough, the
response of the US refning industrys representative bodies was that the
beginning of 2017 for the enactment of Tier 3 gasoline regulations was too soon.
Not that the industrys lobbyists were in any way approving of the basic
premise of the new limit: the Tier 3 rule will unnecessarily raise costs and
actually increase CO
2
emissions at US refneries because of the energy-
intensive hydrotreating equipment needed to meet the new standard, was
the American Petroleum Institutes (API) response to the EPAs announce-
ment. More specifcally, according to the API, the rule could require $10
billion in capital costs. The annual compliance cost is $2.4 billion, equating to
a potential cost increase of between six cents and nine cents per gallon of
gasoline produced. The new sulphur standard would yield only very small
additional improvements in air quality.
Besides the enormous costs and negligible environmental beneftthe
rushed timeframe leaves little opportunity for refners to design, engineer,
permit, construct, start up, and integrate the new machinery required. This
accelerated implementation only adds costs and potentially limits our indus-
trys ability to supply gasoline to consumers.
For its part, the American Fuel and Petrochemical Manufacturers (AFPM)
pulled no punches, describing Tier 3 in a statement as the most recent
example of the agencys propensity for illogical and counter-productive rule
making. Tier 3 not only lacks scientifc justifcation, but in fact will lead to
higher greenhouse gas emissions due to the greater energy-intense refning
process required to reduce sulphur in gasoline.
A new hydrotreater is a very costly item of equipment but it seems likely
that many refners are already equipped to meet Tier 3 requirements, subject
to some expensive, but not so prohibitive, revamping costs. Whether or not
refners need to install new units or upgrade existing equipment, the issue of
whether notice to comply with Tier 3 is adequate or not comes into play in a
practical rather than rhetorical sense. Building anew and revamping both
demand lengthy periods for planning and engineering studies albeit there
has already been a years notice before the EPA announcement that gives
almost three years notice of action.
Then there is the matter of permitting, by the EPA no less, to ensure that
the design phase meets technical requirements and emissions regulations.
Permitting can be one of the lengthiest steps in the whole sequence of pro-
cesses ahead of actual construction. By the time construction is ready to go
ahead and provided a suitably qualifed workforce is available to carry it
out a large portion of the period of notice will have passed.
Construction of a new unit, or a signifcant revamp, also calls for a major
plant turnaround. Shutdowns are scheduled to take place every four or fve
years so that the appropriate preparations are made to avoid undue loss of
production time. Refners who carried out major turnarounds during the past
12 months may have to think about further signifcant downtime before the
start of 2017.
CHRIS CUNNINGHAM
p
t
q
Y L R E T R A U Q Y G O L O N H C E T M U E L O R T E P
ed com copy 4.indd 2 07/03/2014 13:32
You might say that Thomas Weber, a lead scientist at CRITERION, is a know-it-all. Not because hes
proud, but because he sincerely wants to know it all. To satisfy his insatiable drive to understand
the whys behind every reaction, Weber immersed himself in academia. While earning his post-
doctorate, a professor (and former Shell scientist) exposed Thomas to the world of catalyst
research, leading him to a career with CRITERION. Today, when Thomas isnt in the laboratory,
hes likely teaching chemistry at one of the top universities in catalysis in the Netherlands.
Like his mentor before, Thomas is developing the next generation of catalysts and next generation
of scientists as well.
Industrial and academic research
the best of both worlds.
Meet Thomas Weber:
Knowledge Seeker, Paying it Forward
Leading minds. Advanced technologies.
www.CRITERIONCatalysts.com
criterion.indd 1 04/03/2014 10:52
Q
What are the chief issues we should address in a project
to debottleneck a crude unit pre-heat network?
A
Eva Andersson, Market Manager Renery, Alfa Laval, evae.
andersson@alfalaval.com
Debottlenecking of crude pre-heat train
Crude pre-heat train bottlenecks
Crude pre-heat trains can become bottlenecks when
capacity increases or if crude feed is being changed. It
can result in:
Hydraulic limitations: pressure drop over heat
exchanger network is higher than capacity of pumps
Thermal limitations: heat recovered from hot fractions
plus external heat supplied by fred heater is not enough
to bring the crude to a suffcient inlet temperature to the
distillation tower or to cool P/As and R/Ds to required
temperature for on-spec production or storage.
Replacing pumps with higher capacity ones would
be very expensive and it would require increasing the
pressure rating of all pressure vessels downstream of
the new pump. Likewise, investing in a higher capac-
ity fred heater would be a major capital expenditure.
Therefore, most of the time, a revamp of the pre-heat
train is considered the most feasible solution to over-
come pre-heat train bottlenecks.
Revamp of crude pre-heat train
Due to the complexity of the pre-heat train heat
exchanger network, this is not always an easy, straight-
forward exercise.
First, it must be kept in mind that modifying one
position/heat transfer service in the pre-heat train will
affect the performance of all other exchangers in the
network, so a major energy integration study would
probably be required to identify best solutions with
minimal capital investment.
To off-load a hydraulically limited pre-heat train,
parallel lines could theoretically be added. On the
other hand, the reduced fow rates through the original
heat exchangers would affect the thermal performance
of those exchangers negatively but, even worse, would
increase the fouling tendency.
With increased fouling tendency, the heat transfer
effciency will be further reduced over time while pres-
sure drop will increase. In the end, it might lead to
more frequent and unplanned shutdowns for cleaning
of such exchangers.
To offoad a thermally limited pre-heat train, more
(effcient) exchangers are required. In some cases, exist-
ing exchangers can be modifed with more passes on
the tube side or by using mechanical inserts, but most
of the time, more heat transfer area is required.
Adding heat exchangers in parallel to existing
exchangers will not always be successful for reasons
explained above and adding heat exchangers in series
might not be possible for hydraulic reasons.
Finally, the matter of availability of plot space should
not be neglected. Most of the time, these installations
are already very crammed and it is physically not
possible to add the amount of heat transfer area
required to debottleneck the pre-heat train.
Use of welded plate heat exchangers
Many refners have already recognised the advantages
of using welded plate heat exchangers (WPHE) when
revamping their crude pre-heat train.
The corrugated plates give three to fve times higher
heat transfer effciency than traditional shell & tube
heat exchangers (S&T), meaning that the required heat
transfer area is reduced with that same factor.
In addition, the compact assembly of those corru-
gated plates means that the plot space required for
installing such WPHEs is minimised.
Combining these two parameters gives the possibil-
ity to add 4000 m
2
of traditional S&T heat transfer area
(which equals around 850 m
2
WPHE heat transfer area)
on a plot space of only 1.5 x 1.5m.
In addition, as these WPHEs are designed with a
more or less fully counter-current fow, a temperature
cross can be handled in one single exchanger. This
means that several S&T heat exchangers installed in
series can be replaced by one single WPHE, thereby
minimising both plot space and pressure drop required
for a certain service.
The corrugated plates also give rise to a high shear
stress in the heat transfer channels. This is a major
factor when minimising the fouling tendency of a heat
exchanger in a crude pre-heat train. Both hot and cold
heat transfer channels in the WPHEs can be designed
with shear stress levels of up to 10 times the levels
normally found on the tube side in a S&T heat
exchanger, thereby safeguarding the longest possible
run length in-between cleanings of the exchanger.
WPHEs in operation have shown up to three times
longer up-time compared to S&Ts in similar services.
If many heat exchangers need to be added or
replaced in the original pre-heat train, it could also be
worth looking into an improved energy integration
study, based completely on WPHE technology. With
the high heat transfer effciency and close to counter
current fow, it is possible to achieve a temperature
www.eptq.com PTQ Q2 2014 5
ptq&a
Additional Q&A can be found at www.eptq.com/QandA
Q&A copy 15.indd 1 12/03/2014 11:59
approach of only 5C in a WPHE. This means that it is
also possible to economically recover energy from low
grade sources, such as overhead vapour from the
distillation tower.
With a better energy integration, the inlet tempera-
ture to the furnace increases and the fuel requirements
and emissions from the process are thereby minimised.
Studies have shown that more than 25% of fuel
savings is a possible scenario.
So to summarise, the advantages of using WPHEs
when debottlenecking a crude pre-heat train are:
Three to fve times less heat transfer area
Minimal plot space required
No need for several units in series for one heat
recovery service
Reduced pressure drop over the heat recovery
service (when several S&Ts in series are replaced by
one WPHE)
Up to 10 times higher shear stress in heat transfer
channels and thereby up to three times longer opera-
tion interval in-between cleaning
Possibility to improve energy integration by up to
25%.
A
Marcello Ferrara, President, ITW Technologies, mferrara@
itwtechnologies.com
When fouling of a pre-heat network is limiting overall
performance, you might consider performing Online
Cleaning of the CDU. This will clean the entire
pre-heat network in as little as 24 hours on an oil-to-oil
basis.
When applied proactively, this is particularly useful
to improve energy savings of the unit. Furthermore,
fouling is an issue when processing opportunity
crudes and/or when running the unit at a lower
capacity. All of the above problems can be addressed
by Online Cleaning without the need for capital
expenditure for a revamp.
Q
Can you recommend an antifouling programme for the
heater network to a delayed coker? What are the particular
fouling issues related to heating a resid feed?
A
Berthold Otzisk, Consulting Engineer, Kurita Europe,
otzisk@kurita.de
Delayed coking is a thermal destruction process. Low
value residual fuel oils are converted into petroleum
coke, fuel gas, LPG, naphtha and heavy gasoils. Main
areas of concern are the bottom product line to the
coke drum and the coke drum overhead vapour line.
Fouling is caused by precipitation of inorganic materi-
als, asphaltenes or formation of coke.
Inorganic materials like sodium, calcium, silicon and
iron may be introduced as suspended solids in the oil
as part of the upstream processes. Inorganic fouling is
typically observed in the lower part of the convection
section and in the top of the radiant section of the
furnace. Organic materials like asphaltenes are very
complex molecules. The asphaltenes contain thermally
stable polynuclear aromatic cores. Their pendant
groups are connected to the core by thermally unstable
bonds, which break above 350C. The loss of the
pendant groups results in a rapid precipitation of the
destabilised asphaltenes. At higher temperature, free
asphaltene radicals combine to form insoluble coke.
Coke formation is the traditional coker fouling mecha-
nism. It is typically observed in the lower radiant
section of the furnace.
Kuritas delayed coker antifoulants control fouling
caused by deposition of inorganic materials, asphaltenes
and coke formation. They have a high thermal stability
and excellent dispersing power, which avoids precipita-
tion of inorganic materials and coke particles. These
antifoulants do not contain any metals, phosphorus or
oxygenates. They work by surrounding asphaltenes
molecules similar to natural resins. This shifts the criti-
cal resin concentration to a higher level. The asphaltenes
are stabilised and kept in a disperse phase. This signif-
cantly reduces coke build-up on the furnace tubes and
coke formation in the coke drum overhead vapour line.
The viscosity of the more viscous varnish type polynu-
clear aromatics is lowered.
A
Bruce Wright, Senior Technical Engineer, Baker Hughes,
Bruce.Wright@bakerhughes.com
Antifoulant programmes exist for delayed coker
furnaces. These programmes require careful investiga-
tion of the causes of the fouling, both from a chemistry
and a plant operation perspective. Products have to be
selected that will survive the extreme conditions pres-
ent in delayed cokers, and that will specifcally attack
the mechanisms that are contributing to fouling.
Treatment programmes may consist of multiple prod-
ucts to target specifc fouling issues; and the program
injection locations and dose rates have to be appropri-
ate to control all aspects of fouling.
Fouling in a delayed coker furnace is a result of
multiple chemical and operational issues such as:
Deposits result from destabilisation and agglomera-
tion of asphaltenes that attach to furnace tube walls
and gradually dehydrogenate to petroleum coke
Concentrated inorganics also agglomerate and
contribute to the deposits
Thermal cracking of hydrocarbons and subsequent
recombination and rearrangement of cracked molecules
form another class of organic molecules that contribute
to fouling. The recombined cracked molecules become
ring compounds and further lose hydrogen to become
aromatic. These molecules act similarly to asphaltenes
because they readily agglomerate
Once agglomerated hydrocarbons become adhered
to the boundary layer on the furnace tube walls, they
further dehydrogenate to become petroleum coke.
Inorganics become embedded in the hydrocarbon
deposit and may catalyse some of the reactions
High heat fux and low velocities in the coker
furnaces will result in higher fouling rates.
A
itwtechnologies.com
Fouling of a coker heater is related to asphaltene
6 PTQ Q2 2014 www.eptq.com
Q&A copy 15.indd 2 10/03/2014 12:55
cbi.indd 1 10/03/2014 11:10
Red is not a common colour observed in hydrotreated
diesel. It has been reported in cases with a high amount
of cracked feeds, especially LCO. It can be reduced by
higher hydrogen partial pressure, lower feed rate,
reduced cracked feed endpoint, lower reactor tempera-
ture, and from high activity NiMo hydrotreating catalyst.
Q
Suggestions please for an effective means of separating
oil/water emulsions in our tank farm.
A
Luis Diaz, Business Manager Oil & Gas, Alfa Laval, luis.
diaz@alfalaval.com; Emelie Arvinius, Regional Business
Manager Oil & Gas, Alfa Laval, emelie.arvinius@alfalaval.com;
Lasse Hurnasti, Business Center HSS Oil & Gas, Alfa Laval,
lasse.hurnasti@alfalaval.com
Mechanical separation
Modern slop oil treatment plants use centrifugal sepa-
ration equipment, typically decanter centrifuges and
disc stack centrifuges. The benefts of this equipment
are widely known in the market, such as more effcient
separation, smaller footprint, less dependence on
pre-treatment and chemical addition, and higher treat-
ment fows.
Decanter centrifuges are designed to offer a frst
stage of separation in the process, focusing normally
on the separation of the solids and compressing them
to achieve high oil recovery. This fexible and reliable
equipment is ready to split a waste stream into differ-
ent phases: two (liquid and solids) or three
(hydrocarbons, water and solids). Although different
setups are possible, for those cases where the fuid is
emulsion friendly process experts normally recom-
mend a three-phase separation as soon as possible in
the process to prevent new emulsions formation.
Disc stack centrifuges are the most common centrifu-
gal equipment when it comes to polishing each of the
separated liquid phases. They are normally placed
downstream a decanter increasing the quality of the
fuid by reducing the level of impurities to on spec
requirements.
Other types of equipment are moving into waste oil
treatment applications: bio-membranes, ultra fltration,
evaporation, and so on. These technologies are
normally used in what could be called a tertiary sepa-
ration stage to achieve the increasing regulations
towards a lower environmental impact, ultimately
creating zero discharge systems.
Multiple plant setups are possible, depending on
feed and required outlet conditions. Thus it is impossi-
ble to defne a standard.
While many manufacturers have tried to introduce
their equipment into this diffcult application, only
premium brands have been able to achieve consistent
and stable process operation.
Use of chemicals
Crude oil emulsions comprise fnely divided droplets
of water (dispersed phase) suspended in oil (continu-
ous phase). The water droplets would normally have a
destabilisation from residue feed. Following thermal
destabilisation, asphaltene will associate and after-
wards degrade to coke, which will deposit inside the
coils. Destabilisation can be mitigated by introducing
into the feed a very specifc asphaltene stabiliser, which
is active at the very high temperatures of the coker
heater. ITW has developed asphaltene stabilisers which
are effective in avoiding coke formation inside furnace
coils. Furthermore, yield increase can be achieved in
coker units with a proper management of ITW
patented asphaltene stabilisers.
Q
We are experiencing an unwanted red discoloration in our
diesel product stream. What are the likely causes and possible
solutions?
A
Antoine Fournier, Business Line Group Manager, Catalyst
Technology Department, Axens and Jacinthe Frcon,
Technology Group Manager, Middle Distillates & Conversion
Business Line, Axens, jacinthe.frecon@axens.net
Diesel colour deterioration is typically experienced in
units that are operating under low to medium pressure
and increases with catalyst ageing. This phenomenon
is strongly linked to the presence of polyaromatics in
the product. In order to compensate for catalyst deacti-
vation during the run, the unit operates at a higher
temperature and faces thermodynamic limitations
impacting the HDA performances of the system.
In order to minimise colour deterioration, the best
option is to maximise the HDA activity of the system.
This can be achieved through different options. Some
of them are listed below:
Selection of a high hydrogenation activity catalyst
Fine tuning of the operating bed temperature profle
in order to minimise the WABT in the bottom section
while meeting the product specifcations
Installation of a fnishing section operating at lower
temperature.
A
Paul Zimmerman, Principal TS Specialist, UOP, Paul.
Zimmerman@Honeywell.com
Diesel colour can occur due to polynuclear aromatics
or heterocyclic nitrogen compounds which absorb light
in the visible region. These are usually a very small
amount of the product, but can have a large effect on
colour. Colour bodies can occur in the feed or be
formed in the reactor if conditions are favourable.
Some feeds have been more susceptible to colour,
especially with high nitrogen content, multi-ring
aromatics, endpoint, or sulphur. These often occur in
cracked feeds, particularly in the heaviest fractions.
Operating conditions contributing to colour are high
reactor temperatures, high severity, low hydrogen
partial pressure, and low recycle gas rate. Colour most
often occurs at end of run when the reactor tempera-
tures are the highest. Flow channelling can also cause
colour due to localised higher temperature. NiMo
hydrotreating catalysts are preferred for colour suscep-
tible hydrotreaters due to better nitrogen conversion
and aromatic saturation.
Q&A copy 15.indd 3 12/03/2014 12:01
limiting factors love
limitless possibilities
Overcome limiting factors affecting refnery capacity and
operating fexibility with BASF innovative FCC products,
services and solutions. Our products deliver value to enhance
sustainability and performance.
At BASF, we create chemistry for a sustainable future.

Catalysts

Co-Catalysts

Additives

Services and Solutions
www.catalysts.basf.com/refning
Untitled-1 1 11/03/2014 14:40
pcs myth.indd 1 8/3/12 09:55:39
tendency to agglomerate and coalesce, forming ever
larger droplets which would separate from the oil by
gravity. In crude or slop oil this process is retarded or
prevented by the presence of natural emulsifers
absorbed at the water/oil interface.
The emulsifers are commonly asphaltenes, oil resins,
naphthenic acids, paraffn waxes, or inorganic materi-
als such as clay, carbonates, silica and metal salts. They
form a stable flm around droplet interfaces, prevent-
ing contact and coagulation. Paraffn waxes will
form extremely strong coherent flms of semi-solid
material below critical temperatures (cloud point),
often encapsulating other solids in the matrix. Such
emulsions may be impossible to treat except at high
temperature.
Additional facts affecting emulsion stability include
viscosity of the oil, the viscosity of the oil/water
mixture, droplet size, gravity difference between oil
and water, and ageing (increasing stability).
Demulsiers promote coalescence of the water droplets in
the emulsion to form larger droplets
From the above it can be seen that the functions of an
effective demulsifer should:
Have high mobility in the emulsion fuid oil phase
to transport the chemical and seek out droplet surfaces
Disrupt the stabilising emulsifer flms
Assist focculation of droplets by increasing attrac-
tive forces
Achieve coalescence or combination of droplets
together
Wet solids which may be contributing to emulsifca-
tion and allow them to transport into the bulk water
phase
Reversed demulsiers will cause the oil droplets suspended
in the produced water to coalesce
The water-soluble reverse demulsifers can be used to
treat contaminated oil-in-water emulsions, as can be
encountered in various industries. Cleaning of the
water phase is the main objective. Coalescence agents
are used to destabilise this type of emulsions by
increasing the size of the oil droplets present.
A
itwtechnologies.com
Separation of oil/water emulsions in tanks can be
easily achieved by injecting an emulsion breaker into
the tank. ITW has developed highly effective emulsion
breakers, capable of working under very severe operat-
ing conditions, namely high water content, high
solids content and low temperatures. In particular our
emulsion breakers proved effective in slop oil treat-
ment, for separating oil from water. ITW is also
specialised in tank cleaning and can offer this service
on a turn-key basis up to the achievement of gas-free
conditions.
A
Larry Kremer, Technology Advisor, Baker Hughes, Larry.
Kremer@bakerhughes.com
Baker Hughes manufactures chemicals designed to
dehydrate the crude oil in storage so that the oil
charged to the crude unit is low in basic sediment and
water (BS&W). This procedure requires that the refn-
ery must be capable of draining brine from the bottom
of the tank prior to charging the tank to the crude unit.
There are several good reasons why a refnery would
want to dehydrate crude in tankage. Some refneries
do not have desalters, so tank dehydration is the only
method available for removing corrosive salts from the
crude charge. Even in refneries with desalters, tank
dehydration can reduce the salt going to the crude unit
enough to enable desalters to reach desired salt levels
in the desalted crude oil.
In some cases, custody transfer costs are established
by gauging shipments in storage. Any emulsion layer
in the bottom of the tank is paid for at crude oil prices.
By dehydrating this material, the refnery no longer
pays for brine water at crude oil prices.
The water draining process is often diffcult for oper-
ators. Some crude oils hold up water, so it takes
excessively long periods of time to settle the tanks.
When draining the water, it is often oily and black,
making it diffcult for the operator to judge when to
stop draining water. In-line water monitors can be very
helpful in determining when the transition from water
to oil occurs. Properly formulated dehydration chemis-
tries will also result in a clean oil/water break and
reduce the oil under carry in the drained water. This
makes it much easier for the operators to determine
when to stop draining tanks.
The data in Table 1 shows the results of tank
dehydration that illustrates several of these points.
A cargo of diffcult to dehydrate crude was received
Settling Without Chemical Tank 1 Baker Hughes Treatment Tank 2
time, A Point B Point A Point B Point
hrs Top Middle Bottom Top Middle Bottom Top Middle Bottom Top Middle Bottom
0 >100 >100 >100 >100 >100 >100 95 >100 >100 >100 >100 >100
6 21 >100 >100 22 >100 >100
24 >100 >100 >100 36 >100 >100 12 12 >100 13 16 >100
30 25 >100 >100 30 >100 >100 13 14 88 12 14 >100
48 19 >100 >100 24 >100 >100 12 12 70 13 11 20
72 22 >100 >100 26 >100 >100 8.5 8 8 9 14
96 19 20 >100 30 60 >100 8 8 8 8.2 8.2 8.2
Crude oil salt content reduction with chemical pretreatment
Table 1
Q&A copy 15.indd 4 10/03/2014 12:56
pcs 1.indd 1 8/3/12 09:50:20
challenges to recovery system design. Depending
on the quality of the residue, a propylene yield poten-
tial of 8-12+ wt% is common with a similarly wide
range in overall light ends (C
4
-) to be recovered.
However the feed quality can vary so widely that
one needs to consider each case separately as a straight
run residue from one crude can be much higher
in quality than hydrotreated residue from a different
crude.
In addition to the feedstock yield potential, one also
needs to consider the contaminants in the different
feeds which include sulphur, nitrogen, metals and CCR
as these affect product quality and RFCC heat balance
(the ability to control the amount of catalyst delivered
to the reaction zone and ultimately the conversion).
The design of the RFCC or R2P is different from that of
a conventional FCC to specifcally deal with these
higher contaminant levels and licensors have devel-
oped technology to maximise feedstock fexibility.
However, when looking at the marketplace and how
refners are choosing to operate RFCC units for maxi-
mum proftability, one can see that feeds with a CCR
of about 4-8 wt% are common with feed metals of
10-20 ppm and equilibrium metals on the FCC catalyst
of about 5-10 000 ppm. While these are not limits, they
defne typical industrial practice.
There are also operating cost challenges when
processing resid at high severity for propylene produc-
tion in terms of catalyst deactivation due to
metals contamination of both the main catalyst and
ZSM5 additives. While we employ specifc designs
such as two stage catalyst regeneration to minimise
catalyst deactivation, the catalyst cost is an important
consideration when designing a full feed fexibility
RFCC for high severity operation to maximise
propylene.
A
Mark Schnaith, Director FCC & Alkylation Technology,
UOP, Mark.Schnaith@Honeywell.com
Propylene yield from the FCC unit starts with the feed-
stock properties. Since FCC does not involve any
hydrogen addition, a more hydrogen rich feedstock
has a greater propylene potential. This is why a
lighter, more paraffnic feedstock has more potential
for propylene than a heavier, aromatic feedstock.
Hence, VGO can be expected to achieve a higher
propylene yield than resid. Hydrotreating the FCC
feed increases the propylene potential via the addi-
tional hydrogen added to the feed, along with the
increased conversion potential of that feed. For an
existing FCC unit, maximising propylene is typically
constrained not by the feedstock, but by the size of the
product recovery section. Increasing propylene carries
with it a corresponding increase in total LPG and light
gas, and it is this total volume increase that will typi-
cally limit the propylene yield for any existing
operation. For a new unit, the fractionation and recov-
ery sections should be designed with suffcient size
and fexibility for the anticipated range of propylene
yield, and there is a cost associated to incorporate
such fexibility.
by ship and transferred into two tanks in the refnery.
The crude in tank 1 was not chemically treated.
The same crude oil in tank 2 was treated with a
Baker Hughes tank pretreatment additive. Samples
were taken from the top, middle and bottom of both
tanks from two different locations and analysed for
quality parameters. Table 1 shows the salt content
of the crude oil samples in pounds per thousand
barrels (PTB).
It is clear that, after 24 hours of settling, the tank is
not uniform in the salt distribution with the good oil
in the samples from the top. As time progresses, the
salt content decreases as brine droplets fall to the
bottom of the tank. This process is greatly accelerated
by the addition of the Baker Hughes tank pretreatment
additive. It can also be seen that the salt level is ulti-
mately lower in the treated tank at all levels than it is
in the untreated tank.
Chemical treatment produces a more uniform distri-
bution of salt and BS&W in the tanks, as well as better
quality, thus the desalter can expect improved, more
uniform performance.
A
Paul Zimmerman, Principal TS Specialist, UOP, Paul.
Zimmerman@Honeywell.com
The most common oils to form an emulsion are high
molecular weight or high viscosity oils, such as with
heavy crude oils. Oils with high asphaltenes or partic-
ulates are the most susceptible. It has been reported
that high water pH (>8) can contribute to emulsions.
Once formed, emulsions are diffcult to break.
Generally the frst step in breaking an emulsion is to
raise temperature. Oil source and water quality may
need to be evaluated.
Q
Can we effectively maximise propylene from the FCC with
either hydrotreated resid or straight run resid using the same
process setup?
A
Nicolas Lambert, Technologist, Middle Distillates &
Conversion Business Line, Axens, nicolas.lambert@axens.net
The question is very broad with many caveats about
feed quality and it is important to avoid generalisation,
however I will try to summarise some key points.
While the FCC process is very fexible, the yield
potential for a relatively hydrogen rich product such as
propylene is naturally limited by the quality of
the feed. Residue FCC or RFCC units have been
designed to process a wide range of feed qualities from
straight run resid to hydrotreated resid using the
same design. In this context we usually refer to resid
as a feed containing a Conradson carbon residue
(CCR) value greater than 2-3 wt% as this requires
specifc technology features to address heat balance
and contaminant control. When maximum propylene
from residue feed is the objective, we make specifc
unit modifcations and utilise a setup or confguration
we call Resid to Propylene or R2P. The overall
yield structure, and particularly the light products
including propylene, will vary signifcantly posing
Q&A copy 15.indd 5 10/03/2014 12:56
Copyright 2013 Merichem Company
To win a race you need not only speed, but a timely start. Merichem Company has a proven track record
of providing hydrocarbon treating solutions that meet the stringent timelines of our customers. Get your
treating project off to a timely start by choosing Merichem Company as your technology licensor.
Midstream Solutions
Proven Results
Think Fast...Think Merichem
Midstream Solutions with Industry-Leading Speed to Market
Merichem Companys portfolio of midstream treating technologies
includes our patented FIBER FILM

Contactor that offers a variety of
caustic, amine and acid treating processes. Merichem also offers
solutions for gas treating, including our patented LO-CAT

wet scrubbing,
liquid redox system that converts H2S to innocuous, elemental sulfur. In
addition to our large portfolio of licensed treating technologies,
Merichem offers industry leading spent caustic management services that
utilize Beneficial Reuse options to create a non-waste solution for your
spent caustics. For fast, effective treating solutions, choose Merichem.
www.merichem.com | 5455 Old Spanish Trail | Houston, Texas 77023 | 713.428.5000
C
M
Y
CM
MY
CY
CMY
K
Merichem Company-FY14-MID.pdf 1 2/24/2014 8:19:32 AM
merichem.indd 1 28/02/2014 10:28
Reformer monitoring via in-tube
temperature measurement
L
arge-scale steam reformers in
syngas plants (including ammo-
nia, methanol and hydrogen
production) are vulnerable to
over-fring, especially during plant
transient conditions. This vulnera-
bility is due to the fact that external
temperature instrumentation lags
during normal operation, and that
human monitoring cannot be
conducted on a suffciently frequent
basis. The consequence is that, over
the past decades, many of the
catastrophic reformer failures docu-
mented have occurred during the
most common transient operation
start-ups.
Despite repeated education as to
the risks of over-fring, these events
continue to happen. It is considered
even more important to ensure that
regular and frequent visual inspec-
tions of the reformer tubes are
made during critical periods associ-
ated with transients. Most reformer
instrumentation is designed for
monitoring and control of opera-
tion during normal operation.
During plant transients, when fow
rates are well below design values,
reliance on standard plant instru-
mentation measurements (of the
header or pigtail exit temperature)
alone is dangerous. In low fow
situations, the measured reformer
gas exit temperature is not a good
indication of the metal tempera-
tures of the reformer tubes.
A relatively new temperature
measurement technology, called
CatTracker, is available. This tech-
nology allows in-tube temperature
monitoring of the catalyst tube,
enabling the reformer to be more
reliably monitored and protected
from over-fring. It can also give
A temperature monitoring technology applied to steam reformers in synthesis gas
plants provides protection against over-ring plus detailed catalyst monitoring
OLIVER J SMITH IV Air Products and Chemicals, Inc
BILL COTTON Johnson Matthey
additional information during
normal operation that can be used
to better monitor the performance
of the catalyst and optimise
reformer operation.
This article reviews the benefts
of CatTracker for safety and moni-
toring of catalyst performance that
was gained after installation in a
number of reformers. These instal-
lations include several hydrogen
reformers operated by Air Products
and Chemicals.
Temperature measurement
Previous papers
1
have discussed
the details of the various techniques
for tube wall temperature measure-
ment. While these benefts will not
be addressed in this article, Table 1
summarises the advantages and
disadvantages of the various meas-
urement techniques for reference
purposes.
Optical and gold cup pyrometers
do not allow accurate temperature
measurement at all points along the
tube. It is therefore likely that by
using this techniques alone, the
maximum tube wall temperature
may not be measured.
The reformer imager allows for
the measurement of the tube wall
temperature profle along the full
length of the reformer tube and
hence will allow for the measure-
ment of the maximum tube wall
temperature. A further beneft with
the reformer imager is that it
provides a high resolution image
rather than a single point measure-
ment. It would be impractical to
take the same number of measure-
ments as the reformer imager does
using an optical pyrometer.
For other tube wall temperature
measurement devices, to fully
understand the tube wall tempera-
Type Optical Gold Cup
1
Reformer Imager Cat-Tracker
Number of tubes 80-95% 10-30%
2
All up to 8
Tube length Majority
3
10% All All
Accuracy Excellent Excellent Excellent Excellent
Time for one row 10 minutes 5 minutes
4
15 seconds On-line
1 Estimates provided for a top red reformer.
2 On a Foster Wheeler Reformer, it is possible to measure 80-90% of the tubes.
3 Although an optical pyrometer can measure the temperature of the majority of the tube length, there is
incrementally more background radiation as the path length is increased.
4 This is to measure a limited number of tubes on a top red reformer.
Available tube wall measurement techniques
Table 1
The reformer
imager allows for
the measurement
of the tube wall
temperature prole
along the full length
of the reformer tube
j matthey.indd 1 10/03/2014 13:05
Design
A CatTracker comprises a number
of sensing points (4, 9 or 11) sepa-
rated by an inert material and
enclosed in a metal sheath. The 11
point design is illustrated in
Figure 1.
The system provides highly accu-
rate temperature measurements
within a catalyst bed (2F or 1C)
with precise locations (within 2in
or 50 mm). These properties allow
for the process gas temperature
within the reformer to be measured
very accurately at a known posi-
tion. The hermetically sealed
individual probes are fed to a
common junction box, which can
be appropriately located on the
reformer.
Johnson Matthey can specify the
optimal location of the sensors
within the tube for a given reformer
design using its Reform
2
modelling
program. These sensing positions
can be located at points of interest
such as predicted carbon formation
zone or in line with peephole
heights.
CatTracker can be installed in
new plants or retrotted to an
existing plant. For the reformer
application, it is installed via a
modied top ange and placed in
the centre line of the reformer tube
before catalyst loading. A typical
installation in hydrogen service is
shown in Figure 2.
The CatTracker requires no exter-
nal sheath, which minimises the
diameter. For a reformer applica-
tion, it is in (6 mm) in diameter.
The small size minimises the
impact on the catalyst loaded
density in the tubes with the
system installed. As a result, all of
the tubes in the reformer will see
the same process gas ow rate,
regardless of the presence of a
CatTracker or not.
For other applications, the
CatTracker can enter the vessel via
any suitable nozzle (typically 1 or
2in diameter). In addition, its
highly exible nature allows it to
be bent in order to place the sens-
ing point precisely where
temperature measurements are
required. This exibility makes
it highly suitable for radial ow
reactors where temperature meas-
ture prole either an advanced
modelling technique
2
or an alterna-
tive temperature measurement
system is required. One such
system is the CatTracker.
CatTracker
The catalyst temperature tracking
system patented by Daily
Thermetrics was rst introduced
commercially in 2001. It is now a
proven technology with over 500
installations across ve continents
proling hundreds of vessels. With
25 000 sensing points it has shown
superior operational service. At
present there are 33 systems
installed in a range of synthesis gas
plants, which will increase to 81
installations this year.
The CatTracker, employing
patented aerospace thermocouple
technology, offers the syngas indus-
try the most rugged yet exible
temperature probes designed to be
in direct contact with the process.
Its temperature sensors are engi-
neered to withstand the harshest
environments and the most strenu-
ous temperature demands making
them highly suitable for installation
in a reformer.
Each probe consists of an insu-
lated cable which can sense
temperature at predetermined loca-
tions along its length. Multiple
temperature sensing points are
independent and isolated from one
another, while at the same time
ungrounded from the sheath. The
systems design eliminates any
possibility of signal interference
(due to the fact that there is no
common leg) and provides unsur-
passed reliability for the most
demanding applications.
There are many potential bene-
ts, including the following:
Continuous on-line monitoring of
the in-tube temperature prole,
with the prole available in real
time through the plant DCS system
Interlocks or trips, based on
process gas temperatures within
the catalyst tubes, preventing
potential injury to plant personnel
and costly reformer damage
Tube temperature prole meas-
urement leading to more accurate
prediction of the operating
proximity of the catalyst to the
carbon formation zone thus
allowing operation avoiding carbon
formation
Operation of the reformer at
lower steam to carbon ratios with-
out the risk of carbon formation by
using the CatTracker process gas
temperatures to calibrate a reformer
process simulation
Improved understanding of cata-
lyst parameters for an improved
understanding of catalyst perfor-
mance and change out scheduling.
Improved understanding of the
catalyst performance can be deter-
mined by using the measurements
in conjunction with a detailed
process simulation of the reformer
Monitoring real reformer tube
outlet temperature to determine
actual pigtail temperatures leading
to a reduction in pigtail failures
Calibration of pyrometers when
used in conjunction with process
simulation of the reformer.
0.02 in wire diameter
11 point design, OD = in
Wall thickness = 0.02 in
310 stainless steel sheath
Figure 1 CatTracker sensing points
Courtesy: Daily Instruments Corp.
Figure 2 CatTracker installation
j matthey.indd 2 10/03/2014 13:05
RELIABLE SWISS QUALITY
API 618
Rod load up to 1'500 kN/335'000 Ibs
Power up to 31'000 kW/42'100 HP
FULL RANGE:
YOU GET MORE THAN JUST A
PROCESS GAS COMPRESSOR
Lubricated up to 1'000 bar, non-
lubricated up to 300 bar

For highest availability: We recom-
mend our own designed, in-house
engineered compressor valves and
key compressor components

Designed for easy maintenance
We are the competent partner
with the full range of services
worldwide
www.recip.com/api618
YOUR BENEFIT:
LOWEST LIFE CYCLE COSTS
burckhardt.indd 1 04/03/2014 10:55
urements are required both axially
and radially.
Tube failures
Reformer tubes fail due to a
number of reasons including the
following:
1. Failure due to micro-void coales-
cence (known as creep fssures) that
grow together through the thick-
ness of the tube and cause rupture.
2. High tube metal temperature
from localised over-fring of a tube
or tubes due to burner problems
(such as fame impingement) that
cause tensile overload exceeding
the tube yield stress and tube
failure.
3. High tube metal temperature
due to poor catalyst performance
(for instance catalyst poisoning,
bridging or carbon formation),
again causing tensile overload and
tube failure.
4. Over-fring of an entire reformer
due to transient conditions.
In cases 1, 2 and 3, generally
there is only a single tube failure.
While a single tube failure does
impact the reliability and proftabil-
ity of the plant, the shutdown time
to replace the failed tube is not
catastrophic.
Figure 3 illustrates a single tube
failure due to Case 3, above, where
the catalyst has been poisoned
while the reformer is still in opera-
tion. Figure 4 shows the same tube
after removal
Case 4 generally occurs when the
plant is undergoing a transient
operation such as a plant start-up
or shutdown, or is under steaming
conditions. During such plant tran-
sients, the process feed to the
reformer is typically only steam
and is generally operated with a
low reformer outlet temperature.
Without the reforming reaction, the
heat load on the reformer is much
lower than during normal opera-
tion, and hence there is a relatively
small heat sink. If too much fuel
is passed to the reformer, then
the reformer tubes will be over
heated. In the worst case, it will
lead to multiple catastrophic tube
failures.
The consequence of such a failure
can be very severe, as new tubes
will need to be procured and
installed in the reformer. The
downtime associated with such a
reformer tube replacement can be
measured in months rather than
days, leading to large losses in
proftability.
Transients measurement issues
One of the major issues that opera-
tors face during plant transients is
that the instrumentation on the
reformer will not provide a true
representation of what is happen-
ing within the reformer. Due to the
lower process fow conditions (and
constant heat loss between the tube
exit and the reformer process gas
outlet thermocouple), the differen-
tial between the actual and the
indicated temperature is greater
than for normal operation. For
example, consider the case where
heat loss in normal operation
results in a differential of 25F
(14C) between the tube and the
measurement point. Since this heat
loss is not a function of plant rate,
when operating at half production
rate the differential is 50F (28C).
If the reformer is controlled at a
constant outlet temperature and
this effect is not considered, over-
heating of the tubes can be a
consequence.
Fluegas temperatures at the exit
of the radiant section of the
reformer can be used as a guide,
but these thermocouples operate in
a severe duty. As such they are
prone to complete failure or drift.
Furthermore, since they operate in
the radiation heat transfer regime
and are frequently located close to
heat recovery coils, the low fow
conditions present during tran-
sients mean that they may read
lower than the radiant section exit
temperatures.
Start-up tube failures
It is rare that a year passes without
a catastrophic reformer melt down.
This fact is illustrated in a number
of papers
.3-9
These events present a
signifcant hazard to plant person-
nel and a large loss in proftability.
Reformer monitoring
Start-up monitoring
Installing a CatTracker in a
reformer tube makes it possible to
measure the process gas tempera-
ture down the reformer tube. By
linking this temperature data into
the DCS, the plant operator has a
real time measurement of the
process gas temperatures. These
temperature measurements can be
alarmed to provide protection
against catastrophic tube failures
due to high temperature during
plant transients. Having this real
time data reduces reliance on plant
personnel to routinely check the
reformer fre box and tubes.
Figure 5 illustrates how the vari-
ous temperatures (process gas
inlet/outlet, fuegas outlet and
CatTracker) on a reformer vary
during a reformer start-up,
As Figure 5 shows, the CatTracker
is highlighting that the process gas
temperature within the reformer
tube is for the majority of the
start-up hotter than the reformer
inlet, exit and fuegas temperatures.
Throughout the whole start-up,
the CatTracker measurement is
Figure 3 Failed reformer tube in service
Figure 4 Failed tube out of service
j matthey.indd 3 10/03/2014 13:06
Advanced Solutions for
the Worlds Toughest
Energy Challenges
ExxonMobil Global Leader in Fuels and
Lubes Process Technologies
Enabled by Proprietary Catalysts and Solvents
ExxonMobil Technologies are applied across our corporation and by licensees worldwide in
a growing list of process industry applications. Leverage our vast experience and ongoing
commitment to continuously improve our industry-leading technologies.
www.exxonmobil.com/tsl
Resid Upgrading
Premium Diesel
Acid Gas Clean Up
Methanol to Gasoline
Premium Lubricants
KEY TECHNOLOGIES:
FP.Technology_4PTQ.indd 1 3/6/14 3:22 PM exxon.indd 1 10/03/2014 13:03
issues, such as internal fuegas
recirculation patterns within the
reformer.
1
By comparing measurements
from the CatTracker against the
results from a process simulation
and tube wall temperature meas-
urements, it is possible to provide
additional confdence in the process
simulation results. Figure 6
compares the average of the meas-
urements (a total of eight installed
in this reformer) against the
predicted results from Johnson
Mattheys Reform process
simulation.
There is a good match between
the process simulation and meas-
urements from the CatTracker. This
agreement between actual and
simulated allows increased conf-
dence in the results of what if
process simulations. Such simula-
tions could include operation at
lower steam to carbon ratio,
improved understanding of the
impact of operation at higher plant
rate, or determining the remnant
life of the catalyst.
Clearly the CatTracker only
provides temperature data for those
tubes in which they are installed. In
order to understand the overall
balancing of a reformer, a three
dimensional, or plan, temperature
map (see Figure 7) is a good opera-
tional aid as it allows for
identifcation of the hot and cold
sections of the reformer.
Using the tube wall temperatures
in the three dimensional plot, in
conjunction with the on-line
CatTracker, it is possible to under-
stand what the tube wall
temperatures could be in real time.
Minimisation of the temperature
spread across the reformer is impor-
tant to maintain high reliability and
high effciency. An accurate tube
wall temperature map is a necessary
prerequisite when attempting to
minimise this spread. For optimisa-
tion of effciency, a good rule of
thumb is that a decrease in tempera-
ture spread by 10F (6C) allows
savings of about $100 000/y for a
large scale hydrogen reformer.
Catalyst monitoring
Installation of the CatTracker in
a reformer tube provides a
Steaming
Again, the installation of the
CatTracker provides the ability to
monitor the tube temperature in
detail during this non-routine
activity.
Normal operation monitoring
Accurate tube wall temperature
measurement is paramount when
assessing the performance of the
reformer and the reforming cata-
lyst. It is therefore recommended
that temperatures within the
reformer be monitored on a regular
basis to provide early identifcation
of any problems. This type of
monitoring will also assist in iden-
tifying other reformer related
almost always higher than the
reformer outlet temperature. This
point reinforces the message that
reformer outlet temperatures
do not represent how hot the
process gas (and hence the reformer
tubes) actually are within the
reformer.
Shutdown monitoring
The system offers similar benefts
to the operator during a shutdown
as during a reformer start-up.
When the hydrocarbon feed is
removed, and the reforming reac-
tion ceases to remove a great deal
of heat removal, the tube tempera-
ture profle can be monitored in
real time to avoid overheating.
800
900
1000
700
600
500
400
300
T
e
m
p
e
r
a
t
u
r
e
,

C
200
0 10 20 30 40 50 60 70 80 90 100
Time, hours
Flue gas
Reformer inlet
Reformer outlet
CatTracker
Figure 5 Reformer temperatures during start-up
800
900
850
750
700
650
600
550
T
e
m
p
e
r
a
t
u
r
e
,

C
500
0 2 4 6 8 10 12 14 1 3 5 5 9 11 13
Distance down tube, m
Process simulation
CatTracker measurement
Figure 6 Comparison of CatTracker measurements against process simulations
j matthey.indd 4 10/03/2014 13:06
reformer can be operated close to
the carbon formation regime with a
positive impact on overall plant
protability.
By articially aging the catalyst
within the process simulation (to
reect additional time on-line or
the impact of poisoning), the point
at which carbon formation can then
be predicted and therefore the
timing for the change-out for the
catalyst can be made on a more
educated basis. This may allow for
more efcient operation or an
extension of catalyst life.
Steam to carbon
One option for reducing the impact
of carbon formation is to operate
with an increased steam to carbon
ratio. By the same process detailed
above for carbon formation, the
poisoned. A shutdown to steam the
catalyst may be required. The
steaming operation is usually
followed by operation at a higher
steam to carbon ratio to mitigate
the effects.
The main issue with carbon
formation, compared to all the
other limits that have been
discussed, is that carbon formation
cannot be measured directly. In fact
in an operating plant when carbon
is not forming it is impossible to
detect how far away the reformer
operation is from carbon formation.
With the additional condence
generated by a CatTracker and tube
wall temperature measurements, it
becomes possible that the process
simulation will more accurately
estimate the onset of carbon forma-
tion. With this information, the
measurement of the process gas
temperature within the tube. Until
now, this process parameter has
been very difcult to measure. By
using these measurements, in
conjunction with reformer process
simulation based plant data, the
activity and heat transfer character-
istics of the installed catalyst can be
determined far more accurately
than has historically been possible.
This increase in accuracy allows
for an improved understanding of
the present performance of the
installed catalyst and hence more
accurate prediction of future perfor-
mance. The future performance can
be predicted in terms of expected
tube wall temperatures, carbon
formation, and methane slip. In
addition, the allowable minimum
steam to carbon can be predicted
more precisely.
If as part of a what if study, the
operator is interested in tube wall
temperatures within the reformer,
then this can be determined using
the CatTracker calibrated process
simulation. An example of this is
illustrated in Figure 8, which high-
lights the process gas temperatures
(from the CatTracker and the
process simulation), against the
tube wall temperatures (measured
using a pyrometer and from the
process simulation).
As Figure 8 shows, there is excel-
lent agreement between the
corrected tube wall temperature
measurements and the simulated
tube wall temperature prole. Since
the systems temperature prole
matches that of the simulated
process gas temperature prole,
there is a high degree of condence
in being able to use this process
simulation for future catalyst
performance projections. Future
catalyst projections are based on
the catalyst parameters derived
from the measurements and the
known changes with time for
activity and pressure drop
characteristics.
Carbon formation
In most cases, carbon formation
prevents or limits operation of the
plant towards the end of catalyst
life, or in situations when the
reforming catalyst has been
2
1
3
4
5
6
7
R
o
w

n
u
m
b
e
r
8
4 8 12 16 20 24 28 32 36 40 44 48 6 10 14 18 22 26 30 34 38 42 46 50
Tube number
830-840C
880-890C
840-850C
870-880C
890-900C
900-910C
860-870C
850-860C
Figure 7 Reformer temperature map
800
1000
950
900
850
750
700
650
600
550
T
e
m
p
e
r
a
t
u
r
e
,

C
500
0 2 4 6 8 10 12 14
Distance down tube, m
Process simulation
CatTracker measurement
Process simulated TWT
Bottom TWT
Top TWT
Figure 8 Comparison of tube wall temperatures against CatTracker measurements
j matthey.indd 5 10/03/2014 13:06
predicted by this simulation is
predicted to be 0.003 between a
standard tube and one with a
CatTracker installed.
While the CatTracker has suf-
cient strength to hold itself in
position during catalyst loading,
the catalyst is loaded using a
Johnson Matthey patented method.
This technique ensures that its posi-
tion down the centre line of the
reformer tube is maintained while
ensuring that the catalyst is loaded
effectively.
The small diameter of the
CatTracker and patented loading
technique ensures that there is no
impact on the reformer operation
or on the gas ow distribution
through the catalyst lled reformer
tubes. This uniform ow has been
conrmed by comparing measured
temperature data from the system
and tube wall temperature meas-
urements taken by pyrometers and
reformer imager.
Other applications
In addition to the reformer installa-
tions, CatTracker has been installed
in a number of other catalytic unit
operations in synthesis gas plants
including medium and low temper-
ature shift converters, methanol
synthesis converters and, shortly, in
sour shift converters. Although
these applications are different to
the reformer application, its
features (high accuracy in terms of
temperature measurement and
precise sensing point location)
make it a useful measurement tech-
nique for a variety of applications.
These applications include the
following:
1. Axial ow converters that have
been short loaded
2. Axial ow converters that are to
be converted to radial ow
3. New radial ow converters
where both axial and radial temper-
ature measurements are required.
For existing reactor bed tempera-
ture schemes (point 1), the existing
sensing points within a conven-
tional thermowell are unlikely to be
in the correct position. Some are
very likely above the surface of the
bed. Replacement of the existing
unit with a CatTracker allows for a
redesign such that all of the sensing
Catalyst loading
As part of the development of this
application, Johnson Matthey
conducted a number of computer
simulations of catalyst loading in a
reformer tube. These simulations
included loading with and without
a body position on the centre line
of the reformer tube. The focus was
to determine whether a CatTracker
would impact on the loaded void-
age, and hence affect the process
gas ow through the tubes with the
system installed in comparison to
the normal tubes. Figure 10 high-
lights the achieved loading based
on the simulations.
The variation in voidage
time when this will become an
operational issue for the plant can
be determined. If this date is before
the next turnaround, then the
process simulation can be used to
determine what steam to carbon
should be used to maintain the
reformer in a carbon free
environment.
Operation at higher steam to
carbon ratios will clearly have a
negative impact on plant efciency,
but applying the above technique
limits this impact until the catalyst
can be changed.
Tube wall temperatures
The same approach can be taken to
predict tube wall temperatures as
the catalyst ages. This information
can also be used to determine the
timing of the catalyst change-out.
Installation
A CatTracker is relatively simple to
install in a reformer tube. The only
modication required is that a in
hole must be drilled through the
centre line of the top tube ange.
The modications to the ange are
illustrated in Figure 9.
The system is provided with
extension lines that are run back
from the end of the tube to a junc-
tion box placed in a suitable
location in the reformer penthouse.
The signal can then be transferred
from the junction box to the plant
DCS to allow for real time monitor-
ing of in-tube temperatures.
Figure 9 Tube ange with CatTracker
Standard tube CatTracker tube
Figure 10 Catalyst loading effects
j matthey.indd 6 10/03/2014 13:06
primary reformer tubes during 30 years: failures,
avoiding problems, and improving performance.
4 Nielsen N K, Is hot restart jeopardizing plant
safety?, AIChE 2012.
5 Cotton B, Broadhurst P, Common problems
on primary reformers, AIChE 2004.
6 Webb G M, Taylor W K, Reformer tubes: not a
commodity, AIChE 2006.
7 Rani B, Catastrophic failure of reformer tubes
at Courtright ammonia plant, AIChE 2006.
8 Rogers M, Lessons learned from an unusual
hydrogen reformer furnace failure, AIChE 2006.
9 Brightling J, Roberts M, Catalyst catastrophes
II, AIChE 2005.
Oliver J Smith IV is the Global Process Lead
for the Hydrogen and Syngas business with Air
Products and Chemicals, Inc. He holds a BChe
from the University of Delaware and a PhD in
chemical engineering from Carnegie Mellon
University. He has authored over six patents
and numerous journal articles in the areas of
process optimisation and asset utilisation.
Bill Cotton supports Johnson Matthey Leading
Edge Customers by direction of technical
support and development programmes. He has
worked in a variety of technical and techno-
commercial roles covering methanol, ammonia
and hydrogen plants covering technical
support, plant operations and plant revamps.
He holds a BEng in chemical engineering from
Loughborough University of Technology.
Conclusions
A CatTracker offers a range of bene-
ts to the operators of synthesis gas
plants across many unit operations
including reformers, shift and
synthesis converters. For the rst
time, operators can now accurately
measure the process gas tempera-
tures within a reformer tube and
gain an insight into something that
has been previously hidden from
view. The unique features of system
also allow for application to radial
ow converters.
There are endless possibilities for
application of single and multiple
units with a variety of designs and
sensing point locations.
CatTracker is a registered trade mark of Daily
Instruments Corporation. Daily Thermetrics is a
division of Daily Instruments Corporation.
References
1 Cotton W J, Fisher B R, A complete analysis
of your reformer, AIChE 2002.
2 Davies M, Fisher B, Cotton W J, Advanced
modelling tools and techniques for primary
reformers, AIChE 2004.
3 Atef Hassan Saad El-Din, A.F.C. experience in
points are in the bed. Furthermore,
the location of the sensing points
can be biased to allow for highly
accurate remnant life calculation
and therefore maximisation of cata-
lyst charge. For poison-susceptible
applications (for instance, low
temperature shift) some of the sens-
ing points can be biased to the inlet
of the bed, allowing for close moni-
toring of the rate of poisoning to
help determine the poisoned
volume and hence catalyst change-
out timings.
For radial/axial ow applications
(points 2 and 3), it is important to
understand whether there is good
ow distribution across both the
ow planes within the bed. A
CatTracker can be designed such
that a single unit can provide
temperature measurements in both
the radial and axial planes.
Conventional thermocouples strug-
gle to provide such data as they
cannot be bent into the complex
shapes required for such monitor-
ing. A CatTracker can be bent into
any appropriate shape.
We make technology work.
www.hoerbiger.com
Like musicians set the tone,
HOERBIGER sets standards in
the Oil and Gas Industry.
Image_ad_KT_2014.indd 1 06.03.2014 15:46:22
j matthey.indd 7 10/03/2014 13:06
hidden potential
UOPs proven skid-mounted gas processing equipment
can help you monetize your unconventional gas resources.
A global leader in gas processing technologies, UOP has been creating
world-class solutions for the natural gas industry for decades. Just as
you are developing new sources of natural gas, UOP is continuing to
innovate to provide the most efficient contaminant removal and liquids
recovery solutions. UOP offerings are smaller, lighter and pre-packaged
to help lower your capital and energy costs.
With UOP Separex Membrane Systems, youre up and running with faster delivery and lower installation
and operating costs. Separex Membrane Systems are ideal for remote locations where transportation
logistics are challenging. In many cases, our latest Twister Supersonic Gas Separation technology can
help provide gas liquids recovery as much as 20% higher than conventional dewpointing applications.
UOP Russells packaged plant systems help reduce field construction time and expenses and overall operating
costs. With modular equipment, monetizing your gas resources anywhere in the world becomes simple.
To learn more about how UOP can help you uncover higher profits, please visit www.uop.com/gasprocessing.
For more information about UOP, visit www.uop.com/gasprocessing
2014 UOP LLC. All rights reserved
A Century of Innovation
in the Oil and Gas Industry
1914 - 2014
UOP-Shale-Gas-ad-Hidden-Potential-Ad-Full-Page_UOP Ad 8.5x11 UOP6457 1/13/14 9:03 AM Page 1
uop hidden potential.indd 1 04/03/2014 11:16
Revamping the hydrogen generation unit
D
emand for refnery hydrogen
is increasing rapidly due to
the processing of heavy sour
crudes and stringent fuel quality
requirements. The existing capacity
for hydroprocessing and the associ-
ated hydrogen network limits
refnery throughput and operating
margins. Innovative process
schemes and reforming technolo-
gies are available for revamps of
the hydrogen generation unit to
enhance production substantially
with minimum capital cost and
energy consumption.
In this article, a simulation
model is developed to study the
impact of such upgrading schemes
with optimum utilisation of exist-
ing reformer and downstream
equipment. Installation of a pre-
reformer and gas heated reformer
in series or parallel combination
with the existing reformer can
enhance hydrogen generation
capacity by 10-27% and reduce
fuel consumption by 12%. For
capacity increases up to 40%, an
auto thermal reformer is more
attractive but this scheme requires
Hydrogen plants can be modied for higher production by employing innovative
process schemes
K V PATWARDHAN, V D DHUMAL and P BALARAMAKRISHNA
R&D Centre, L&T Hydrocarbon Engineering, India
major modifcations in the down-
stream sections.
An optimum revamp process
scheme can always be devised
depending on the plants specifc
requirements and capital/opera-
tional constraints.
Future hydrogen balance
Hydrogen is usually regarded as a
utility in oil refning. A modern
10 million t/y refnery will often
require 150 000 Nm
3
/hr or more of
hydrogen. Current trends in the
refning industry are reducing the
availability of hydrogen to the
point where most refneries are
concerned about their future hydro-
gen balance. New specifcations for
low sulphur fuels require increased
hydrogen consumption in hydro-
treaters. The use of heavier crude
oils and more bottom of the barrel
processing increase hydrogen
demand in hydrocracking and
heavy oil hydrotreating units. At
the same time, limits on the
aromatics content of gasoline, and
requirements for oxygenates, have
led to lower severity in the catalytic
reformer. As a result, less hydrogen
is produced in this unit. Demand
for gasoline and middle distillates
means that crude oil will require
greater processing and thus more
hydrogen.
In future, practically all fractions
of sour crude will be subjected to
catalytic processes that involve
hydrogen. The total quantity of
fractions directed to hydrotreating,
hydrocracking and hydrodesul-
phurisation will amount to 90% of
the total crude run. This demand
can be met by applying a system-
atic, cost effective approach to
hydrogen management. Options for
the supply of hydrogen include
increasing hydrogen plant
production, revamping existing
equipment, building a new hydro-
gen plant, purchasing hydrogen
from outside suppliers, or recovery
of hydrogen that was going to fuel
by installing a hydrogen purifca-
tion unit, or the lower cost
alternative of optimising and
revamping the hydrogen distribu-
tion network.
For many years, hydrogen has
Feed
Process
steam
HP
steam
Offgas
Hydrogen
DM water in
HDS
WHB
Steam
drum
DMW
preheater
DMW
preheater
Reformer
HTS LTS
Figure1 Schematic of hydrogen production by SMR
l&T.indd 1 10/03/2014 13:29
nickel catalyst contained in vertical
tubes situated in the radiant
furnace of the reformer
The third section is syngas heat
recovery and incorporates CO shift
reactor(s) to increase the hydrogen
yield. In the adiabatic CO shift
reactor vessel, the moderately
exothermic water gas shift reaction
converts carbon monoxide and
steam to carbon dioxide and
hydrogen
The fnal section is raw hydrogen
purifcation, in which modern
plants employ a pressure swing
adsorption (PSA) unit to achieve
the fnal product purity. The PSA
purifcation unit removes CO, CO
2
and CH
4
gases by adsorption from
the hydrogen.

Modern reforming technologies
Pre-reforming
Pre-reforming is used for low
temperature steam reforming of
hydrocarbon feedstocks ranging
from natural gas to heavy naphtha.
Converting higher hydrocarbons in
the pre-reformer results in stable
and mild operating conditions for a
downstream tubular reformer and
thus ensures reliable operation of
the tubular reformer. The pre-
reformer is placed upstream of the
tubular reforming unit. In order to
obtain the required steam to carbon
ratio, feedstock is mixed with
process steam before entering the
pre-reformer. In this, all higher
hydrocarbons are converted to a
mixture of carbon oxides, hydrogen
and methane at equilibrium based
on the methanation and water gas
shift reactions. Pre-reforming allows
operation at low steam to carbon
(S/C) ratios and thereby reduces
overall energy consumption. The
pre-reformer also increases the life-
time of the tubular reformer and the
shift catalysts as the sulphur pres-
ent in the hydrocarbon feed and
process steam is absorbed by the
pre-reforming catalyst.
Gas heated reformer (GHR)
GHR utilises the heat available in
the process gas at SMR exit for
steam reforming in a heat
exchanger type of reactor. This
scheme is available in series or
parallel combination with SMR. In
been produced commercially via
steam methane reforming (SMR) of
hydrocarbons. The technology has
now reached a mature state and is
the basis of a majority of plants built
for generating hydrogen in refning
and petrochemical complexes.
This article focuses on the revamp
of the existing SMR based hydro-
gen plant, either for higher
production or for lower specifc
energy consumption using modern
reforming technologies available
commercially from licensors.
Steam methane reforming
These plants consist of four basic
sections (see Figure 1):
The frst is feedstock treatment
where sulphur and other contami-
nants are removed
The second is the steam methane
reformer, which converts feedstock
and steam to syngas at high
temperature and moderate pres-
sure. The reforming reaction
between steam and hydrocarbons is
highly endothermic and is carried
out using a specially formulated
l&T.indd 2 10/03/2014 13:29
a parallel combination with the
available heat, up to 20% of the
feed can be split and taken to GHR.
In both combinations, a proportion
of heat is utilised in the process
side which reduces the steam
production from the plant. GHR
can also be used in series/parallel
arrangement with a SMR-ATR
combination of steam methane
reforming and autothermal reform-
ing. Other advantages of GHR
technology compared with the fred
steam reformer are:
Small size and footprint, which
can be essential when increasing
the capacity of existing plants
Lower investment costs, particu-
larly in plant expansions
Reduced steam utility size and
cost
The complete unit is transported
to the processing plant for installa-
tion, which saves costs when the
plant is located far from infrastruc-
ture or in a high wage country
Debottlenecks the existing
reforming section.
Autothermal reforming (ATR)
The ATR unit is a refractory lined
pressure vessel containing a burner,
a combustion chamber and a cata-
lyst bed. The hydrocarbon feedstock
is mixed with steam and pure
oxygen/enriched air/air at the top
of the reactor. As the name
suggests, the unit operates without
addition of any external heat. In the
combustion chamber, partial oxida-
tion reactions take place and the
heat generated is utilised for endo-
thermic steam reforming reactions.
This can be used in parallel or series
combination with the existing SMR.
Innovative process schemes
for revamps
The reforming technologies
discussed above are focused
towards incremental savings in
energy as well as lower emissions
from the fue gas stack. These tech-
nologies can be combined with the
existing SMR with the aim of either
capacity enhancement or reduction
in specifc energy consumption.
Some of the modifcations required
due to the addition of an individual
unit in the existing plant are
discussed as follows.
Pre-reformer
This imposes minimum modifca-
tion in the present scheme. The
reformer convection section will
have to be modifed to incorporate
an additional pre-heating coil for
the pre-reformer. With the addition
of a pre-reformer, the reformer can
be operated at a lower S/C ratio,
thus a reduction in fring energy
can be observed. The S/C ratio
(1.52.0) in the reformer is limited
by the high temperature shift cata-
lyst. Hence the high temperature
shift (HTS) catalyst needs to be
replaced with medium temperature
shift (MTS) catalyst operating at
lower temperatures.
GHR
GHR schemes require connection
modifcations between feed purif-
cation and the outlet of the
reformer. The gas fow in the shift
section is approximately 20-25%
higher. To obtain the same CO slip
from the exit of the shift section, an
additional catalyst volume is
needed either in MT/LT shift reac-
tors or in an additional CO guard
bed.
ATR
Modifcation with ATR is possible
only in the case of availability of a
cheap source of pure oxygen. This
oxygen needs to be compressed to
the operating pressure of the
reformer, hence a dedicated oxygen
compressor is required. This unit
runs in parallel to the reformer
which will require a parallel waste
heat boiler and steam drum system
to compensate for additional steam
production from the plant. The
shift section as well as the PSA unit
will require major modifcation in
terms of additional volumes of
catalyst and adsorbents. The
economics of the schemes employ-
ing ATR should be compared with
that of a new parallel line of SMR.
All three process modifcations
impose higher pressure drop in the
system and need a higher pressure
of natural gas at the battery limit.
These technologies lead to vari-
ous process scheme combinations
in the reforming section.
Model development and simulation
A simulation model was developed
in Aspen Hysys V7.3 for two
options:
Hydrogen production (Option A):
increase hydrogen capacity by
keeping fxed reformer duty. This
may require additional modifca-
tion in downstream equipment.
Specic energy consumption (Option
B): reduce the specifc energy
consumption keeping the same level
of hydrogen production. This option
will not require any modifcation in
Feed
Process
steam
HP
steam
Offgas
Hydrogen
DM water in
HDS
GHR
WHB
Steam
drum
DMW
preheater
DMW
preheater
Pre-reformer
Reformer
MTS LTS
Figure 2 Schematic of hydrogen production by GHR-p
l&T.indd 3 10/03/2014 13:29
case is limited by the reformer as
higher hydrocarbons have a
tendency to deposit carbon on the
catalyst. The steam produced at the
steam drum is utilised: for heating
natural gas, as process steam, in the
process condensate stripper, and in
steam turbines required to run ID/
FD fans and a boiler feed water
pump. The net steam exported
from the hydrogen plant is the
difference between that generated
and consumed in the plant. Though
the amount of steam produced in
the plant is high, the majority of it
is consumed as process steam in
the plant.
Case 1: pre-reformer + reformer +
MTS + LTS + PSA
Hydrogen production
The pre-reformer converts all
higher hydrocarbons to methane at
lower reforming temperatures, thus
reducing the risk of carbon deposi-
tion in the reformer. Hence in this
scheme S/C ratio can be reduced to
between 1.5 and 2.0. The bottleneck
of operating at this lower S/C ratio
is shifted to the shift section where
For a pre-reformer scheme
S/C ratio at the inlet of the
primary reformer is 1.5-2.0.
For an ATR scheme
Oxygen purity is considered as
>99.9%
ATR outlet temperature is fxed
at 930C
S/C & O
2
/C ratio is kept at a
minimum and adjusted to match
the required reformer outlet
temperature.
For a GHR scheme
Natural gas feed split (10-20%)
GHR outlet temperature for the
combined gas is maintained around
770C
S/C ratio is maintained at 2.8.
Results and discussion
Base case
In this scheme, steam to carbon
ratio is kept to 3.0, following
convention for reformers in opera-
tion. A higher S/C ratio leads to
higher steam production from the
steam drum at the cost of natural
gas fuel fring. The S/C ratio in this
downstream equipment. In addition,
reformer conditions will be less
severe which will enhance the life of
catalyst and tube. The following
parameters are considered as the
basis for simulation.
Base case
Natural gas conditions at the
battery limit: pressure 35 kg/cm
2
.g,
temperature 40C, net calorifc
value 11 222 kcal/Nm
3
, and CH
4
=
72% (mole%)
Hydrogen production for this
scheme is 50 kNm
3
/hr
The scheme adopted is Reformer
followed by HTS & LTS combina-
tion in shift section followed by
PSA for fnal purifcation
Reformer: heat duty = 45.8 Gcal/
hr, outlet temperature = 910C,
with 10% excess air in the reformer
fue gas
S/C ratio at the inlet of reformer
is maintained at 3.0
PSA recovery effciency is 89%
and off-gas generated from PSA is
fred in the reformer
High pressure steam production
at 45 kg/cm
2
.g and 425C.
With broad experience in the management
of treatments applied to bunker fuel, his
main areas of activity include ow improvers
for middle distillates and residues. He holds
a masters degree in chemistry from Rome
University. Email: ggrande@chimec.it
Alessandra Berra is Technical Development
Manager with Chimec Fuel Additives
Technological Units. She has over 15 years of
experience in fuel additives development and
marketing in the downstream and renewable
fuel areas and holds a masters degree in
chemistry from Rome University.
Email: aberra@chimec.it
key to avoiding unwanted off-specs
and unnecessary extra consumption
of chemicals.
To achieve these targets, collabo-
ration between customer and
supplier is fundamental to identify
critical variables, available tools
and options and areas for improve-
ment. And it is only through such
collaboration that new ideas and
tailored solutions can come alive.
Giacinto Grande is Product Manager with
Chimec Fuel Additives Technological Units.
are confdential information but the
trend may be easily seen. The aver-
age dosage rate during our
management was reduced by 37.5%
as a result of implementation of the
predictive software.
As a result of the good results
achieved, the industrial test turned
into a continuous application to the
full satisfaction of the customer and
the supplier. The model algorithm
and parameters are controlled and
updated when necessary by Chimec
personnel, to maintain the effective-
ness of the software.
Conclusions
The case study presented here
shows how an H
2
S scavenger appli-
cation may need particular care to
provide real cost optimisation. Not
all refneries have conditions simi-
lar to those discussed and the
necessary solutions may differ but
it is the approach that must be the
same in all the cases.
Good additive selection is impor-
tant, but generally not suffcient. A
proper treatment setup and tailored
monitoring procedures provide the
R
a
t
i
o

a
d
d
i
t
i
v
e
:

H
2
S
Time
Treatment start-up Further optimisation
achieved after software
implementation
Average dosage rate:
37.5% relative to our
current start-up
Figure 3 Dosage trend
potentially to the revamp of exist-
ing facilities. The zero gasoline
production refnery is a practical
and viable reality.
References
1 Gillis D, Unique opportunities with proven
technologies to maximise residue conversion
& renery margins, Asia Technology forum,
Bangkok, 10-11 Oct 2012.
2 Edwards S M, et al, Relative economics of
mixed C
4
s processing routes, PTQ, Q1 1998, 1.
Blasis Stamateris is Downstream Business
Consultant in the Business Solutions Group
of Foster Wheeler, UK, participating in renery
conguration studies, feasibility studies and
FEED work for grassroots ren eries, CTL
facilities and renery upgrades. A graduate in
chemical engineering, he has over 25 years
experience in oil rening and upgrading.
Email: blasis_stamateris@fwuk.fwc.com
Dan Gillis is Director, Rening for Foster
Wheeler, US. He provides technical direction
and applications development for Foster
Wheelers proprietary heavy oil technology. He
is an engineering graduate of the University of
Saskatchewan, Canada.
Email: dan_gillis@fwhou.fwc.com
and competitive confguration.
Certainly, if the objective is to
maximise revenues with proven
technology, in a reliable environ-
ment, the confguration that has
been developed by Foster Wheeler
utilising its SDA + DCU combina-
tion offers a competitive advantage.
How to make the right decision
The integration of refning and
petrochemical facilities offers
several opportunities for adding
value to refnery streams by
exploiting synergies between differ-
ent process units while rationalising
investments that can positively
impact project economics.
Different refnery confgurations
can be tailor-made to suit a range
of processing objectives, including
variations to the propylene-to-
ethylene ratio, and the production
of middle distillates and aromatics.
Capacities of the process units can
be adjusted to accommodate differ-
ent crude oils.
The concepts outlined are appli-
cable to grassroots projects and
of magnitude total installed cost
(TIC) estimates that have been
prepared, based on Foster
Wheelers in-house cost estimating
database.
As can be seen, the investments
for the upgrading section of the
refnery confgurations are of the
same order of magnitude. With
similar product slates and invest-
ments, the differences in economic
indicators between the refnery
confgurations will not be signif-
cant enough to clearly favour one
confguration over another.
Therefore, when it comes to the
decision of which refnery confgu-
ration to choose, other factors
besides the product slate and
investments, such as operating
complexity and associated reliabil-
ity of the units, should be taken
into consideration.
Ultimately, the selected option is
based on an individual refners
objectives. The SDA + DCU options
represent well-proven technologies.
If there is a suitable outlet for the
coke produced, this is a viable
customer conveyed the need to
process more oil sands-derived
coker gas oil. Consequently, the
rst initiative was to plan a transi-
tion to a more metals-tolerant
catalyst system, and Ascent
DN-3551 catalyst was selected for
Cycle 2. This catalyst has been
shown to have superior feed poison
tolerance and was considered to be
better suited to accommodate the
expected poisons that would be
associated with processing incre-
mentally higher percentage of oil
sands-derived coker gas oils.
The second initiative included
optimisation of the demet catalyst
package to enhance catalyst stabil-
ity. Prior to Cycle 1s completion
and the DN-3551 loading, we
advised the customer that the opera-
tion was being subjected to a higher
feed endpoint and the catalyst
system was at risk of accelerated
aging due to feed resid and metals
entrainment. Subsequent detailed
feed component analysis at
Criterions Research facility
conrmed elevated feed poisons
by DN-3551. The points enclosed in
the highlighted areas labelled
DN-3651 represent samples tested
as part of scale-up and manufactur-
ing optimisation. Use of high
throughput experimentation
allowed rapid development of a
reliable manufacturing process
yielding the highest practical cata-
lyst performance.
Applying technology improve-
ments such as DN-3651 requires a
holistic joint approach to ensure
that the full benets of the technol-
ogy can be taken advantage of by
the operator, taking into account
the site-specic constraints and
objectives.
In this example, Criterion utilised
a collaborative approach with its
customer to improve catalyst
performance and facilitate more
difcult oil sands-derived feed
processing capabilities. Three initia-
tives were implemented.
The original cycle 1 utilised
Criterions Centinel DN-3100 cata-
lyst as the primary conversion
system. During the cycle, the
tubular ow reactors with auto-
mated process control and sampling.
The use of this multi-tube reactor
system allows signicant accelera-
tion of catalyst development relative
to conventional testing techniques.
Leads generated with the high
throughput equipment were
conrmed by conventional-scale
pilot plant testing.
The target performance level for
DN-3651 was an improvement in
HDN activity of at least 10F
(5.5C) relative to DN-3551 (~20
RVA) with equivalent or better
HDS activity with a challenging
design feed containing bitumen-
derived vacuum gas oil and heavy
coker gas oil. Figure 14 illustrates
the high throughput experimenta-
tion HDN and HDS activity data
obtained during the prototype
development phase as well as
during the scale-up and commer-
cialisation phase. The activity as
presented in these plots is the
decrease in temperature required to
achieve the target S or N level in
the product relative to that required
Kurita Europe GmbH
.
Industriering 43
.
41751 Viersen
.
Germany
Phone:+49 2162 9580110
.
www.kurita.de
We keep a close eye on your plant effciency.
The trouble-free operation of your plant is the precondition for maximum economic effciency. Kuritas patented ACF technology inhibits the formation
of chloride and ammonium salts, thereby preventing fouling and corrosion. This is only one example of how Kurita protects your production unit against
unscheduled shutdowns, extends its runtime and lowers process costs while at the same time increasing plant and operational safety. Particularly
for refneries and petrochemical plants, Kurita offers innovative and sustainable process technologies as well as a broad portfolio of products for water
management. From consultation through to implementation and maintenance, we support you with a comprehensive range of services.
As you can see, we keep a close eye on the effcient operation of your plant.
Process technologies for refneries and petrochemical plants
Water treatment technologies for cooling water, boiler water, waste water and reverse osmosis
RZ_Anzeige_Quer_24052013_MQ.indd 1 24.05.13 15:24
cri.indd 8 10/06/2013 16:38
f wheeler.indd 8 12/09/2013 13:41
chimec.indd 4 11/12/2013 12:39
l&T.indd 4 10/03/2014 13:30
it is required to produce lower
steam from the plant and save
natural gas.
ATR-s + MTS + LTS + PSA
Hydrogen production
Increasing the load on the existing
reformer affects the methane slip at
the outlet of the reformer as
reformer heat duty is the limiting
parameter. In this case, the addition
of ATR in series with the primary
reformer is very useful to control
methane slip. The reformer outlet
temperature drops when the
reformer load is increased, resulting
in higher methane slip
which can be reduced in ATR by the
addition of process air. This scheme
(see Figure 4) can increase hydrogen
production capacity by 40% with a
fall in steam production of 17%.
Specic energy consumption
Direct fring in ATR saves approxi-
mately 4-5% of the fuel which
would otherwise have gone to
up to 23% without changing steam
production.
This scheme provides a reduction
in specifc natural gas consumption
by 8%. This can be opted for when
it is required to produce lower
steam from the plant and save
natural gas.
Case 3: pre-reformer + GHR-s (series)
+ reformer + MTS + LTS + PSA
Hydrogen production
This scheme (see Figure 3) provides
an additional 27% hydrogen
from the existing plant at the
expense of ~20% less steam
production. The pressure drop
observed in this case is higher by
2-3 kg/cm
2
which needs to be
accounted for at the natural gas
battery limit.
This scheme provides a reduction
in specifc natural gas consumption
by 13%. This can be opted for when
HTS catalyst is replaced by MTS
catalyst. Simulation indicates
hydrogen production can be
increased by 6-8%.
Operating with a lower S/C ratio
has two opposing effects: reformer
fuel is saved as less fring is
required to achieve the same
temperature at the reformer outlet
and lower energy in the reformed
gas available for steam production.
In this scheme, a reduction in
specifc natural gas energy in the
range 5-7% can be observed.
GHR-p (parallel) + MTS + LTS + PSA
Hydrogen production
In this scheme (see Figure 2), addi-
tion of GHR in parallel combination
with the primary reformer gives
the added advantage of increasing
plant capacity without disturbing
the existing reformer. Simulation
results show that hydrogen produc-
tion capacity can be increased by
Feed
Process
steam
HP
steam
Offgas
Hydrogen
DM water in
HDS
GHR
WHB
Steam
drum
DMW
preheater
DMW
preheater
Pre-reformer
Reformer
MTS LTS
Figure 3 Schematic of hydrogen production by GHR-s
Feed
Process
steam
HP
steam
Offgas
Hydrogen
DM water in
HDS
ATR
WHB
Steam
drum
DMW
preheater
DMW
preheater
Pre-reformer
Reformer
MTS LTS
Figure 4 Schematic of hydrogen production by ATR-s
l&T.indd 5 10/03/2014 13:30
provides more heat for GHR. This
will result in a 33% increase in
hydrogen production and a 10%
increase in steam production for a
20% feed split in GHR..
A reduction of 8% in specifc natu-
ral gas consumption can be
achieved in this combination (see
Figure 6).
The results of the above discus-
sions are summarised in Table 1 for
relative hydrogen, steam produc-
tion, and natural gas consumption
for Option A. Table 2 summarises
the specifc energy consumption
results for Option B.
Conclusion
The existing hydrogen facility is
becoming a bottleneck because of
stringent emission control legisla-
tion for transportation fuels. Steam
methane reforming is a well proven
technology conventionally used in
refneries for hydrogen generation.
Plants can be modifed for
higher production by employing
reformer increases hydrogen
production by 27% with a fall in
steam production by 13.5%.
A reduction of 15-16% in specifc
natural gas consumption can be
achieved.
ATR-s + GHR-p + MTS + LTS + PSA
Hydrogen production
As conversion in GHR is based on
heat available in reformed gas from
the primary reformer, ATR in series
atmosphere via the fue gas stack.
The combination of these two tech-
nologies, reforming and ATR,
improves the H
2
/CO ratio achieved
at the exit of the reforming section
and the energy required for reform-
ing. A reduction of 15-16% in
specifc natural gas consumption
can be achieved.
ATR-p + HTS + LTS + PSA
Hydrogen production
In this scheme (see Figure 5), ATR
kept in parallel with the existing
Case No. S/C ratio % H
2
% NG % Steam
(reformer inlet) production consumption production
Base case 3.0 100 (50 kNm
3
/hr) 100 (3.93 Gcal/knm
3
) 100 (75 TPH)
1 1.5-2.0 108-106 95-93 75
2** 2.8 112-123 94-92 108-100
3 2.8 127 87 82
4 1.5-2.0 141 85-84 82
5 1.5-2.0 127 92-94 86
6 2.8 133 92 110
*Reformer heat duty is xed at 45.8 Gcal/hr for all cases. **Feed split for GHR is 10-20%.
Process scheme comparison for relative hydrogen production (Option A)
Table 1
Feed
Process
steam
HP
steam
Offgas
Hydrogen
DM water in
HDS
ATR
WHB
Steam
drum
DMW
preheater
DMW
preheater
Pre-reformer
Reformer
HTS LTS
Figure 5 Schematic of hydrogen production by ATR-p
Feed
Process
steam
HP
steam
Offgas
Hydrogen
DM water in
HDS
ATR
GHR
WHB
Steam
drum
DMW
preheater
DMW
preheater
Pre-reformer
Reformer
HTS LTS
Figure 6 Schematic of hydrogen production by ATR-s+GHR-p
l&T.indd 6 10/03/2014 13:30
V D Dhumal is a Process Engineer at the R&D
centre of Larsen & Toubros Hydrocarbon IC.
He has been working in process simulation,
design and commissioning for the last two
years and holds a masters degree in chemical
engineering from Institute of Chemical
Engineering, Mumbai.
P V Balaramakrishna is head of the Process
Engineering Group at the R&D Centre of Larsen
& Toubros Hydrocarbon IC. He has more
than 20 years experience in process design,
advanced process control, commissioning,
troubleshooting and optimisation of process
plants, and holds a masters degree in
chemical engineering from Indian Institute of
Technology, Kanpur.
that it is possible to reduce specifc
energy by up to 15%.
The selection of an optimal
process scheme depends on the
plants specifc requirements of
capacity enhancement and energy
reduction, taking into consideration
layout and operational constraints.
K V Patwardhan is a senior Process Engineer
at the R&D Centre of Larsen & Toubros
Hydrocarbon IC. He has seven years experience
in process design, modelling, troubleshooting
and optimisation of process plants, and holds a
PhD in chemical engineering from Institute of
Chemical Technology, Mumbai.
innovative process schemes such as
pre-reforming, autothermal reform-
ing and gas heated reforming in
series/parallel combination with
SMR.
A process revamp with a pre-
reformer provides the fexibility to
handle changes in the feedstock
composition as well as delivering
8-10% higher production from
existing facility.
GHR in parallel/series arrange-
ment can increase plant capacity by
12-27% but at the expense of steam
production from the plant. This
scheme offers the best solution for
capacity enhancement and for
reducing energy consumption, espe-
cially for refneries where steam
export is not a major constraint.
For large capacity enhancement
(~50%), a process scheme with ATR
can be considered when a low cost
oxygen source is available. However
this requires major modifcations in
all sections of the plant and the cost
impact should be compared with a
new hydrogen plant.
Simulation studies indicate, in
addition to capacity enhancement,
Case No. Reformer Specic NG Specic steam Net specic NG Net specic
heat duty, (feed + fuel), (A) export, (B) energy, (AB) consumption, energy,
Gcal/hr Gcal/kNm
3
Gcal/kNm
3
Gcal/kNm
3
% %
Base case 45.8 3.93 0.5 3.43 100 100
1 41.8-41.4 3.61-3.77 0.62-0.59 2.99-3.18 92-96 87-93
2* 37.7-38.5 3.69-3.44 0.45-0.40 3.24-2.99 94-88 95-87
3 36 3.41 0.39 3.02 88 88
4 38.6-39.7 3.53-3.57 0.60-0.58 2.92-2.99 90-91 85-87
5 36.3-36.1 3.57-3.69 0.64-0.70 2.93-2.99 91-94 85-87
6* 35.6-36.3 3.47-3.39 0.46-0.42 3.01-2.97 88-86 88-87
* Feed split for GHR is 10-20%; hydrogen production is 50 kNm
3
/hr **
Process scheme comparison for specic energy consumption (Option B)
Table 2
TESTED RIGOROUSLY, TRUSTED IMPLICITLY.
FEATURES
163 LUMENS
360
o

ROTATING HEAD
2 METER IMPACT TESTED
FLOOD
INTEGRATED
HOOK
POWERFUL
MAGNET
DURABLE
CLIP
SPOT
30 EAGLEVILLE ROAD, EAGLEVILLE, PA 19403 | 800.523.7488
CONNECT WITH US WWW.STREAMLIGHT.COM
No fashlight is more versatile in hazardous
conditions than the new Knucklehead
HAZ-LO

from Streamlight. Its head articulates 210
and rotates 360, respectively. Its removable
magnet delivers 135 pounds of pull strength.
And its available as either a Flood with an
integrated hook or a Spot with a spring-loaded
clip. Either way, this Class 1, Div. 1 light is the
perfect addition to any toolbox of intrinsically
safe equipment. When you
need a light you can trust,
the Knucklehead
HAZ-LO
is
at the top of its class.
KNUCKLEHEAD

HAZ-LO
FLOOD AND SPOT
l&T.indd 7 10/03/2014 13:30

We didnt build the rst Boiler.
But in all your born days, you wont nd a manufacturer today that makes a boiler that performs
better than a RENTECH boiler. Its no yarn. Each of our boilers is custom-designed by RENTECH
engineers and built in state-of-the-art facilities to operate efciently in its unique application in a
variety of industries. Our innovative, cost-effective technology will add value to your day-to-day
operations with lasting benets for the competitiveness of your business. Dont wait another day,
call us about your next boiler project.
325.672.3400
WWW.RENTECHBOILERS.COM
rentech.indd 1 04/03/2014 11:09
Enhance processing with capacity control
of reciprocating compressors
I
ndustrial compression of gases
such as hydrogen or hydrocar-
bons in refneries is growing due
to increased demand for high
purity fuels, bottom-of-the-barrel
conversion technologies for heavy
oil feedstock and increasing
volumes of shale-based paraffnic
crudes. In combination with low or
even varying molecular weight
hydrocarbons, reciprocating
compressors are often the best
and most economic solution
for compression. Reciprocating
compressors (recips) are fexible,
energy effcient and suitable for
high-pressure applications.
In the past, reciprocating compres-
sors were often considered
unreliable, with diffculties in
achieving precise control of their
output capacity. Electric power was
wasted because the compressor
capacity was typically controlled by
recycle valves, often referred to as
bypass or spillback valves, in combi-
nation with step control. Step
control is realised by means of
compressor suction valve unloading
systems or fxed volume clearance
pockets. Due to the large control
steps provided by such systems,
compressors were not operated eff-
ciently in terms of the electrical
energy consumption of the main
driver. Furthermore, the generally
slow reaction speed of recycle
valves does not allow for the most
effective optimisation of the process.
With todays increasing capacity
and power ratings of new recips,
advanced control systems can
signifcantly reduce energy
consumption by avoiding excessive
recycling of process gas. On top of
this, reciprocating compressors are
Improvements to renery, petrochemical and gas processing are available from
the full-range stepless capacity control of reciprocating compressors
KLAUS STACHEL and MARKUS WENISCH
Hoerbiger Compression Technology
frequently integrated into complex
processes, with side streams or
multi-stream compression, which
requires precise control and opera-
tional fexibility.
Todays renery challenges
The refning industry is undergoing
continuous technical changes in
order to comply with stringent
environmental regulations for clean
fuel products against a backdrop of
a wider variety of crude feedstocks
ranging from heavy asphaltenic
crudes to high API gravity shale
crudes.
Processing of highly contami-
nated crude feedstock, with high
sulphur content, represents one of
the biggest challenges for existing
plants. On the other hand, the
industry has to face a clear market
trend towards the quality improve-
ment of light fuels for road, air and
marine transportation, while
demand for heavy fuel products for
industry and power generation is
declining. The conversion of such
sour crude into light and mid-
distillate products through hydra-
tion processes is a vital part of the
production chain of a refnery, but
it is also afficted with high capital
and operating costs and signifcant
investments in the erection of new
facilities or conversions of existing
plants. Hence, fexible and sustaina-
ble production is of utmost
importance.
As substantial parts of many
refning processes, such as hydroc-
racking, hydrotreating, isomerisation
and reforming plants, multi-stage
reciprocating compressors are used
to compress hydrogen at high pres-
sure (up to 220 bar) for essential
parts of the process.
A change in feedstock quality
requires fexible plant operation
involving advanced capacity
control concepts for any reciprocat-
ing compressor. Therefore, many
end users, compressor manufactur-
ers (OEM), engineering companies
(EPC) and process licensors are
already taking advantage of such
advanced control systems, either as
an integrated system for new
equipment or as compressor
upgrade solutions for existing
machines.
Conventional control concepts for
reciprocating compressors
In previous years, many existing
refneries were upgraded to produce
clean fuels or process feedstock with
higher sulphur content. Before the
modifcation, the sole capacity of
one compressor, running at full
load, was suffcient to meet hydroc-
racker or hydrotreater hydrogen
demand. A second compressor was
kept in stand-by mode only to
ensure equipment redundancy.
With todays growing demand for
hydrogen, the sole capacity of a
Processing of highly
contaminated crude
feedstock, with high
sulphur content,
represents one of the
biggest challenges
for existing plants
hoerbiger.indd 1 10/03/2014 13:39
ing compressor is the recycle or
spillback control. Here, the
compressor itself runs at full load
or at dened load steps. Stepwise
capacity variation is achieved by
additional control devices, such as
pneumatic compressor suction
valve unloaders or xed volume
cylinder clearance pockets. In order
to regulate the gas ow according
to process demands, part of the
compressed gas is re-expanded and
recycled to the suction side, result-
ing in signicant energy losses.
Figure 1 shows a typical load
scenario of a hydrogen make-up
recycle application. The production
requires a hydrogen (H
2
) ow,
which corresponds to 80% of the
compressors rated capacity.
Consequently, the compressor
delivers 100%, while 20% of the
excess process gas is recycled to the
suction side.
Another important aspect of recy-
cle control is the risk that the
process gas can become contami-
nated with, for example, catalyst
debris, which is entrained in the
compressor, causing performance
degradation, higher wear rates and
reduced lifetime of performance
determining components such as
compressor valves, rider bands and
main packings.
Step control
Step control, also known as on/off
control, is another widespread
method to adjust the output of a
reciprocating compressor. Capacity
variation is achieved by perma-
nently unloading the compressor
suction valves of one or more cylin-
der ends. The possible variation of
load steps is dened by the number
of cylinders per compression stage.
Looking, for instance, at a two-cyl-
inder double-acting compressor
with two stages, 0%, 50% and 100%
capacity steps can be realised.
Suction valve unloading is achieved
by means of pneumatic actuators
that deactivate the compression in
the specic compression chamber.
It is an inherent disadvantage of
step control that this method is
only efcient as long as the
required process gas ow is equal
to the adjusted load step on the
compressor. Otherwise, excessive
single compressor is no longer suf-
cient, and therefore many stand-by
compressors have now become
main units in order to satisfy the
additional hydrogen demand. For
new greeneld projects, the plant
layout has also changed. For new
plants, a stand-by unit is often not
foreseen, as a result of the desire to
limit capital cost and also taking
into account the higher reliability of
modern recips.
Typically, the output capacity of
two compressors running in parallel
is higher than the maximum antici-
pated process requirement at full
production output. Therefore, gas
ow into the process itself has to be
reduced by a control system to
match precisely the prevailing
process demand. Conventional
control systems do not allow for
effective process control. Some of
the systems enable only incremental
adjustment of the capacity (for
example, 0%, 50%, 100%), while
others drastically decrease the ef-
ciency of the plant by the recycling
of compressed gas back to the
suction side of the machine. Various
conventional methods used to
control compressor capacity include
recycle valve control, step control
and clearance valve control.
Recycle valve control
The most common way of
controlling the ow of a reciprocat-
Gas feed
e.g. H
2
Feed
Treated product,
e.g. low sulphur
diesel, gasoline, etc.
20% recycling of H
2

to the suction side.
Since recycling rate is
proportional to consumed
power, energy is wasted by
using the bypass valve for
continuous load control.
The spillback valve adjusts
the required recycle
gas-flow by recycling
either to the make-up feed
or to the refinery network.
Compressor supplies
100% H
2
, the remaining
20% of the gas has to be
recycled through the
bypass.
Required H
2
load, e.g. 80%
of rated compressor
capacity.
Load depending on feed
quantity and quality (sulphur
content or other gas impurities).
Depending on the
feedstock quantity and
quality, more or less
recycle gas is
generated and needs to
be re-compressed.
Bypass/spillback
valve
Recycle valve
Process /
Reaction
Recycle gas
separation
Feed
compressor
Feed pump
Recycle / net gas
compressor
(Controlled with
pneumatic unloaders)
20%
recycle
80% load
required
100% capacity
Figure 1 Schematic layout of a compressor running on recycle-/spillback control
hoerbiger.indd 2 10/03/2014 13:40
enersul.indd 1 10/03/2014 11:01
gas has to be recycled through a
spillback control valve, resulting in
energy losses due to the re-expan-
sion of already compressed gas.
Whenever the compressor is
switched to a higher or lower load,
sudden fow changes are generated.
These fow changes cause pressure
fuctuations in the process, since
the recycle control valves are not
fast enough to compensate this
disturbance effectively. Pressure
fuctuations have a signifcant
impact on the process and can lead
to operational problems with
upstream or downstream equip-
ment. Furthermore, with lubricated
compressors, continuous cylinder
unloading leads to an accumulation
of lubrication oil inside the
unloaded compression chamber,
which is then discharged together
with the process gas as soon as the
specifc cylinder end is loaded
again. The resulting oil slug can
damage the compressor valves and
consequently lead to an unplanned
compressor stop for maintenance.
Due to the low degree of automa-
tion of many existing step control
systems, changes of compressor
load often have to be initiated
manually by the operators. In
practical terms, these devices are
rarely used and the compressor
runs at full load. In many cases,
experienced plant personnel are
required to be available for
compressor starts, stops and
switch-over procedures in order to
maintain stable process conditions
at all times.

Clearance volume control
If the compressor is equipped with
fxed or variable clearance pockets,
additional clearance volume can be
added to the compression chamber.
The compressor output reduces if
the clearance volume of the cylin-
der increases. Figure 2 shows a
fxed volume clearance pocket
installed on the cylinder cover. The
fxed clearance pocket enables a
75% load step in addition to the
0%, 50% and 100% load steps per
cylinder, which can already be
achieved by suction valve unload-
ers. However, recycle valve control
is still required to precisely match
the process requirement.
Stepless capacity variation can be
achieved by means of variable
clearance pocket control, which
enables a capacity turndown down
to around 60% per cylinder.
However, variable volume clear-
ance pockets for make-up and
recycle gas compressors can only
be adjusted when the compressor is
stopped and de-pressurised.
Therefore, this technology is not
suitable for refning and petro-
chemical applications. In most
cases, the operators do not utilise
variable clearance pockets and,
consequently, the capacity is again
controlled by the recycle valve.
Advanced control system
A solution to overcome the previ-
ously discussed obstacles is to make
use of a stepless, full-range control
system based on the reverse fow
principle. Here, compressor output
is regulated steplessly by means of
hydraulically actuated and electron-
ically controlled actuators (see
Figure 3), which are directly
installed on the suction valve covers
of the compressor (stepless valve
unloaders).
The reverse fow principle
achieves capacity variation by
actively controlling the start of the
Figure 2 Compressor cylinder equipped with a xed clearance volume pocket (in front of
the cylinder)
Figure 3 Electro-hydraulic reverse ow control system
hoerbiger.indd 3 10/03/2014 13:40
dynamics and contributes signi-
cantly to improved process
exibility. In case of process distur-
bances, such as those caused by
problems with up- or downstream
equipment, the system immediately
manages to control pressure uctua-
tions and most likely avoids a plant
shutdown. Provided such a system
is considered for new compressor
projects, it can considerably reduce
the capital costs for the compressor
where energy-efcient operation is
required (see Figure 6).
By using an electro-hydraulic
control system as an integrated part
increasing suction valve durability
and maintenance intervals, and
reducing maintenance costs. Since
the run time between compressor
stops is often limited due to the
lifetime of the compressor valves,
service intervals can thus be
increased, and production periods
can be extended.
By using advanced capacity
control technology, the recycle valve
remains fully closed during normal
compressor operation with process
gas. Compared to conventional
systems, stepless reverse ow
control features the highest control
compression stroke. The working
principle is outlined in Figure 4. At
full load, compression starts at
point C when the piston is in
bottom dead centre (BDC). At part
load, the compression starts at C
r
.
Since the volume in the cylinder
gets reduced from C to C
r
, a
portion of the gas ows back to the
suction chamber. Hence, only the
gas ow required by the process is
compressed.
Since the indicated power of the
compressor is proportional to the
gas ow rate being compressed, the
power consumption of the electric
driver decreases at part load, result-
ing in signicant energy savings.
Figure 5 shows the potential of
power cost savings for a mid-size
hydrogen compressor with 3.0 MW
indicated power (power required
for compression). The graph shows
annual savings in power costs as a
function of annual average
compressor output and various
costs for electric power (70, 100 and
120 per MWh). This reveals huge
savings potential, thus allowing for
a high return on investment and
very short payback periods for
the stepless control system.
Furthermore, active control of the
suction valve closure velocity
reduces the impact forces of the
valve sealing element, hence
D
r
D
A
P
V
B
C
r
C
D
r
D
A
P
V
B
C
r
C
Energy savings
compared to
recycle valve control
TDC BDC TDC BDC
Figure 4 Working principle of the stepless reverse ow control: left: 100% capacity; right: part load
1500
2500
3000
2000
1000
500
A
n
n
u
a
l

s
a
v
i
n
g
s
,

1
0
0
0
0
100 90 80 70 60 50 40 30 20 10 0
Average compressor load, %
70 /MWh
100 /MWh
120 /MWh
Figure 5 Possible annual power cost savings for a 3.0 MW compressor based on 70 100
and 120/MWh energy costs as function of required compressor capacity to match
process demand
hoerbiger.indd 4 10/03/2014 13:40
stages into a process that typically
runs at 70-90% load of the rated
capacity. When considering pneu-
matic suction valve unloaders and
spillback valve control, a four-
cylinder compressor with two
cylinders per compression stage
would be required to enable at
least 25% load increments to run,
relatively effciently, at loads
between 70% and 75%. However,
between 100% and 76% load, the
dispensable process gas still needs
to be recycled through the spill-
back valve. In this worst case, up
to 24% of the rated compressor
power is wasted. Additionally,
installed fxed clearance pockets
would improve the situation
slightly, but the control logic
would become more complicated
and access to the compressor for
maintenance impeded. However,
two cylinders per stage are still
required as compared with stepless
control.
The same application with a step-
less full-range fow control system
would require only one cylinder
per compression stage (two cylin-
ders in total for the compressor)
of automation allows for precise
plant control and fne-tuning of
process parameters, such as gas
pressure, hydrogen/feed ratio and
gas fow rate.
The following example explains
the impact on the compressor
arrangement based on the chosen
capacity control system. A recipro-
cating compressor shall compress
process gas in two compression
of the compressor, the number of
cylinders per stage can be reduced.
Hence, the capital costs and foot-
print of the compressor are
minimised. Furthermore, fewer
cylinders mean reduced mainte-
nance costs, shorter shutdown
periods for overhauls, and fewer
spare parts to be kept in stock. The
higher effciency reduces electric
power costs, and the high degree
60
100
90
80
70
50
40
30
20
10
I
n
d
i
c
a
t
e
d

p
o
w
e
r
,

%
0
0 10 20 30 40 50 60 70 80 90 100
Capacity, %
Step control
Reverse flow control
Ideal control system
Energy savings
Figure 6 Comparison of the compressor power consumption between step control and
reverse ow control as a function of the compressor capacity
3 Respini, Jones, Spanu, Sesselego, Avoiding foul
play, Hydrocarbon Engineering, Nov 2006.
Matteo Virzi is Senior Technology Manager
with ISAB Priolo renery in Italy. He is an
expert in distillation and thermal conversion
processes. With more than 20 years of
experience in technology, operations and
automation, he holds a degree in chemical
engineering from the University of Palermo.
Email: mvirzi@isab.com
Marco Respini is a Senior Technology Expert
with Baker Hughes Downstream Chemicals,
specialising in renery and petrochemical
process improvements in fouling control.
He has 15 years of rening experience and
is currently involved in developing new
technologies for improving renery conversion
processes. With extensive experience in
asphaltene related problems in oil production
and rening, he is an inventor of ve US patents
and has published 10 technical papers and
seven conference papers on visbreakers and
heavy fuel oil stability problems. A graduate
of Milan University with a degree in industrial
chemistry, he has been a Research Fellow in
the eld of organometallic catalysts and is a
registered professional chemist in Italy. He is
also a member of ACS and NACE.
Email: marco.respini@bakerhughes.com
which provided refnery personnel
with the right information to be
able to constantly keep the unit at
the best process severity for any
processed feed.
VisTec is a mark of Baker Hughes Incorporated.
References
1 Petralito G, Respini M, Achieving optimal
visbreaking severity, PTQ, Q1, 2010.
2 A Phase Separation Kinetic Model for Coke
Formation, Preprints ACS, Div. Pet Chem, 38,
428-433, 1993.
result in better handling of fouling,
and at the same time increase
HVGO directly (less gasoil recycled
as wash oil). Figure 17 shows the
trend of wash oil rate.
The decrease in wash oil rate
was continuously optimised. The
HVGO was analysed on a daily
basis with respect to the level of
contaminants, using the VCI tech-
nique that measures the coke
particles entrained within the
HVGO (thus not removed by wash
oil).
Also, the rate of antifoulant on
the wash oil was optimised and
increased when needed.
Conclusions
This article describes an example of
high conversion visbreaking with
increased run length when
compared to typical visbreaker run
lengths.
This successful result was made
possible by coupling Baker Hughes
VisTec anticoke/antifoulant treat-
ments with monitoring technology,
A controlled decrease
in wash oil can result
in better handling of
fouling, and at the
same time increase
HVGO directly
B
e
s
t
V
a
lv
e
s
s
in
c
e
1
8
6
7
baker hughes.indd 8 13/12/2013 11:38
hoerbiger.indd 5 10/03/2014 13:40
sor and mixed with the fresh
hydrogen stream coming from the
make-up section.
Prior to the plant upgrade, one
compressor was operated between
90% and 95% and the second
machine was kept in stand-by
mode. In the course of the plant
modication, to enable better treat-
ment of highly contaminated
feedstock, hydrogen demand
increased to about 120% of the
rated capacity of one compressor.
Thus, both compressors were
required to run in parallel
are integrated into a hydrogen
desulphurisation unit. Each
compressor has four cylinders and a
motor rating of 5.0 MW. The plant
scheme is provided in Figure 7. The
make-up section of the compressor
includes two stages, with one cylin-
der per stage, and discharges
hydrogen at a pressure of around
80 bar. The control requirement is
to automatically maintain pressure
in the high-pressure separator. The
ow of the recycle gas, coming from
the amine absorber, is recompressed
by the recycle stage of the compres-
since the load steps for the
complete required turndown range
are negligibly small.
New compressors are normally
designed with 10-15% surplus
capacity to provide a margin for
performance losses and degrada-
tion. Therefore, new compressors
are running at part load even at full
production output.
Advanced control system cases:
rening applications
Hydrogen desulphurisation unit
Two make-up/recycle compressors
Fast response time of
control system allows
for precise control of
hydrogen and feed
(H
2
-to-feed ratio) for
optimised reactor
operation and extended
catalyst cycle time.
Possibility of
heater power
savings and
emissions
reduction.
Possibility of additional
compressor power
saving due to stepless
control of recycle stage
and accurate adjustment
of recycle gas stream
Increased product quality
and more process
stability due to advanced
compressor control.
Possibility of compressor power
savings due to stepless control of
make-up stages. Multistage operation
is perfectly balanced due to automatic
interstage pressure control.
Hydrogen
Feedstock
Product
low sulphur diesel
Bypass/spillback
valve
Heat exchanger
Fired
heater
Reactor
HP
separator
Compressor A
make-up
Pump
Bypass/spillback
valve
Compressor B
make-up
Bypass/spillback
valve
Compressor A
recycle
Bypass/spillback
valve
Compressor B
recycle
Recycle gas
Figure 7 Schematic HDS plant layout with integrated stepless ow control on make-up/recycle compressor
hoerbiger.indd 6 10/03/2014 13:40
Multi-service hydrogen compressor
Reverse fow control is capable of
controlling each compression stage
individually. Therefore, multi-
stage compressors or applications
with side streams between
compression stages are perfectly
controlled and pressure balanced.
Furthermore, reverse fow control
fully supports integrated machin-
ery concepts and plant layouts
where one machine handles
multiple processes (multi-service
compressors).
The next case study shows the
application of reverse fow
control to three 14MW hydrogen
compressors, which were part of a
large refnery project in Asia (see
Figure 9). Two compressors run in
parallel and the third one serves as
a stand-by unit. Each compressor
consists of three stages and
compresses hydrogen into three
different process sections of the
refnery.
The frst stage supplies fresh
hydrogen into a desulphurisation
plant for light gasoline products.
The hydrogen fow is mixed with
the feed and has to be accurately
balanced, depending on feed quan-
tity and quality. The second stage
receives the hydrogen recycle gas
fcations to process gas piping were
necessary. The existing control logic
and operator interface in the
distributed control system (DCS),
which were originally based on
recycle valve control, were slightly
modifed by adding split range
function blocks for each stage (see
Figure 8). The programming efforts
for the distributed control system
were low and the entire installation
was carried out within a shutdown
period of two weeks.
ideally, each at 60% part load.
Therefore, the proposed plant
layout included a stepless capacity
control system, enabling effcient
part load operation on both
machines, without the use of the
recycle control valve.
The existing compressors were
upgraded with electro-hydraulic
unloaders and their associated
auxiliaries, such as the control
signal interface unit and hydraulic
unit. Neither civil works nor modi-
MIN
selector
Split
range
Make-up
1
st
stage
Make-up
2
nd
stage
HP
separator
Amine
scrubber
Recycle
stage
Split
range
MIN
selector
PIC
001
PIC
002
8PIC
003
8FT
001
FIC
001
8PT
001
8PT
002
Split
range
8PT
003
H
2
feed
0-100% 50-100% 0-50% 0-100%
0-100%
New signals / DCS logic
Existing signals / DCS logic
Figure 8 Control concept integration based on existing spillback control
Figure 9 Multi-service compressor with stepless reverse ow control on test bench
hoerbiger.indd 7 10/03/2014 13:40
from the amine scrubber of the same unit and recom-
presses it into a treat gas unit together with fresh
hydrogen coming from a side stream. The third stage
forwards the hydrogen mix as make-up gas for the
hydrotreater.
Due to the integrated machinery concept and the
requirement to control each stage individually, the
licensor and the engineering company jointly decided
to implement an electro-hydraulic capacity control
system. But the resulting energy savings are also worth
mentioning: the compressors were operated at 60%
load on average for the frst few years, until the fnal
expansion phase of the plant was fnished. During this
time, the refner saved approximately 3.5 million/y
in power costs, allowing a return on investment within
two weeks. Stepless full-range capacity control systems
are an enabler for cost-effective integrated machinery
arrangements.
Conclusion
With todays growing compressed hydrogen demand
for treating lower quality crude oil, plant fexibility has
become more and more important. In order to control
refning processes more precisely, stepless reverse fow
control on reciprocating compressors allows for
fne-tuning of process parameters, such as gas fow or
pressure. Compared to conventional traditional control
systems, the stepless reverse fow principle enables
energy savings because recycle or spillback operation
is avoided, since just the required gas amount is
compressed and forwarded to the plant. Furthermore,
pressure peaks due to sudden load changes caused by
cylinder unloading can be avoided and the consequen-
tial risk of plant shutdown is reduced to a minimum.
The high degree of automation allows for individual
control of compression stages, thus enabling balanced
compression of multistage machines, as well as the
implementation of complex control concepts (multi-ser-
vice compressors).
With more than 1000 advanced control systems
installed, the system has proved to be a reliable and
sustainable solution for new and existing compressor
applications in the refning, chemical, natural gas and
shale gas industry.
This article is based on a presentation by Alberto Vargas, Hoerbiger
Corporation of America, Inc., at the 5th World Rening Technology &
Shale Processing Summit 2013 (5th WRTS) on 4-5 December 2013 in
Houston.
Klaus Stachel is Product Manager, Monitoring and Controls, with
Hoerbiger Compression Technology in Vienna, Austria.
Email: klaus.stachel@hoerbiger.com
Markus Wenisch is a Senior Product Expert with Hoerbiger Compression
Technology in Vienna, Austria.
Email: markus.wenisch@hoerbiger.com
DAMPF TRIX
The Best for the
transport of steam
ContiTech
Fluid Technology
info@fluid.contitech.de/ih
www.contitech.de
Steam hoses from ContiTech have been designed to meet
the highest requirements on hose material for the safe
transport of steam and hot water in the chemical and
petrochemical industry, in refineries, the building industry,
in shipyards. Three types were designed according to DIN
EN ISO 6134 for specific applications:
q DAMPF TRIX 5000 according to DIN EN ISO 6134-1A
q DAMPF TRIX 6000 according to DIN EN ISO 6134-2A
q DAMPF TRIX 6000 OIL according to DIN EN ISO 6134-2B
Continental Anzeige 270x86 GB DampfTrix+OL 03-2014.indd 1 07.03.14 12:24
hoerbiger.indd 8 12/03/2014 12:01
Our competition would have
you believe we arent a
global EPCM contender.
They would say were small and
domestic competition. We see
ourselves as quicker, more agile,
less risk averse and more
likely to be underestimated.
To learn more visit http://marketing.woodgroupmustang.com/contender2
mustang.indd 1 07/03/2014 13:06
Removing salts and contaminants in
glycol regeneration
F
or the transport of liquefed
natural gas (LNG) in pipelines,
monoethylene glycol (MEG) is
added as corrosion inhibitor and to
reduce the freezing point of the
water in the LNG. The MEG is then
reclaimed.
The rich MEG (high water content
up to 70% and containing dissolved
salts) is run through a MEG recla-
mation unit and the lean MEG
(water content reduced to 10%) is
produced by evaporating the major-
ity of the water. The MEG is
recycled and reused in the system,
leading to an increase of salts
concentration during the evapora-
tion process. Although the majority
of the salts are monovalent salts
such as sodium chloride, there are
also divalent salts, such as
carbonates (calcium, sodium) and
hydroxides (magnesium, iron).
These salts along with pipeline
debris and oxidation products end
up in the heat exchangers and jeop-
ardise their performance.
As a solution to the above, the
reclamation unit normally contains
a pH adjustment step, allowing the
divalent salts to precipitate after
adding sodium hydroxide and
seeding crystals. The salts and
other contaminants can be removed
from the rich MEG prior to enter-
ing the thermal regeneration
section.
This article discusses the use of a
combination system for solid-liquid
fltration as an alternative to static
thickeners, centrifuges, flter
presses and cartridges. In the frst
stage, the low concentrated slurry
is pre-thickened with the use of
candle flters to achieve a solids-
free fltrate and a concentrated
A new process combining candle lters and pressure plate lters enables more
glycol to be safely recovered from natural gas than by conventional methods
BARRY A PERLMUTTER BHS-Sonthofen Inc
CHRISTIAN GASSEN BHS-Sonthofen GmbH
sludge of approximately 4-5%
solids. In the second stage, the
sludge is then washed to produce
an MEG-free cake for disposal and
then dried for non-hazardous
disposal.
This article begins with the
bench-top laboratory tests for pres-
sure, flter media, precoat material
and similar process parameters. The
article concludes with a case history
and installation details including
maintenance and reliability.
Typically, MEG is added to natu-
ral gas for transportation in
pipelines to absorb the water
contained in the gas and thus lower
the freezing point. At the same
time, it serves as a corrosion inhibi-
tor. The glycol is then extracted
from the gas in MEG reclamation
units and regenerated.
Primarily, this involves removal
of the previously absorbed water,
which is loaded with various salts,
depending on the location of the
production well. Beyond this, MEG
contains oxides and hydroxides due
to piping abrasion (as rust particles)
as well as other impurities. The
resulting deposits clog the heat
exchangers, piping and other plant
components and lead to high main-
tenance costs.
Due to increasing environmental
awareness and rising operating
costs, the treatment and recovery of
glycol is gaining importance. The
goal of many operations is for the
dried solids for disposal to contain
no more than 10% glycol.
Particularly for offshore applica-
tions, compact systems are required
that operate automatically as well
as have low maintenance
requirements.
Overview of MEG regeneration
Initially for MEG regeneration, the
divalent salts form a precipitate and
the solids are subsequently extracted
from the glycol. Depending on the
location of the treatment plant, vari-
ous systems are used for this
solid-liquid separation.
For onshore regeneration plants,
for example, static thickeners
(settling tanks) are often used.
They are cost-effcient, but the
settled sludge still contains
large amounts of glycol. Additional
separation systems for example,
such as high-speed separators
and decanter centrifuges, are
subject to considerable wear due to
the abrasive properties of the solids.
In offshore MEG regeneration
plants, maintenance of separators
and centrifuges is a signifcant cost
factor since it is frequent and
requires trained specialists. This has
often led to the use of back-fush or
cartridge flters equipped with
deep-bed fltration elements. These
produce a sludge that contains
large amounts of glycol and must,
therefore, be processed further. The
deep bed flter elements also only
have limited self-cleaning proper-
ties and require time-consuming
disposal. This in turn results in high
operating costs as salt content
increases.

Conventional single-step
precoat ltration
Candle flters, as an alternative to
separators and centrifuges, have
proven to be particularly suitable
for both onshore and offshore use.
They achieve high-quality fltration,
operate fully automatically and are
low maintenance.
q2 bhs.indd 1 10/03/2014 13:43
water or recirculated fltrate. A
typical system operates with three
candle flters, divided into 3 x 50%.
The flters are offset in operation
since each flter is precoated and
flled anew about every eight to 12
hours.
Use of the precoat material,
however, increases the amount of
solids for disposal. The precoating
also leads to increased costs for
handling, storage and disposal
costs. While this type of fltration
leaves a small amount of glycol in
the flter cake, there is a need to
further reduce the residual glycol
in the cake and reduce the amount
of precoat material used.
BHS combination process
In order to further reduce the resid-
ual amount of glycol and
simultaneously offer compact and
low-maintenance plants, BHS has
developed a new combination
fltration and washing-drying
process that does not use precoat-
ing or chemicals. It reduces glycol
loss by displacing the residual
glycol in the flter cake with water.
BHSs lab and pilot plant testing
developed the process and operat-
ing parameters.
BHS had installed a combination
process with concentrating candle
flters followed by pressure plate
flters for cake washing and drying
A flter cake forms in the flters
and subsequently pre-dries. It is
then automatically discharged into
containers. Since the MEG suspen-
sion only contains a small amount
of very fne solids, a precoat mate-
rial, usually perlite, is applied. It
ensures a high fltration rate
regardless of water quality and
operating state. This precoat layer
prevents the flter medium from
plugging with fne solids (hydrox-
ides, dust or abrasive particles) and
allows the medium to be less
impacted by the hydrocarbon
constituents in the MEG.
The use of candle flters results in
recovering signifcantly more glycol
than by conventional separation and
has proven reliable under an
extremely wide range of operating
conditions. The BHS candle flter
systems with precoat has been
installed for these applications. A
typical installation is shown in
Figure 1.
While there are benefts to the
candle flters with precoat, there are
also increased operating costs. The
precoat material is delivered in
large bags and must be suspended
into a precoat slurry tank with
Figure 2 MEG application showing BHS combination of concentrating candle lters and pressure plate lters
Figure 1 Conventional single-step precoat ltration
q2 bhs.indd 2 10/03/2014 13:43
For more information,
visit us at gewater.com or
scan the code above.
Reducing costs and waste, while increasing process uptime is not
only stressful, its critical to operating a protable renery today. GE
can help boost your plants efciency by identifying prot-robbing
problems, applying the most cost-effective treatments, as well as
monitoring and maintaining the results. Keep your renery operating
safely and reliably, while lowering maintenance and energy costs
with GEs integrated solutions; encompassing water and process
chemicals, equipment, and services.
GEs offerings include:
Comprehensive water and process
chemical solutions
Real-time remote monitoring
and diagnostics
Expert technical assistance for
integrated solutions throughout your renery
Let us help you manage the stress and
uncertainty of todays rening challenges . . .
and deliver operational calm.
Delivering operational calm
ge.indd 1 07/03/2014 13:19
for a similar application for the
treatment of specialty amines. In
this process, the solids in the amine
suspension are thickened (concen-
trated) using candle lters without
precoating and then further
processed using a plate lter. Figure
3 shows the amine process while
Figure 2 shows the complete BHS
skid. Figure 4 shows the process
ow for MEG.
Since the suspension contains
signicantly less than 1% solids,
BHS applies the combination
process. In the rst stage, the
suspension is ltered in BHS candle
lters without lter aids or chemi-
cals, and pre-thickened. In this
approach, similar to the process in
settling tanks, a solids concentration
is achieved that is capable of being
further processed. Depending on
the solids content, the batch time
for concentrating is between two
and eight hours.
As soon as a lter cake of a few
millimetres has built up through
ltration, the suspension feed is
stopped and the cake is displaced
from the candles. The cake settles
for approximately 30 minutes and
collects in the cone of the lter
vessel. The concentrated slurry that
is discharged has a solids content
of around 36% from an initial
concentration of less than 1%.
This material is pumped into an
agitated vessel. The pre-thickened
suspension is then transferred to a
BHS pressure plate lter. In the
pressure plate lter, the circular
plates are arranged in a vertical
plate stack; this results in a stable
cake for washing and drying of the
glycol without the risk of the cake
sliding off.
A residual glycol content of less
than 5% is achieved with a low wash
ratio. This residual content is lower
by a factor of 30 compared with the
conventional ltration precoating
approach. Water consumption can
be minimised by using closed-cycle
washing or counter-current washing.
After washing, the lter cake is
dried by blowing with nitrogen.
For cake discharge, the plates are
vibrated with 1 HP motors for
several minutes. The nal product
is then crumbly and of low volume.
It can be disposed of as a non-
hazardous cake as there is only a
small trace of glycol and precipi-
tated salts.
Discussion and cost benets of the
combination process
The combination process provides
operators with additional benets.
These include a decrease in operat-
ing and maintenance costs as well
as improved glycol recovery. A cost
analysis is described below.
For an average glycol regeneration
process with an hourly capacity of
40 m, producing suspensions with
an average salt content of about 130
ppm, the conventional single-step
precoat ltration method requires a
lter area of approximately 200 m
(three lters each with 65 m lter
area). With a normal cycle time of
about eight to 12 hours and a
precoat thickness of about 6-8 mm,
the annual average consumption of
perlite is almost 100 tonnes.
By comparison, the BHS combina-
tion process requires only three
candle lters each with a lter area
of 19 m as well as only two
Figure 3 Amine application showing BHS combination of concentrating candle lters and pressure plate lters
Slurry
Filtrate
Solids
(washed and dried)
Candle filter
Agitated
feed tank
Wash water
tank
Pressure
plate filter
Concentrated
slurry discharge
Pre-thickening
H
2
O
Filtrate
Figure 4 MEG process ow for BHS combination system
q2 bhs.indd 3 12/03/2014 12:02
and you will learn everything you need to know to
make the right decision for recovering and refning PGMs
from your spent process catalystsquickly, safely,
conveniently, and proftably. Visit our new website now,
and let us tell you what we can do for you.
Platinum Palladium Rhodium Ruthenium Rhenium Gold Silver
Processing facilities and technical service offices around the world
Sabin Metal OneClick.indd 1 2/15/14 11:00 AM sabin one click.indd 1 04/03/2014 11:13
the savings from the reduced nitro-
gen consumption for drying. The
power requirements for both
designs remain at a comparable
level. Overall, the ongoing operat-
ing costs for the new combination
system are only about 20% of those
for the conventional single-step
precoat fltration (see Figures 6 and
7). Finally, the new system saves on
consumables, requires no chemicals
and no flter aid while the loss of
MEG is reduced by a factor of 30.
Summary and take-aways
The single-stage precoat fltration
using BHS candle flters allows for
signifcantly more glycol to be safely
recovered from natural gas than by
using conventional methods of sepa-
ration and centrifugation. The new
combination process with concen-
trating candle flters followed by
pressure plate flters for cake wash-
ing and drying further increases the
glycol recovery effciency. The
combination approach saves on
investment costs and operating costs
with the elimination of precoating
and chemicals as well as a very
compact design, saving space for
offshore platforms.
The take-away is that process
engineers must be creative with
their approaches to problem solving
and evaluate all fltration possibili-
ties. Laboratory and pilot testing is
an effcient technique for determin-
ing the optimum fltration solution.
Barry A Perlmutter is President and Managing
Director of BHS-Sonthofen Inc., a subsidiary of
BHS-Sonthofen GmbH. He has published and
lectured extensively worldwide on theory and
applications for the chemical, pharmaceutical
and energy/environmental industries. He
holds a BS degree in chemistry from Albany
State (NY) University, a MS degree from
the School of Engineering at Washington
University, St. Louis, and a MBA from the
University of Illinois.
Email: barry.perlmutter@bhs-ltration.com
Christian Gassen is a Project Manager for
Filtration Application Engineering with BHS-
Sonthofen GmbH. He has experience with
chemical and biological applications at nuclear
power plants, pharmaceutical cleanroom
installations and municipal waste treatment
plants, and holds an advanced engineering
degree in chemical and biological engineering
from the University of Erlangen/Nuremberg,
Bavaria.
Email: christian.gassen@bhs-sonthofen.de
design requires only about 70% of
the investment costs compared with
the conventional design.
As for the additional water
consumption costs for the cake
washing, this is balanced out with
pressure plate flters each with a
flter area of 1 m. The total area for
the BHS combination process is less
than 60 m compared with 200 m
for the conventional approach. As
Figure 5 shows, the combination
30%
60
100
90
80
70
50
40
30
20
10
I
n
v
e
s
t
m
e
n
t

c
o
s
t
s
,

%
0
precoat filtration
Figure 5 Summary of investment costs
60
100
90
80
70
50
40
30
20
10
C
o
s
t
,

%
0
Conventional single-step precoat filtration
Perlite MEG Nitrogen Power Water Total
Figure 6 Operating costs
80%
60
100
90
80
70
50
40
30
20
10
O
p
e
r
a
t
i
o
n
a
l

c
o
s
t
s
,

%
0
precoat filtration
Figure 7 Summary of operating costs
q2 bhs.indd 4 10/03/2014 13:43
JOHNSON SCREENS
SHAPED SUPPORT GRID (SSG)

designed to be installed into the bottom head of hydroprocessing
or gas dehydration vessels, allowing better liquid and gas fow,
bed utilization, distribution and an overall more effcient process
than traditional fat surface grid assemblies. Patented design.
JOHNSON SCREENS
INLET DIFFUSER BASKET

designed to control velocities of gas or liquid distribution over
media, providing improved performance over traditional plate disc
type distributor designs as well as even distribution and minimal
scouring at the top of the bed. Patented design.
INNOVATIVE SOLUTIONS FOR THE
HYDROCARBON PROCESSING INDUSTRY
Bilfngers engineering department constantly works to offer the best and innovative solutions to the Hydrocarbon Processing Industry:
BILFINGER WATER TECHNOLOGIES

www.water.bilfnger.com
Australia - Asia Pacifc
Phone +61 7 3867 5555
Fax +61 7 3265 2768
asiapacifc.water@bilfnger.com
France
Phone +33 5 4902 1600
Fax +33 5 49021616
france.water@bilfnger.com
North & South America
Phone +1 651 636 3900
Fax +1 651 638 3171
usa.water@bilfnger.com
PTQ_Issue Q2.indd 1 03/03/2014 17:50:54
biflinger.indd 1 04/03/2014 10:57
Sulzer Chemtech
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 Chemtech, USA, Inc.
8505 E. North Belt Drive | Humble, TX 77396
Phone: (281) 604-4100 | Fax: (281) 540-2777
TowerTech.CTUS@sulzer.com
www.sulzer.com
Background
For many decades, people have noted the destruction of
their distillation trays without the typical deformation of the
trays due to a column upset. Numerous instances have been
published (and also not published) on this topic since 1973.
The cause of this destruction is harmonic vibrations, where
the trays natural frequency equals the bubble generation
rate. Avoidance of this phenomenon can be easily achieved
with the proper tray design and operation, but you have to
know what to look for.

Design Considerations
Just to be clear, most tray designs do not experience this
phenomenon. Under normal operating conditions, all trays
are sufciently loaded with vapor to keep a tray operating
with a Net Upward Force. However, in cases where a
tower is operating at turndown conditions, the trays
are potentially subject to harmonic vibrations and their
destructive effects (see Figure 1). One of the characteristic
telltale signs of harmonic vibration destruction is the
presence of numerous cracks in the tray decks without any
bends in the tray material itself, as shown in Figure 2.
Trays are always designed to have sufcient strength to hold a
person (or several people) standing on the tray deck and also
designed to avoid dislodgement due to random operational
forces. With these forces in mind, larger diameter trays are
typically designed with tray panel lengths between 66 to
120 (1.7 3.0m). These are also the ideal lengths that result
in natural frequencies of 30 Hz or less, depending on the
depth of the integral truss. Bubble generation rates are also
typically less than 30 Hz.
To avoid harmonic vibration destruction of your trays, two
things can be done. The rst is to keep the tray sufciently
loaded with vapor such that the tray never operates for
any length of time without a net upwards force. Of course,
plants do operate at turndown for extended lengths of time
because it costs energy to keep a tower articially loaded.
The second way to avoid vibration is to manipulate the
natural frequency of the tray panels. This can be done by
changing the panel direction, deepening the integral truss,
or changing the panel length. Each of these changes add a
little extra cost to the tray design, but the benet (of avoiding
harmonic vibration damage) will easily outweigh this cost.
Important Tips
As mentioned earlier, not all towers experience this
problem. Those towers that:
Are 36 (0.9m) in diameter or less
Have vapor densities greater than 0.1 lb/ft3 (1.6 kg/m3)
Have 3 or more tray passes
have never been seen to experience this phenomenon.
For more detailed information, please reference: Summers, D.R.,
Bad Vibrations, Hydrocarbon Engineering, March 2005, pp 51-54
Tower Technical Bulletin
Avoid Harmonic Vibration Tray Destruction
The Sulzer Renery Applications Group
Sulzer Chemtech 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.
Figure 2: Complete Tray Failure From Harmonic Vibrations
Figure 2: Complete Tray Failure From Harmonic Vibrations
sulzer.indd 1 10/03/2014 11:15
Accurate prediction of tower relief
R
efneries and upgraders are
often required to increase
throughput beyond their
initial nameplate capacity. Existing
bottlenecks must be identifed and
mitigated before the unit can
process more feed. With increased
throughput, relief loads are recalcu-
lated to ensure that the
corresponding process units are
protected. When conventional
calculation methods are used, cases
arise where the newly calculated
relief loads exceed the existing
capacity of pressure safety valves
(PSV) on columns, or the capacity
of the existing fare system. For a
debottlenecking project, adding
PSVs or modifying an existing fare
system can be costly and impracti-
cal due to constraints on available
downtime. Conventional methods
for calculating relief loads are
generally conservative
1
and lead to
over-sizing of the relief and fare
system. Dynamic simulation
provides an alternative method to
better defne the relief system and
improve understanding of what
happens during relief.
2
Dynamic simulation is rigorous
and can model many processes,
contributing to a reduction in relief
loads that the steady state-based
unbalanced heat method cannot.
The dynamic model simulates the
fuid inventory hold-up and
predicts the time before relief pres-
sure is reached. This time before
relief can be used to take credit for
operator intervention to prevent/
mitigate the relief load. Unlike the
steady state based models, which
assume an unlimited amount of
light components being relieved,
the dynamic model accounts for
A comparison of conventional and dynamic simulation methods applied to a
project to debottleneck a deisobutaniser
HARRY Z HA, ABDULLA HARJI and JONATHAN WEBBER
Fluor Canada Ltd
the depletion of light components
and estimates changes in heat of
vaporisation and temperature with
time, which can lead to a partial
loss of reboiler duty due to
decreased log mean temperature
difference (LMTD) or depletion of
liquid inventory in the column
sump.
Dynamic simulation can take
credit for the column overhead
condenser duty before the accumu-
lator foods, reducing the relief load
through condensation. Typically,
credits are not given to the control
valve responses on over-pressure
protection in either conventional or
dynamic calculations. In cases to
consider favourable responses of
control valves to reduce relief
loads, dynamic simulation has the
advantage of simulating the
response over the time of incident,
to ensure that any credits taken are
appropriate within that time frame.
Dynamic models also simulate
72
64
33
1
FC
FC
FC
Flare
LC
PC
PI
HH
TC
Feed-2
from OSBL
Feed-1
from
alkylation
reactors
Non-condensable
to flare
Condensers
Accumulator
Re-boilers
Reflux/
Product
pump
De-iC4
FC
Isobutane
product
LC
FC
MP steam
Condensate
De-iC4
bottoms to
debutaniser
PX
Figure 1 System diagram of a deisobutaniser column
fluor.indd 1 10/03/2014 13:45
is controlled by the overhead liquid
product fow.
A dynamic simulation in Hysys
was set up based on the confgura-
tion shown in Figure 1 and on the
latest information on existing equip-
ment, piping, and instrumentation
(datasheets, P&IDs, mechanical
drawings, and isometric drawings
of piping). The dimensions of equip-
ment and piping were input into the
simulation with the exact elevations.
Control valves were simulated with
reported volume fows and open-
ings from the datasheet. PID
controllers were used to compute
the controller actions. Relief valves
are installed on the overhead line
and a spreadsheet function was
used to calculate the required relief
area as a function of time. The PSV
sizing was based on the API
Standard.
2
In particular, baffed
thermosiphon reboilers were
installed with an imaginary column
fash zone and a sump to achieve
the designed boil-up ratio and circu-
lation (see Figure 2). The details of
this confguration were important in
calculating the detailed hydraulics
between the column and the ther-
mosiphon reboilers. This allowed
simulation of the circulation and
boil-up changes resulting from the
loss of static head during column
relieving. Condensate levels in the
reboilers were not simulated and the
reboilers were assumed not to be
fooded, allowing for a conservative
reboiler duty for relief load
estimation.
The dynamic model was then
tuned to match the performance of
steady state simulation for opera-
tional conditions (P/T), feed and
product rate and quality, and the
duties of reboilers and condensers.
In the evaluation of the load from
an individual relief device, no
favourable response was allowed
by any control instrument that
would reduce the relief load
according to API standard guide-
lines. If the normal controller
response could act to reduce the
relief load, then the control valve is
assumed to remain in its last posi-
tion before the upset.
Based on relief analysis using
conventional methods, two relieving
scenarios were identifed as the
interactions among the processing
units and provide a more realistic
overall picture of the response in
various relief scenarios. Dynamic
simulation can also help identify
potential design modifcations to
reduce the relief load potential
which conventional methods
cannot address.
In this study, relief loads for an
existing deisobutaniser column
have been calculated using both the
conventional unbalanced heat
method and dynamic simulations
under various relief scenarios. As a
result of debottlenecking, the feed
to the column was increased by
18%. Calculated relief loads from
both methods are compared to each
other and against the existing PSV
capacity. The advantages of
dynamic simulations are
pronounced when an existing
column relief system is being
rechecked for debottlenecking.
Dynamic simulation model set-up
The deisobutaniser is one of several
columns in the refnery alkylation
unit that remove the light ends
from alkylates, to enhance the
octane number of gasoline. As
Figure 1 shows, the deisobutaniser
has 72 trays; two feed streams enter
the column at trays 64 and 33. The
column is equipped with two
baffed, once-through thermosiphon
reboilers at the bottom and
air-cooled condensers on the over-
head. The heating medium for the
reboilers is medium pressure steam
(150 psig).
The bottom tray temperature is
cascaded to the fow controller on
the condensate from the reboilers.
The column sump level is
controlled by the alkylate product
fow from the column bottom. The
column pressure rides on a push-
pull pressure controller on the
accumulator. The accumulator level
72
64
33
1
FC
FC
FC
Flare
LC
PC
PI
HH
TC
Feed-2
from OSBL
Feed-1
from
alkylation
reactors
Non-condensable
to flare
Condensers
Accumulator
Re-boilers
Flash zone Sump
Reflux/
Product
pump
De-iC4
FC
Isobutane
product
LC
FC
MP steam
Condensate
De-iC4
bottoms to
debutaniser
PX
Figure 2 Dynamic simulation conguration of the deisobutaniser column
fluor.indd 2 10/03/2014 13:45
INNOVATION
PROFLUX
TM
severe service
grid packing
YOU CAN RELY ON US.
Engineered to endure harsh conditions.

Tests show that PROFLUX
severe service
grid packing has:
Superior resistance to compression
compared to conventional grid
packings
Higher fouling resistance compared
to conventional grid and structured
packings
Better de-entrainment performance
compared to conventional grid
packings with similar surface area
United States (316) 828-5110 | Canada (905) 852-3381 | Europe +39 035 2273411 | Asia +65-6831-6500
PROFLUX and K KOCH-GLITSCH are trademarks of Koch-Glitsch, LP and are registered in the U.S.A. and various other countries.
YOU CAN RELY ON US is a trademark of Koch-Glitsch, LP. PROFLUX
technology is protected by patent in the U.S.A.; other patents pending.

www.koch-glitsch.com
Lower level of
compression
equals improved
mechanical rigidity
koch.indd 1 11/03/2014 19:12
FROM FEEDSTOCK
TO END PRODUCT
You wont nd a more comprehensive
offering of separations and phase
contacting process internals than
AMACS. Whether its demothballing,
debottlenecking or grassroots,
AMACS can provide the components,
hardware, know how and field
support to streamline your process.
From your process inlet to its outlet,
AMACS can provide the engineered
solutions to meet your desired
performance requirements or build to
meet your specications. In short, if
it is separations or phase contacting
related contact AMACS!
AMACS 24-7 Emergency Service Available at (281) 716 - 1179
2013 Amacs Process Towers Internals. All Rights Reserved.
www.amacs.com
MESH MIST
ELIMINATORS
INLET
DIFFUSERS
TRAYS RANDOM
PACKING
STRUCTURED
PACKING
LIQUID
DISTRIBUTORS
amacs.indd 1 11/03/2014 09:17
potential governing relief cases for
the deisobutaniser: total power fail-
ure (TPF) and refux failure (RF).
The assumptions for each scenario
are summarised as follows:
Total power failure (site-wide power
failure)
Feed from the alkylation reactors
stops
Feed from OSBL (mixed isobu-
tanes) stops
Overhead product/refux pump
stops
Column bottom product contin-
ues on level and pressure as LC
fails in position
Medium pressure steam to
reboiler, fow controls fail in posi-
tion (no credit taken)
25% of the normal duty is
assumed available for air cooling
condensers based on natural draft;
however, the duty of overhead
condensers is considered to be zero
once the accumulator is fooded
PSV back pressure is assumed
40% of the PSV set pressure as per
preliminary hydraulic calculation of
the fare header
PSV set pressure is 140 psig and
110% overpressure is used for PSV
sizing
Liquid level in the accumulator is
at 70 vol% before the incident
(normal operation).
Reux failure (refux pumps fail)
Feed from alkylation reactors
continues as normal
Feed from OSBL (mixed isobu-
tanes) continues as normal
Overhead product/refux pump
stops
Column bottom product contin-
ues on level and pressure as LC
fails in position
Medium pressure steam to
reboiler, fow controls fail in posi-
tion (no credit taken)
Air cooler fans are assumed avail-
able until the accumulator is fooded
(then the condenser will be lost)
PSV back pressure is assumed
40% of the PSV set pressure
PSV set pressure is 140 psig and
110% overpressure is used for PSV
sizing
Liquid level in the accumulator is
at 70 vol% before the incident
(normal operation).
The scenarios and actions were
set up in a scheduler in Hysys
dynamic simulation. Relief cases
were studied by running the sched-
uler over one hours running time
(until the relief load was reduced to
an approximately constant fow
rate). The results were recorded in
small steps (0.25 second) and
summarised in plots (discussed in
the following section).
Column relief loads
For comparison, the relief loads of
the deisobutaniser were also calcu-
lated based on conventional
methods for the selected cases.
Conventional methods for relief
load calculation are largely based
on steady state simulations. In
these methods, an accumulation
term is introduced to determine the
relief load based on heat and mate-
rial balance at a particular instant
during the relieving process. In this
study, the unbalanced heat (UBH)
method is referenced as the conven-
tional method.
3
In this widely used
method, a heat and material
balance is produced at relief condi-
tions to determine whether there is
excess (unbalanced) heat. The accu-
mulation term (mass fow) is then
determined by dividing the excess
heat by the latent heat of vaporisa-
tion of the accumulated stream at
relief conditions. The refux or the
top tray liquids at steady state are
usually taken as the accumulated
stream to obtain the physical
properties for relief load calcula-
tions. This approach simplifes
calculation of the relief load and is
designed to give conservative
results. The relief loads calculated
by the UBH method and dynamic
simulations are discussed below for
the total power failure and refux
failure cases.
Total power failure case
Figure 3 shows the results of using
the conventional method and
dynamic simulation as a function of
time. The conventional method is
not time-dependent and the relief
load appears as a constant relief
load over time (a horizontal line in
Figure 3), while the dynamic simula-
tion develops the transient
behaviour of the system. In conven-
tional calculations, the latent heat of
refux at dew point and relieving
pressure is employed and an endless
supply of vaporising components is
assumed. As expected, a conserva-
tive relief load was obtained by the
conventional UBH method with the
calculated load as high as 950 000
lb/h, which was signifcantly higher
than the peak relief load predicted
by dynamic simulations. According
to the dynamic simulations, relief
does not occur until 6.5 minutes
after the TPF event occurred after
the column overhead pressure
reached the PSV set pressure. The
peak relief load is reached approxi-
mately fve minutes after the relief,
beyond which the relief load
decreases to below 100 000 lb/h
within 21 minutes. The calculated
relief area corresponding to the peak
relief is 30.6 in
2
which requires a T
600
1000
900
800
700
500
400
300
200
100
R
e
l
i
e
f

l
o
a
d
,

1
0
0
0

l
b
/
h
0
30
40
35
25
20
15
10
5
P
S
V

s
i
z
e

r
e
q
u
i
r
e
d
,

i
n
2
0
0 5 10 15 20 25 30
Time, minutes
UBH load
PSV Size-Dyn
Relief load-Dyn
Figure 3 Total power failure relief loads by the unbalanced heat method and dynamic
simulation at 70% initial volume
fluor.indd 3 12/03/2014 12:03
product pumps failure. The relief
load calculated by the UBH method
remains high (757 000 lb/h), while
the dynamic simulation predicted a
peak relief load of 562 000 lb/h 7.3
minutes after the pumps failed. The
required relief area for the relief
loads calculated by the UBH
method and dynamic simulation
are 40 in
2
and 30.1 in
2
respectively,
both exceeding the available area of
the existing PSV (26 in
2
). Compared
to the dynamic simulation of the
TPF case, the refux pump failure
case has a similar peak relief load
but occurs three minutes later than
the TPF case. This is due to the
continuous feeds to the column and
the liquid on the trays being heated
up more slowly than in the TPF
case. (The feed temperatures are
lower than the temperature of the
liquid inside the column.)
However, the relief load decays
much more slowly than in the TPF
case after the peak; the relief load
was still signifcant at 200 000 lb/h
even after 60 minutes. The fatter
relief load profle is attributed to
more light components available
from continuous feeds under the
refux pump failure case.
Similar mitigations to the TPF case
were also considered for the refux
pump failure case (see Figure 6). As
a result of reducing the initial liquid
level in the overhead accumulator
(from 70% to 55% by volume), the
column relief condition was delayed
by 2.3 minutes and the relief load
was reduced to 462 000 lb/h which
required a relief area of 24.8 in
2
. The
existing PSV (26 in
2
) will be able to
take the reduced peak relief loads.
Considering the challenges and risk
in modifying the existing relief and
fare systems, the mitigation
approach (lowering the initial liquid
level in the overhead from 70%
volume to 55%) is recommended to
minimise the impact on the debot-
tlenecking project.
Summary
Conventional methods for column
relief load calculation are typically
more conservative than in dynamic
simulation due to the conservative
assumptions inherent in the calcu-
lations. Conventional methods do
not account for the compositional
pump failure scenario can be simu-
lated and credit taken for the time
it takes the overhead drum to food.
This time is related to the liquid
level in the drum at the beginning
of the relief scenario. The original
normal liquid in the accumulator
was 70% by volume. An initial
liquid level of 55 vol% was studied
in dynamic simulation and the
result is shown in Figure 4. With
the lowered liquid level in the
drum at the beginning of relief, the
time to food the accumulator is
postponed by two minutes
compared to the 70% initial level
case. As a result, the peak relief
load was reduced to about 430 000
lb/h. The corresponding required
relief area is 22.6 in
2
, and the exist-
ing T sized PSV is now suffcient.
Reux pump failure case
Figure 5 shows the results of relief
analysis for the case of refux/
and a P sized API PSV, while the
relief load calculated by the UBH
method requires two T sized PSVs.
However, the existing deisobutan-
iser only has one 8T10 API PSV (26
in
2
) on its overhead line.
Considering the challenges and
substantial costs to add an addi-
tional PSV and modifying fare
system, mitigations were investi-
gated by lowering the initial liquid
level at the overhead accumulator to
minimise the impact on the project.
The initial liquid level in the
overhead accumulator was a key
parameter for column relief calcula-
tion. In conventional methods, it is
common practice to assume the
overhead drum and condenser will
be fooded as a result of losing the
refux and product pumps (unless
there is signifcant hold-up volume
allowing adequate operator inter-
vention time in the overhead
drum). In dynamic simulation, the
300
450
400
350
250
200
150
100
50
R
e
l
i
e
f

l
o
a
d
,

1
0
0
0

l
b
/
h
0
30
40
35
25
20
15
10
5
P
S
V

s
i
z
e

r
e
q
u
i
r
e
d
,

i
n
2
0
0 5 10 15 20 25 30
Time, minutes
PSV Size
Relief load
Figure 4 Total power failure relief loads by dynamic simulation at 55% initial volume
600
800
700
500
400
300
200
100
R
e
l
i
e
f

l
o
a
d
,

1
0
0
0
l
b
/
h
0
0 5 10 15 20 25 30
Time, minutes
30
35
25
20
15
10
5
P
S
V

s
i
z
e

r
e
q
u
i
r
e
d
,

i
n
2
0
UBH load
PSV Size-Dyn
Relief load-Dyn
Figure 5 Reux failure relief loads by unbalanced heat method and dynamic simulation
at 70% initial volume
fluor.indd 4 12/03/2014 12:03
www.eptq.com
dynamic simulation exceeded the
available capacity of the existing
PSV. Mitigation approaches were
studied by lowering the initial liquid
level in the overhead drum to 55%
volume. As a result, the relief loads
were further reduced by 26% and
18% for the TPF and refux pump
failure cases, respectively. With the
recommended mitigation approach,
the new predicted relief loads are
within the capacity of the existing
PSV. Therefore, the risk and cost of
modifying the existing relief and
fare systems are minimised and
potentially avoided.
References
1 Chittibabu H, Valli A, Khanna V, Calculating
Column Relief Loads, PTQ, 55-65, Q2 2010.
2 API RP 520: Recommended Practice for the
Design and Installation of Pressure Relieving
Systems in Reneries, Part I (Sizing and
Selection, 2008) and Part II (Installation, 2003),
American Petroleum Institute, Washington D.C.
3 Sengupta M, Staats F Y, A new approach to
relief valve load calculations, 43rd Proceedings
of Rening Section of American Petroleum
Institute, Toronto, Canada, 1978.
Harry Z Ha is a Senior Process Engineer/
Specialist with Fluor Canada Ltd, Calgary,
Alberta, Canada. He holds a masters degree
in environmental engineering from Hong
Kong University of Science and Technology
and a PhD in chemical engineering from the
University of Alberta.
Email: Harry.Ha@Fluor.com
Abdulla N Harji is an Executive Director of
Process Technology, at Fluor Canada. He holds
a BSc degree in chemical engineering from
Loughborough University, UK.
Jonathan Webber is a Process Engineer with
Fluor Canada. He holds a PhD in process control
from Dalhousie University and a masters in
biotechnology from McGill University.
changes on the trays and the
system volumes contribution to the
transient availability of cooling and
heating. They also ignore the time,
temperature and pressure depend-
ency of heat input or heat removal
from the system. In the cases of
plant revamp or debottlenecking,
conservative relief loads from
conventional methods will likely
call for the addition of PSVs and/or
the modifcation of the existing
fare system, which adds substan-
tial cost and risk to the project.
Dynamic simulation simulates the
actual composition and inventory
changes inside the column and
accumulator and provides more
accurate predictions on relief loads.
Typically, the relief loads predicted
by dynamic simulation are less
than those calculated by conven-
tional methods. Depending on the
system, the reduction in calculated
relief loads can be signifcant (by
>50%). For plant revamp or debot-
tlenecking projects, this reduction
in relief loads can result in signif-
cant savings for the project.
This case study illustrated such an
example. Employing dynamic simu-
lation for an existing deisobutaniser
has reduced the calculated relief
loads by more than 35% for both
TPF and refux pump failure cases,
compared to the conventional meth-
ods. This study also demonstrated
that the initial liquid level in the
overhead accumulator is a key
parameter in relief load calculation.
When the initial liquid level is set at
70% volume (as it is in current oper-
ation), the relief loads predicted by
300
500
450
400
350
250
200
150
100
50
R
e
l
i
e
f

l
o
a
d
,

1
0
0
0

l
b
/
h
0
30
35
25
20
15
10
5
P
S
V

s
i
z
e

r
e
q
u
i
r
e
d
,

i
n
2
0
0 5 10 15 20 25 30
Time, minutes
PSV Size
Relief load
Figure 6 Reux failure relief loads of dynamic simulation at 55% initial volume
Our new clean-fuels
plant is straining the
auxiliary units!
How can we boost their capacity
without major construction?
Read more on this topic at
www.amacs.com
TO MEET CLEAN FUELS
requirements for gasoline, a medium-
sized Midwestern renery added a low-
sulfur fuels technology plant. Gasoline
throughput was unchanged. However,
the sharp increase in sulfur removal
required more hydrogen from the
hydrogen unit and sent more sulfur
gases to the amine treaters and
downstream sulfur units. These
auxiliary units became bottlenecks,
overdriven at the cost of product
purity and amine consumption.
On studying the hydrogen, amine,
and sulfur units, AMACS found many
opportunities for improving separation
efciency and capacity. The problems
were solved without major construction
by applying modern technology to mist
eliminators, liquid-liquid separators,
and tower trays and packing. Results
included haze-free product and
reduced amine consumption. Now a
diesel clean-fuels plant is being added.
Phone: 713-434-0934 Fax: 713-433-6201
amacs@amacs.com
24-hr Emergency Service:
1-800-231-0077
Visit our new website at
www.amacs.com
fluor.indd 5 12/03/2014 12:04
heurty.indd 1 07/03/2014 13:01
Myth of high cutpoint in dry
vacuum units
R
efnery vacuum units maxim-
ise higher value vacuum gas
oil (VGO) and minimise
vacuum residue (VR). Dry unit
designs (see Figure 1) are being used
to process a variety of crude blends
from very light crudes to heavy
Canadian blends. Designers claim
these dry units can achieve VGO
true boiling point (TBP) cutpoints of
1050-1105F (565-595C). Dry units
have lower operating costs and
marginally lower capital investment
versus those using steam in the
heater coils and/or steam stripping
the VR; therefore, they are often
mistaken for the most proftable
design. Dry unit operation may be
justifed when processing light
crude blends or when the crude
contains less VGO boiling range
material making lower TBP
cutpoints acceptable. When process-
ing medium or heavier crudes or
thermally unstable crudes, dry
vacuum units produce much lower
VGO yield and higher VR produc-
tion. VR is often coker feed where
10-20% of the VGO boiling range
material is turned into coke; this
reduces refnery liquid volume
yield.
Vacuum unit performance
Over the last 18 years, rigorous feld
testing of more than 100 CDU/
VDUs processing crudes ranging
from West Texas Intermediate
(WTI), Urals, Arab Light, Canadian
bitumen, Venezuelan extra-heavy to
everything in between shows high
TBP cutpoint claims are rarely
achieved. Nearly every unit tested
had an actual VGO TBP cutpoint
40-100F lower than design or had
run lengths as short as six to 24
Accurate measures of true boiling point cutpoints are essential to correctly
evaluate vacuum unit performance
SCOTT GOLDEN, TONY BARLETTA and STEVE WHITE
Process Consulting Services
months. High VGO cutpoint claims
continue, but actual performance
has shown that rarely are these
claims valid across a four to six year
run. It is possible to run a high
heater outlet temperature and oper-
ate with little or no overfash for a
short period of time. However this
combination usually results in a
shortened vacuum heater run
length or coked wash bed. There
are many documented cases where
actual VGO yields were 3-6 vol%
lower on whole crude than claimed,
and heater and column run lengths
less than two years are common. If
crude is worth $100/barrel, destroy-
ing VGO in a coker and taking
unscheduled shutdowns to pig
heaters or replace the wash section
reduces refnery revenue.
Low cost is a simple measure, but
Reduced
crude
Wash zone
Flash zone
Vacuum
residue
Fired heater
Figure 1 Dry vacuum unit
Reduced
crude
Coil
steam
Wash zone
Flash zone
Stripping
trays
Vacuum
residue
Steam
Fired heater
Figure 2 Vacuum unit with steam
q2 pcs.indd 1 10/03/2014 13:51
process many crudes, therefore
crude blend percentages are often
estimates, and crude assays are not
useful in measuring operating
performance.
Crudes contain signifcantly
different amounts of VGO boiling
range material; the slopes of the
distillation curves in the 975-1100F
(524-593C) range are very differ-
ent. For example, Middle East Arab
Light contains much less VGO than
many of the heavy crudes, there-
fore incentives to cut deep vary by
crude type. Designing for high TBP
cutpoint on Arab Light yields less
incremental VGO than other
crudes. Table 1 shows the VGO
volumes for Arab Light, Arab
Heavy, Merey, Maya, and Canadian
bitumen blends between 975-1100F
(524-593C).
Table 1 shows that VGO TBP
cutpoint is a poor measure of incre-
mental VGO recovery because VGO
content is a function of crude type.
The steeper the distillation curve,
the less the incremental VGO
content for a given cutpoint change.
Higher slope TBP curves require
incrementally higher heater outlet
temperature, lower column operat-
ing pressure or more steam to
recover incremental VGO product.
Crudes with less VGO have less
incentive to cut deeper than crudes
containing more recoverable VGO
and each incremental 1 vol% yield
increases costs.
Many refners and some technol-
ogy providers confuse TBP cutpoint
with other VGO yield metrics. VGO
whole crude TBP cutpoint is not
equal to the HVGO product
95 vol% temperature or an average
of HVGO and VR product 95%/5%
temperatures. It also is not the
atmospheric equivalent temperature
(AET) calculated by the Maxwell-
Bonnell method using the measured
fash zone temperature and pres-
sure. For a given feed, fash zone
pressure and temperature are a
relative measure of VGO product
yield. They do not correspond to
whole crude VGO TBP cutpoint.
The Maxwell-Bonnell conversion of
ASTM D5236 Potstill measurements
bears little relationship to VGO
cutpoint on an operating unit.
Crude type and upstream atmos-
it is often not the most proftable.
Dry vacuum units should be used
when the operating cost savings are
worth more than lost revenue from
reduced VGO and refnery liquid
volume yields. Simple metrics that
use ejector steam consumption as a
measure of performance are very
misleading. Refners are in the busi-
ness of selling liquid product, not
producing coke or fuel oil. Dry
vacuum units always produce less
VGO than units designed with
steam (see Figure 2); no competent
designer can argue that dry units
cut deeper. Processing moderate to
heavy crude blends and thermally
unstable crudes requires steam to
achieve moderate to high TBP
cutpoint throughout a four to six
year run. Dry vacuum units have
been used with heavy crudes and
several have TBP cutpoints less
than 930F (498C), even though the
stated designs claims higher than
1000F (537C).
Denition of VGO TBP cutpoint
VGO TBP cutpoint is a refning
industry metric to compare vacuum
unit performance. Higher cutpoint
means more VGO and less VR
production. Unfortunately, reported
TBP cutpoints are often not correct.
TBP cutpoint is based on the whole
crude TBP distillation curve; the
temperature on the Y-axis of the
curve corresponds to yield of VR as
a volume percent of crude (see
Figure 3). The whole crude TBP
curve must be accurate and the unit
material balance good to determine
actual cutpoint, otherwise the calcu-
lated value is not correct. While
TBP cutpoint is a simple concept,
the data needed to calculate it
correctly is not readily available
from typical refnery product analy-
ses. Consequently, shortcut
methods and other practices have
been developed to estimate TBP
cutpoint. Unfortunately, these meth-
ods are very misleading.
Generating an accurate whole
crude TBP curve requires good
temperature and volume data in the
VGO boiling range. For an operat-
ing unit, VGO and VR distillations
must be accurate but are generally
not available or are often inaccurate
due to inherent diffculties with VR
distillation tests. Most refneries
1200
1600
1400
1000
800
600
400
200
T
B
P
,

F
0
200
0 10 20 30 40 50 60 70 80 90 100
Volume, %
VGO cutpoint
(73.2%, 902F (487C)
Figure 3 VGO TBP cutpoint
Crude TBP F (C) Arab Light Arab Heavy Maya Merey Canadian bitumen
975-1000 (524-538) 1.5 1.7 1.6 2.5 2.3
1000-1025 (538-552) 1.4 1.8 1.6 2.5 2.3
1025-1050 (552-566) 1.3 1.8 1.5 2.5 2.2
1050-1075 (566-579) 1.2 1.6 1.5 2.0 2.2
1075-1100 (579-593) 1.1 1.6 1.4 2.0 2.3
Whole crude volume % by boiling range Whole crude volume percent by boiling range
Table 1
q2 pcs.indd 2 11/03/2014 14:24
fast returns
A Century of Innovation
in the Oil and Gas Industry
1914 - 2014
Quickly maximize the value of natural gas assets with NGL recovery
solutions from UOP.
Get cash flow and profits sooner with hydrocarbon management solutions from
UOP, including the UOP Russell line of complete, modular, factory-built systems
that are designed to have you up and running six months quicker than other
options. The reduced on-site installation time, proven designs and pre-fabricated
solutions can deliver an earlier startup at a lower cost than stick-built facilities.
Or choose the UOP-Twister
Supersonic Gas Separation System, with a small

footprint and low maintenance thats often ideal for remote and offshore facilities.
Or get maximum NGL/LPG recovery an industry-leading 99+% with solutions
from UOP and Ortloff Engineers, Ltd. In short, whether you need fast, compact,
efficient or all of the above, UOP has a solution for your NGL-recovery needs.
For more information about UOP Russell solutions, visit www.uop.com/uoprussell or
visit www.uop.com to learn about all of the UOP hydrocarbon management solutions.
2014 Honeywell International, Inc. All rights reserved.
UOP-Russell-Fast-Returns-Ad-Full-Page_UOP Ad 8.5x11 UOP6467 1/10/14 3:22 PM Page 1
uop fast returns.indd 1 04/03/2014 11:15
pheric column distillate yield
determine vacuum unit feed
composition. Heavier feeds result in
lower cutpoint for the same fash
zone temperature and pressure,
thus using fash zone pressure and
temperature to calculate cutpoint is
incorrect. Vacuum units stripping
VR can yield 3-10 vol% additional
VGO on whole crude even when
operating at higher fash zone pres-
sures than a dry vacuum unit. The
Maxwell-Bonnell AET method is
not the same as whole crude TBP
cutpoint.
Dry vacuum unit:
fundamental principles
Dry vacuum units rely solely on
operating pressure and temperature
to vaporise the feed. They do not
use coil or stripping steam to infu-
ence oil partial pressure. Heater
outlet pressure is set by fash zone
pressure and transfer line pressure
drop. Maximum outlet temperature
is set by oils thermal stability and
heater design. Low stability crudes
such as northern Athabasca mined
bitumens crack at 680-690F (360-
365C) outlet temperature, even
with a well-designed double-fred
heater that minimises oil flm
temperature rise, whereas Arab
Light can be operated at much
higher temperatures with a well-
designed conventional heater. Poor
heater design lowers maximum
heater outlet temperature, inde-
pendently of crude type, because it
is results in higher peak tempera-
tures. Dry vacuum designs do not
use coil steam to control radiant
section oil residence time, therefore
heater outlet temperature must be
lower than in designs using coil
steam to control oil residence time.
Heater oil mass fux should always
be optimised to control oil flm
temperature rise to minimise
cracking.
Heater outlet temperature and
pressure, fash zone pressure and
overfash rate determine VGO yield.
Liquid drawn below the wash zone
(sometimes called slop wax, dirty
gas oil [DGO], or wash oil recycle)
can be recycled to the heater inlet
but rarely does this improve yields
(see Figure 4). Many vacuum units
continue to operate with little or no
overfash; nearly 100% of the wash
oil is revapourised. Short wash beds
using only grid can be operated dry
without coking with thermally
stable crude oils. Deeper and/or
more effcient wash beds require
suffcient wetting to prevent coking.
Wash bed coking is more common
in deeper wash beds.
Many designers process models
are not structured properly; they
overestimate VGO yield and under-
estimate wash oil and overfash
rates. Wash oil is the fow to the top
of the wash bed. The liquid with-
drawn from the column below the
packing is the overfash (wash oil
that did not revapourise) and fash
zone liquid entrainment. Many
refners and designers continue to
confuse slop wax (liquid withdrawn
below the wash bed) with overfash.
Many vacuum units do not operate
with suffcient wash oil; nearly all
of it revapourises. The material
withdrawn from the column is
nearly all resid entrainment. Many
feld tests have shown the distilla-
tion and API gravity of this material
nearly equal to that of VR.
Recycling slop wax containing little
or no overfash generates no incre-
mental VGO yield. When
processing unstable feeds, recycle
contributes to heater tube coking
and coking in the slop wax circuit.
Transfer line design has been the
subject of much discussion over the
years. Our experience, having meas-
ured many heater outlet pressures
and modelled more than 50 operat-
ing transfer lines, is that they all
have a critical fow expansion and
generally have more than one.
Nearly all heater outlet tube-to-
transfer line expansions and most
column inlets are at or near critical.
Experience has shown that it is
nearly impossible to design a trans-
fer without critical expansions, let
alone design one at 80% of
two-phase critical mass velocity. It
may be possible to design at 80% of
critical velocity assuming vapour
phase sonic velocity, but two-phase
critical mass velocity is much lower
than calculations based on the
vapour phase alone. Large transfer
line diameters force the critical
expansion back toward the heater
outlet and often into the heater
because outlet tubes are seldom
larger than 10.
Ultimately, transfer line design
sets heater outlet pressure. The
lowest measured pressure over the
last 18 years was 3.0 psia, with most
at or above 6.0 psia and a few above
atmospheric pressure. Transfer line
modelling requires a program that
accurately calculates critical mass
velocity and pressure loss through a
complex piping system consisting
of both vertical and horizontal runs
and several line size expansions.
Having measured many heater
outlet pressures and rigorously
modelled these lines, our calcula-
tions compare very well with the
measured pressure drop. In nearly
all cases, there are one or more criti-
cal expansions.
Vacuum column fash zone pres-
sure is a function of the ejector inlet
pressure and column pressure drop.
Minimum inlet pressure is a func-
tion of the intercondenser cooling
water temperature. Ejectors can
operate at up to 15/1 compression
ratio but the authors have seen
many instances of poor jet perfor-
mance when compression ratios
exceed 10/1 with large ejectors. If
the inter-condenser cooling water
temperature is suffcient to
condense at 70 mmHg on the US
Gulf Coast, the minimum suction
pressure would be about 7 mmHg.
Some small dry vacuum unit ejec-
tors operate at 15/1, resulting in an
inlet pressure closer to 4 mmHg.
Overflash
and
entrainment
Overflash
Wash oil
Entrained
Vapour
VTB
Feed
Wash
bed
Figure 4 Slop wax recycle
q2 pcs.indd 3 10/03/2014 13:51
ing has been performed by a
specialty laboratory that are experts
in ASTM D7169.
Once accurate data is collected,
only then can a process model be
developed. The model must be
capable of matching the actual oper-
ating conditions, otherwise it is not
a useful design tool for future
vacuum unit design. Heater outlet
pressure must be measured or
calculated using the right model,
and the heater outlet temperature
measurements location must be
verifed. Many times, the measured
heater outlet temperature was not
in the heater outlet tubes but in the
transfer line after a line expansion.
Often these are critical expansions
reducing temperature by 10-15F,
therefore the actual heater outlet
temperature is higher.
Model structure
Designers continue to assume that
the liquid/vapour from the heater
to the fash zone are in equilibrium.
This is not correct because the
phases separate in the larger hori-
zontal runs. The locations where
this occurs depend on the transfer
line. Structuring the process model
correctly requires rigorously model-
ling the transfer line and
determining where the vapour and
liquid separate. Heater outlet tubes
are relatively small so vapour and
liquid are in equilibrium. A transfer
plant pressure measurements and
refnery laboratory data suffcient
and/or accurate enough to use for
process modelling. Poor input
ensures poor results although often
little effort is spent ensuring good
data because of the cost and
personnel time involved. Vacuum
column top measurements are
rarely correct often due to large
span instruments being used to
measure an operating pressure at
5% or less of full scale. Flash zone
measurements are often poor due to
instrument line plugging or liquid
flling. In some cases refners have
two or three thermocouples meas-
uring the same point with more
than 10F variation.
Using outmoded methods such as
ASTM D1160 tests, extrapolating
distillation curves or relying on
crude assays alone does not allow
TBP cutpoint to be calculated accu-
rately. Presuming the actual crude
oil feed matches the crude oil assay
data should never be done. VGO
and VR product ASTM D7169 TBP
distillations are diffcult to do,
hence all test run ASTM D7169 test-
Most ejectors measured suction
pressures are above 8 mmHg and
many of the older designs have
20-40 mmHg inlet pressures. It is
not possible to achieve a 1050F
(565C) VGO TBP cutpoint with a
dry vacuum unit when operating at
fash zone pressures above 20
mmHg, even with the lightest crude
oils.
Field data: the real world
Do calculations match actual unit
temperatures, pressures and process
stream fow rates? Many process
engineers have never feld tested
their vacuum unit designs to deter-
mine how actual results compare
with their model calculations. They
rely on offce-based guidelines and
hearsay. Process Consulting
Services has tested more than 100
crude/vacuum units in the last 18
years. In all cases, considerable
effort was made calibrating fow
meters prior to testing. Crude and
product stream sampling was
witnessed by the engineers doing
the modelling, and stream analysis
was performed by the same labora-
tory on the heavy streams (diesel
and heavier, including atmospheric
and vacuum residue). Operating
pressures and temperatures were
measured with dedicated, high
accuracy instruments to ensure the
data were accurate. Field instru-
ments are calibrated or replaced
when out of calibration, hence the
pressure and temperature data used
to model are accurate. In no cases
were the refnery instruments alone
relied on to confrm actual vacuum
column operating pressures. In
several cases, refnery instrument
measured process temperatures
were incorrect on fash zone and
heater outlets. Metering problems
are very common, therefore consist-
ency checks must be done to ensure
that the synthesised crude based on
unit material balance and stream
analysis match the measured crude
TBP distillation. These consistency
checks ensure that the crude TBP
distillation is correct and calculated
VGO TBP cutpoint is correct.
Vacuum unit feed must be char-
acterised properly and operating
temperatures, pressures and fow
rates must be correct. Rarely are
Liquid
Vapour
Figure 5 Phase separation in transfer line
Wash oil
rate
Overflash
Transfer
line
Stripping
section
Wash
section
Steam
Flash
Flash
Flash
Flash
Entrainment
Splitter
Furnace
outlet
Transfer
vapour line
Figure 6 Modied model structure
q2 pcs.indd 4 10/03/2014 13:52
analysis of oil residence time and
peak oil flm temperature. Both
determine the rate of cracking
inside the heater tubes, which sets
the amount of cracked gas fowing
to the ejectors. Heater cracked gas
production is highly dependent on
oil stability and heater design.
Cracked gas rates exceeding
0.3 wt% of feed cause rapid coke
laydown inside the heater tubes.
Coil steam rate can be adjusted to
control oil residence time and
reduce thermal cracking.
Heater and ejector system perfor-
mance are integrally linked. Ejector
system capacity sets the column
fash zone pressure. Ejector suction
pressure depends on process gas
load which consists of coil and
stripping steam, cracked gas and air
leakage. Increasing coil steam
allows a higher heater outlet
temperature for a given amount of
cracked gas, but the process steam
also raises the frst and second stage
ejector gas load. This will increase
fash zone pressure unless the ejec-
tor has suffcient capacity.
Estimating the heater and transfer
line pressure profle accurately
requires a model capable of rigorous
tube-by-tube heat transfer and accu-
rate two-phase fow calculations.
Heater pressure drop is a function
of the tube size, tube length, coil
steam rate and heater tubes coke
layer thickness. Pressure in the last
two to three tube rows of the radi-
ant section decreases rapidly and oil
vaporisation begins. Accurate feed
heating-cooling curves are essential
to properly size the outlet tubes, size
transition locations and transfer line.
Incorrect sizing of the last two to
three tube rows can result in low oil
vaporisation in the larger tubes,
high oil residence time and rapid
coking. These larger transition tubes
often have both high oil flm
temperature due to lower oil mass
fux rate and high oil residence time
caused by the increase in tube size.
It is not unusual for coke to form in
transition tube locations.
Dry vacuum units or vacuum units
with steam
VGO TBP cutpoints continue to be
overestimated and wash oil sections
continue to coke from insuffcient
when designing vacuum units. It is
not possible to separate the design
of the transfer line from the heater
or column.
Vacuum unit pressure and
temperature
Heater outlet temperature and pres-
sure determine VGO product yield
at fxed fash zone pressure and
overfash rate. For any given pres-
sure as the crude blend becomes
heavier, heater outlet temperature
needs to be increased or VGO yield
decreases. Maximum heater outlet
temperature is limited by the rate of
coke formation inside the heater
tubes. Vacuum column fash zone
pressure is largely set by ejector
inlet pressure. Therefore, vacuum
heater and ejector system design are
inter-dependent because the ejector
system must be able to handle
cracked gas and coil steam rate
leaving the heater. Operating at
very high heater outlet tempera-
tures, or when processing unstable
crude blends, requires heater coil
steam to avoid premature tube
coking. As crude gets heavier, VR
must be stream stripped to meet
VGO yield and reliability objectives.
Increasing the heater outlet
temperature raises the amount of
cracked gas produced. When
designing a reliable heater, it is
necessary to ensure that the peak oil
flm temperature and oil residence
time do not increase the rate of
cracking above the ejector capacity,
or result in rapid coke laydown
inside the tubes. Rigorous heater
modelling will include tube-by-tube
line consists of a complex piping
run with large diameter piping
where the liquid and vapour sepa-
rate (see Figure 5) in the horizontal
sections, vapour fowing along the
top of the pipe and liquid across the
bottom. Describing the heat, mass,
and component balances at all
points in the transfer line is not
practical. However, a method to
estimate it is important. As long as
the model approximates what actu-
ally happens in the equipment, as
opposed to being theoretically
perfect, it is a useful tool for process
design engineers.
A modelling approach (see Figure
6) that has been validated with
many feld tests is used to approxi-
mate non-equilibrium in the
transfer lines and vacuum columns.
It better predicts VGO yields, wash
rate and other critical operating
parameters. Using multiple unit
operations allows the non-equilib-
rium nature of the system to be
estimated. This technique, although
only an approximation of the actual
non-equilibrium process, more
accurately predicts VGO and VR
product yields and better estimates
the wash fow rate needed to avoid
coking in the wash section (see
Figure 7). Operating pressure at the
heater outlet and where the phases
separate in the transfer line must
also be estimated. An accurate
heater through transfer line model
is needed to estimate the pressures
at the heater outlet and location
where phase separation occurs.
Both process and rigorous equip-
ment system models must be used
Figure 7 Wash bed coking
q2 pcs.indd 5 10/03/2014 13:52
heater and results in increased
coking in wash oil sections, heater
tubes, slop wax and VR circuits. To
avoid these problems, Process
Consulting Services specifes a low
speed, low head pump with slop
wax returned to the stripping
section or bottom of the column.
The improved reliability of this
system compensates for the theoret-
ical yield improvement that heater
recycle may offer.
Scott Golden is a Chemical Engineer with
Process Consulting Services, Inc, in Houston,
Texas. He specialises in front-end process
engineering for renery unit revamps and has
authored more than 100 technical papers.
Email: sgolden@revamps.com
Tony Barletta is a Chemical Engineer with
Process Consulting Services, Inc., in Houston,
Texas. His primary responsibilities are
conceptual process design and process design
packages for large capital revamps.
Email: tbarletta@revamps.com
Steve White is a Chemical Engineer with
Process Consulting Services, Inc, in Houston,
Texas. He handles large capital grassroot
process design packages. He has more than 30
years of process design experience for renery
revamps and grassroots units.
Email: swhite@revamps.com
one example, a dry column operat-
ing at low pressure was expected to
produce a 1050F (566C) VGO
product cutpoint, based on stand-
ard modelling techniques. When the
model was revised to account for
transfer line phase separation and
provide proper overfash, the same
heater outlet conditions achieved a
TBP cutpoint of only 1004F (540C)
which matched the actual feld
measured performance. To achieve
the 1050F (566C) VGO cutpoint
and an acceptable overfash to meet
a four to six year run, a steam
stripped vacuum column is
required. VGO product yield and
reliable operation throughout a four
to six year run determine the proper
vacuum unit design.
Recycling slop wax is a standard
design procedure for many engi-
neering contractors. In order to
recycle to the front of the heater, a
low volume high head pump is
required. This has resulted in poor
seal and bearing performance and
signifcant maintenance cost. In
addition, the entrained resid pres-
ent in slop wax is continually
exposed to high temperatures in the
wash oil fow rate. Poor feed char-
acterisation and improper model
structure are the leading causes of
under-performance. Extensive feld
testing, sampling and simulation
shows of equilibrium conditions
from the heater outlet through the
transfer line and column fash zone
are not valid. Calculated wash oil
fow rates are too low, leading to
rapid wash bed coking from low
overfash. Low effciency, short
wash zones can be operated at
essentially no overfash, but beds
requiring reasonable effciency will
coke. A revised modelling scheme
accounts for non-equilibrium opera-
tion and has been validated in
numerous test runs and feld
surveys over the past 18 years.
Meeting VGO yield and unit run
length is essential. Process model-
ling structure needs to account for
the non-equilibrium leading to
higher wash oil vaporisation and
lower overfash rates at a given
wash oil rate than is calculated by
many designers. As a result, higher
heater outlet temperatures are
required for a given VGO TBP
cutpoint and desired overfash. In
Now, we can predict corrosion rates
in our amine and sour water systems. YES!
Only with ProTreat
ProTreat

Simulation Software
Subscribe to our FREE monthly technical bulletin,
The Contactor
, at www.ProTreat.com.
The industrys most reliable gas treating simulator
100% mass transfer rate based since inception
Reducing risk and cost for 20 years
www.ProTreat.com +1 281 970 2700 info@ProTreat.com
125_OGT_ProTreat_Half_Horizontal_Q2_PTQ.indd 1 3/5/14 10:48 AM
q2 pcs.indd 6 10/03/2014 13:52
Introducing ThyssenKrupp Industrial Solutions, a new company that
combines the strengths of ThyssenKrupp Uhde with those of ThyssenKrupp
Resource Technologies, itself the product of last years merger between
ThyssenKrupp Frdertechnik and ThyssenKrupp Polysius.
We now provide expertise in the engineering and construction of chemical
plants, refineries, cement works and other industrial complexes as well as
advanced technologies for mining, mineral processing and bulk materials
handling systems.
Engineering Excellence
3
: a new dimension of true expertise for customers
everywhere. Why not contact us today?
www.thyssenkrupp-industrial-solutions.com
ThyssenKrupp
Industrial Solutions
Engineering Excellence has now become
3
ThyssenKrupp Uhde and ThyssenKrupp Resource Technologies are now
ThyssenK
rupp Industrial Solutions
tkrupp.indd 1 10/03/2014 11:09
Coordinating a complex automation
project
T
o meet the latest European
fuel specifcations, INA has
upgraded its Rijeka refnery,
including building a new
hydrocracking/hydrodesulphurisa-
tion unit, hydrogen generation unit
and sulphur recovery unit. Critical
to the projects success was the
careful coordination of a complex
automation project by Emerson
Process Management, which helped
ensure on-time start-up and eff-
cient operations.
INA d.d. is an oil and gas
company with exploration and
production activities in the Adriatic
Sea, Angola and Egypt, and refner-
ies producing fuels in the Croatian
cities of Rijeka and Sisak. The Rijeka
refnery, located next to the Adriatic
Sea, has a nominal crude oil distilla-
tion capacity of 4.5 million t/y.
Production is centred on white
products such as naphtha, gasoline
and kerosene, but with a signifcant
amount of fuel oils produced.
In 2007, INA began the process of
upgrading the Rijeka refnery
including installation of a new
hydrocracking complex to enable
production of Euro V grade prod-
ucts and meet the latest EU
environmental legislation. The
complex consists of a new
hydrocracking/hydrodesulphurisa-
tion unit, hydrogen generation unit
and sulphur recovery unit. A new
centralised control facility for all
three units was also part of the
upgrade.
Because of the signifcant size of
the overall project, the construction
of each process unit was contracted
out to a different engineering,
procurement and construction
(EPC) company. The lack of availa-
Installing an integrated control system in support of a major renery upgrade
meant coordinating the work of multiple EPC contractors
IGOR EPI
C and BORIS EELJ

INA Rijeka renery, Croatia
ble EPCs in Croatia at the time of
the start of the project led INA to
turn to companies from outside the
country. The hydrocracker process-
ing unit was constructed by an EPC
from the Czech Republic, the
sulphur recovery unit by a
Hungarian company and the
hydrogen generation unit by an
Italian group.
The modernisation project was
split into two phases. Phase one,
which fnished in mid-2010, aimed
solely to enable Rijeka to produce
Euro V quality gasoline and diesel
fuels in compliance with EU envi-
ronmental standards. Phase two
will be a residual upgrade project,
which looks to improve yield and
proftability from the new units.
Integrated control and safety system
In 2009, during construction of the
new hydrocracker complex, INA
approached Emerson Process
Management to provide an inte-
grated control system and safety
solution for all three of the process
units in the hydrocracking complex.
Operated from a central control
room, the integrated system had to
offer a common user interface for
operators of all three units. This
would streamline training and
enable the refnerys operators to
move from one unit to another. It
was also important that operators
could view both the process data
and the safety system data within
the same user interface. This would
improve the effciency of operators
and help to improve plant safety.
Digital plant architecture
With the three processing units
having different contractors and
completion dates, the integrated
control and safety system project
represented a considerable chal-
lenge. Critical to the success of this
signifcant project was Emersons
ability to coordinate activities
between the EPCs. Emerson
provided a single point of contact
for them, took sole responsibility for
the automation systems, and coordi-
nated the engineering approach.
This coordinated approach ensured
that fxed scopes, time schedules,
budgets and warranties for each
project were not adversely affected
by the others.
INA selected Emerson to be the
main automation contractor (MAC)
for the sulphur recovery unit,
responsible for front-end engineer-
ing and design (FEED) for the
integrated control and safety
system, and providing services that
included automation engineering,
installation, acceptance testing,
commissioning, confguration, and
start-up support for all three units.
The integrated solution for the
hydrocracking complex is based on
Emersons PlantWeb digital plant
architecture including the DeltaV
automation system, DeltaV SIS
safety instrumented system and
AMS Suite predictive maintenance
software. As an integrated control
and safety system, DeltaV and
DeltaV SIS systems provide the
refner with an integrated engineer-
ing, maintenance and operations
environment, enabling the new
units to operate safely and at opti-
mum levels to maximise
proftability.
INA found that all safety-related
information is easily accessible
through familiar and intuitive
q2 emerson.indd 1 11/03/2014 16:41
was that there were no delays for
any of the unit projects. Having
installed the most advanced equip-
ment and systems, Emersons
technical experts are now helping
the refner to maximise perfor-
mance, ensuring the plant becomes
one of the most effcient facilities of
its type.
The suppliers project manage-
ment services helped bring the
plant on stream quickly and
smoothly. Prior to start-up, INAs
operators were trained off-line
using Emersons OTS simulator.
This enabled them to understand
the capability of the new system
and how best to respond to process
changes and unusual situations
whilst in a no-risk environment
prior to actual plant operation.
Integrated advantages
The integrated process control and
safety system has created a number
of opportunities for procedural
improvements. The centralised
system operated from one control
room (see Figure 1) has enabled INA
to optimise the number of operators
and improve communication
between individual unit operators.
The entire operating team now
works very much as a single team.
Because all subsystems are visible
through one common user inter-
face, with an overview schematic
graphic on a central screen, it is
much easier to use than the old
system. United graphics on the
dashboards make things easy to
fnd and work with. Identifcation
of process errors is subsequently
faster and training operators is now
much quicker. Because of the
success of this project, INA will
now standardise on a centralised
control room and an integrated
process control and safety system
for all future expansion projects
and upgrades to existing facilities.
Igor epi
c is Rijeka Renery Sector Director

at INA d.d. He holds a Chemical Technology
Engineer (BSc) degree from the Faculty of
Chemical Engineering and Technology in
Zagreb and an EMBA from the IEDC Bled
School of Management, Slovenia.
Boris eelj is Head of the Process Automation
Department at the INA Rijeka Renery. He
holds an IT Expert Specialist masters degree
from the Polytechnic of Rijeka.
applications. The integrated but
separate architecture complies with
standards for physical separation
and independence of safety and
control, delivering the benefts of
total integration and total separa-
tion without the trade-offs
associated with the two extremes.
The refner also required an inte-
grated gas detection system within
the new safety instrumented
system. Emersons fully integrated
gas detection system simplifed the
implementation, requiring only an
additional separate cabinet to be
connected to the DeltaV SIS. The
integrated system requires less
engineering, offers faster alarming
and visualisation, and creates a
single point of responsibility for
upgrades and maintenance.
In addition to the integrated
control and safety system, Emerson
provided the majority of the feld
instrumentation, tank gauging and
control valves. Now installed, this
huge system incorporates nearly 50
process controllers, 180 logic solv-
ers and approximately 5000 process
and safety I/O devices.
Maintenance improvements
INA is using Emersons AMS Suite
predictive maintenance software to
support improved maintenance
procedures at the plant. By using
predictive intelligence about the
performance and condition of valves
and instruments, we are now able to
improve the availability and perfor-
mance of these key assets.
The central control room is
located approximately 1.5km from
the process units. A video surveil-
lance system with remotely
operated cameras and a radio
system were installed as a founda-
tion for remotely operated systems.
A secure write mechanism within
the DeltaV SIS enables maintenance
overrides to be authorised for
specifc areas of the plant or group
of equipment. This is performed via
a key switch that is accessible by
authorised operators using an HMI
in the control room. This simplifes
our override procedures, which
would normally require a trip into
the feld and activation of a hard-
wired override key switch located
within the ESD system cabinet.
Project implementation success
The approach taken by Emerson to
the project and the implemented
solution was successful. An auto-
mation project of this size requires
capabilities not only to implement
the suppliers own equipment but
also to manage multiple EPCs from
different countries with different
standards, specifcations and sched-
ules. A critical element to the
success of the project was the
ability to deploy locally based
resources to coordinate these activi-
ties and ensure that the integrated
system offered a common user
interface for all three units.
During the course of this project
Emerson helped to solve any issues
INA had and provided on-going
service and support. It took sole
responsibility for the automation
systems and coordinated the engi-
neering approach, and the result
Figure 1 The centralised system is operated from one control room
q2 emerson.indd 2 10/03/2014 14:43
www.jmprotech.com
UK Tel +44 (0)1642 553601 Fax +44 (0)1642 522542
At Johnson Matthey we dont just provide world
class hydrogen catalysts, we provide knowledge
and understanding of the hydrogen process
and catalytic interactions through a unique
range of services to enhance plant operation
and improve efciency and reliability.
service,
support,
solutions.
j mathey.indd 1 10/03/2014 16:31
global expertise. local focus. personal drive.
Innovation drives our combustion experts. Zeeco pushes
combustion technology forward, developing and implementing
petrochemical solutions like our Next Generation Ultra-Low NOx
GLSF Free-Jet burners. We understand the challenges our
customers face. We meet them every day in every corner of the
world with better by design flares, burners, and thermal oxidisers.
We design and manufacture worldwide so the solution or
service you need is never far away. Global expertise, local focus,
and 1,000 people personally committed to every project.
Choose the company where promises matter and people deliver.
Choose Zeeco.
burners flares thermal oxidisers aftermarket parts, service & engineered solutions
Zeeco, Inc.

22151 E 91st St., Broken Arrow, OK 74014 USA
+1-918-258-8551 sales@zeeco.com zeeco.com
Zeeco, Inc. 2014
ZEECO
GLSF Free-Jet Burner

PTQ_2nd Quarter_Mar2014_8.5x11.indd 1 3/11/14 8:06 AM
zeeoco.indd 1 11/03/2014 14:27
Dynamics of operation for are systems
F
lares are often used in petro-
leum refneries as control
devices for regulated vent
streams as well as to handle
non-routine emissions. Flares are
used to dispose of unwanted gases
or liquids. Rarely, liquids are sepa-
rated from the gas and burned in
the liquid state or vaporised and
burned as a gas. The unwanted
material is generally composed of
hydrocarbon gases.
1
The faring process can produce
different pollutants SO
2
, NOx,
CO, VOCs, CH
4
and CO
2
their
constituents depending on two main
factors: the waste gas composition
sent to the fame and the combus-
tion effciency.
Besides waste gas composition
and combustion effciency, main
factors affecting faring availability
and reliability are operating and
maintenance practices. For opera-
tional safety in faring, there should
be:
Well defned operating proce-
dures and training
Planned and routine maintenance
of fare components.
3
Smokeless aring and combustion
efciency
The fare is usually required to be
smokeless for the gas fows that are
expected to arise from normal day
to day operations. This is usually
15-20% of the maximum gas fow.
Various techniques are available
for producing smokeless operation,
most of which are based on the
premise that smoke is the result of a
fuel-rich condition and is eliminated
by promoting uniform air distribu-
tion throughout the fames.
4
One of the main challenges for
Detailed studies of problems with aring performance led a renery to dene
and implement a scheme for optimum are performance
YAHYA AKTA and ZGE ZARIK
Tpra Izmir renery
faring is initiating and maintaining
ignition systems. Initiation of burn-
ing requires a pilot that can
withstand strong winds, rain and
inert surroundings.
2
The primary function of a fare is
to use combustion to convert fam-
mable, toxic, or corrosive vapours to
less objectionable compounds.
Briefy, one theory suggests that
steam separates the hydrocarbon
molecules, thereby minimising
polymerisation, and forms oxygen
compounds that burn at a reduced
rate and temperature and are not
conducive to cracking and polymeri-
sation. Another theory claims that
water vapour reacts with the carbon
particles to form carbon monoxide,
carbon dioxide and hydrogen,
thereby removing the carbon before
it cools and forms smoke.
5
Smokeless fares eliminate any
noticeable smoke over a specifed
range of fows. Smokeless combus-
tion is achieved by utilising air,
steam, pressure, energy, or other
means to create turbulence and
entrain air within the fared gas
stream.
Typically, smoking tendency is a
function of the gass calorifc value
and of the bonding structure of the
hydrocarbons. The paraffnic series
of hydrocarbons has the lowest
tendency to produce smoke,
whereas olefnic, diolefnic and
aromatic hydrocarbons have a much
higher tendency to produce smoke.
Smokeless fares can be provided
with a steam assist or air assist
system to improve combustion. Gas
system hydraulics (that is, the gas
pressure drop available for the fare
equipment) can infuence the
method chosen for smoke suppres-
sion. The pressure (kinetic energy)
of the fare gas can, if suffcient, be
used to make the fare operate with-
out smoke. The smoke suppression
method is dependent on the utility
availability and cost of the utility.
5
Steam assisted fares are single
burner tips, elevated above ground
level for safety reasons, which burn
the vented gas in essentially a diffu-
sion fame. They reportedly account
for the majority of fares installed
and are the predominant type of
fare found in refneries and chemi-
cal plants.
6
In Tpra Izmir refnery, except
for acid gas fares all the fares are
steam assisted with a single burner
tip.
The quantity of steam required for
smokeless burning is a function of
the gas composition, the fare burn-
ers size and design, the steam
injector design, operating pressure
and the environmental conditions.
While steam assist enhances the
combustion of relief gases that
smoke, it adversely affects the
combustion of relief gases with a
high level of inerts. Relief gases with
a high level of inerts, when fared
from a steam assisted fare, can
require a greater calorifc value to
sustain the required fame stability
and the effciency of hydrocarbon
destruction. Steam is often injected
into the relief gas discharge at the
top of a fare burner. Typically, a
steam ring that has a number of
injection nozzles or slots is
employed.
The upper steam injection func-
tions to draw in air and to force the
air mixture into the relief gas
discharging from the fare burner.
The steam injection pattern is
tupras.indd 1 10/03/2014 13:56
rate is dependent on the carbon to
hydrogen ratio of the gas being
fared. A high ratio requires more
steam to prevent a smoking fare.
Liquid in the vent stream can
extinguish the fame or cause irregu-
lar combustion and smoking. In
addition, faring liquids can gener-
ate a spray of burning chemicals
that could reach ground level and
create a safety hazard.
In addition to burner tip design,
gas density also directly affects the
minimum purge gas required to
prevent fashback, with lighter gases
requiring more purge.
6
Poor combustion by low atomisa-
tion steam and a bad fame in the
fare system is caused by:
Poor mixing at the fare tip
Oxygen defciency
Liquid in the vent stream
Poor effciency of fare burners.
If there is an oxygen defciency
and if the carbon particles are
cooled to below their ignition
temperature, smoking occurs. In
large diffusion fames, combustion
product vortices can form around
burning portions of the gas and shut
off the supply of oxygen. This local-
ised instability causes fame
fickering, which can be accompa-
nied by soot formation.
As in all combustion processes, an
adequate air supply and good
mixing are required to complete
combustion and minimise smoke.
6
In contrast to the situation arising
from inadequate steam, using too
much steam in a fare can reduce
fare performance, Given that many
steam assisted fares are designed to
have a minimum steam fow rate in
order to protect the fare tip,
over-steaming has resulted; in
particular during base load condi-
tions the excessive consumption of
steam for a steam assisted fare can
be visualised in Figure 1. The fame
is cleaner, but operating a fare with
such an amount of steam is not
feasible and the possibility of the
fame extinguishing is high.
It can be concluded that steam
consumption must be optimised by
monitoring the amount and compo-
sition of the waste gas. If possible,
the sources of faring from the
process side, for example safety
valve openings, must be controlled
intended to enhance fuel-air mixing
and can add to the momentum of
the relief gas discharge. The steam
and air acts to dilute the hydrocar-
bon fuel content, which also reduces
the smoking tendency. The steam
vapour can also participate in the
combustion kinetics, assisting in the
conversion of carbon to carbon
monoxide.
The effective addition of steam
from an upper steam ring increases
the turbulence of combustion. The
overall noise level of the fare burner
increases due to both the additional
combustion noise and the jet noise
from the steam nozzles. Steam
assisted, smokeless fares can have
signifcantly increased overall noise
levels in comparison to fares with
no steam assist.
In addition to the operating
considerations mentioned earlier for
the pipe fare, attention should be
given to the rate of steam injection.
If too little steam is added to the
fare burner, a smoking, softer, more
wind defected fame is produced.
Proper steam injection proportions
the steam injection rate to the relief
gas fow rate. The lowest cost opera-
tion for steam injection is to operate
just above the incipient smoke point
for the gas composition and fow
rate.
Higher steam injection rates make
the fame harder, cleaner and less
wind defected. Higher steam-injec-
tion rates also increase the noise
levels. Excessive steam injection
rates produce combustion instability
accompanied by excessive fare
noise. At the extreme, over-steaming
can extinguish the fame.
Steam fows from an upper steam
ring can condense and create water
and blockage problems for a fare
burner.
If excessive condensate occurs
with soot formation, it causes block-
age problems in the molecular seal
drum, and plugs the waste gas
route. This operational failure
encountered in Tpra Izmir refn-
ery is explained in detail later.
The use of a defned minimum
fow of dry or, preferably, slightly
superheated steam together with the
proper design and layout of steam
trapping facilities are critical to
minimising these problems.
It is necessary for the refractory
used for fare burners to be resistant
to high temperature and thermal
shock. Flare burners experience
rapid changes in temperature.
Internal refractory should be well
anchored and it is recommended
that the refractory material should
include metal needles to help hold it
in place. With any refractory instal-
lation, it is necessary to consider the
consequences of a refractory failure.
For dense refractory material, this
consideration focuses on where the
failed refractory will fall. The refrac-
tory can fall inside the fare and
potentially obstruct the relief gas
fow, or fy out with the excess gas
fow to grade and create potential
risks for the operators.
6
If the refractory particles fall
inside the molecular seal, there is
also a risk of plugging of the seal
drain and consequent obstruction of
the relief gas fow.
Analysing the practices of
are operation
To ensure an adequate air supply
and good mixing, a steam assisted
type of fare system injects steam
into the combustion zone to
promote turbulence for mixing and
to induce air into the fame, as was
mentioned earlier.
Poor mixing at the fare tip is the
primary cause of fare smoking
when burning a given material.
Streams with high carbon to hydro-
gen mole ratios (greater than 0.35)
have a greater tendency to smoke
and require better mixing for
smokeless faring. For this reason,
one generic steam to vent gas ratio
is not necessarily appropriate for all
vent streams. The required steam
Figure 1 High consumption of steam
tupras.indd 2 10/03/2014 13:56
continuously. Through improved
communication with the fare opera-
tors, steam injection must be
increased or decreased immediately
without formation of excessive
smoke.
Another important factor affecting
smokeless faring is the introduction
of liquid and the condensation of
hydrocarbons in the fare header or
fare riser. If this liquid phase is
carried to the combustion zone, the
smokeless performance of the fare
decreases proportionally.
These hydrocarbon droplets in the
gas stream usually cause the fame
to be smoky with the formation of
soot. In order to prevent this distur-
bance to combustion effciency, the
knock-out and liquid seal drum
design and heat tracing systems
become more critical.
If there is still liquid entrainment
to the burner, and it cannot be
drained due to plugging of the
molecular seal drain via soot and/or
refractory malfunction, the relief gas
cannot be vented and the pressure
at the fare header starts to increase
slowly.
The case that occurred in Tpra
Izmir refnery is explained below.
Due to a pressure increase on the
fare header, the related process
units were faced with gas relieving
diffculties. When the pressure
transmitter on the header after the
liquid seal drum was checked, we
observed a sharp increase in the
pressure, but in the real case it was
not a sharp increase. It was seen
that the pressure increased slowly
over two months with a total
increase of 0.6-0.7 bar. When the
pressure increased to 1.7 barg (see
Figure 2), the fare blew out a large
quantity of black water containing
soot from the tip. The photos from
this case can be seen in Figure 3.
The total blow-out period was 10
minutes with subsequent smoky
burning.
After this experience, thankfully
without injury, the pressure trans-
mitters on the main header started
to be monitored and controlled
continuously.
A short time afterwards, the fare
header pressure started to increase
for a second time. As a result of this
repeating increase, the decision was
taken to go to detailed maintenance.
During the maintenance period, we
observed that the drain of the
molecular seal was plugged with a
large quantity of soot and refractory
particles (see Figure 4).
Flare burners are among the
important mechanical components
of faring in order to ensure safe and
reliable combustion. The integrity
and reliability of this component has
a direct effect on operability and the
run length between maintenance
intervals for the facility.
Although many design alterna-
tives for fare burners exist, the
following are the basic operational
requirements observed and experi-
enced in Tpra Izmir refnery. For
all fare burners:
Steam injection should not disrupt
the basic fame stabilisation mecha-
nisms of the fare burner
Flare burners in combination with
their system of pilots/ignition
systems should be capable of main-
taining stable combustion of the
main fame for the specifed service
conditions, including the environ-
mental conditions
Steam condensate should be
drained from the internal steam/air
injection point.
4
Figures 5 and 6 show the
disrupted fare burners. The main
reasons for the disruption are the
leakage of cooling steam and routine
maintenance.
Conclusions
The observations presented in this
article are a result of the analysis of
Figure 2 Header pressure increase due to blockage in the molecular seal drum
1.4
1.8
1.7
1.6
1.5
1.3
1.2
1.1
1.0
0.9
H
e
a
d
e
r

p
r
e
s
s
u
r
e
,

b
a
r
g
0.8
1
/
8
/
2
0
1
2
1
1
/
8
/
2
0
1
2
2
1
/
8
/
2
0
1
2
3
1
/
8
/
2
0
1
2
1
0
/
9
/
2
0
1
2
2
0
/
9
/
2
0
1
2
3
0
/
9
/
2
0
1
2
1
0
/
1
0
/
2
0
1
2
2
0
/
1
0
/
2
0
1
2
3
0
/
1
0
/
2
0
1
2
First
blockage
Second
blockage
Figure 3 Blow-out from the are tip
tupras.indd 3 10/03/2014 13:56
KALDAIR
j zink.indd 1 07/03/2014 13:05
turbulence, all are systems except
acid gas ares are steam assisted.
These two components are very
important in order to reach high
combustion efciency.
4
References
1 Baukal C E JR, The John Zink Combustion
Handbook, Ch I, 5, Process Industries, CRC Press
LLC., 2001, USA.
2 Perry R H, Green D W, Perrys Chemical
Engineers Handbook, Heat Generation part:
27-25, Combustion Background, McGraw-Hill,
1997.
3 Martin J, Lumbreras J, Rodriguez M E, Texting
aring emission factors for aring in reneries.
4 API Standar 521, Guide for Pressure-
Relieving and Depressuring Systems, 15 Feb
1999.
5 ANSI/API Standard 537, Flare Details for
General Renery and Petrochemical Service,
Dec 2008.
6 EPA/452/B-02-001 Section 3, VOC Controls.
Yahya Aktas is Process Superintendent with
Tpra, Izmir renery, Turkey. He holds a BSc in
chemical Engineering and a MBA.
Email: yahya.aktas@ tupras.com.tr
zge zarik is Production Superintendent
with Tpra, Izmir renery, Turkey. She holds
a double major BSc degree in chemical and
environmental engineering.
Email: ozge.demirel@ tupras.com.tr
of new spare parts will result in
increased efciency of the ares,
hence reliability and safety will
increase and maintenance shutdown
periods will be reduced.
For each are, a suggested
approach for development and
Implementation should include:
Physically check the ares for
performance on high ame. Flare
spare parts should be ready for
shutdown and should be well main-
tained and carefully operated
Prepare and implement are poli-
cies and strategies as well as
technical arrangements affecting
industrial products in accordance
with the principles, objectives and
policies in the development plans
Taking into consideration all of
the objectives: determine, implement
and ensure the enforcement of tech-
nology and innovation policies for
all are systems.
To conclude, liquids in waste
gases can cause an irregular ame.
To prevent these liquids from reach-
ing the tip of the are, normal
aring systems have a knock-out
drum. All the stacks in Izmir ren-
ery have this equipment. Besides
providing correct mixture and
several experimental studies of are
efciency and performance. Since
aring is a critical operation in a
renery, the are and related
mechanical components should be
designed to operate properly for the
specied service conditions for a
minimum of ve years without the
need for an outage of the operating
facility.
Barriers to success for are
systems include:
Lack of support for the are
systems by operating staff
Improper are maintenance
Several ares being operated and
controlled by different staff, each
unaware of the others experiences
on the site.
In order to dene actual and func-
tional operational parameters for the
are system, we have observed that
the following parameters are critical
and must be well dened:
Complete composition of the gas
Flow variation of the gas
Temperature range.
4
In addition to daily operations, if
the parameters listed above are
clearly dened and followed, the
design of a new ue gas recovery
system will function well. As a
result, we have installed ultrasonic
ow meters on each are header,
with related pressure and tempera-
ture transmitters. Along with ow,
temperature and pressure measure-
ment, we have started to take
samples twice a week from each
are knock-out drum in order to
dene the complete composition of
the gas streams.
The renery are system needs to
be monitored in real time.
Additionally, eld inspections
should be carried out every shift by
eld operators to decrease gas and
steam consumption.
The very poor condition of refrac-
tory in the ares at Tpra Izmir
renery resulted in high casing
temperatures and numerous are
gas leaks through the casing. Safety
on platforms around the are is not
very secure as a result of unburned
are gas leakage and hot are tip
casing. If the are pilots, are tip
and refractory are repaired, the ef-
ciency of the are should increase
signicantly due to reduced atomi-
sation steam losses. The installation
Molecular seal drum
Drain
Soot formation
Figure 4 Molecular seal plugging with soot and refractory
Figure 5 Flare tip Figure 6 Flare, pilots view
tupras.indd 4 10/03/2014 13:56
7130 South Lewis, Suite 335, Tulsa, OK 74136, call 1-918-496-7599 or visit our website www.callidus.com
2014 Honeywell International Inc. All rights reserved.
combustion leader
Callidus Technologies, LLC is an established world leader in
process burner, flare and thermal oxidizer technology.
We have earned that reputation by providing solutions that set new
standards in performance and reliability. Our approach to combustion
science continues to break new ground every day with innovative new
products and solutions for the worlds most difficult combustion
challenges. This commitment to technical excellence is driven by our
focus to provide a custom designed solution for every project based on each customers specific
requirements. One example is our flare gas recovery system, designed to achieve zero flaring, reducing
process fuel costs while eliminating visible flame, odors and auxiliary flare utilities.
Having one of the largest installed bases of process burners, production flares and thermal oxidizers in
refining and petrochemical facilities around the world. Callidus should be your choice for next generation
combustion equipment and solutions for new construction or retrofit operations with installation
capabilities anywhere in the world.
uop callidus.indd 1 10/03/2014 11:13
Air pre-heater improves energy efciency
I
n the search for energy effciency
in the refning industry, waste
heat recovery from fue gas is one
of the most interesting sources of
hidden energy to look at. There are
three reasons why a stack should
be one of the frst locations to look
at in the search for energy savings.
The frst reason is that at a stack
location signifcant amounts of
primary energy are converted into
heat at one single location. The
second reason is that fue gas is
discharged into the atmosphere at
relatively high temperatures
between 150C and 250C. And
fnally, the outlet for the recovered
waste heat is nearby as combustion
air with a suffcient low tempera-
ture to absorb the excess heat.
Energy effciency improvements by
up to 5% can be realised at all
kinds of industrial steam boilers,
furnaces and fred heaters, even in
the case when air pre-heating is
already applied to some extent.
The current metal air pre-heaters
are designed for a minimum fue
gas exit temperature of approxi-
mately 160C in order to prevent
corrosion and subsequent high
maintenance cost. Flue gas originat-
ing from sulphur containing fuel
(for instance, refnery gas, fuel) has
an acid dew point at around 130C.
For this reason, many existing air
pre-heater systems incorporate a
steam heater in order to pre-heat
ambient combustion air to a mini-
mum temperature that prevents
corrosion in the air pre-heater.
The Dutch company HeatMatrix
Group recently developed a new
generation air pre-heater that
enables heat recovery from corro-
sive and/or fouling fue gas
A corrosion resistant air pre-heater enables valuable heat recovery from
corrosive and fouling renery ue gas streams
BART VAN DEN BERG
HeatMatrix
streams. This exchanger contains
polymer tube bundles that are
resistant to corrosion by concen-
trated sulphuric acid at elevated
temperature. The characteristics of
this exchanger and its case study
based performance are outlined in
this article.
Acid dew point related corrosion
During combustion, the sulphur
component of sulphur contami-
nated fuel is converted into sulphur
dioxide and trioxide. The sulphur
trioxide condenses in the presence
of water vapour at a dew point
temperature, which is a function of
the partial pressure of sulphur
trioxide and water (acid dew
point). At this dew point, a frst
small amount of highly concen-
trated sulphuric acid precipitates,
for example on the air pre-heater
surface. In an air pre-heater, the
skin temperature of the heat
exchanging surface at the fue gas
side (wall temperature) is leading
in this process. The bulk tempera-
ture of the fue gas can still be
signifcantly higher. Detailed infor-
mation on acid dew point
temperature calculations can be
found elsewhere.
1
When the temperature of the fue
gas is further reduced beyond the
acid dew point the concentration of
sulphuric acid is also reduced, as
well as the corrosiveness. Below
90C, the corrosiveness of the fue
gas is signifcantly lower compared
to the corrosiveness just below the
acid dew point temperature (see
Figure 1). From a variety of exotic
metals, only tantalum can with-
stand acid dew point corrosion at
acid dew points higher than 150C
(see Figure 2).
2
The polymer that is
applied for the HeatMatrix polymer
tubes is resistant to acid dew point
concentrations up to 150C and has
a design temperature of 200C.
High temperature acidic fue gas
crosses the acid dew point close to
the tip of the stack as result of cool-
ing by ambient air. This lost energy
can be recovered when the right
R
a
t
e

o
f

c
o
r
r
o
s
i
o
n
40 60 80 100 120 140 160
Metal surface temperature, C
Water dew point
Acid dew point
Figure 1 Relative corrosiveness of sulphuric acid along the boiling curve
heatmatrix.indd 1 10/03/2014 13:58
shell. The smaller size exchangers
have a cylindrical shape, as Figure 3
shows, and the larger size air
pre-heaters have a container shape
to accommodate fue gas fows up
to 500 000 kg/hr.
Installation options
For grass roots installations with a
fue gas temperature below 200C,
integration of the polymer air
pre-heater is straightforward. For
installations with a fue gas temper-
ature above 200C, a combination
between a metal air pre-heater and
polymer air pre-heater in series is
required. This hybrid design has
the following advantages:
Increased heat recovery over a
wide temperature range
The polymer air pre-heater
protects the metal air pre-heater
against low air temperatures that
lead to cold spot corrosion
problems
The metal air pre-heater protects
the polymer air pre-heater against
high temperatures
A steam air pre-heater for raising
the temperature of the combustion
air is no longer necessary with this
hybrid air pre-heater design.
Addition of a polymer air
pre-heater to existing installations
will be a proftable investment as
well. Existing civil and steel struc-
tures frequently have suffcient
over-design to accommodate an
additional lightweight exchanger.
Also, for boilers with a large
distance between stack and
combustion air induced draft fan
solutions can be provided. For this
case, a twin coil system comprising
a polymer fue gas exchanger and a
simple fnned tube exchanger is
recommended.
Case study
The following typical case is based
on the performance of multiple
projects that have been realised
over the past years. A fue gas fow
from a large steam boiler of 100000
kg/hr at a temperature of 170C
enters the polymer air pre-heater
and is cooled to 85C by 95000 kg/
hr combustion air at 15C. The
recovered energy is 2.6 MW, which
is approximately 5% of the steam
boiler duty.
heat exchanger materials are
applied leading to an improved
energy effciency.
Polymer heat exchanging
tube bundles
The HeatMatrix air pre-heater
consists of multiple corrosion resist-
ant tube bundles contained in a
single metal shell or housing, which
is made corrosion resistant through
a coating or polymer liner. The
proprietary polymer bundle design
consists of multiple tubes that are
connected to each other over almost
the full length of the tube. This
structure creates a strong rigid
matrix grid that is able to resist
high gas velocities and ther-
mo-shocks. As opposed to polymer
hose or glass tube designs, the
connected polymer tube bundles
are not sensitive to breakage or
rupture. The connector between the
individual tubes creates simultane-
ously a counter-current fow
confguration between the two
fuids. This confguration improves
the heat transfer by up to 20%
compared to cross fow type
exchangers (see Figure 3).
Flue gas fows from top to bottom
through the tubes (red arrow) and
combustion air fows in the oppo-
site direction around the tubes
(blue arrow). The top end of the
polymer tube bundles is fxed to
the upper tube sheet and the lower
end is allowed to expand in a seal-
ing system connected to the lower
tube sheet. The extra tube sheet in
the middle of the exchanger
prevents bypassing and directs
combustion air into the polymer
tube bundles.
The inlets and outlets of the
exchanger are located at the side of
the heat exchanger in order to
allow easy access to the polymer
tube bundles. These lightweight
bundles are retractable from the top
and can be cleaned or replaced
without demounting the complete
exchanger. In the case of fouling
fue gas, each bundle can be
equipped with a spraying nozzle,
which thoroughly cleans each
bundle in an alternating cleaning
sequence (see Figure 4).
The capacity of the air pre-heater
is fully scalable by placing several
polymer bundles in parallel in a
150
250
300
200
100
50
T
e
m
p
e
r
a
t
u
r
e
,

C
0
0 10 20 30 40 50 60 70 80 90 100
H
2
SO
4
concentration, wt%
Zirconium
Titanium
Boiling point Hastelloy B-3
Hastelloy B
Tantalum
Niobium
Hastelloy C
Figure 2 Corrosion resistance of several high alloy metals
2
Figure 3 Cross section of a HeatMatrix air
pre-heater
Iso-corrosion chart - H
2
SO
4
5 mpy (0.13 mm/y)
Conclusion
Energy efciency and carbon abate-
ment are currently hot topics and
energy efciency is seen as a most
important contributor in every
governmental strategy to reduce
fossil fuel consumption. In that
light, ue gases should be seen as
an important source of hidden
energy because ue gas still
contains 5 to 10% of the primary
energy used to drive the combus-
tion process. New technologies like
this polymer air pre-heater can
contribute to improve energy ef-
ciency throughout the rening
industry.
References
1 Hijbrechts W W M, Leferink R, Latest
advances in the understanding of acid dew
point corrosion: Corrosion and stress corrosion
cracking in combustion gas condensates, Anti-
Corrosion Methods And Materials, 2004, Vol 5,
issue 3,173-188.
2 Gambale D, Heat exchangers for hot acids:
material selection, Chemical Engineering, July
2011.
Bart van den Berg is a co-founder of
HeatMatrix Group and has spent 18
years in various positions throughout the
petrochemical industry. He holds a masters
degree in chemical engineering from Twente
University, The Netherlands.
Figure 4 HeatMatrix air pre-heater equipped with an in situ cleaning system
To keep pace with the demanding
quality requirements of modern
fuels, advanced, precise and easy to
use analytical technology is required.
With a complete range of XRF and
ICP spectrometers, SPECTROs
unique solutions for at-line and
laboratory elemental analysis
are capable of meeting the most
demanding product specification
testing requirements.
Elemental Analysis
of Fuels and Oils
Discover more exciting details at
www.spectro.com/fuels
Determination of Sulfur and other
elements at-line and in the laboratory
Discover more at www.metso.com/oilandgas/fowcontrol
Neles Jamesbury Mapag
Look what goes into a Metso valve. It starts with a long track record of delivering engineered
performance and legendary reliability with premier products such as Neles, Jamesbury
and Mapag. But the numbers really paint the picture. In almost 90 years, Metso has delivered
globally millions of valves, control valves and on-off valves. We have also become one
of the leading suppliers of smart positioners. All backed by feld service expertise from over
55 automation service hubs and over 30 valve service centers around the world. We see it
this way: keeping oil and gas producers working safely and reliably protects investments,
people and the planet.
There is more
to Metso than
meets the eye.
metso.indd 1 06/03/2014 13:52
Maximising the use of process energy
T
he crude distillation unit has
the highest processing capac-
ity in a refnery. It is highly
energy intensive and consumes
around 2% of the total crude
processed to meet its energy
requirement.
1
Various methods
have been reported for better heat
recovery, such as the application of
pinch analysis and exergy analysis,
and for achieving energy savings,
for instance the installation of a
fash drum or pre-fractionation
column.
2,3-8
The energy requirement
of a pre-fash integrated crude
distillation unit is slightly lower
than for the design without a
pre-fash device.
5,6
However, yields
of valuable distillate products from
the atmospheric distillation column
are decreased
3,5
and the yield of
residue is increased. Processing of
more residues in a vacuum column
requires a signifcantly larger
Integrating a ash drum in an atmospheric and crude distillation unit offers the
potential for savings and capacity enhancement
SUNIL KUMAR, SHRIKANT NANOTI and M O GARG
CSIR-Indian Institute of Petroleum
diameter column as one pound of
vaporised oil at 10 psi pressure
occupies 70 times more space than
at atmospheric pressure. Energy
consumption is also increased due
to greater residue fow and the
requirement to generate more
vacuum.
9

Benali et al
7
reported a decrease
in exergy destruction up to 14% as
a whole and more than 21% in the
furnace alone due to integration of
a pre-fash drum with the atmos-
pheric distillation unit. It is also
reported that the overall change in
the energy requirement for a
pre-fash integrated design is
marginal; however, the level of this
energy changes signifcantly.
8

Substantial amounts of energy can
be saved utilising low level heat.
However, the fnding that low level
heat utilisation can increase the
crude temperature to the furnace is
to be read with utmost caution.
This will be true only if the pinch
temperature is below the tempera-
ture of this low level energy. It is
essential to note that application of
low level energy below the pinch
point will rather increase the cold
utility requirement without saving
high level energy (furnace duty).
2,10-12

In view of the energy intensiveness
and carbon footprint of a crude
distillation unit, it is important to
emphasise that a small improve-
ment in its energy effciency will be
of great importance.
This study focused on the quanti-
tative evaluation of savings in
furnace duty, carbon dioxide emis-
sions and fuel expenditure, and the
capacity for potential enhancement
of an atmospheric and vacuum inte-
grated crude distillation unit by
installation of fash drum, without
compromising on quality and yields
Crude
oil
Sour
water
Water
Water
Naphtha
Steam
Heavy
naphtha
Steam
Kerosene
Steam
Steam
Diesel
VD
Steam Steam
LVGO
HVGO
Vapour
to ejector
Vacuum
residue
HEN
HEN
HEN
PA1
PA2
PA3
HEN
Condenser
Desalter
Furnace
Furnace
Vacuum
distillation
column
Atmospheric
distillation
column
HEN
HEN
Flash
zone
Flash
zone
Residue
Figure 1 Simplied schematic of atmospheric and vacuum column integrated crude distillation unit
IPP.indd 1 10/03/2014 14:05
is fed to the desalter to remove
water soluble salts from the crude.
The desalted crude enters another
heat exchanger network and
receives heat from hot streams. The
preheated crude enters the furnace.
Partially vaporised crude from the
furnace is fed to the fash zone of
the atmospheric column. The
vapour from the fash zone moves
upwards in the column whereas
liquid falls to the bottom of the
column. The vapour is then frac-
tionated into distillate products
such as naphtha, kerosene and
diesel in the upper section of the
column. To recover heat at a differ-
ent temperature, heavy naphtha,
kerosene and diesel pumparound
circuits are used along with the
overhead condenser. The distillate
products withdrawn from different
trays of the column are then
stripped by steam in their respec-
tive side strippers for removing
lighter components to meet ASTM
D-86 distillation specifcations. The
liquid falling downwards from the
fash zone is stripped out using
steam at the bottom of the column
to remove lighter, diesel-range
material from the residue.
The residue from the atmospheric
distillation column is fed to the
furnace along with the coil steam.
Partially vaporised crude from the
furnace is fed to the fash zone of
the vacuum column. The vapour
from the fash zone is then fraction-
ated into distillate products such as
vacuum diesel, light vacuum gas
oil and heavy vacuum gas oil. The
liquid from the fash zone is
stripped using the steam at the
bottom. Heat at a different temper-
ature from the vacuum column is
recovered using the vacuum diesel,
light vacuum gas oil and heavy
vacuum gas oil pumparound
circuits. The combined fow of
pumparound and product is taken
from the specifc column tray and
cooled to the desired pumparound
return temperature. The combined
stream is then divided into product
and pumparound streams. The
pumparound stream is returned to
the upper tray in the column. The
product stream further exchanges
heat with the cold streams and is
fnally cooled in a water cooler to
of distillate products. Pinch technol-
ogy, a well-proven tool to minimise
utility consumption in a crude distil-
lation unit, was used to estimate the
utility targets and excess process
heat available in the process. Insight
from the principles of pinch technol-
ogy was applied to create the scope
for utilising excess process heat to
save furnace duty. Area targeting
for each case considered was also
carried out to understand the quali-
tative effect of energy savings on a
heat exchanger network.
Conventional crude distillation unit
It is essential to establish a base
case to understand the quantitative
benefts of any improvement or
modifcation in the process. In this
study, an atmospheric and vacuum
column integrated crude distillation
unit with a capacity of 5.0 million
t/y was selected to represent the
base case. This base case was used
to establish the basis for yields of
distillate product and their quality,
and for separation criteria for prod-
ucts obtained in other cases to be
studied.
Description
Figure 1 is a simplifed schematic of
a conventional crude distillation
unit. Crude is frst heated in a heat
exchanger network using the hot
products and pumparound streams
before entering the desalter. Water
Specic gravity @ 15C 0.8282
Light end analysis
Component Volume% Component Volume%
Methane 0.000 n-Butane 1.242
Ethane 0.090 i-Pentane 1.000
Propane 0.414 n-Pentane 1.850
Iso Butane 0.414

TBP distillation range of crude oil
Volume% Temperature, C Volume% Temperature, C
5.0 40.7 50.0 267.7
10.0 75.6 70.0 379.3
30.0 162.1 90.0 475.9
40 217.1 95.0 535.2
Crude properties
Table 1
Crude ow, tonne/hr 62.5 Total number of actual trays 40.0
Crude entry at tray number 6.0 Striper trays efciency 0.4
Flash zone pressure, kg/cm
2
a 3.254 Column top pressure, kg/cm
2
a 2.50
Efciency for different sections of column
From To Efciency From To Efciency
40.0 32.0 0.65 18.0 15.0 0.55
31.0 29.0 0.30 12.0 14.0 0.30
28.0 22.0 0.55 7.0 11.0 0.55
21.0 19.0 0.30 6.0 1.0 0.40
Pumparound detail
Draw stage Return stage Flow, m
3
/h Temperature drop, C
Heavy naphtha 29__Main TS 31__Main TS 349 976.6 60.0
Kerosene 19__Main TS 21__Main TS 539 188.7 40.0
Diesel 12__Main TS 14__Main TS 614 544.8 50.0
Strippers and stripping steam detail
No of actual Liquid draw Vapour return Stripping steam,
stages in stripper stage stage kg/hr
Heavy naphtha 6.0 29__Main TS 31__Main TS 1624.0
Kerosene 6.0 19__Main TS 21__Main TS 2117.0
Diesel 4.0 12__Main TS 14__Main TS 2066.0
Column bottom 8500.0
Parameters used in simulation of atmospheric distillation column
Table 2
IPP.indd 2 10/03/2014 14:06
attain the desired run-down
temperature.
Simulation
An atmospheric and vacuum inte-
grated crude distillation unit,
selected in the study to represent
the base case, is shown in Figure 1.
Simulation was carried out using
Aspen Hysys. The Grayson-Streed
thermodynamic model was used in
a simulation for the prediction of
stream properties.
3,13-14
The proper-
ties of crude used in the study are
given in Table 1. The details of
important information/parameters
used in the simulation of atmos-
pheric and vacuum distillation
columns are given in Table 2 and
Table 3 respectively. Crude was
heated to a typical furnace coil
outlet temperature of 364.5C.
Partially vaporised crude was fed
to the fash zone of the atmospheric
distillation column containing 40
trays. Stripping steam (8500 kg/hr)
was used at the column bottom to
obtain liquid distillate yields of
66.11 vol% predicted from the true
boiling point curve of the crude,
with a fnal boiling point tempera-
ture of 370C.
The temperatures corresponding
to 95 vol% of top product, heavy
naphtha, kerosene and diesel were
fxed at the values 110C, 160C,
245C and 370C respectively, to
predict the distillate yield of each
product. The values of stripping
steam used in the strippers to
remove the lighter material are
given in Table 2.
Residue obtained from the
bottom of the atmospheric distilla-
tion column was mixed with steam
and an ethane-propane mixture
(4:1). This mixed stream was heated
in the furnace to 365C and fed to
the fash zone of the vacuum distil-
lation column. Vacuum diesel, light
vacuum diesel and heavy vacuum
diesel streams were drawn from
the column.
Energy targeting
Pinch technology is a proven tool
for the estimation of minimum hot
and cold utility targets in a
process.
2,10
The technology is based
on the frst and second laws of ther-
modynamics and is straightforward
to implement in the design and
revamping of plant operations to
minimise energy consumption.
Composite and grand composite
curves are generated in pinch tech-
nology to estimate hot and cold
utility targets and their temperature
level. A pinch point in the compos-
ite curve divides the whole process
into two sections. The section above
the pinch point is in heat defcit and
requires hot utilities to heat the cold
streams to their target temperature,
whereas the section below the pinch
point is in heat surplus and requires
cold utilities to cool the hot streams
to their target temperature. The
pinch point temperature is
governed by the enthalpy of the hot
and cold streams. The grand
composite curve provides an esti-
mate of the amount of excess heat
available below the pinch and its
temperature level.
Table 4 lists the enthalpies,
supply and target temperatures for
cold and hot streams which were
extracted from the simulation fle.
The composite and grand compos-
ite curves shown in Figure 2 were
Residue ow, std. ideal liquid volumetric ow, m
3
/h 257.4
Flash zzone pressure/temperature, mmHg/C 49/365
Column top/ bottom pressure, mmHg 26/65
Residue entry tray number 3.0
Total number of theoretical trays 14.0
Bottom stripping steam, kg/hr 600.0
Furnace coil steam, kg/hr 600.0
Light ends, std. ideal liquid volumetric ow, m
3
/h 5.0
Pumparound detail
Pumparound name Draw stage Return stage
Vacuum diesel 13__TS-1 14__TS-1
LVGO 8__TS-1 9__TS-1
HVGO 6__TS-1 7__TS-1
Pumparound name Flow, m
3
/h Temperature drop, C
Vacuum diesel 100.0 115.0
LVGO 113.0 86.9
HVGO 350.0 63.7
Product detail
Product name Draw stage Flow, m
3
/h
Vacuum diesel 13__TS-1 39.8
LVGO 8__TS-1 92.0
HVGO 6__TS-1 68.9
Parameters used in simulation of vacuum distillation column
Table 3
Stream name Supply T, C Target T, C Enthalpy, million kcal/hr
Crude 35.0 290.0 96.65
Crude 290.0 364.5 53.56
Heavy naphtha 130.1 40.0 2.68
Kerosene 181.0 40.0 7.72
Diesel 265.5 40.0 21.16
Heavy naphtha PA 148.3 88.3 -12.58
Kerosene PA 198.0 158.0 -13.87
Diesel PA 283.0 233.0 -21.20
Vacuum diesel 165.0 50.0 2.14
Light vacuum gas oil 267.9 90.0 8.85
Heavy vacuum gas oil 308.7 90.0 8.43
Slop 357.2 120.0 1.62
Vacuum residue 349.9 120.0 6.46
Vacuum diesel, PA 165.0 50.0 -5.39
LVGO, PA 267.9 181.0 -5.66
HVGO, PA 308.7 245.0 -13.85
Enthalpy and temperature of cold and hot streams of base case
Table 4
IPP.indd 3 10/03/2014 14:06
Flash drum integrated crude
distillation unit
The vaporisation of the lighter frac-
tion of crude and its vapour routing
to the column above the fash zone
can be facilitated through installa-
tion of a fash drum below the
pinch point. Besides, the loss of
carrier effect due to separation of
the lighter fraction of crude in the
fash drum and the lower energy
input to the column as it bypasses
the furnace can be compensated by
increasing the stripping steam,
increasing the coil outlet tempera-
ture of the furnace, and adjusting
the pumparound return tempera-
ture/duty to maintain the distillate
yields as established in the base
case.
To investigate the quantitative
benefts of fash drum integration
with the atmospheric and vacuum
column in a crude distillation unit
without compromising on the
distillate yields and their quality,
the present study considered two
cases:
Flash drum case 1 (FD-1): the
bottom stripping steam was
increased to the required amount
Flash drum case 2 (FD-2): the coil
outlet temperature was increased
by 5C, then bottom stripping
steam was increased to the required
amount.
Simulation of ash drum cases
Changes made to the operating
parameters in simulations of fash
drum cases, keeping other parame-
ters the same as were used in
simulation of the base case, are
described below.
FD-1 case
Crude was heated to 215C then
fed to the fash drum at 4.2 kg/
cm
2
a pressure. Vapour from the
fash drum was fed to the column
at the 18th tray from the bottom.
The vapour entry location was
selected on the basis of the temper-
atures of the vapour and the feed
tray in the column. Flashed crude
was heated to 364.5C. The strip-
ping steam was increased to a
value of 12 200 kg/hr, against the
8500 kg/hr used in the base case, to
match the distillate yields value
established in the base case.
using the excess energy available
below the pinch point and avoiding
the superheating of vapour of
lighter distillates in the furnace.
The heat of vaporisation of crude is
around 100 times its specifc heat.
Thus, crude vaporisation below the
pinch point will decrease the
enthalpy of the cold stream above
the pinch and the enthalpy of hot
stream duty below the pinch.
Moreover, direct routing of vapour
to the column also eliminates the
energy requirement for superheat-
ing in the furnace.
generated using pinch software.
The composite curve indicates that
36.1 million kcal/hr of excess heat
is available below the pinch point
during the process. The grand
composite curve shown in Figure 3
indicates that out of 36 million
kcal/hr of heat, 17.8 million kcal/
hr is available in the temperature
range 188-234C. This suggests use
of the plus minus principle of
pinch technology to reduce the util-
ity targets. The plus minus
principle suggests that the utilities
target of a process can be changed
by modifying the enthalpy of hot
and cold process streams above
and below the pinch point. That is,
the hot utility target reduces if the
cold stream enthalpy above
the pinch decreases, and the cold
utility target reduces if the hot
stream enthalpy below the pinch
decreases.
This desired change in enthalpy
of the hot and cold streams above
or below the pinch point is made
possible by vaporising the crude
0 5 10 15 20 25 30 35 40 45 50 55 60
Enthalpy, MMkcal/hr
300
400
350
250
200
150
100
50
T
e
m
p
e
r
a
t
u
r
e
,

C
0
T=374.8
Q=54.7
T=233.7
Q=17.9
T=188.3
Q=17.8
300
400
350
250
200
150
100
50
T
e
m
p
e
r
a
t
u
r
e
,

C
0
0 20 40 60 80 100 120 140 160 180 200
Enthalpy, MMkcal/hr
T=364.5
Q=186.4
T=286.3
Q=131.5
T=357.5
Q=131.5
T=309.0
Q=130.0
T=283.3
Q=122.6
T=233.7
Q=82.6
T=199.1
Q=72.7
T=283.3
Q=130.0
T=263.6
Q=122.6
T=158.4
Q=82.6
T=132.1
Q=72.8
T=35.0
Q=36.2
T=134.5
Q=36.2
T=89.1
Q=11.4
Hot
Cold
T=29.9
Q=36.1
Figure 2 Composite and grand composite curves for the base case
Direct routing of
vapour to the column
eliminates the energy
requirement for
superheating in the
furnace
IPP.indd 4 11/03/2014 14:31
the heat exchangers were extracted
from the hot composite curve.
Temperatures for the cold streams
for process-to-process heat
exchangers were extracted from the
process-to-process heat exchangers
and the subsequent two sections
represent cold utility heat exchang-
ers. The temperatures of the hot
streams and the enthalpy values for
FD-2 case
In this case, fashed crude was
heated to 369.5C, then stripping
steam was increased to 10 200 kg/
hr, keeping the crude temperature
to the fash drum, the pressure of
the fash drum and the vapour
entry location the same as in the
FD-1 case.
In both cases, the return tempera-
tures of the pumparounds and the
stripping steam in the kerosene and
diesel stripper were adjusted to
meet the separation criteria for
distillate products. The parameters
of the vacuum column simulation
were not changed because the fow
rate of residue feeding the vacuum
column was the same as established
in the base case.
Energy targeting of ash drum cases
The enthalpies of the cold and hot
streams and their supply and target
temperatures taken from the
conversed simulation fle for each
case are given in Table 5. Hot and
cold utility targets were estimated
from composite and grand compos-
ite curves generated using pinch
software with the DT minimum
value of 20C. The composite
curves generated for the FD cases
are shown in Figure 3.
Area targeting
Figures 2 and 3 indicate that vapori-
sation of crude below the pinch
point decreases the delta tempera-
ture difference between the cold
and hot composite curves; this will
lead to a decrease in the tempera-
ture gradient across the heat
exchangers. This also results in an
increase in the area required in the
heat exchangers network in the FD
cases. The effect of this negative
impact on the capital cost of the
heat exchanger network can be
captured by estimating the UA
values for each case using the
assumption of the same heat trans-
fer coeffcients for hot and cold
streams. The composite curves
were divided into seven sections
depending upon the change in
slope of the hot and cold composite
curve to represent heat exchangers
with vertical and countercurrent
heat transfer. The fve sections on
the right hand side represent the
300
400
350
250
200
150
100
50
T
e
m
p
e
r
a
t
u
r
e
,

C
0
0 20 40 60 80 100 120 140 160 180 200
Enthalpy, MMkcal/hr
T=280.9
Q=123.2
T=357.5
Q=123.2
T=279.2
Q=113.3
T=308.4
Q=121.7
T=194.9
Q=68.7
T=239.1
Q=81.1
T=278.0
Q=121.7
T=260.0
Q=113.3
T=196.7
Q=81.1
T=161.4
Q=68.7
T=35.0
Q=23.6
T=115.4
Q=23.6
T=96.8
Q=13.6
Hot
Cold
300
400
350
250
200
150
100
50
T
e
m
p
e
r
a
t
u
r
e
,

C
0
0 20 40 60 80 100 120 140 160 180 200
Enthalpy, MMkcal/hr
T=280.4
Q=122.2
T=356.9
Q=122.1
T=279.2
Q=112.0
T=308.4
Q=120.6
T=195.5
Q=69.2
T=241.5
Q=81.5
T=277.4
Q=120.6
T=258.8
Q=112.0
T=200.3
Q=81.5
T=165.6
Q=69.2
T=35.0
Q=22.7
T=113.0
Q=22.7
T=96.2
Q=13.4
Hot
Cold
Figure 3 Composite curves for FD-1 and FD-2 cases
FD-1 case FD-2 case
Stream name Supply T, Target T, Enthalpy, Supply T, Target T, Enthalpy,
C C million kcal/hr C C million kcal/hr
Crude 35.0 215.0 64.40 35.0 215.0 64.40
Crude 203.5 223.7 75.42 203.5 220.1 77.90
Heavy naphtha 129.0 40.0 2.64 129.5 40.0 2.66
Kerosene 176.0 40.0 7.49 176.4 40.0 7.53
Diesel 253.4 40.0 19.70 254.2 40.0 19.89
Heavy naphtha PA 146.0 96.0 -10.53 147.0 97.0 -10.55
Kerosene PA 194.0 154.0 -13.79 195.0 156.0 -13.47
Diesel PA 278.0 241.0 -15.73 279.0 239.0 -17.00
Vacuum diesel 154.6 50.0 1.96 154.6 50.0 1.96
Light vacuum gas oil 267.9 90.0 8.85 267.9 90.0 8.85
Heavy vacuum gas oil 308.7 90.0 8.43 308.7 90.0 8.43
Slop 357.2 120.0 1.62 357.2 120.0 1.62
Vacuum residue 349.9 120.0 6.46 349.9 120.0 6.46
Vacuum diesel, PA 165.0 50.0 -5.39 165.0 50.0 -5.39
LVGO, PA 267.9 181.0 -5.66 267.9 181.0 -5.66
HVGO, PA 308.7 245.0 -13.85 308.7 245.0 -13.85
Enthalpy and temperature of cold and hot streams of FD-1 and FD-2 cases
Table 5
IPP.indd 5 10/03/2014 14:06
sandvik.indd 1 10/03/2014 11:12
requirement, to ensure smooth
operation of the column.
The vapour profle indicates that
maximum vapour fow rate in the
base case is signifcantly higher
than the maximum vapour rate
obtained in the FD cases. This
implies that fash drum integration
creates scope for increasing the
stripping steam fow rates in an
existing atmospheric distillation
column without disturbing the
columns operation. Moreover, in a
grass roots scenario this also
provides the opportunity to reduce
the diameter of crude distillation
columns.
maximum value of vapour fow
rate throughout the column. In the
FD cases, either stripping steam or
the coil outlet temperature along
with stripping steam were
increased to maintain the yield and
quality of distillate products
obtained in the base case. The
increase in stripping steam, furnace
coil outlet temperature gives an
impression that it can lead to food-
ing in the column. Therefore,
vapour traffc profles of the atmos-
pheric distillation column for all
cases were generated (see Figure 4)
to understand the effect of fash
drum integration on the diameter
cold composite curves whereas for
utility exchangers it was assumed
that the cold stream is heated from
32C to 40C. Equations 1 and 2
were used to estimate the UA value
for process-to-process and utility
heat exchangers. The overall UA
for each case was calculated using
Equation 3:
(1)
(2)

(3)
Where:
LMTD is log mean temperature
difference
p represents process-to-process heat
exchangers
c represents cooling stream and
process heat exchangers
I is the number of a heat exchanger.
Comparison of base and FD cases
Distillate products
Table 6 lists the yields, D-86 5% and
95% points and (5-95) gaps for
distillate products obtained for all
cases. D-86 5% and -95% points
imply the quality of distillate prod-
ucts and the (5-95) gap indicates
the quality of separation between
adjacent products.
The results shown in Table 6 indi-
cate that there is only slight
variation in the distillate product
yields for the proposed cases in
comparison to the base case. Almost
the same value of atmospheric resi-
due fow rate in all cases ensures no
change in the yields of distillate
products obtained from the vacuum
column. The lower values of the
D-86 (5-95) gap obtained in the FD
cases suggests better fractionation
among the products. It can be
concluded that product yields and
their quality were not compromised
in a fash drum integrated crude
distillation scenario.
Vapour trafc prole
The diameter of the distillation
column is estimated using the
610
4
710
4
510
4
410
4
310
4
210
4
10
4
A
c
t
u
a
l

v
a
p
o
u
r

v
o
l
u
m
e
t
r
i
c

f
l
o
w
,

m
3
/
h
r
0
0 5 10 15 20 25 30 35 40
Number of trays
FD-1 case
Base case
FD-2 case
Figure 4 Vapour trafc prole of atmospheric distillation column
Distillate products yields, M
3
/hr
Products Base case FD-1 case FD-2 case
Top product 129.9 130.2 130.2
Heavy naphtha 68.2 68.1 68.1
kerosene 118.7 120.0 119.2
Diesel 183.1 181.7 182.6
Atmospheric residue 257.4 257.3 257.3
D86 5% point, C
Heavy naphtha 110.2 110.2 110.2
kerosene 155.4 155.5 155.4
Diesel 232.0 232.2 231.7
D86 95% point, C
Heavy naphtha 110.0 110.0 110.0
kerosene 160.0 160.0 160.0
Diesel 245.0 245.0 245.0
Product quality: D86 (5-95 gap), C
Top product-heavy naphtha 0.2 0.2 0.2
Heavy naphtha-kerosene -4.6 -4.5 -4.6
Kerosene-diesel -13.0 -12.8 -13.3
Diesel-atmospheric residue -34.2 -30.4 -31.6
Products yields and their separation gaps
Table 6
UA
!"
=
LMTD
!"
Q
!"

UA
!"
=
LMTD
!"
Q
!"

!"
!"#$%&&
= UA
!"
!
!!!
+ UA
!"
!
!!!

UA
!"
=
LMTD
!"
Q
!"

UA
!"
=
LMTD
!"
Q
!"

!"
!"#$%&&
= UA
!"
!
!!!
+ UA
!"
!
!!!

UA
!"
=
LMTD
!"
Q
!"

UA
!"
=
LMTD
!"
Q
!"

!"
!"#$%&&
= UA
!"
!
!!!
+ UA
!"
!
!!!

IPP.indd 6 10/03/2014 14:06
GET A RETURN ON INVESTMENT
WITHIN 24 HOURS
Using ITW Online Cleaning on a proactive way will help improve
TURNAROUND OPERATIONS and OPERATIONAL EXCELLENCE
An entire Process Unit can be ITW Online Cleaned in as low as 24 hours on an oil-to-oil basis. Our cleaning technology will
clean all of the Units equipment quickly, safely and environmental friendly.
To recover a Units performance, ITW Online Cleaning can be applied in a Turnaround during shutdown operations, to
reduce/eliminate the need for mechanical cleaning, or during the run.
Furthermore, use of our cleaning technology eliminates the need to open or enter hazardous process equipment.
ITW Online Cleaning is the only technology which has an immediate Return On Investment (ROI).
By using ITW Online Cleaning during the run, to recover Units performance, at Unit start-up (just after 24h, ITW
Technology has been applied), all the benefits are immediately monetized in terms of:
- improved throughput
- improved energy efficiency
- improved product quality
- reduced emission taxes
Regular application of ITW Online Cleaning will target an increased run length under clean conditions, with a far higher
ROI.
For Turnaround applications, a further ROI increase can be added by applying ITW Improved Degassing/Decontamination
to achieve quick and effective safe entry conditions.
EVALUATE ITW ONLINE CLEANING 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
Join ITW Team worldwide and send your Curriculum Vitae to : jobs@itwtechnologies.com
ITW
Innovative
Technologies
Worldwide
Now hiring
Professionals
Worldwide
itw.indd 1 11/9/12 11:45:38
for natural gas and from $3.287
million to $5.479 million for fuel oil
(see Table 7).
Area requirement
The UA values of heat exchangers
from $/t to $/million Btu.
15
The
sensitivity analysis is shown in
Figure 6. Fuel cost savings increased
with increase in fuel costs. The
saving in fuel expenditure varied
from $2.739 million to $4.565 million
Hot and cold utilities and
CO
2
emissions
The estimated minimum hot and
cold utility targets for all cases are
given in Table 7. The savings in
utilities and CO
2
emission were
estimated using their normalised
values (see Figure 5). Normalised
values were estimated by selecting
the hot and cold utility targets for
the base case (BC) at a value of
100%. There is a substantial reduc-
tion, ranging from 17-19% in hot
utilities, 14-15% in cold utilities and
17-18% in CO
2
emission targets in
the FD cases. The savings in utili-
ties are attributed to utilising excess
process heat below the pinch point
by vaporising the crude and avoid-
ing the superheating of vapour
from lighter distillate products in
the furnace. The estimated saving
in hot utility leads to fuel savings
equivalent to 92 039-82 982 million
Kcal/y in the furnace and a reduc-
tion in CO
2
emissions to the value
of 28 76225 932 t/y for the FD
cases.
Fuel expenditure
Natural gas and fuel oil are the
most commonly used fuels in refn-
eries. The choice of fuel is basically
governed by its price and availabil-
ity. Sometimes, costlier fuel is used
in a refnery due to availability and
disposal issues. Therefore, fnancial
savings due to a reduction in hot
utilities were estimated for both
fuels. The base prices of natural gas
and fuel oil were used at the rate of
$6.0/million Btu and $300/tonne
respectively. The thermal effciency
of the furnace was assumed to be
80%.
The respective estimated values of
savings in fuel expenditure for natu-
ral gas and fuel oil are $2.739
million and $3.287 million.
Considering the fuctuation in fuel
prices over a period of time, a sensi-
tivity analysis of savings in fuel
expenditure with respect to fuel
price was also carried out. The price
of fuel was varied from $6 to $10/
million Btu for natural gas and from
$300 to $500/t for fuel oil. The price
range was divided into four sections
using fve equidistant points. A
calorifc value of 10 500 Kcal/kg for
fuel oil was used to convert its cost
80
100
95
90
85
75
70
65
60
55
N
o
r
m
a
l
i
s
e
d

l
e
v
e
l
50
Hot utility Cold utility CO
2
emissions
FD-2 case
Base case
FD-1 case
Figure 5 Normalised percent savings in hot utility, cold utility and CO
2
emission
4.0
6.0
5.5
5.0
4.5
3.5
3.0
2.5
2.0
1.5
F
u
e
l

c
o
s
t

s
a
v
i
n
g
s
,

U
S
$
1
0
6
1.0
1 2 3 4 5
Points representing the fuel-price range
Natural gas
Fuel oil
Figure 6 Sensitivity analysis of fuel prices for nancial savings
Utility type Base case FD-1 case FD-2 case
Furnace duty, MMkcal/hr
b
Hot utility 54.6 40.4 42.6
a
Stripping steam, Kg/hr Heavy naphtha 1624 1624 1624
Kerosene 2117 2617 2617
Diesel 2066 3066 3066
Bottom 8500 12 300 10 200
Total 14 307 19 607 17 507
Total hot utility requirement, millionkcal/hr 61.9 50.4 51.5
Cold duty, million Kcal/hr Condenser duty 34.1 37.1 35.7

c
Cold utility 35.5 22.5 23.5
Total cold utility requirement, MMkcal/hr 69.6 59.6 59.2
a 1000 kg of MP steam (temperature: 240C; pressure: 12kg/cm
2
a) is equivalent to 0.5085 MMkcal/hr.
b Minimum hot utility predicted by pinch analysis for heating the crude to its target temperature in furnace.
c Minimum cold utility predicted by pinch analysis for cooling the hot streams to their target temperature.
Detail of energy consumption
Table 7
IPP.indd 7 10/03/2014 14:06
the higher side and adjusting the
pumparound duty
Though the above changes
require additional energy, the
results still indicate 17% and 16%
reductions in hot and cold utility
requirement with ash drum
integration
The 17% drop in furnace duty
alone results in a reduction in CO
2
emissions to the environment by
17%
The reduction in furnace duty
requires more area in the heat
exchanger network due to the
reduced temperature gradient
across process-to-process heat
exchangers
There is a reasonable reduction in
maximum vapour trafc load in the
atmospheric column which
provides an opportunity for capac-
ity enhancement in the existing
crude unit without a column
revamp, or a reduction in the diam-
eter of the atmospheric distillation
column in a new crude unit design
The substantial reduction in
furnace duty also provides a solu-
tion to debottleneck the furnace to
enhance the capacity of an existing
crude unit
For a typical crude processing
capacity of 5 million t/y, savings in
hot utility alone lead to fuel savings
equivalent to 92 039-82 982 million
Kcal, a reduction in CO
2
emissions
to the value of 28 76225 932 t/y
and nancial savings of $2.739-
4.565 million and $3.2875.479
million for natural gas and fuel oil
respectively, depending on the
price of fuel available in the ren-
ery. These savings will increase
further with a rise in the processing
capacity of the crude unit
In view of the above, it seems
very attractive to revamp the exist-
ing conventional atmospheric and
vacuum column integrated crude
distillation units by adding a new
drum with revised operating
parameters.
References
1 Bagajewicz M, Ji S, Rigorous procedure for
the design of conventional atmospheric crude
fractionation units. Part I: targeting, Ind. Eng.
Chem. Res., 40, 2001, 617626.
2 Milosevic Z, Rudman A, Brown R, Are you
using pinch analysis effectively in your daily
company after factoring out the
change in the installation cost
of a furnace with reduced duty
and an exchanger network with
increased area requirement.
However, at the prima facie level
nancial savings due to the
substantial reduction in furnace
duty seems to be more than the
additional expenditure required
for the additional area of heat
exchanger network.
Conclusion
A rigorous simulation and pinch
analysis of integrated crude distilla-
tion units, with and without a ash
drum, is presented. From this
study, we reached the following
conclusions:
Loss of distillate yield due to
ash drum integration can be offset
by adjusting the stripping steam,
furnace coil outlet temperature to
were estimated for process-to-
process heat exchangers and cool-
ing utility heat exchangers to
understand the effect of utilising
excess process heat on the heat
exchanger area, which is in turn
responsible for the cost of heat
exchanger network. Figure 7
summarises the UA results for all
cases.
It is clear that the total area of
process-to-process heat exchangers
follows the reverse trend of
furnace duty, whereas the trend of
area for cooling stream to process
heat exchangers is the same as the
furnace duty in the cases studied.
The order of the combined area
requirement of the overall heat
exchanger network is FD-1 case >
FD-2 case > base case. This implies
that the area requirement increases
with the utilisation of excess
process heat below the pinch
point. Hence, more investment in
the heat exchanger network is
required for a case with a target of
minimum furnace duty. However,
it should be noted that furnace
costs will also decrease as duty
reduces. Very accurate nancial
savings due to reduced furnace
duty and fuel expenditure can be
estimated by a detail engineering
Base case
54.6
FD-1 case
40.4
FD-2 case
42.6
Base case
1933
FD-1 case
2676
FD-2 case
2644
Base case
833
FD-1 case
677
FD-2 case
689
Base case
2766
FD-1 case
3354
FD-2 case
3334
A
C
B
D
Figure 7 a) Minimum furnace duty, MMkcal/hr; b) UA values for process-to-process heat
exchange; c) UA values for process to cooling stream heat exchange; d) UA values for
combined (b+ c) heat exchange
The total area of
process-to-process
heat exchangers
follows the reverse
trend of furnace duty
IPP.indd 8 10/03/2014 14:07
in the Master of Technology Programme
in Chemical Engineering at IIT Kanpur and
received Ambujas Young Researchers Award.
Shrikant Nanoti is Head of Separation Processes
Division at the Indian Institute of Petroleum,
Dehradun. He has 26 years experience in
the development and scale-up of separation
based technologies, process designing, process
integration and pinch analysis for the petroleum
rening and petrochemical industries, has
published more than 35 research papers, and
has eight patents to his credit. He holds a
bachelors degree in chemical engineering
from Laxminaryan Institute of Technology,
Nagpur, and a doctorate from Indian Institute
of Technology, Roorkee.
M O Garg is Director of the Indian Institute of
Petroleum, Dehradun, a constituent laboratory
of the Council of Scientic & Industrial
Research, and specialises in liquid-liquid
extraction, simulation and modelling, process
integration, advance control and process
conceptualisation. He has 33 years experience
in the rening industry, has developed and
commercialised several technologies and has
received two CSIR Technology Awards as well
as a CSIR Shield for his commercialisation
efforts. He graduated from Indian Institute
of Technology, Kanpur, in the Research &
Development Division of Engineers India Ltd,
holds a PhD in solvent extraction from the
University of Melbourne, Australia, and has 207
papers and 26 patents to his credit.
for the Efcient Use of Energy, Second edition,
2007.
11 Linnhoff March, Introduction to Pinch
technology, www.ou.edu/class/che-design/a-
design/Introduction%20to%20Pinch%20
Technology-LinhoffMarch.pdf
12 Linnhoff B, Use pinch analysis to knockdown
capital cost and emission, Chemical Engineering
Progress, Aug 1994, 6-57.
13 Aspen Physical property system physical
property methods and models, 11.1.
14 Edwards J E, Process modeling selection
of thermodynamic methods, Technical report,
Mar 2011, www.chemstations.com/content/
documents/Technical_Articles/thermo.pdf
15 Fuel and combustion, www.beeindia.in/
energy_managers_auditors/documents/guide_
books/2Ch1.pdf
Sunil Kumar is a Scientist in the Separation
Process Division of Indian Institute of
Petroleum, Dehradun, working in the areas
of modelling and simulation of petroleum
renery separation processes, process
integration, pinch analysis, development
of solvent extraction processes, molecular
modelling and simulation application in
adsorption and solvent extraction. He holds a
masters in chemical engineering from Indian
Institute of Kanpur, India, has been awarded a
Certicate of Merit for Academic Excellence
operations? Part 1, Hydrocarbon Processing, Jul
2013, 99-102.
3 Errico M, Tola G, Mascia M, Energy saving
in a crude distillation unit by a preash
implementation, Appl. Thermal Engineering, 29,
2009, 1642-1647.
4 Feintuch H M, Peer V, Bucukoglu M Z, A
preash drum can conserve energy in a crude
preheat train, Energy Prog., 5, 1985, 165-172.
5 Ji S, Bagajewicz M J, Design of crude
distillation plants with vacuum units. Part I:
targeting, Ind. Eng. Chem. Res., 41, 2002, 6094-
6099.
6 Ji S, Bagajewicz M J, Design of crude
fractionation units with preashing or
prefractionation: energy targeting, Ind. Eng.
Chem. Res. 2002, 41, 3003-3011.
7 Benali T, Tondeur D, Jaubert J N, An
improved crude oil atmospheric distillation
process for energy integration: Part I: Energy
and exergy analyses of the process when a
ash is installed in the preheating train, Applied
Thermal Engineering, 32, 2012, 125-131.
8 Benali T, Tondeur D, Jaubert J N, An improved
crude oil atmospheric distillation process for
energy integration: Part II: New approach for
energy saving by use of residual heat, Applied
Thermal Engineering, 40, 2012, 132-144.
8 Ji S, Bagajewicz M J, Methods of increasing
distillates yield in crude oil distillation, United
States Patent, 7172686_B1, 2007.
10 Kemp I C, Pinch Analysis and Process
Integration A User Guide on Process Integration
4CLEANING
4FABRICATION
4REPAIR/RETUBING
4BAKE-OUT
4DECOMMISSIONING
4SAFE TRANSPORT
DUNN HEAT EXCHANGERS, INC.
www.dunnheat.com
CALL: (281) 337-1222 or (409) 948-1704. FAX: (409) 945-9873
Dunns specialized facility
offers complete services for
shell and tube type heat
exchangers and related
process equipment.
24 hours a day
7 days a week
Contact us at rfq@dunnheat.com
for more information.
410 21st St. SouthTexas City, TX 77590
IPP.indd 9 10/03/2014 14:07
four quest.indd 1 04/03/2014 11:04
Real time online energy management
I
t is Kuwait Petroleum Companys
(KPC) and its subsidiaries policy
to ensure that all energy should
be managed in accordance with best
engineering environmental practices
and within regulatory compliance
at all times.
1
All operations must be
committed to institute an energy
effciency programme and minimise
emissions of pollutants including
global warming gases. As part of
the activities to follow those prem-
ises, energy management cells
(EMC) have been formed at each of
the Kuwait National Petroleum
Company (KNPC) refneries: Mina
Al-Ahmadi (MAA), Shuaiba (SHU)
and Mina Abdulla (MAB). Such
cells implemented an energy
management programme based on
best international practices and on
real time, online optimisation.
This article describes the tasks
performed to develop and imple-
ment a real time, online model for
energy cost minimisation and
energy management at KNPCs
MAA refnery.
After a description of the main
project implementation tasks, the
Real Time Energy Management
Systems (RTEMS) functionalities
are described and the optimisation
implementation procedures are
commented on. Finally, several
obtained results are presented.
Project implementation tasks
The main stages of the project have
been:
Defne and installation stage
Model building and confguration
stage
Training stage
Site acceptance stage
Commissioning stage.
Creating a model of a renery energy system validated with real time data
increased energy efciency and reduced total energy costs
MOHAMMAD ERSHAID KNPC Mina Al-Ahmadi Renery
CARLOS RUIZ, DIEGO RUIZ and NICOLS VISUARA Soteica Visual MESA
DHANASEKAR PERIYASAMI EBS
Dene and installation stage
The frst stage of the project
includes the tasks of data collection,
control system review, software
installation and detailed defnition
of the model scope. Data collection
consists basically of gathering
system diagrams, real time data-
base tags and equipment
information, plus price related
information.
The control system reviews main
goals are: to develop a good under-
standing of how model
optimisation handles and process
constraints are related through the
site control system; and to identify
any new control strategies or
changes to existing strategies that
are needed to implement the opti-
misation. As a result of such a
review, there is an alignment of the
model with the real plant, making
sure that suggestions for optimisa-
tion can be properly implemented
using the sites operating proce-
dures, control strategies and control
structure.
Software installation is done on a
computer server, accessing the
historian data via a standard OPC
DA/HDA interface. Software
installation at PC clients is also
designed for client server access to
the model and to facilitate working
with the model in standalone
mode.
During this stage, a document is
prepared and approved with a
detailed scope of the model to be
implemented.
Model building and
conguration stage
A detailed model of the steam,
fuels, electricity, emissions, boiler
feed water and condensate systems
has been built, including all interac-
tions among these systems, real
plant constraints and degrees of
freedom of their operation. Other
utility systems have also been
modelled for monitoring purposes
such as air, nitrogen, faring system,
cooling water, sea water and
hydrogen networks.
Such a model is scheduled to
perform automatic executions for
the optimisation of the entire
system and is continually popu-
lated with validated, live plant
data. Figure 1 shows a screenshot
general view of MAA refnerys
RTEMS model graphical user
interface.
The main optimisation handles of
the energy system are:
Steam production at boilers
Fuel gas/fuel oil to the fred
boilers
Pump swaps (steam turbines/
electrical motors switches)
81 T-M possible switches above
100 HP; the steam turbines take
and discharge steam at different
pressure levels
17 T-M possible switches below
100 HP; the steam turbines work
from UMP/MP to LP steam pres-
sure levels
Condensing turbines
Extraction-condensing turbines,
UHP (ultra high pressure) to UMP
(ultra medium pressure) and/or
condensate
Electricity importation
Natural gas make-up to the fuel
gas system
Steam letdown, excess condens-
ing and vents
Deaerator water make-up preheat-
ing with low pressure (LP) steam.
soetica.indd 1 10/03/2014 14:22
Real time optimisation
What-if planning
Auditing and monitoring.
Real time optimisation
One of the main objectives of the
programme is to operate the energy
system at minimum cost, taking
into account all available resources
and real operational constraints.
2
Since the energy system is highly
non-linear and includes continuous
and discrete optimisation variables,
it requires a mixed integer, non-
linear optimisation technique.
The calculation core of the system
is MESA (Modular Energy System
Analyzer). For the solution of
continuous non-linear optimisation
problems, MESA employs an
implementation of the Sequential
Quadratic Programming (SQP)
algorithm, developed by Dr Leon
Lasdon.
3
Lasdons SQP solver has
been fne-tuned specifcally for the
characteristics of energy system
optimisation with particular
emphasis on robustness and speed.
MESA uses a technique loosely
based on relax and cut to deal
with mixed integer, non-linear opti-
misation problems resulting from
the start/stop decision of equip-
ment. In this technique, all integer
The main constraints to be met
are:
Process energy demand
Burners capacities
Emissions limits (NOx/SO
2
)
Contractual constraints (for
instance, natural gas and electricity
supply contracts)
Maximum and minimum operat-
ing limits for boilers.
Overall optimisation problem
fgures are:
180 optimisation variables
44 continuous variables
136 discrete variables
64 constraints.
Recommendations given by the
model are taken into account by
operations on a daily basis. The
same model used for the online,
real time optimisation is also used
in standalone mode, populated
either with current or historical
data, to perform case studies for
planning, or to evaluate alternatives
for better operation of the energy
system.
Training stage
A mid-point review is carried out
to review the model and optimisa-
tion with users; at this time, the
users training takes place. A
burn-in period corresponds to a
fne tuning of the model, operating
report and optimisation based on
operations feedback in order to
achieve day-to-day use of the
application.
Operators usually receive a set of
recommendations through a
custom report. Figure 2 shows an
example of such a report. The EMC
group contacts the shift supervisor
every day for coordination and to
follow up changes.
Site acceptance stage
This stage corresponds to the site
acceptance test, a check-list review
of the functionalities required by
the RTEMS.
Commissioning stage
During this stage, a full document
is prepared and delivered with all
details of the implemented system,
including troubleshooting and
maintenance procedures. Benefts
already obtained are also analysed
and reported. After commissioning,
the sustainability period begins
with the main objective of sustain-
ing the benefts.
RTEMS functionalities
The RTEMS has three main
capabilities:
Table 1
Figure 1 Graphical user interface of RTEMS model, general view for KNPC MAA renery
soetica.indd 2 10/03/2014 14:22
Process
Ideal Stage or Mass Transfer...
Which Model Should Be Used?
Insight:
The design and optimization of separation processes is
carried out using process simulators, which utilize various calcula-
tion approaches. Two techniques that are widely used for modeling
distillation are the ideal stage model and the mass transfer model.
IDEAL STAGE MODELS
The ideal stage model requires a minimum amount of
dataonly equilibrium relationships and enthalpy data for the heat
balance. The assumptions of the ideal stage approach are: 1) that
the vapor and liquid are both perfectly mixed so that the vapor and
liquid leaving a stage are at the same composition as the material
on the stage and 2) that thermodynamic equilibrium is obtained on
each stage. The equilibrium assumption also means liquid and va-
por leaving a stage are at the same temperature. Ideal stage mod-
els can also account for non-ideal column performance through the
use of reaction kinetics as is done for amine sweetening columns.
Obviously, the main disadvantage of the ideal stage
approach is just thatthe use of ideal stages to model real trays
or packing depths. However, for most processes encountered in
gas processing and other industries, the overall effciencies are
well established for proper operating conditions of the column. For
systems that are unavailable, similar systems often exist to allow for
effciency estimation. If not, the mass transfer approach is available
as an option.
MASS TRANSFER MODELS
For the end user, the notable feature made available via
the mass transfer approach is the ability to model a column with the
actual number of trays in the unit or the actual depth of packing.
However, there are still several assumptions that are made in this
approach that can have a signifcant impact on results. Two that are
worth mentioning include the mixing model for trayed columns and
the discretization of the packing depth for packed towers.
Application of the mass transfer model to random or
structured packing requires the column height to be discretized into
vertical segments or stages. For trayed columns, various mixing
models can be used for the liquid and vapor phases. The most
basic assumption is that of complete mixing in both the liquid and
vapor phases. However, the concentration gradients that develop
on a tray can signifcantly impact the predictions made by this model
since this gradient is the driving force for mass transfer. As the
column diameter becomes larger, the perfectly mixed fow model is
less applicable.
For modeling both liquid phase chemical reaction and
mass transfer, the use of the enhancement factor technique may be
considered. The enhancement factor describes the increased rate
of absorption due to the effect of a chemical reaction. The material
balance requires kinetic rate expressions for all chemical reac-
tions occurring in the system. As with equations for a non-reacting
system, an appropriate model for interface behavior must be used.
Mass transfer models require data necessary to calculate
interphase mass and heat transfer coeffcients and interfacial area
based on correlations of the following transport and thermal proper-
ties: diffusivities, viscosities, densities, heat capacities, thermal
conductivities, etc. Furthermore, mass transfer models require
detailed information on the column internals. For trays, this includes
information such as weir heights and fraction active area. For pack-
ing, this includes surface area per unit volume and void fraction.
If the simulator allows the user to select from various
alternatives for these parameters, knowing the correct selection
may be problematic. Further, the prediction of multicomponent
mass transfer coeffcients is of questionable accuracy. These facts
prompt the recommendation that columns modeled with the mass
transfer approach be checked against an ideal stage model with
an expected effciency until suffcient experience with the particular
application is achieved.

CONCLUSIONS
When performed properly, both the ideal stage and mass
transfer approach as implemented in ProMax 4.0 can calculate ac-
curate results for a variety of separation processes with and without
reactions. The ideal stage approach can be used initially to deter-
mine appropriate equipment sizes and operating conditions. More
detailed studies can be performed using the ideal stage approach,
the mass transfer approach, or both. Although signifcant operat-
ing experience provides reasonable effciency estimates for most
processes, the empiricism in scaling up from ideal to real stages or
ideal stages to real bed lengths can be a disadvantage when ac-
curate overall effciencies or HETPs are unavailable.
The mass transfer approach requires more equipment
design details and does not make use of overall effciencies or
HETPs. More detailed composition and temperature profles are
produced by this method at the expense of longer calculation time.
The mass transfer approach may appear more predictive in nature,
but is not necessarily more accurate. It relies on more parameters
that must be estimated, as both require thermodynamic data to
model equilibriumfor the tray composition in the ideal stage
approach and for the interface composition in the mass transfer
approach. Many of these mass transfer parameters are of limited
accuracy but also may be of limited sensitivity in some systems.
Both techniques are useful tools in process simulation.
For more information about this study, see the full article at
www.bre.com/support/technical-articles.aspx.
ProMax process simulaton sofware by Bryan Research & Engineering, Inc.
EngineeringSolutonsfortheOil,Gas,Refning&ChemicalIndustries
sales@bre.com
www.bre.com
979 776-5220
US 800 776-5220
bre.indd 1 12/09/2013 11:21
decisions are converted to continu-
ous variables by relaxing the
bounds of the associated variables.
This reduces the problem to a
non-linear continuous optimisation
problem that can be solved using
Lasdons SQP. This method has
been extensively adapted to suit the
characteristics of steam, fuel and
power optimisation on a real time
basis, especially taking into consid-
eration the time it would take to
solve the mixed integer problem.
The objective function to be mini-
mised is the total energy cost which
is calculated as:
FuelCosts+PowerCosts+ OtherCosts
Under FuelCosts the following
costs are included in the different
refnery energy systems models:
Natural gas cost
Fuel oil cost
LPG cost.
Under PowerCosts the following
Purchased electricity cost
Export electricity cost.
Under OtherCosts the following
Cooling water, sea water, distil-
late and fresh water costs.
The following example corre-
sponds to a case of real time
optimisation at MAA refnery. The
set of optimisation recommenda-
tions in this case were:
Steam production ratio among
boilers
Pump swaps (swappable turbines
and motors in different plant
areas).
A screenshot from the RTEMS
graphical user interface, corre-
sponding to the execution of a
stand-alone optimisation, is shown
in Figure 3. The hexagonal blocks
with mountain icons inside are
optimiser blocks. Such items are
used to defne and confgure opti-
misation variables or degrees of
freedom linked to the correspond-
ing equipment. Furthermore, when
the hexagonal blocks are coloured
inside, they indicate that the opti-
misers are at a limit; this is a means
to highlight visually the constraints
the optimisation is hitting during a
given optimisation case. In this
particular case, optimiser blocks are
tied to the low pressure steam vent
(LP Steam Vent) and to the low
pressure steam condensate (25-E-
401). In Figure 3, both blocks have a
red background because they are at
the limit during optimisation; that
is, both will be completely closed as
a consequence of applying the
calculated optimisation actions.
As the optimisation results begin
to be applied, the potential beneft
is reduced, meaning that the over-
all utilities system is being operated
at a higher effciency and with
lower costs. Figure 4 illustrates the
variation of potential savings that
could be obtained from the real
time optimisation of the KNPC
MAA refnery from 11 April to 25
July 2012. Three different periods
are shown: base line period,
commissioning period and optimi-
sation applied on a regular basis
period.
What-if planning
Using the model in standalone
mode helps to evaluate potential
capital or operating changes to
the complex and to assess the
Figure 2 Custom report example with RTEMS optimisation recommendations
soetica.indd 3 10/03/2014 14:22
report summarising and highlight-
ing the results.
Auditing and monitoring
Auditing and monitoring the
energy system are important tasks
which the RTEMS carries out by
helping in the follow-up of key
parameters such as equipment
The corresponding effect on the
KNPC MAA renery is a decrease
in the steam letdown system and a
reduction in the renerys power
imports due to the new power
generation. Figures 5 and 6 show
the new STG included in a Visual
MESA model for what-if case anal-
ysis, as well as the customised
economics and operability of such
changes.
During a what-if study, an actual
site status (as well as the corre-
sponding calculated optimum) can
be compared with the optimum
case study and easily analysed to
show how this change impacts on
the site and how to operate the
energy system after that change.
All imbalances are maintained
constant and steam production is
allocated to the new demand/
requests proposed scenario.
An example of these kinds of
studies is as follows. This study
evaluated the impact of installing a
new steam turbo generator (STG)
unit to provide power to the site
and to reduce the UHP-HP steam
letdown on the utilities system. The
STG includes a back pressure
turbine operating within ultra high
pressure (UHP) and high pressure
(HP) steam levels. In this case, 5.3
t/h more of UHP steam needs to be
generated in the boilers and a
reduction of 195.1 t/h in the
UHP-HP steam letdown can be
achieved. The estimated power
production is 4.56 MW.
Figure 3 Example of an execution of the RTEMS optimisation in standalone, displaying the optimised results
300
500
450
400
350
250
200
150
100
50
C
o
s
t
,

$
/
h
0
1
1

A
p
r

2
0
1
2
2
1

A
p
r

2
0
1
2
1

M
a
y

2
0
1
2
1
1

M
a
y

2
0
1
2
2
1

M
a
y

2
0
1
2
3
1

M
a
y

2
0
1
2
1
0

J
u
n

2
0
1
2
2
0

J
u
n

2
0
1
2
3
0

J
u
n

2
0
1
2
1
0

J
u
l

2
0
1
2
2
0

J
u
l

2
0
1
2
Baseline period Optimisation on a regular
basis period
Figure 4 RTEMS optimisation potential savings during Visual MESA project implementation
soetica.indd 4 10/03/2014 14:22
continuous basis, the potential gap
will start decreasing over time. If in
the future the system is in a new
condition, a new potential gap will
appear until the operators take opti-
mising action and the cycle is
repeated. If the projected savings are
consistently high, this should
prompt management, supervision
and/or engineering staff to identify
any issue that is impeding the
capture of cost reduction.
Another important KPI is related
to equipment effciencies and plant
energy costs. Figures 7 and 8 show
an example of a customised report
for a set of boilers and plants.
The model calculates the incre-
mental cost for fuel burnt. Such a
report allows prioritisation of
actions to improve combustion
based on economic impact.
By auditing the energy system,
imbalances can be identifed and
actions can be taken to reduce it.
Therefore, the data can be more
and more relied upon to evaluate
the value of energy production and
usage, and waste can be
eliminated.
The methodology is based on
getting inaccurate meters fxed
because imbalances are inherent to
every real world steam system
header calculation. The way the
current and actual operating costs
and the minimum and optimal
operating costs is also calculated
and historised. The savings associ-
ated with this gap will be
materialised only if all the suggested
optimisation movements are imple-
mented simultaneously. These
predicted savings constitute one of
the main economic key performance
indicators (KPI). It is expected that,
if the operators implement all the
proposed recommendations on a
effciencies and energy cost savings,
and also in the validation of data.
At every automatic run, the
RTEMS calculates and historises:
Fired boiler effciencies
Process heater effciencies
Imbalances at every steam pres-
sure level header
Energy consumption per unit rate
for every process unit
Actual and optimal total energy
cost.
The potential gap between the
Figure 5 New STG is added; case study example
Figure 6 Customised report optimisation summary; case study example
soetica.indd 5 10/03/2014 14:22
To learn more about Lewis
Pumps
or Weir Minerals products, visit:
www.weirminerals.com/Lewis
LEWIS
PUMPS
Vertical Chemical Pumps
Weir Minerals Lewis Pumps
8625 Grant Rd.
St. Louis, MO 63123
T: +1 314 843-4437
F: +1 314 843-7964
Email: sales.lewis@weirminerals.com
Copyright 2014, EnviroTech Pumpsystems, Inc. All rights reserved. LEWIS and LEWIS PUMPS are trademarks and/or registered trademarks of Envirotech Pumpsystems, Inc.; WEIR is a trademark and/or registered trademark of Weir Engineering Services Ltd.
Weir Minerals Lewis Pumps is known internationally in the
sulphur, sulphuric acid and phosphoric acid industries with
equipment installed in more than 120 countries worldwide.
With new product innovations and a dedicated group of
employees, Weir Minerals Lewis Pumps is the recognized
world leader for pumps and valves in difcult applications.
weir.indd 1 06/03/2014 13:53
RTEMS deals with imbalances is as
follows:
Wherever there is a closed mass,
or mass and energy balances, the
RTEMS calculates a real time
balance offset or bias (the so-called
balloons, see Figure 9) as the differ-
ence between the inlet and outlet
streams to/from a given header
This offset is then held constant
for optimisation or case studies
(considered as an unmeasured load).
Several actions can be taken
based on the magnitude of the
imbalance:
Big imbalances are visually iden-
tifable (the size of the balloon
increases and/or the colour dynam-
ically changes)
Either visually or in a report, the
RTEMS will fag these large imbal-
ances to the user who can then
focus maintenance attention on the
meters associated with that balance.
If a particularly important sensor
fails in the validation test, the
RTEMS could automatically shut
down optimisers, or force those
constraints to which it is related, or
even shut down the entire site
optimisation.
Optimisation implementation
procedures
Operations apply, on a daily basis
under EMC supervision, the actions
calculated and recommended by
the RTEMS in order to minimise
the total operating cost, following
the implementation procedure
shown in Figure 10.
Figure 7 Example of boilers efciencies monitoring report
Figure 8 Example of KPIs report for different units
Figure 9 Example of balloons location
RTEMS gathers the sensor
data from DCS/Plant
Information System
through OPC server
RTEMS processes the online
data values, and it is
automatically executed for
the recommendation
The recommendations are
automatically generated
every half hour and shown
in the web-based report
Energy Management Cell
(EMC) evaluates the
recommendations and
passes them to Operations
for implementation
New process condition will
occur and the results shown
in DCS/Plant Information
System after implementation
Figure 10 RTEMS optimisation implementation procedures
soetica.indd 6 10/03/2014 14:23
Lasdon L, Journal of Optimization Theory and
Applications, Vol. 56, No. 3, Mar 1998.
Mohammad Ershaid is Specialist, Process
Engineer at Kuwait National Petroleum
Companys (KNPC) Mina Al-Ahmadi renery in
Kuwait.
Carlos Ruiz is Energy Management Systems
Director with Soteica Visual MESA He holds
a degree in chemical engineering from
Universidad Tecnolgica and masters and
doctors degrees in chemical engineering from
Universidad del Sur, Argentina.
Email: carlos.ruiz@soteicavisualmesa.com
Diego Ruiz is Professional Services Director
with Soteica Visual MESA. He holds a
chemical engineering degree from Universidad
Tecnolgica Nacional - Facultad Regional
Rosario and a doctorate in chemical engineering
from Universitat Politcnica de Catalunya.
Nicols Visuara is a Consultant with Soteica
Visual MESA. He holds a degree in chemical
engineering from Universidad Nacional del
Litoral, Santa Fe, Argentina.
Dhanasekar Periyasamy is Team Leader for
Process Modeling and Energy Management
with EBS Group WLL, Kuwait.
optimisers are taken into account
by operations on a daily basis.
As a result of the project, informa-
tion for each refnery energy system
has been organised into models and
reports to which everyone has
access. Having a model of the whole
refnery energy system validated
with real time data, in one environ-
ment that everyone has access to, to
understand all the decision variables
and the associated constraints, helps
coordination among plant areas in
order to increase energy effciency
and reduce total energy costs.
References
1 KPC HSE Standard Doc-18 (Management of
Energy & Resources).
2 Use of online energy systems optimization
models, Ruiz D, Ruiz C, Book: Energy Management
Systems, ISBN 978-953-307-579-2, edited by
Giridhar Kini P, Ch 4, 85-100. InTech, 2011.
3 Experiments with successive quadratic
programming algorithms, Fan Y, Sarkar S,
Cost savings and CO
2

emissions reduction
Figure 11 shows the reported results
from one of the refneries (KNPC
MAA). The base case was estab-
lished when the model was
completed in May 2012 with an
energy cost savings opportunity of
about $680/hr. By the end of
December 2012, the potential for
energy cost savings was reduced to
about $180/hr, refecting obtained
savings of about $500/hr ($4.4
million/y).
The cost savings and CO
2
emis-
sion reductions came from two
main sources:
1. Reduction of excess steam vent/
condensing (less fuel gas used for
high pressure steam production
because of the reduction in steam
waste)
2. Let-down reduction (less power
purchase for driving pumps and
compressors because of the use of
turbines instead of motors).
The estimation basis for reduction
of CO
2
emissions is shown in
Figure 12.
Conclusions
The tasks performed to develop
and implement real time, online
models, to help with KNPC MAA
refnerys energy cost minimisation
and energy management and
scheduled to perform automati-
cally, have been described. They
enable clear identifcation of the
different economic trade-offs that
challenge the operation of the site-
wide energy systems at minimum
cost while reducing energy waste
and, therefore, the CO
2
footprint.
Recommendations given by the
480
800
720
640
560
400
320
240
160
80
S
a
v
i
n
g
s

p
o
t
e
n
t
i
a
l
,

$
/
h
0
90
150
135
120
105
75
60
45
30
15
t
C
O
2
/
d
a
y

C
O
2

e
m
i
s
s
i
o
n
s

r
e
d
u
c
t
i
o
n

p
o
t
e
n
t
i
a
l
0
M
a
y

2
0
1
2
J
u
n

2
0
1
2
J
u
l

2
0
1
2
A
u
g

2
0
1
2
S
e
p

2
0
1
2
O
c
t

2
0
1
2
N
o
v

2
0
1
2
D
e
c

2
0
1
2
Savings potential
CO
2
emissions reduction potential
Figure 11 CO
2
emissions and economic potential savings, KNPC MAA renery; their
reduction corresponds to captured savings and improvement in the efciency of the
utilities system
Figure 12 Estimation of CO
2
emission reductions as a result of optimisation action
soetica.indd 7 10/03/2014 14:23
Onis France
1 av. Fernand Julien
13410 Lambesc, FR
33 (0)4 42 92 93 20
commercial@onislineblind.com
www.onislineblind.com
Onis Inc.
One Riverway, Suite 1700
Houston, Texas 77056 USA
1 (713) 840 6377
sales@onislineblind.com
Operators can blind in less than
3 minutes zero tools
Replace spectacle blinds, slip blinds
& figure-8 blinds
H2S Flare application Offshore platforms Furnace isolation
Sizes : 50
Temperature : -155F 1400F
Flanges : 150# - 2500#
Built to customer requirements:
ASME, NACE, API607
Reduce downtime &
improve safety
100% positive isolation:
gases, powders, liquids
& viscous solutions
onis.indd 1 07/03/2014 13:08
Advanced solutions for efcient
crude blending
I
n the past, refneries were
constructed to distill conventional
light crude oils. Current econom-
ics, variations in the price of crude
oils and shifting demand for distil-
lates have forced refneries to
reduce the cost of their distillation
feedstock. Commonly, this is
achieved by blending high-value
light crude oils with heavy (uncon-
ventional) crude oils of inferior
quality, or by buying ready-made
blends. Low quality crudes include
heavy crudes from known loca-
tions, as well as opportunity crudes
that are brought on the market by
traders worldwide. These crudes, of
lower quality, can be purchased at
low cost. Blending of these with
costly crudes is inevitable to
produce crude blends that bear
optimal properties to be processed,
and at minimum cost.
Refneries worldwide are
constructed from an engineering
point of view and from materials
that enable the distillation of
well-defned types of crude oils.
These refneries were built based
on the availability of certain types
of crude oils in their neighbour-
hoods, the cost of certain crude oils
on the market, and demand for
predominantly light distillates for
gasoline production.
Distillation of crude oil was
mainly targeted to produce gaso-
line components, such as light and
middle distillate. More recently
and especially in the US and in
Europe, demand for fuels has
shifted from gasoline towards
diesel fuels. This means that while
in the past predominately light
crudes were distilled, today refn-
eries must be able to distil heavier
The use of nuclear magnetic resonance based process analysers supports the
production of blends at lowest cost
GREGORY SHAHNOVSKY, TAL COHEN and RONNY MCMURRAY
Modcon-Systems Ltd
crude oils to increase the amount
of middle and heavier distillates.
Refning margin for many refner-
ies which were not able to adapt to
the changing situations decreased.
Technological limitations caused
many refneries to buy expensive
light crudes that do not produce
specifcally those distillates that
are most needed in the market. For
many refneries, the losses were
too large. Many closed or changed
their activities from distilling
toward blending.
Nowadays, crude blending is
performed either by blenders or by
refneries themselves which buy
various types of low cost crude oils.
They upgrade their chemical and
physical properties to produce a
synthetic crude oil at lowest cost,
which can be processed in refnery
equipment and will yield high
value distillate.
Characteristics of crude oils
Quality properties determine the
market value of each type of crude.
The most important quality charac-
teristics are the density, the total
acid number (TAN) and the
sulphur content. The API ranges
from light crudes (high API, low
density) to heavy crude oils (low
API, high density). Sulphur is pres-
ent in crude oils as hydrogen
sulphide and as polysulphides.
These sulphur containing molecules
will partially decompose during
distillation, while hydrogen
sulphide evolves. The sulphur
content and other acidic compo-
nents in crude oil, such as
naphthenic acids, are highly corro-
sive, and responsible for crude oil
to be of a sour or sweet character.
These characteristics mostly lead
the price paid for different types of
crude oils.
Renery equipment
High TAN crude oils are character-
ised by fewer light components,
high density and viscosity, low
solidifcation point, high nitrogen
content, high gel-asphalt content,
high salts and high heavy metals
contents and a low yield of light oil
distillates. Oil separation in the
desalter is more diffcult than in
conventional crudes. These proper-
ties also cause these crudes to give
low quality products and they are
very corrosive. Commonly, high
TAN crude oils are called opportu-
nity crude oils. The price is about
80% that of conventional crude oil.
The additional cost of processing
high TAN crude is within the range
$1.1510.73/bbl, but the savings
compared to conventional crude
processing are $43.5462.7/bbl.
Utilising these crude oils in any
way possible is therefore very
attractive to refners.
1
In the past, most refneries were
designed and constructed from
materials according to the crude oil
The additional cost of
processing high TAN
crude is within the
range $1.1510.73/
bbl, but the savings
are $43.5462.7/bbl
q2 modicon.indd 1 11/03/2014 14:32
opens opportunities for oil blend-
ers, oil trading companies and
terminals to bring low cost blends
onto the market. These blends can
be sold to refneries with a high
market value and quality.
Crude mixing can be applied
throughout the entire supply chain
of crude oil, from its well enabling
transportation, through terminal
blending to the refneries. The fnal
crude supply to the distillation unit
may be a combination of these
activities.
The strategy of crude oil blending
includes several parameters. Each
of them contributes to the overall
fnal cost of crude oil entering the
crude distillation unit as well as the
refning margin:
The engineering limitations of
crude distillation units to refne any
type of crude oil
Cost differences of crude oils
according to their location of origin,
and their chemical and physical
properties. An increased ability to
process unconventional crudes
leads to improved refnery margin
Product shifting in the market
from gasoline towards diesel fuels.
Increased demand for diesel fuels
in the European market caused
refneries to increase diesel yield
over naphtha yield
High viscosity, especially in
heavier crude oils, affects the fow
properties of crude during trans-
portation. Blending these types of
crude oils with diluents or conven-
tional crudes may be required to
reduce viscosity and to improve
fow properties.
Crude oil economics
Previously, refneries distilled crude
oil from single locations, but nowa-
days refnery profts are a direct
result of the ability to create blends
that include lesser quantities of high
value crude oils and higher quanti-
ties of unconventional crudes, such
as heavy and extra heavy crudes,
sour crudes and bitumen extracted
from oil sands. However, these
blends should still have those phys-
ical and chemical properties that are
required to enable smooth and
continuous operation of the
distillation unit at the lowest possi-
ble cost.
available and its ease of purchase.
This limits the versatility of many
refneries to purchase other crude
oils of different qualities. Many of
those refneries that are constructed
to distil light and low sulphur
crude oils are restricted as regards
processing heavy fuels.
Critical differences in their physi-
cal and chemical properties make
heavier crudes oils more diffcult to
distil than light crudes. Heavy
crude oils are sour and more corro-
sive than light crudes. Higher
viscosities, fouling tendencies and
different fow streams make it more
diffcult to maintain stable crude
charge rates, which are required for
stable product yields, quality and
reliability. Differences in boiling
points between light and heavy
crudes require different process
temperature requirements such as
pre-heating, different distillation
temperatures, overheads and so on.
Heavy fuels are rich in asphaltenes
and metals and other contaminants
which cause poorer desalting
performance.
Product shifting
Demand for certain distillates and
refnery products is shifting. It is
expected that middle distillates will
comprise some 45% of global
demand per barrel by 2015, which
is a rise of 10% compared to 2005.
It can be expected that production
of diesel gas oil in developing
nations will increase by 10 million
b/d from 2009 to 2030.
2
The US is the largest consumer of
crude oil. While US demand
remains stable, today China is in
second place with an annual 4%
increase in demand for crude oil.
New refneries, built nowadays, are
already designed as such that they
are not limited to a small range of
crude oils.
The major operational cost of the
refnery is contributed by the price
of the crude oil, an estimated by
80-90% of cash fow. Reducing the
cost of the crude feedstock, without
changing the range and volumes of
high valued distillates, increases the
refning margin. Refnery profts are
a direct outcome of the strategy
applied by the refnery to purchase
low cost crudes and to produce
distillates with a high market value.
To increase refning margin and
remain competitive, refneries are
obliged to minimise the cost of their
crude feed, without affecting their
capacity to produce high value
distillates. As heavier crudes are
more diffcult to process, and with
the increase of consumption of
diesel oil as compared to gasoline,
light sweet crude oils are marketed
at a higher price than heavy crudes.
Reducing the cost of the crude
input, without changing the range
and volumes of high valued distil-
lates increases the refning margin.
Potential crude blenders
Two broad types of organisation
deal with the business of crude
blending, refneries and blend
producers/trading companies.
Crude blending is applied directly
by refneries to prepare low-cost
and compatible blends for internal
consumption or for trading in the
market. Effcient crude blending
80
120
140
100
60
40
U
S
$
/
b
b
l
20
2008 2009 2010 2011 2012 2013
WCS
WTI
Edmonton Par
Brent
Prices quoted are monthly averages
Figure 1 World oil price according to benchmarks
q2 modicon.indd 2 10/03/2014 14:37
The Emerson logo is a trademark and a service mark of Emerson Electric Co. 2012 Emerson Electric Co.
See more, do more and be more profitable with the most trusted partner in wireless Emerson.
Emerson is your proven partner with Smart Wireless in more customer sites and with more operating
hours than anyone else in the process industry. Smart Wireless has the widest range of technologies
to expand your vision into more places across your operations. And its self-organising mesh network
delivers the highest reliability available. It is simply the most intelligent, secure and cost-effective
operation-wide wireless option available. See how Smart Wireless can empower your bottom line
at www.EmersonProcess.com/SmartWireless
Using wireless here and there is one thing.
But using it across my entire operation?
Theres no one I could trust to do that.
ITS
IMPOSSIBLE
115802 ins17_itsPossClimbers_PTQ.indd 1 31/08/2012 16:10
emerson.indd 1 04/03/2014 10:56
Basically, crude oils can be divided into four major
groups:
Light low sulphur (API 30-40, S 0.5% mass)
Light, moderate sulphur (API 30-40, S=0.5-1.5%
mass)
Heavy, high sulphur (API 1-30, S 1.5-3.1% mass)
Extra heavy high sulphur (API = 15, S 3% mass).
The ratio of a component in a blend is actually
limited by the physical properties required for produc-
tion of the highest valued distillates to the largest
extent, and by the construction of equipment to process
the blend.
Current values based on the benchmark of light
Brent crudes and the somewhat heavier WTI crudes
show differences in price of around $10/bbl a barrel
(see Figure 1).
4
Opportunity crudes are of course much cheaper.
Various crude oils, such as some Venezuelan and
Canadian crudes, are very heavy and are attractive for
bitumen production. Their processing is limited by
their very low API gravity. To produce other distillates
from these crude oils, they must be upgraded by dilu-
tion with light crudes or kerosene.
The high viscosity of many of these heavy crudes is
another drawback. Blending with light crude oils,
kerosene or other diluents is also required to give them
fow properties that enable their transport through
pipelines without heating.
It is the aim of each refnery to maximise the
consumption of opportunity crudes.
Heavy oils are hydrogen defcient and have high
levels of contaminants such as sulphur, nitrogen,
organic acids, vanadium, nickel, silica and asphaltenes.
The method for upgrading heavy oils at relatively low
cost is to dilute them with hydrogen rich, higher qual-
ity light crude oils or by using hydrogen rich diluents
to increase the H/C ratio.
Blending processes
Crude oil blending can be performed by two
technologies.
In-tank blending (batch blending)
Specifc volumes of different kinds of crude oils stored
in separate tanks are loaded into a blending tank
where they are mixed until a homogenous composition
is achieved. The tanks are mechanically stirred.
Samples must be withdrawn to determine whether the
blend is homogeneous and whether it conforms to its
predetermined specifcation. In the event of discrep-
ancy, correction of the blend must be conducted. The
entire procedure of in-tank blending is very time
consuming and expensive.
In-line blending
In contrast to tank blending, in-line blending is
performed by simultaneously transferring different
crude oils through an on-line static mixing device to the
fnal blend tank. The predetermined fow ratio between
the different crudes will provide a blend of the required
quality. In-line blending enables on-line correction of
the quality of the blend, by changing the ratio between
www.eptq.com
Is your plant
ready for the future?
Please visit us at www.yokogawa.com/eu
or contact us at info@nl.yokogawa.com
Making critical plant information fully visible is just the
beginning of the vigilant cycle. Seeing clearly gives
you the knowledge necessary to anticipate the changes
required in your process. Knowing in advance brings you
the speed and exibility to optimize your plant in real time.
And by acting with agility, you are able to adapt to the
ups and downs of your business environment. VigilantPlant
excels at bringing out the best in your plant and your
people - keeping them fully aware, well informed, and
ready to face the next challenge.
Is your plant
ready for the future?
Life-cycle optimalization
Control_86x270_2014.indd 1 06/03/14 11:36
q2 modicon.indd 3 10/03/2014 14:37
ware should also focus on the
potential prot derived from the
blend. This requires software also
to relate to:
Cost of various crude oils and
crude oil blends
Prices of nal distillates and
other renery products.
Volumes of nal distillates
required by the market
Calculation of the blend compo-
nents and their ratios
Ratio limits
Predicted fraction temperatures
Constraints in the properties of
the blend
Properties of the fractions
Constraint limits.
Next to the chemicophysical
properties of the blend, the soft-
feeds. The blend is produced instan-
taneously and no stirred blending
tanks are required.
To operate the blending process
efciently and without error, on-line
process analysers are required to
instantaneously measure the blend
downstream and to feed the blend-
ing operators with the required
quality details of the blend in
production. This enables real-time
and on-line correction during the
blending process, providing the
blend of predetermined properties.
This reduces corrective re-blending
of an entire tank, as well as unneces-
sary giveaways.
Determination of blending recipes
Simulation software, such as linear
programming (LP) modelling, is
widely used to predict the ratio
between individual components to
prepare a blend. Based on the
composition data of crudes used,
and using the proper algorithm,
this software is commonly applied
to calculate and predict the physi-
cal properties of blends.
The software calculates the ratio
of different crudes, resulting in a
crude blend with the appropriate
properties, leading to the desired
distillates at optimal yields.
Incorporation of a large database
which covers a broad range of crude
oils is required to predict accurately
a blend of predetermined physical
properties and with the potential to
maximise production of high-value
distillates. Adequate blending simu-
lation models should include not
only the chemistry of crude oil
distillation but also its economics. It
must be able to calculate the compo-
sition of different crudes to provide
the best economic blend at the
lowest cost. Such blends contain
maximised volumes of those crude
oils of lowest cost, but still bear the
most attractive rening properties.
This strategy will minimise variable
costs and maximise prot.
5
LP is based on the assay of differ-
ent crudes oils to be blended. Any
changes in the assay will affect the
LPs predicted blend.
Fundamentally, effective crude
blending simulation software
should include the following
features:
6
15
20
10
5
0
0 20 40 60 80 100 120 140
CH
2
Aromatic
rings
Hydrogen
Alpha to
CH
3
Figure 2 Typical NMR spectra of crude oils
930
980
880
830
N
M
R
,

k
g
/
l
780
R
2
= 0.9554
780 830 880 930 980
Laboratory, kg/l
2
4
5
3
1
0
N
M
R
,

%
1
1 0 1 2 3 4 5
Laboratory, %
R
2
= 0.9672
Density 15 C
Water content
Figure 3 Correlation between NMR predicted results and laboratory measurements
q2 modicon.indd 4 10/03/2014 14:37
defned chemical shifts represent
the chemical structure of molecular
species. Linear correlation between
the intensity of the signal and the
hydrogen concentration makes it
possible to quantify the different
hydrogen nuclei.
Physical properties in crude oils
and in distillates correlate with their
chemical compositions. This allows
chemometric methods to correlate
between the measured spectral data
and the physicochemical proper-
ties of crude oil or other distillates.
In contrast to other chemometric-
based spectral technologies, such as
Raman and NIR spectrometry,
which are based on fngerprints,
due to its molecular specifcity and
its linear quantitative correlation
NMR technology requires far fewer
reference samples to establish a
chemometric model.
NMR-based on-line spectrometers
are not limited to transparent fuids,
but can be applied to transparent
and opaque liquids alike. Crude
oils contain water heteroatom
electromagnetic (EM) energy at a
specifc resonance frequency. The
basics of NMR process analysers
are the alignment of nuclei in a
magnetic feld. An external radio
frequency (RF) pulse is applied;
this distorts the alignment of the
nuclei in the magnetic feld. The
resonance frequency depends
mainly on the strength of the
magnetic feld. When the RF pulse
ends, the protons relax and align
back to their initial equilibrium
position, which generates a decay
signal, the free induction decay
signal (FID).
Crude oil is a mixture of organic
chemical compounds, mainly
carbon and hydrogen-based mole-
cules. Neighbouring atoms, such as
carbon, oxygen and sulphur, and
neighbouring chemical bonds,
infuence the strength of energy
absorption and emission by hydro-
gen nuclei in a magnetic feld.
Accordingly, the signal of each
hydrogen atom shifts differently in
the NMR spectrum. These well-
Time consuming and costly labo-
ratory analyses are required to
verify the real physical properties
of the blend. Re-blending is
required if these properties are not
achieved.
Effcient blending requires on-line
monitoring of the blend properties
throughout its entire production.
Chemical compositions differ from
crude to crude. Notwithstanding
whether the crude oil is pure, or a
blend of crude oils, on-line correc-
tions are continuously conducted to
maintain stable product quality.
This requires real time collection
and validation of physical
properties from the blend through-
out its entire production process.
Among all of the analysers availa-
ble in the market, nuclear magnetic
resonance (NMR) process analysers
are the most suitable for that
purpose.
The frst NMR process analysers
were launched in the late 1990s. In
NMR, hydrogen nuclei in a
magnetic feld absorb and re-emit
3
5
4
2
1
N
M
R
,

%
0
0 1 2 3 4 5
Laboratory, %
Initial
boiling
Point
R
2
= 0.9561
30
50
40
20
10
N
M
R
,

%
0
0 10 20 30 40 50
Laboratory, %
R
2
= 0.9558
80
100
90
70
60
50
40
30
N
M
R
,

%
20
20 40 60 80 30 50 70 90 100
Laboratory, %
R
2
= 0.9530
70
100
120
80
60
50
N
M
R
,

%
40
40 50 60 70 80 90 100
Laboratory, %
R
2
= 0.9584
TBP 350
TBP 220
TBP 520
Figure 4 Correlation between NMR predicted boiling points and laboratory measured boiling points
q2 modicon.indd 5 10/03/2014 14:37
and partially to physical properties
such as the distillation curve, and an
excellent overlap between simulated
distillation and NMR analytical
results. Taking into account the time
accuracy in correlation between
NMR predicted results and labora-
tory measurements. Partially these
measurements relate to chemical
matter such as water and sulphur,
molecules, which are easily distin-
guished by NMR spectrometry.
NMR on-line spectrometry with
appropriate chemometrics has the
ability to determine the following
properties in crude oil:
Specifc gravity
True boiling point yield
Aromatic content, %
Olefn content, %
Pour point
Water, %
Sulphur, %.
Following these parameters is
most important during crude
blending. Their on-line measure-
ment makes it possible to blend
synthetic crude to deliver prede-
fned properties, either from a
physico-chemical point of view or
from an economic point of view.
On-line monitoring of the blend-
ing process prevents the production
of blends that do not comply with
the requirements of the refnery.
Blending errors and giveaways can
be prevented which can lead to
annual savings of millions of dollars.
Precision of NMR process analytics
High accuracy in the correlation
between a NMR process analysers
results and laboratory results char-
acterises the new generation of
NMR process analysers. NMR
magnets are sensitive to tempera-
ture differences. Earlier generations
of NMR process analysers were
especially sensitive to temperature
differences due to the accumulation
of heat produced by their electron-
ics and heat-conducting measuring
probes. In the new generation of
NMR process analysers, the overall
design excludes any accumulation
of heat in the magnet or in its
surroundings by uncontrollable
fuctuations in temperature. This
increased analysers stability to
heat fuctuation from 2C to
10C. This means that any heating
of crude oil required prior to blend-
ing, or after the desalter, is possible
without affecting the analytical
results, as long as a temperature
deviation of 10C is maintained.
Figures 3 to 5 show the correla-
tion between an NMR analysers
predicted results and laboratory
analyses of different crude oils.
The fgures demonstrate the high
IBP Cum
150
Cum
220
Cum
350
Cum
520
60
100
90
80
70
50
40
30
20
10
W
e
i
g
h
t
,

%
0
SimDis
NMR
IBP Cum
150
Cum
220
Cum
350
Cum
520
60
100
90
80
70
50
40
30
20
10
W
e
i
g
h
t
,

%
0
SimDis
NMR
Crude A
Crude B
Figures 5 Overlapping NMR and simulated distillation curves
Components
Control system
Online blend
optimisation
Online
analysers
Blender
Blended
crude
Figure 6 Set-up of a crude blending station with incorporated NMR process analyser,
simulation modelling and blending control. (Components can be high and low quality
crude oils, diluents and/or gasses (NLG)
q2 modicon.indd 6 10/03/2014 14:37
to the application of NMR process
analysers for blending different
crude oils.
Crude oil compatibility
during blending
Blending different crudes, espe-
cially when unconventional crudes
are involved, may cause precipita-
tion of asphaltenes, which causes
fouling in the pipes and process
units. Asphaltenes are soluble in
polar aromatics, such as toluene,
but insoluble in paraffnic non-
polar solvents. On-line analyses of
the SARA content (saturates, resins,
aromatics and asphaltenes) can be a
potential tool for on-line determina-
tion of quantitative ratio between
different crudes to be blended, or
between crude oils and polar
solvents, without causing
asphaltenes to precipitate.
Natural gas liquids in crude
oil blending
Natural gas liquids (NGL) are
produced by refrigeration and
distillation processes in gas plants
and refneries and are considered
byproducts in the oil and gas
industry. Gas plants extract NGLs
for proft and/or to ensure produc-
tion of pipeline quality natural gas.
NGL prices are relatively low.
They and other off-spec materials
from natural gas production are
used by refneries and blending
companies to upgrade heavy crude
oils. Another application is lower-
ing the viscosity of heavy crudes to
make them fow more easily
through pipelines. Implementation
of on-line NMR process analysers
provides an effective tool for eff-
cient blending of NGLs and crude
oil to deliver the required physical
properties at lowest cost.

Conclusions
Different blending options exist to
upgrade unconventional crude oil
into synthetic crudes of higher
values. An automatic crude blend-
ing station integrates LP with
on-line NMR process analytics. It
can be used either by traders who
offer blending services, or directly
by refners. Cost, market value,
availability and choice of technol-
ogy are the main factors to be
feedback control of raw materials
based on the quality of product
samples obtained from the blend
header.
Both LP simulation and NMR
process analysis can operate stand-
alone. However, for best
optimisation of the crude blending
process, it is essential to integrate
the two technologies.
Effcient blending optimisation is
a dynamic process involving
mixing, continuous blend analysis,
simulation model adjustment and
process control. All of these
elements should be taken into
account (see Figure 7). Any missing
link in this chain of operations will
impact the effciency of the entire
process and reduce its revenue.

Alternative applications of on-line
NMR process analytic
Other uses of NMR on-line process
analytics are of interest, in addition
required for laboratory analyses, the
cost to perform crude oil assays, or
purchase and maintenance costs
justifes the use of NMR process
analysers in crude blending
processes, especially in cases of
in-line blending. NMR makes it
possible to monitor precisely the
quality of the blend in production
and, if required, to change the ratio
between different crude oil feeds to
establish and maintain the quality of
the fnal blend.
Optimised crude oil blending station
A crude blending station consists of
a blending skid to receive liquid or
gas streams, optimisation software
and analytical equipment. Analytical
equipment should be able to
provide online measured data of
component and product streams
involved in a blending operation.
This data is transferred to optimisa-
tion software whose target is to
support production of a blended
product with minimum product
cost, minimal quality giveaway, and
minimal deviation from individual
raw material properties. To achieve
this objective, the optimisation
system continuously receives quality
feedback of the fnished product
using on-line analysers.
Using the inputs from on-line
analysers, the optimisation system
performs either feed-forward or
Calculated ratio and
predicted physical
properties
Real ratio
Real physical properties
Crude oil
availabilities
Simulator
LP
Blend calculation
Physical blending
Automatic adjustment
of streams
(CDU)
NMR process analyser
Quality analyses
Blend re-calculation
Crude oil cost
Crude oil quality
parameters
Distillate
requirements
Final blend
Figure 7 Dynamic process of mixing, continuous blend analyses, simulation model
adjustment and process control
Implementation of
on-line NMR process
analysers provides
an effective tool for
efcient blending of
NGLs and crude oil
q2 modicon.indd 7 10/03/2014 14:37
Tal Cohen is Executive Vice President of R&D
and Business Development with Modcon-
Systems Ltd and previously held positions of
CEO, CTO and VP R&D in high tech companies.
Email: talc@modcon-systems.com
Ronny McMurray is an Application Scientist
with Modcon-Systems Ltd, with experience
in applied R&D and as chief chemist in the
chemical and petroleum industries. He holds a
PhD in chemistry.
Email: ronnym@modcon-systems.com
predicting crude oil blending properties,
4th Mercusor Congress on Process System
Engineering.
Gregory Shahnovsky is the Chief Executive
Ofcer of Modcon-Systems Ltd, a developer
of process analysers, optimisation and control
solutions. With over 25 years experience in
the chemical and petrochemical industries, he
holds a MSc in process control engineering and
a MBA in business administration.
Email: gregorys@modcon-systems.com
considered in planning a congura-
tion to be used for upgrading
unconventional crude oil.
Two principal technologies are
required for optimised crude
blending:
On-line process analytics moni-
tors crude oil and blend quality at
any time and at any stage
Dynamic simulation modelling
(blending simulation models) are
commonly used to determine the
required blend composition.
Highest blending optimisation can
be achieved only by updating the
simulation program with real time
analytical data for crude oil and
blend quality.
NMR-based process analysers can
be used to determine chemical
composition and physical proper-
ties in dark and opaque streams.
The benet of NMR spectrometry
lies in its linear correlation between
the hydrogen atoms of the mole-
cules present in crude oil and the
chemical nature of its components.
Chemometrics transforms spectro-
metric measurements into the
characteristic physical properties of
crude oils and blends.
This technology provides real
time data and information about
the physical and chemical proper-
ties of the blend in process. On-line
adjustments and changes between
blend components can be
performed accordingly until the
required physical properties are
achieved.
Thanks to Aspect imaging for their cooperation
in NMR analysing crude oils.
References
1 Wu Qing, Processing high tan crude: part
1, Jan 2010, www.digitalrening.com/article/
1000524,Processing_high_TAN_crude__part_I.
html#.UmUziHBkOsQ.
2 Prospectus Opportunities Crude Report
2011, Hydrocarbon Publishing Company, 2011
www.hydrocarbonpublishing.com/ReportP/
Prospectus-Opportunity%20Crudes%20
II_2011.pdf
3 Stratiev D, et al, Evaluation of crude oil
quality, Petroleum & Coal, 52 (1), 35-43, 2010.
4 Drilling down on crude oil price differentials,
Observation, TD Economics, 14 Mar 2013.
5 Aldescu M, Challenges of heavy crude
processing, Sour and Heavy, 2012.
6 da Cruz D, et al, Optimization tool for
150 YEARS
No matter how outstanding the machine requirements
in the chemical and pretrochemical industry are, the
solutions ofered by AERZEN are always reliable, tailor-
made and efcient!
The 3 ALLROUNDER machines from Aerzen - blowers,
oilfree and oil- injected compressors can successfully
cope with almost any challenge in the process gas
technology - all over the world.
CHALLENGE US !
www.aerzen.com
RELIABILITY
3
.
ARE YOU READY FOR MORE EFFICIENCY
IN YOUR PROCESS GAS CONVEYANCE?
PTQ_120x190.indd 1 11.03.14 12:41
q2 modicon.indd 8 11/03/2014 12:13
O
ryx GTL slurry phase distillate
(SPD) raw wax catalyst slurry is
the system where Everlasting Valves are
installed. This unique valve was designed
nearly 100 years ago for blowing down steam
locomotive boiler solids. The U.S. Department
of Energy in development of the clean coal
technologies selected this valve as state of the art
valve for dry solids and slurries.
A major licensor of rening processes now
species the Everlasting rotating disc metal seated
valve for Fluid Catalyst Crackers handling fresh
catalyst and for hot CAT withdrawal. It is also listed in
the Best Practices of a world class petroleum rener
for continuous catalyst reforming. A preeminent
catalyst additive supplier has standardized on this
valve concept for all their chemical injection units.
Spent catalyst
recyclers use
them on their
vacuum truck load
out systems. The
flat rotating disc
renews the sealing
surfaces each time
the valve is cycled. No
other valve is similar, with vacuum through 10,000
psig (689b) and temperatures to 1500F (815C). See
more specications on the reverse side.
Whether its fresh, spent, dry or slurry use
the Everlasting Rotating Disc Valve and see your
CAT valve problems disappear. There are many
ways of skinning the CAT, but one best valve for
catalyst slurries.
Call us today at 1-908-769-0700 or visit our website at (www.everlastingvalveusa.com)
American Boiler Manufacturers Association
Theres more
than one
way to skin the
CAT!
VALVE COMPANY, INC.

Everlasting
Valve Manufacturers Since 1906

Boiler Blowoff Valves
Bulk Material Valves
Process Valves
everlasting.indd 1 09/12/2013 11:27
Performance of waxy crudes as FCC feeds
W
axy crudes from forma-
tions in the US state of
Utah were considered
intractable just fve years ago, but
the relatively higher prices of
conventional crudes make the
economics of these Utah crudes
transportation by truck or rail car
acceptable.
In this area Newfeld Exploration
has a long-term commitment to
supply approximately 38 000 b/d of
the waxy crude to refneries in the
area. Newfeld is the largest oil
producer in Utah with current
production of 39 000 barrels of oil
equivalent per day (boe/d). Total
waxy crude production from this
region is 65 000 b/d, with
Newfelds share comprising 45%.
Deposits of 700 million boe have
been reported in the region, which
seems to indicate a reliable and
long-term supply of the crude.
At the onset, the transportation
logistics to refnery processing
centres were complex and there was
a steep learning curve for the
midstream infrastructure and refn-
ers processing these waxy crudes,
such as fouling at all locations of
the transportation network, plug-
ging of pipes and rail car unloading.
Nonetheless, there are compelling
reasons for refners wanting to
process these waxy crudes.
The high paraffnic content of
these waxy crudes makes them
suitable for certain lubes plants. For
those refning facilities with coker
units, the waxy crudes low metals
content also makes them suitable
for anode grade coke feedstocks.
The crudes low content of light
ends allows for blending with
conventional FCC feeds, allowing
Advantages and challenges: determining the optimal blending ratios with waxy
crude and VGO combinations
ERICK GAMAS, CLIFF AVERY, GERBRAND MESU and DAVID DE VILLIERS Albemarle Corporation
ALAN YAHEV Neweld Exploration Company
for complete bypassing of the crude
distillation unit (CDU). The low
sulphur and low metals (Ni and V)
contents results in lower sulphur
FCC fuels and longer catalyst life.
Crude feed sources and their frac-
tions (vol%) are shown in Figure 1
for conventional gas oils, tight oils,
waxy crudes and others. Small
amounts of resid and some distillate
fractions are present in the conven-
tional gas oils (KVGO and PVGO).
Tight oils present signifcant frac-
tions of light ends that require
processing through the crude unit.
In comparison, the waxy crudes
(yellow wax, black wax and Uintah
crude) contain less than 10% naph-
tha and can therefore be blended
with FCC/RFCC feeds.
Advantages and challenges
Waxy crude prices are lower than
conventional crudes and additional
energy savings are incurred by
bypassing the CDU for direct crack-
ing in the FCC. The low Ni and V
in these waxy crudes results in rela-
tively low rates of FCC catalyst
deactivation and easier cracking.
This process advantage is why these
feeds can be potentially used for
blending with much heavier cheap
feeds. Processing waxy crudes
through the FCC produces higher
volume gains and higher percent-
ages of high value added products,
such as higher LCO cetane.
Although waxy crudes character-
istics are such that they provide
refners with the fexibility to
bypass certain process steps, such
as circumventing the CDU for
direct upgrading in the FCC unit,
these crudes are not without their
challenges, such as lower RON in
FCC gasoline. Beginning with a
lack of compatibility when blend-
ing with other streams, their high
wax fractions lead to asphaltene
precipitation and low carbon resi-
due for delta coke. FCC engineers
60
100
90
80
70
50
40
30
20
10
V
o
l
u
m
e
,

%
0
Distillate
Gas oil
Resid.
Naphtha
LPG
Crude feed
A
r
a
b

L
i
g
h
t
W
T
I
Y
e
l
l
o
w

w
a
x
B
l
a
c
k

w
a
x
U
i
n
t
a
h
T
i
g
h
t

o
i
l
R
C
C
K
V
G
O
P
V
G
O
Figure 1 Distillation of feed sources
shale albemarle.indd 1 12/03/2014 12:16
nonetheless a consistent increase in
conversion and gasoline yields. In
addition, the overcracking conver-
sion limit can be expanded by using
feeds with higher parafn content.
There is a consistent loss of LCO
with conversion through the FCC
unit and a corresponding LPG
increase. With higher feed
parafnicity comes the potential for
higher cetane LCO, but with lower
yields. With these combinations of
feeds (YW, BW, Uintah-Wax,
PVGO, KVGO, and so on), an
economic LCO/LPG optimum has
been identied with an apparent
trend underlying the yield patterns.
The underlying yield patterns indi-
cated in Figure 2 reveal an optimal
blending ratio to balance LCO/LPG.
This apparent trend shows conver-
sion of LCO to LPG without loss of
gasoline. In this case, LCO cetane
and gasoline RON, in addition to
yields, are optimisation indexes.
Higher yields of LPG C
3
and C
4
fractions shown in Figure 3 follow a
trend consistent with increased feed
parafnicity, where the impact of
catalyst combined with feed effect
are revealed at highest severity. The
gure shows a 5% C
4
s delta with 1.5
higher CTO at 932F with neat
waxes. The same underlying trend
with feed quality is observed in
propylene yields, where propylene
yields show a strong response to
feed quality (for example, +5.3 wt%
from PVGO) and cracking severity
(+1.88 wt% for 63F higher
temperature).
A linear decrease in C
3
and C
4
olenicity is observed as feed
improves. For example, a 3.0%
reduction in C
3
olenicity and a
22% reduction in C
4
olenicity are
observed with lowest values corre-
sponding to yellow wax.
Parafnic bottoms can be
converted to extinction at high
severity. It can be demonstrated
that feed synergy is observed with
coke yields of waxy crudes and
VGO blends, where coke yield of
blends noted in the previous
gures is similar to VGO coke. For
example, coke yield from PVGO is
higher than coke yield from PVGO-
Wax blend, with CCR at a higher
magnitude than the PVGO (by a
factor of ve [5X]).
must be cautious of gas production
constraints and occlusion of exter-
nal FCC catalyst pore structure
from the crudes high Fe content,
resulting in loss of catalyst activity.
Careful selection of high accessibil-
ity catalyst technology is key for
activity and selectivity control.
From lab to commercial operations
Laboratory testing programs are an
accurate and cost effective tool
when process similarities at the
commercial scale are considered.
Information obtained from lab test-
ing provides directional information
to infer commercial performance,
such as determining the yield struc-
ture response in an FCC reactor
from patterns of feed and catalyst
contacting. Supporting process
models yield estimates for individ-
ual operations. In addition, use of
E-cat in lab testing for feed screen-
ing provides better representation of
commercial conversion than lab
deactivated catalysts.
In designing a testing program to
assess the potential of waxy crudes
as FCC feeds, a proprietary
Albemarle Fluid Simulation Test
(FST) lab reactor was used for waxy
crude cracking tests. In these
instances, processing through the
CDU was bypassed. Catalytic
cracking FST tests of neat waxy
crudes and blends with VGO were
carried out using a high accessibil-
ity E-cat as shown in Table 1.
Feed properties vary from conven-
tional VGOs to neat waxy crudes.
For example, black wax has high
Concarbon of 3.5 as compared to 0.3
for PVGO. High sodium feeds such
as 70/15/15 Uintah/YW/BW (2542
ppm) require an FCC feed desalter.
Very high Fe requires high accessi-
bility and Fe tolerant catalyst.
Cracking yields
While lower RON can be expected
when processing FCC feeds with
higher parafn content, there is
E-cat: Z/M = 1.2; REO = 2 wt%
FST 932F & 995F
conditions: FST CTO: 2, 4, 6
Feeds: Yellow wax (YW)
Black wax (BW)
Purchased vacuum gas oil (PVGO)
Reference vacuum gas oil (KVGO)
Uintah crude
Feed 50/50 YW/BW
blends: 70/15/15 PVGO/YW/BW
70/15/15 Uintah/YW/BW
FST wax cracking testing program
Table 1
UOP K
() (+)
30
50
60
40
20
10
Y
i
e
l
d
s
,

w
t
%
0
40 50 60 70 80 90 100
Conversion, wt%
Gasoline 1:1 Y/Bk W
Gasoline Y W
Gasoline Bk W
Gasoline K VGO
Gasoline P VGO
Gasoline P VGO-Wax
Gasoline Uintah-Wax
LCO 1:1 Y/Bk W
LCO Y W
LCO Bk W
LCO K VGO
LCO P VGO
LCO P VGO-Wax
LCO Uintah-Wax
LPG 1:1 Y/Bk W
LPG Y W
LPG Bk W
LPG K VGO
LPG P VGO
LPG P VGO-Wax
LPG Uintah-Wax
CTO = 7.5
932F/995F
Gasoline
LCO
CTO = 2,4,6
932F
T effect
x = 70 wt%
x = 76 wt%
LPG
Figure 2 Cracking yields: gasoline, LCO and LPG yield patterns
Feed CCR from waxy crudes is
not reected in FST coke yields,
where an inverse relationship is
actually observed. FST coke yields
response to CTO is linear, where a
change in slope and trend is
observed with blends revealing a
possible synergy in behaviours
when combining feeds (see Figure
4). Feed CCR is not reected on
delta coke due to high parafnicity
of feed bottoms in waxy crude mix.
For example, Uintah-Wax mix
behaves like a conventional VGO
despite having a higher parafnic-
ity than KVGO.
Conclusions and recommendations
Results from lab cracking tests with
a high accessibility E-cat support
business opportunities for commer-
cial operations using waxy crudes,
partially replacing VGO/resid frac-
tions. Use of black wax requires an
FCC feed desalter as practised by
major reners buying VGO.
An optimal blending ratio was
identied, driven by LCO/LPG as
well as cetane/RON economics for
individual reners. Waxy crudes in
VGO impart higher cetane numbers
to LCO products, whereas lower
gasoline RON requires a balanced
management approach. Identifying
potential synergies of feeds in wax/
VGO blends is a key benet for
reners. In any event, reners
should maintain close communica-
tion with the FCC catalyst
technology supplier to optimise the
catalyst formulation and particle
accessibility when evaluating oppor-
tunities to process wax/VGO feeds.
FST is a mark of Albemarle Corporation.
This article is based on a presentation for the
5th Annual World Rening Technology & Shale
Processing Summit in Houston, Texas, 4-5
December 2013.
Erick Gamas, PhD, is Senior FCC Technical
Service Representative at Albemarle
Corporation.
Email: erick.gamas@albemarle.com
Clifford Cliff Avery is Technical Services
Manager, FCC Americas, at Albemarle
Corporation.
Email: clifford.avery@albemarle.com
Gerbrand Mesu, PhD, is Applied Product
Research Manager, Heavy Oil Upgrading
Division, Albemarle Corporation.
Email: gerbrand.mesu@albemarle.com
David De Villiers, PhD, is Senior Research
Chemist, Heavy Oil Upgrading Division,
Albemarle Corporation.
Email: david.devilliers@albemarle.com
Alan Yahev is Marketing Manager at Neweld
Exploration Company.
UOP K
() (+)
15
25
30
20
10
5
Y
i
e
l
d
s
,

w
t
%
0
40 50 60 70 80 90 100
Conversion, wt%
CTO = 2,4,6
932F
C
4
s
C
4
C
3
s 1:1 Y/Bk W
C
3
s Y W
C
3
s Bk W
C
3
s K VGO
C
3
s P VGO
C
3
s P VGO-Wax
C
3
s Uintah-Wax
C
4
s 1:1 Y/Bk W
C
4
s Y W
C
4
s Bk W
C
4
s K VGO
C
4
s P VGO
C
4
s P VGO-Wax
C
4
s Uintah-Wax
C
4
olefins 1:1 Y/Bk W
C
4
olefins Y W
C
4
olefins Bk W
C
4
olefins K VGO
C
4
olefins P VGO
C
4
olefins P VGO-Wax
C
4
olefins Uintah-Wax
T effect
CTO = 7.5
932F/995F
C
3
s
Figure 3 FST LPG selectivity
A
B
0 1 2 3 4 5 6 7 8
FST cat/oil
0 1 2 3 4 5 6 7 8
FST cat/oil
3
4
2
1
F
S
T

c
o
k
e

y
i
e
l
d
s
0
CCR
CCR
+
+
1Yw:1Bw (CCR = 1.79)
PVGO (CCR = 0.37)
KVGO (CCR = 0.19)
3
5
4
2
1
F
S
T

c
o
k
e

y
e
l
d
s
,

w
t
%
0
PVGO (CCR = 0.37)
PVGO-Wax (CCR = 0.79)
Uintah-Wax (CCR = 2.45)
Figure 4 CCR effect on coke yields
world congress.indd 1 06/03/2014 13:54
Predicting severity, viscosity and yields
in fuel oil visbreaking
C
rude oils contain a large frac-
tion of heavy products for
which only few useful outlets
exist. Indeed, world demand for
light and middle distillate continu-
ally increases, while at the same
time the available crude oil becomes
heavier.
1
Visbreaking is a non-cata-
lytic thermal process that converts
atmospheric or vacuum residues via
thermal cracking to gas, naphtha,
distillates, and visbroken residue. A
visbreaker thermally cracks large
hydrocarbon molecules in oil by
heating them in a furnace to reduce
their viscosity and to produce small
quantities of light hydrocarbons
(LPG and gasoline). The process
name of visbreaker refers to the
fact that the process reduces the
viscosity of the residual oil.
2,3
Two types of visbreaking technol-
ogy are commercially available: the
coil or furnace type and the
soaker process. In the coil process,
conversion is achieved by high
temperature cracking for a prede-
termined, relatively short, period of
time in the heater. In the soaker
process, a low temperature/high
residence time process, the majority
of conversion occurs in a reaction
vessel or soaker drum, where the
two-phase heated effuent is held at
a lower temperature for a longer
period of time. Therefore its heater
duty and, in turn, its fuel consump-
tion is only 70% of that for the
coil-visbreaking process.
4
Worldwide, about 200 visbreak-
ing units are in operation; Europe
alone accounts for about 55% of
total visbreaking capacity.
4
To have
an effective design and perfect
control of any process, a model is
needed to predict product yields
Neuro-fuzzy technique is used to predict product properties in a
commercial visbreaker
SEPEHR SADIGHI and REZA SEIF MOHADDECY
Research Institute of Petroleum Industry
and qualities versus variables such
as space velocity and temperature.
However, the complexity of
visbreaking feed and product
makes it extremely diffcult to char-
acterise and describe its kinetics at
a molecular level. Modern day
rigorous simulators such as Aspen
Plus or Hysys from Aspen
Technology do not have such
restrictive limits on the total
number of components; therefore, it
is possible to use a unique set of
pseudo components for every
petroleum assay stream.
5
This
approach increases the calculation
time, and characterisation of the
streams and subsequent reports
become unnecessarily complicated.
5
One approach to simplify the
problem is to consider the partition
of the species into a few equivalent
classes, the so-called lumps or
lumping technique, and then
assume each class as an independ-
ent entity. Developing simple
kinetic models (for instance, power-
law model) for complex catalytic
reactions is a common approach,
and it can give basic information for
reactor design and optimisation. In
this feld, many investigations were
reported in which the visbreaking
process was modelled with discrete
two-lump,
6-8
three-lump,
9
four-
lump,
10,11
fve-lump
12,13
and
seven-lump
14
approaches. In all of
these investigation, the experiments
were carried out in a micro or pilot
scale reactor. However, over the last
two decades, soft computing meth-
ods such as artifcial neural
networks and fuzzy logic were
migrated into the realm of refnery
processes for modelling, control,
optimisation and kinetic estimation.
The main advantage of this
approach, named black box model-
ling, is its independence of any
assumption or equation about
kinetic study. Moreover, it is not
needed to ignore the effect of some
parameters to obtain a simpler
mathematical expression.
The aim of this research is to use
the neuro-fuzzy methodology for
predicting the severity, product
yields and viscosity of fuel oil for
an industrial scale visbreaking unit
equipped with a soaker drum. The
performance of the neuro-fuzzy
model with different membership
functions is compared with indus-
trial data obtained from an Iranian
refnery.
Feed characterisation
An industrial soaker-visbreaker
unit was chosen as a case study.
This unit was designed to visbreak
20 000 b/d of a mixture of vacuum
residuum and slop vacuum gas oil,
both taken from a vacuum tower.
The composition of the fresh feed
can vary slightly with time from
Property Value
Sp.Gr. 1.006
Sulphur content, wt% 3.19
Va + Ni content, wt ppm 188
Distillation (ASTM D1160)
Vol% Temperature , C
Initial boiling point 303
5 409
10 457
20 503
30 543
50 585
Feed characterisation
Table 1
q2 ripi.indd 1 10/03/2014 14:47
Neuro-fuzzy modelling approach
The connection of fuzzy systems
with an artifcial neural network
(ANN) is called a neuro-fuzzy (NF)
system. Similar to neural networks
where knowledge is saved in
connection weights, it is interpreted
as fuzzy if-then rules in NF
systems. The most frequently used
neural network in NF systems is
radial basis function neural
network (RBFNN) in which each
node has radial basis function such
as Gaussian and ellipsoidal.
The popularity of NF systems is
due to the simplicity of structure,
well-established theoretical basis
and faster learning than in other
types of neural networks. There are
many developed fuzzy neural
networks (FNN) as NF algorithms
in the literature, but the adaptive
network based fuzzy inference
system (ANFIS) is one of the best
known. In ANFIS, Takagi-Sugeno
type fuzzy inference system is
used. The output of each rule can
be a linear combination of input
variables plus a constant term or
can be only a constant term. The
fnal output is the weighted aver-
age of each rules output. In the
forward pass, the ANFIS uses the
least squares method to identify the
parameters in the network layers.
In the backward pass, the errors are
propagated backward and related
parameters are updated using the
gradient descent method. The
ANFIS can be also trained by a
hybrid algorithm to identify the
membership function parameters.
15
To create a neuro-fuzzy inference
system for the understudy
visbreaking plant, Matlab Fuzzy
Logic Toolbox version 2010a from
MathWorks and ANFIS syntax
were used. This syntax is the major
training routine for Sugeno-type
fuzzy inference systems. ANFIS
uses a hybrid learning algorithm to
identify the parameters of Sugeno-
type fuzzy inference systems. It
applies a combination of the least-
squares method and the
back-propagation gradient descent
method for training fuzzy inference
system membership function
parameters to emulate a given
training data set. The type of
membership functions associated
start of run (SOR) to end of run
(EOR). The specifcation of the
combined feed, which was obtained
during this research, is shown in
Table 1.
Process description
The visbreaking feed is charged to
the coil furnace at a temperature of
about 340C. The visbreaking
furnace is constructed in two
sections, fred independently. After
the coil furnace, the two hot
streams are drained into a transfer
line and the mixed product enters
the soaker drum. The specifcations
of cells and the soaker drum are
presented in Table 2. The output
product from the soaker drum is
quenched by the cold recycle
stream to stop cracking reactions
and inhibit coke formation. Finally,
the combined stream is transferred
to the fractionation tower and side
strippers to separate the visbreak-
ing products. A simplifed process
fow diagram of the described unit
is shown in Figure 1.
During one year of data gather-
ing, 46 sets of data, including
product fow rates, feed inlet
temperature and soaker outlet
temperature, were gathered from
the target visbreaking process.
As Figure 2 shows, light gases
including C
1
, C
2
and LPG, gasoline
and tar are the output streams from
the visbreaking plant. Performing
mass balance around the unit
proved that the error for all experi-
ments was less than 5%, mainly
related to the gross error for the
measuring of the gaseous products
and maybe related to the coke
formation. The boiling range of
VGO feed, fuel and gas oil samples
were analysed according to the
ASTM D1160 standard procedure
whilst the one for the gasoline
sample was analysed according to
the ASTM D86 method.
Additionally, the viscosity of the
visbreaking product (fuel oil) was
measured according to the ASTM
D2170 method.
Coil specication
Number of tubes 128
Number of convection tubes 76
Number of radiation tubes 52
Tube length, m 18.745
Outside diameter, m 0.114
Soaker specication
Outside diameter, m 2.405
Length, m 16.5
Specications of the cell and soaker
of the visbreaking unit
Table 2
Furnace Soaker
Fractionator
Stripper
Stabiliser
Light gas
LPG
Gasoline
Light gas
Tar
Feed
Steam
Figure 1 Block ow diagram of visbreaking process
q2 ripi.indd 2 10/03/2014 14:48
with a visbreaking unit was
selected from all supported types in
Matlab. To train the fuzzy model in
an optimised time, two fuzzy rules
were selected from the ANFIS tool-
box, and the training process was
stopped whenever the designated
epoch number (100) was reached.
To compare the simulated and
measured values, mean average
absolute deviations (AAD%) and
mean squared error (MSE) were
calculated as follows:

(1)

(2)
where N
t
, Y
jn
meas
and Y
jn
pred
are the
number of test runs, the measured
variable obtained from the
visbreaking plant, and the
predicted value using the ANFIS
model, respectively. Moreover, j
designates severity of process (1),
yield of products fuel oil (2),
gasoline (3), and gas (4) and the
viscosity of the fuel oil (5).
Results and discussion
During the feld study, 46 sets of
data composed of product fow
rates, composition of gaseous prod-
ucts, fow rates of feed and steam,
distillation curve of cuts, and soaker
temperature were gathered from
the target visbreaking plant. The
Petro-SIM process simulator from
KBC, Version 3.2, was employed to
lump the feed and products into
components with the specifc boil-
ing point ranges and properties: gas
(C
1
-C
4
), gasoline (180C), gas oil
(180-320C) and fuel oil (320+C).
Due to the low fow rate of gas oil
(<350 kg/h), its yield was not
important for the target refnery
from both economical and opera-
tional standpoints. Moreover, this
product was recycled to the inlet
stream of the reactor and mixed
with the vacuum residue feed.
Therefore its yield was neglected.
From a set of 46 data points
obtained from the visbreaking
plant, 36 points (~78%) were
selected for training the NF model,
and the remaining ones (~22%) for
validating the developed network.
The MSE of trials with different
membership functions, using the
epoch number of 100 and two
fuzzy languages, is shown in Table
3. It was found that for the sever-
ity, yields of process and viscosity
of fuel oil, simulating the results of
the triangular function were in
better agreement with measured
data. Moreover, it was found that
Gaussian membership was approxi-
mately as accurate as the triangular
one. But, for simulating the viscos-
ity of fuel oil, the Gaussian-shape
curve was the best choice.
Additionally, it was confrmed that
increasing the number of epoch
numbers and fuzzy languages did
not considerably decrease the MSE
of simulating. But the time of train-
ing was considerably increased.
Therefore, the epoch number of 100
and two fuzzy languages were
confrmed to build and train the NF
model.
After training the NF inference
system using ANFIS, the output
variables for the remaining data,
unseen data, were predicted by the
Evalfs syntax of Matlab. The parity
plots to compare the measured and
predicted outputs are shown in
Figures 2-6. As was described before,
for simulating the viscosity of fuel
oil, Gaussian membership function
was used, and triangular function
was applied to simulate the severity
of process and product yields. As
can be understood from these
fgures, there is an acceptable agree-
ment between the experimental and
predicted values for all visbreaking
output variables.
Additionally, the AAD% and
MSE of predicted variables is
presented in Table 4. As can be
observed, the predicted values for
the validating data are in good
agreement with the actual values. It
is supposed that the observed devi-
ation is caused by some fuctuations
in the property of the feed (such as
sulphur, nitrogen and aromatic
content), reactor pressure and also
coke deposition on the reactor
tubes. Furthermore, we cannot
Cumulative Expenditure,
expressed as
a % of total Installed Cost (TIC)
Factored off
Equipment
Factored off
Equipment
105
115
120
110
100
95
P
r
e
d
i
c
t
e
d

s
e
v
e
r
i
t
y

90
90 95 100 105 110 115 120
Measured severity
Training
Validating
Figure 2 Comparison of the measured and predicted data points for severity
%
) (
100 %
1
2
2
5
1
t
N
n
meas
jn
pred
jn
meas
jn
j
N
Y
Y Y
AAD
t
= =
=
MSE =
(Y
jn
meas
!Y
jn
pred
)
2
j=1
5
"
n=1
N
t
"
N
t

Sigmoid Bell Gaussian Trapezoidal Triangular
shape shape shape shape shape shape
Severity 6.80E-6 1.65E-7 2.92E-08 3.71E-7 2.84E-08 9.83E-7
Gas yield, wt% 6.85E-9 1.35E-10 1.28E-10 4.39E-10 1.51E-11 8.49E-10
Gasoline yield, wt% 7.84E-09 3.58E-10 7.47E-11 1.23E-09 3.82E-11 1.75E-9
Fuel oil, wt% 1.10E-5 7.07E-8 3.46E-08 1.80E-7 1.11E-8 6.07E-7
Viscosity, CSt 5.42E-6 1.12E-07 1.22E-8 1.21E-7 1.91E-8 3.14E-7
MSE of different membership functions for data simulated by ANFIS for training data
Table 3
q2 ripi.indd 3 10/03/2014 14:48
visbreaker: severity of process,
yield of products, including gas
(C
1
-C
4
), gasoline (IBP 180C) and
fuel oil (320+C), and viscosity of
product. This model considered the
infuence of feed and steam fow
rates, reactor tube outlet tempera-
ture, initial boiling point and
viscosity of the feed. At frst, 36
data points were used to build and
train a NF model using Matlab
Fuzzy Logic Toolbox and ANFIS
syntax. The training results show
that the best membership function
for simulating the severity and
product yields of the process was a
triangular shape, whilst the
Gaussian membership function was
the best choice for predicting the
fuel oil viscosity. Using these func-
tions, the NF model could simulate
the mentioned output variables
with an MSE of 2.84E-08, 1.51E-11,
3.82E-11, 1.11E-8 and 1.12E-07,
respectively.
Next, the remaining test runs (10
data points) were applied to vali-
date the NF model. Results showed
that predicted severity, product
yields and viscosity of fuel oil were
in good agreement with actual data
with an average absolute deviation
(AAD%) of 0.77%, 0.785%, 0.917%,
0.103% and 0.373%. Additionally,
the mean squared error (MSE) of
prediction for these variables was
1.15, 0.002, 0.003, 0.02 and 0.155,
respectively.
References
1 Sadighi S, Arshad A, Rashidzadeh M, 4-Lump
kinetic model for vacuum gas oil hydrocracker
involving hydrogen consumption, Korean. J.
Chem. Eng. 27(4), 2010, 1099-1108.
2 Gary J, Handwerk G, Petroleum Rening
Technology and Economics, Marcel Dekker, Inc.,
1984.
3 Speight J, The Chemistry and Technology of
Petroleum, CRC Press, 2006.
Conclusions
In this research, a NF modelling
approach was applied to predict
momentous output variables in a
commercial scale vacuum residue
neglect the possibility of error
measurements in gathering data
obtained with some faults such as
signal transmission, calibration and
power fuctuation.
4.6
5.0
4.9
4.8
4.7
4.5
4.4
4.3
4.2
4.1
P
r
e
d
i
c
t
e
d

g
a
s

y
i
e
l
d
4.0
4.0 4.1 4.2 4.3 4.4 4.5 4.6 4.7 4.8 4.9 5.0
Measured gas yield
Training
Validating
Figure 3 Comparison of the measured and predicted data points for gas yield
5.2
6.0
5.8
5.6
5.4
5.0
4.8
4.6
4.4
4.2
P
r
e
d
i
c
t
e
d

g
a
s
o
l
i
n
e
4.0
4.0 4.2 4.4 4.6 4.8 5.0 5.2 5.4 5.6 5.8 6.0
Measured gasoline
Training
Validating
91.5
92.5
93.0
92.0
91.0
90.5
P
r
d
i
c
t
e
d

f
u
e
l

o
i
l
90.0
90.0 90.5 91.0 91.5 92.0 92.5 93.0
Measured fuel oil
Training
Validating
Figure 4 Comparison of the measured and predicted data points for gasoline yield
Figure 5 Comparison of the measured and predicted data points for fuel oil yield

MSE AAD%
Severity 1.15 0.77
Gas yield, wt% 0.002 0.785
Gasoline yield, wt% 0.003 0.917
Fuel oil, wt% 0.020 0.103
Viscosity, CSt 0.155 0.373
MSE and AAD% of data simulated by
ANFIS for validating data
Table 4
q2 ripi.indd 4 10/03/2014 14:48
Reaction Engineering Department, Research
Institute of Petroleum Industry (RIPI), Tehran,
Iran. He holds a PhD in chemical engineering
from Universiti Teknologi Malaysia.
Email: Sadighis @ripi.ir
Reza Seif Mohaddecy is Project Manager,
Catalysis and Nanotechnology Division,
Catalytic Reaction Engineering Department,
RIPI. He holds a MS in chemical engineering
from Sharif University of Technology.
Email: Seifsr @ripi.ir
visbreaking, Ind. Eng. Chem. Res., 43, 2004.
14 Bellos G D, Kallinikos L E, Gounaris
C E, Papayannakos N G, Modeling of the
performance of industrial HDS reactors using
a hybrid neural network approach, Chem. Eng.
Process, 44, 2005, 505515.
15 Abonyi J, Babuska R, Szeifert F, Modied
Gath-Geva fuzzy clustering for identication
of Tagaki Sugeno fuzzy models, IEEE Trans. Syst.
Man Cybern, 32, 2002, 612.
Sepehr Sadighi is Assistant Professor, Catalysis
and Nanotechnology Division, Catalytic
4 Sawarkar A, Joshi J, Pandit A, Kataria K,
Kulkarni K, Tandon D, Ram Y, Petroleum residue
upgrading via visbreaking: a review, Can J.
Chem. Eng., 85, 2007, 1-24.
5 Sadighi S, Ahmad A, An optimisation
approach for increasing the prot of a
commercial VGO hydrocracking process, Can.
J. chem. Eng., 13, 2013, 1077-1091.
6 Sou H.Al, Savaya Z , Moahmmed H, Al-
Azami I, Thermal conversion of heavy Iraqi
residue, Fuel, 67, 1988, 1714-1715.
7 Krishna R, Kuchhal Y, Sarna G, Singh I,
Visbreaking studies on Aghajari long residue,
Fuel, 67, 1988, 379-383.
8 Carlo S.Di, Janis B, Composition and
visbreakability of petroleum residues, Chem.
Eng. Sci., 47, 1992, 2695-2700.
9 Benito A M, Martinez M T , Fernandez I,
Miranda J L, Visbreaking of an asphaltenic coal
residue, Fuel, 74, 1995.
10 Del Bianco A, Panartili N, Anelli M, Beltrame
P, Carniti P, Thermal cracking of petroleum
residues 1. Kinetic analysis of the reaction, Fuel,
72, 1993, 75-80.
11 Trauth D, Yasar M, Neurock M, Nasigam
A, Klein M, Kukes S, Asphaltene and resid
pyrolysis: effect of reaction environment, Fuel
Sci. Technol. Int., 10(7), 1992, 1161-1179.
12 AlHumaidan F, Lababidi M, Al-Rabiah H,
Thermal cracking kinetics of Kuwaiti vacuum
residues in Eureka process, Fuel,103, 2013, 923-
931.
13 Kataria K L, Kulkarni R P, Pandit AB, Joshi J
B, Kumar M, Kinetic studies of low severity
91.5
92.5
93.0
92.0
91.0
90.5
P
r
d
i
c
t
e
d

v
i
s
c
o
s
i
t
y
90.0
90.0 90.5 91.0 91.5 92.0 92.5 93.0
Measured viscosity
Training
Validating
Figure 6 Comparison of the measured and predicted data points for the viscosity of
fuel oil
Ivelina Shishkova is R&D Department
Manager with Lukoil Neftohim Burgas. She
holds a MS in organic chemistry engineering
and a PhD in petroleum rening from Soa
Chemical and Technological and Metallurgical
University, and has authored more than 20
technical papers.
Email: Shishkova.Ivelina.K@neftochim.bg
Rosen Dinkov is the Quality Manager in the
Process Engineering department of Lukoil
Neftohim Burgas. His research interests include
crude oil characterisation, bio/conventional
fuels blends characterisation and modelling of
renery distillation processes. He holds a MS
in organic chemistry engineering from Burgas
University and a PhD in the technology of
fossil and synthetic fuels from the University
of Chemical Technology and Metallurgy, Soa.
Email: Dinkov.Rosen.K@neftochim.bg
Vladimir Jegorov is the Sales Development
Manager for Grace in the CIS region. Prior to
joining Grace, he was an FCC process engineer
at the Mazheikiai renery in Lithuania.
Petko Petkov is a full professor and rector
of the Burgas University Assen Zlatarov. He
teaches in the social science department in
the eld of oil rening and lubricants, and has
authored more than 180 scientic papers and
ve books.
Email: PST_Petkov@abv.bg
higher FCC C
4
yield led to higher
production of alkylate, which
resulted in the production of 2%
more premium grade gasoline.
References
1 Watanbe K, Nagai K, Aratani N, Saka Y,
Chiyoda N, Mizutani H, Techniques for octane
enhancement in FCC gasoline, 20th Annual
Saudi-Japan Symposium, Dhahran, December
2010. 17. Montgomery J A, Guide to Fluid
Catalytic Cracking, Part 1, 1993.
Ivan Chavdarov is a Chemical Engineer in
the Process Engineering department of Lukoil
Neftohim Burgas, Bulgaria. His activities are
focused on guiding the operation of the units
of the FCC complex, troubleshooting support
and optimisation of the performance of the
FCC complex.
Email: Chavdarov.Ivan.S@neftochim.bg
Dicho Stratiev is Chief Process Engineer with
Lukoil Neftohim Burgas. He holds a MS in
organic chemistry engineering, and a PhD and
a DSc in petroleum rening from the Burgas
University Assen Zlatarov. He has authored
more than 130 papers.
Email: Stratiev.Dicho@neftochim.bg
Regular A-92
Premium A-95
Super A-98
RVP = 60kPa
1%
28%
71%
RVP = 50kPa
1%
26%
73%
Figure 3 Effect of changing the RVP on
renery gasoline grades produced during
the Resolution catalyst period
MINIDIS ADXpert
Grabner Instruments Messtechnik GmbH | A-1220 Vienna, Austria | Dr. Otto-Neurath-Gasse 1
Phone +43-1-282 16 27-0 | Fax +43-1-280 73 34 | info.grabner-instruments@ametek.at
QUICK, ACCURATE, PORTABLE
DISTILLATION
Portable, True Atmospheric D7344 Distillation
Excellent D86, ISO 3405, IP 123 Correlation
Fully Automated Measuring and Cleaning
6 mL of Sample
No Glassware, no Open Flame
Distillation of Gasoline, Jet Fuel, Diesel, Solvents,
Unknown Samples, Raffinates and Biodiesel
Specified for Use by US Pipelines
ADXpert half page 1.indd 1 06.12.2013 11:46:37
grace.indd 4 11/12/2013 12:49 q2 ripi.indd 5 10/03/2014 14:48
Recovering tungsten from spent
hydrocracking catalysts
V
aldi, located near Limoges in
France, specialises in the
recycling of industrial waste,
more specifcally stainless steel
scrap along with metal hydroxide
sludges and spent catalysts gener-
ated by oil refneries in its
hydrotreating processes.
In partnership with Eurecat and
LElectrolyse, Valdi has developed
an effcient recycling process to
recover tungsten from spent
nickel-tungsten catalysts (NiW cata-
lysts). Among the spent catalysts
which are treated, those containing
nickel and tungsten are used to
provide secondary tungsten for the
European market in the form of a
concentrated tungsten salt, as
opposed to the primary tungsten
extracted from mines.
All of these operations have been
running for more than 10 years in
conformity with European regula-
tions including the Seveso
directives.
Why recycling?
The European market for spent
catalysts amounts to 20 000 tonnes
of material. Of this total, only 1500
tonnes are spent NiW catalysts.
Nonetheless, tungsten is a strategic
metal as defned by Bureau de
Recherches Gologiques et Minires
(BRGM), the French geological
survey (see Figure 1). Only two
countries in the world are net
exporters of tungsten and can
easily control the price of ore.

NiW catalyst composition
Refners use NiW catalysts for
hydrocracking processes. The cata-
lysts are placed in a column into
which the feedstock is introduced
When catalyst is no longer serviceable for hydrotreating, it takes a metallurgical
route to supply tungsten for steelmakers
JEAN-PIERRE DUFOUR and SOPHIE COMTE
Valdi
under pressure with hydrogen.
Different classes of catalysts are
stacked in the column, one above
the other, often separated by
ceramic beds (see Figure 2). In most
instances, NiW catalysts are posi-
tioned just under a layer of molyb-
denum-nickel catalyst (NiMo
catalysts). This arrangement
explains the formation of mixtures
of NiW spent catalysts and NiMo
spent catalysts. The entire recycling
process begins with spent catalysts
being classifed and stored accord-
ing to the characteristics of the
material.

Recovering the tungsten
The NiW catalysts are sorted into
two sub-categories:
NiW catalysts without molybde-
num
NiW catalysts with a molybde-
num content higher than 0.8%.
Valdi is able to treat NiW cata-
lysts and the mixtures containing
molybdenum. Table 1 shows the
average composition of NiW cata-
lysts received in 2013. Depending
on the composition of each catalyst
sample, the process recovers either
molybdenum together with
Ceramics
Ceramics H
2
Ceramics
NiMo
NiMo
NiMo
NiMo
NiW
CoMo
Oil
Petrol
Figure 1 2011 panorama market for tungsten
High criticality area
Precautionary measures
to be taken by the state.
Tracking the evolution
of indicators of criticality.
High criticality area
Active standby recommended
(continuous market
monitoring, alerts, proposed
alternative scenarios.
Average criticality area
Specialised watch recommended
(a report updated annually).
High
Very high
Medium
Low
S
t
r
a
t
e
g
i
c

i
m
p
o
r
t
a
n
c
e

f
o
r

t
h
e

F
r
e
n
c
h

i
n
d
u
s
t
r
y
Very low
Very
low
Low Medium High Very
high
Risks to supplies
W
Figure 2 Hydrocracking column
cat valdi.indd 1 12/03/2014 12:07
together with spent catalyst. In the
roaster, the temperature and rate of
air injection are adjusted in order to
remove sulphur (as sulphur diox-
ide), to burn hydrocarbons, and to
oxidise the metallic elements.
A stream of compressed air sends
the gas into a post-combustion
chamber which treats it for oxides
of carbon. The fumes pass through
three levels of ltration; the gas
rejected is devoid of particles and
does not exceed 50 mg/m
3
/hr of
sulphur dioxide. These three stages
of ltration are performed by sieves
with different cut-off points: the
rst stage retains sublimated metal
particles, the second and third
stages treat the gas. Dust captured
in the rst stage of ltration is
treated at the hydrometallurgy
stage with the roasted calcine. The
tungsten or just tungsten. In the
rst case (including the presence of
molybdenum), the recycling
process is completed by a further
step involving the removal of phos-
phorus and arsenic.
Recovery of both molybdenum
and tungsten from spent catalysts
involves a combination of pyromet-
allurgical and hydrometallurgical
processes (see Figure 3).

The roasting step
First of all, a new roaster that began
operations at Valdi in 2011 is used
Table 1
Composition of NiW(Mo) spent catalysts
Characteristics Wt%
Moisture 3.5
Oil 13
C 14.8
S 4.9
W 7.8
Mo 2.5
Ni 2.9
P 0.5
As 0.02
Fe 0.39
Al
2
O
3
26.3
The average composition of NiW
catalysts received in 2013
CaWO
4
AS + P
Spent NiW
catalysts
Roasting
Roasted
catalysts
Filtration Filtration
Calcination
Purification
Precipitation
Pyro-
metallurgical
process
Hydrometallurgical process
Filtration
Leaching
Product
Slag
Alloy
Garbage
Dusts
Product
Filtration
Liquid
Solids
Gas
<50mgSO
2
/m
3
/hr
Figure 3 Recycling process for spent nickel-tungsten catalysts
grip-type tube fttings will not back
off with thermal cycling or vibra-
tion, unlike the threaded fttings
used with carbon steel.
The third objective maintaining
temperature within a certain range
is achieved by heating the
impulse lines. You can insulate your
impulse lines manually feld
tracing or purchase tubing that
has been insulated in the factory
and encased in a polymeric jacket.
Pre-insulated bundled tubing comes
ready to install in coiled lengths. It
can be heated with steam or left
unheated, as the application
requires. It is important with pre-in-
sulated tubing bundles to follow the
manufacturers instructions for seal-
ing the insulation when splicing into
or cutting into the pre-insulated
tubing bundle.
Manifold
The manifold consists of a set of
valves whose bodies are machined
from a single block of metal,
usually stainless steel. The mani-
fold mounts to the side of the
transmitter and serves a critical
function, enabling calibration or
service of the transmitter.
Quality is especially important in
a manifold. During calibration or
normal operation, at least one of
the valves in the manifold is in the
off position. If the shut-off is less
than complete, the result is an inac-
curate reading from the transmitter.
For example, when calibrating a
transmitter in a standard differen-
tial pressure setup, the technician
closes the two isolation valves and
opens the equalisation valve (see
Figure 4). This procedure results in
the transmitter being zeroed out
for calibration. If one of the two
isolation valves is leaking, however,
the calibration will be unsuccessful
and all subsequent readings will be
inaccurate.
Further, after calibration, the
operator will close the equalisation
valve and open the two isolation
valves, placing the transmitter back
in service. If the equalisation valve
does not achieve a positive shut-off,
high pressure will leak across the
seat to the low pressure side,
rendering the differential reading
inaccurate.
Why would these valves leak?
One cause is debris, either scaling
HP LP
To
transmitter
To
transmitter
Isolation
valves
Equalisation
valve
Figure 4 A manifold, consisting of three
needle valves, enables the technician to
isolate and calibrate the transmitter in
the eld
yield selectivity, in particular the
propylene yield. The higher the
hydrogen content, the higher the
propylene yield.
For the types of crude oil that
have been considered here, the
typical CCR content of atmospheric
residues is about 12-14 wt%, and
metals are 80-90 wtppm. Thus,
pretreatment of the feed is required.
The ARDS unit is a specially
designed hydrotreater that pretreats
the feed to the RFCC unit to reduce
contaminants including sulphur,
nitrogen, organometallic metals and
CCR. Additionally, it increases the
hydrogen content of the atmos-
pheric residue, thus improving the
crackability and selectivity of the
residue. For a typical ARDS HDS
rate of 90%, the CCR reduction is
about 65-70% and the metals reduc-
tion (HDM rate) is about 85-90
wt%. The HDT RFCC feed will
then have a CCR content of less
than 6 wt% and metals content of
less than 10 wtppm.
Due to the feed rate, nature of the
feed, and processing objectives, a
relatively high pressure and low
space velocity design is required
compared to other hydroprocessing
technologies. Even so, the cycle
length of the ARDS unit is typically
only one year. For the case evalu-
ated and ARDS operating
objectives, at least two reactor
trains, each with three to fve very
large and thick-walled reactors, will
be required. Consequently, both the
initial investment and on-going
catalyst replacement cost will be
much higher than for most hydro-
processing units.
The other key features of this
confguration to maximise the
production of petrochemicals are:
Process straight-run light naph-
tha along with light catalytic
naphtha (LCN) in the steam cracker
Process heavy straight-run naph-
tha in a catalytic reforming unit
(CRU) to produce reformate
Process heavy catalytic naphtha
(HCN), pygas and reformate
through the aromatics complex to
produce paraxylene and benzene
The C
4
s with relatively high
olefns content are either sold or
recycled to the second riser of the
FCC (directly or through oligomeri-
sation) to maximise propylene
production
Products from the RFCC unit and
steam cracker are combined in a
single light ends recovery section to
produce ethylene and propylene,
minimising the capital investment
Hydrogen from the CRU and
steam cracker is supplemented by
further hydrogen production from
steam reforming to be used in the
hydrotreaters.
Complex conguration with
HCU + VRDS + RFCC
The intention of this refnery confg-
uration is to increase the production
of middle distillate at the expense of
petrochemicals. Relative to the frst
confguration of ARDS + RFCC, the
LVGO fraction of the crude is
routed to a VGO hydrocracker
(HCU) to shift the yields in the
desired direction. Figure 2 shows a
schematic of the refnery.
The key features of this refnery
confguration are:
A vacuum tower is installed to
produce a diesel-type cut to be
processed in the distillate hydro-
treater, a LVGO cut to be processed
in a two-stage, almost full-conver-
sion hydrocracker unit oriented
towards the production of middle
distillates
The HVGO is combined with
vacuum residue to be pretreated in
the vacuum residue desulphurisa-
tion (VRDS) unit before processing
it in the RFCC unit
Depending on the quality of the
combined vacuum residue and
HVGO feed to the VRDS unit, a
slip stream of vacuum residue
could be taken and blended with
diluents such as RFCC decanted oil
(DCO) and light cycle oil (LCO) to
produce some fuel oil. The VRDS
unit would then be able to reduce
the CCR to the levels required by
the RFCC feed, especially when the
CCR content of the feed to the
VRDS unit exceeds 24 wt%.
Complex conguration with
SDA + DCU + HCU + RFCC
The refnery confgurations illus-
trated above, where residue
hydrotreaters reduce the feed
Max Ovchinnikov is a Senior Research Chemist
with Criterion Catalysts and Technologies
based in Houston, Texas. He is primarily
engaged in the research and development of
catalysts for hydroprocessing applications and
has 12 years of experience in heterogeneous
catalysis and rening technologies. He has co-
authored over 20 technical publications and
holds a PhD degree in organic chemistry from
Iowa State University.
2 Dekkers C, Daane R, Oil & Gas J., 1999, 97,
145.
3 Puri B K, Irgolic K J, Environ. Geochem. Health,
1989, 11, 95.
4 Nielse B, Villadsen, Appl. Cat., 1984, 11, 123.
5 (a) Internal communication, Criterion
Catalysts & Technologies; (b) Merryeld R N,
Gardner L E, Parks G D, Catalyst Characterization
Science, ACS Symposium Series 1985, 1.
6 Bhan O K, Arsenic removal catalyst and
method for making same, US Patent 6759364.
requires increased process unit
capabilities in a number of refning
process units, particularity hydro-
processing assets. The magnitude
of impact, of course, depends on
the particular feed processed, the
inherent fexibility of the facilitys
operations and the ability to
offset this increased processing
requirement.
In making this move, proper
selection requires a good under-
standing and modelling of the
facility as well as a strong knowl-
edge of the sources and processing
requirements of the available
crudes. With the continued devel-
opment and increased availability
of such feeds, this knowledge needs
to be continuously updated to
ensure minimal operational
surprises. Reviewing TBP curves
and bulk properties of whole crude
or even the individual cut-point
ranges does not
necessarily characterise the feed
diffculty or indicate the full impact
on the individual processing units.
References
1 Henke K R, Arsenic, Environmental Chemistry,
Health Threats and Waste Treatment, Wiley,
2009, 186.
30
50
60
40
20
10
N
o
r
m
a
l
i
s
e
d

H
D
S

t
e
m
p
e
r
a
t
u
r
e
,

F
Base
Alternative feed Current feed
Temp. 1 Temp. 2 Temp. 3 Temp. 3 Temp. 2 Temp. 1
check-
back
Temp. 1
Catalyst age
DN-3651 alternative feed DN-3551 alternative feed
DN-3651 current GO DN-3551 current GO
Figure 15 DN-3651 and DN-3551 normalised HDS temperature
increased water in the sulphur
plant feed and thus more condensa-
tion in the sulphur plant feed NH
3

gas knock-out drum (and thus
water that has to be recycled back
to the sour water stripper).
The design shown in Figure 2
was common in the 1960s.
However, it also suffers from the
same heat balance drawbacks and
needless complications as seen in
Figure 1. In other words, a lot of
equipment is added to generate
reux when no fractionation is
required between the feed and
overhead product. Again, the only
purpose of the tower is to strip out
the NH
3
and the H
2
S.

Correct stripper design
In 1969, while working for the now
vanished Amoco International Oil
(1)
(8)
(15)
Sour
water
feed
90F
Low NH
3
water
to desalters
Steam
250F
E-1
P-1
13 PSIG
Ammonia gas
to sulphur plant
190F
Medium content
NH
3
water to
hydrotreaters
Figure 4 Two-stage sour water stripper design without feed preheat
Company in the UK, I designed a
sour water stripper that eliminated
the unnecessary features of the unit
shown in Figure 2.
Figure 3 shows the essentials of a
correct sour water stripper design.
Feed is brought in at ambient
conditions (70-100F, 21-38C) from
the sour water feed tank. To heat
the feed from 90F (32C) to 250F
(120C) requires about 16 wt%
steam ow, or about 1.3-1.4lb of
steam per gallon of stripper
bottoms, which is close to a typical
design stripping steam ratio for
sour water strippers. The E-1 feed
preheater, reux pump (P-2) and
the reux cooler (E-2) shown in
Figure 2 are all eliminated. How,
then, does one know that the
design shown in Figure 3 will
work? Because it was built this way
(at the Amoco renery in Milford
Haven, Wales, UK) in 1970, where
it worked just as well as the
conventional design shown in
Figure 2.

Two-stage sour water stripper
Figure 4 shows a sour water strip-
per with a side draw-off. The partly
stripped sour water is extracted
from tray 8 and directed to the
hydrotreaters for use as make-up
water in the salt (NH
4
HS) removal
step of the reactor efuent.
Completely stripped water from
the sour water stripper bottoms is
sent to the crude desalter. While
communicate where the module
will be installed on the plot plan.
Connections between the modules
are designed to be similar in confg-
uration so that construction is
relatively straightforward. Ventech
estimates that, with modularisation,
approximately 70% of a project is
already complete even before the
modules are shipped from their
facility. This greatly decreases feld
construction time to deliver an
operational facility (see Figure 1).
These methods also facilitate easy
disassembly and relocation, if
necessary, at some point in the
future. For example, a remotely
located gas processing facility could
be easily taken apart and moved to
a new natural gas source if an exist-
ing supply was depleted in its
current location.
Applying modularisation to refn-
ery construction has advantages
with regard to productivity, prod-
uct quality and ensuring the safety
of construction personnel. Since the
modules are built in a well-lit,
climate-controlled environment,
work can continue around the clock
regardless of weather conditions,
for greater productivity and easier
quality control. Since module height
is restricted, safety is enhanced, as
workers build at limited heights
within the fabrication facility.
Modularising GTL
The same advantages of modular
construction of refneries are being
applied to the construction of
distributed GTL plants. The GTL
process involves two operations: the
conversion of natural gas to a
mixture of carbon monoxide (CO)
and hydrogen (H
2
), known as
syngas, followed by a Fischer-
Tropsch (FT) process to convert the
syngas into paraffnic hydrocarbons
that can be further refned to
produce a wide range of
hydrocarbon-based products, includ-
ing clean-burning, sulphur-free
diesel and jet fuel. Speciality prod-
ucts including food-grade waxes,
solvents and lubricants can also be
produced from the paraffnic
hydrocarbons.
Large, commercial-scale GTL
plants, including the Sasol Oryx and
the Shell Pearl plants (both located
in Qatar), have been built at enor-
mous capital cost. The Oryx plant,
designed for production levels of
34 000 b/d, cost around $1.5 billion
to build. The Shell Pearl plant, with
an ultimate design capacity of 140
000 b/d of GTL products and 120
b/d of natural gas liquids, cost
around $18-19 billion. Conventional
GTL plant designs rely on econo-
mies of scale to drive positive
fnancial returns and are viable only
where there are large supplies of
low-priced natural gas.
However, another option being
developed smaller-sized and
distributed GTL plants shows
promise for deriving value from
smaller accumulations of unconven-
tional gas that would otherwise be
left underground, such as shale gas,
tight gas, coal bed methane and
stranded gas (gas felds located too
far from existing pipeline infra-
structure). A small, modularised
GTL plant has the fexibility to be
installed close to the trapped
resource and then used to process
that resource locally. Associated gas
(gas produced along with oil) is
another area of opportunity for
modularised GTL plants. This gas
is typically disposed of either by re-
injection, at considerable expense,
back into the reservoir or by the
wasteful and environmentally
damaging practice of faring, which
is subject to increasing regulation.
Modularised GTL plants enable this
otherwise wasted gas to be
converted into additional revenue.
In the larger economic picture, a
modular GTL capability can be the
key factor that enables the construc-
tion of upstream projects that
would otherwise be cancelled
because of poor results derived
from economic models. For exam-
ple, some shale gas discoveries are
being hampered by high develop-
ment costs, which result in marginal
economics due to gas prices that
are often low. These projects can be
enhanced by converting the gas to
higher-value clean fuels produced
in the GTL process.
ITW
Innovative
Technologies
Worldwide
ITW is a fast growing Company, marketing and implementing unique and patented Production Units Online
Cleaning, Tank Cleaning, Decontamination and Reclamation technologies, along with Specialty Chemicals.
Given the considerable success, ITW is expanding its markets and activities and is looking for experienced
professionals worldwide. The candidates should have minimum a 5 years experience in at least one of these
fields:
refining/petrochemicals process specialty chemicals sales
refining/petrochemicals process technology, operations, maintenance, turnaround
and should be willing to travel nationwide and worldwide, along with being prone to hard working and sales.
Having a technical degree and good market knowledge is also required.
The available positions will cover: technical sales, implementation of ITW technologies on the field,
sales and operations management.
Interesting compensation plans will be given along with serious career possibilities. Please contact:
ITW S.r.l. S.Cusumano 96011 Augusta - Italy jobs@itwtechnologies.com www.itwtechnologies.com
HIRING EXPERIENCED PROFESSIONALS WORLDWIDE
velocys.indd 2 10/12/12 14:27:56
lieberman.indd 3 08/03/2013 16:02
cri.indd 10 10/06/2013 16:38
f wheeler.indd 3 12/09/2013 13:40
from corrosion (as mentioned
earlier) or other debris, such as
burrs. Burrs are a byproduct of the
machining process during manufac-
turing. A critical step during
manufacturing and installation
should be the removal of all burrs
from all wetted surfaces. Otherwise,
they can dislodge when the mani-
fold is in service and catch in or
scratch the valve stem or seat,
preventing a positive shut-off. Burrs
may also be an issue in monofanges
and other globe pattern needle
valves.
Seat seals in needle valves
Another reason why valves leak is
poor design. Most needle valves
employ a metal to metal seal. The
metal tip of the stem moves down-
ward to seal against a metal seat of
matching shape. The tip of the stem
may be shaped like a ball or a vee.
In either case, it is important that
the tip does not rotate with the
stem. If it does, it may grind into
the seat, scoring it and creating a
pathway for leakage.
There are two main design
approaches to achieving a non-
rotating stem tip. In the case of a
vee tip, there may be a knuckle
joint that enables the upper stem to
turn while the lower stem remains
stationary, except for axial (up and
down) movement.
In the case of a ball tip, a ball
foats in the stem tip, like in a ball-
point pen. Among ball tips, there
swagelok.indd 3 11/12/2013 14:52
to produce calcine, which provides
feedstock for the following step.
This roaster is specic; it was
developed by Valdi to deliver
highly effective desulphurisation
and decarbonation, good produc-
tivity and very low cost of treat-
ment. It was designed to be capable
of treating all categories of spent
catalyst. The conditions of roasting
are controlled and the gas treat-
ment is also closely monitored.
Screening at the head of the furnace
and at its output enables the extrac-
tion of ceramic materials found
cat valdi.indd 2 12/03/2014 12:07
dust obtained after the second and
third fltration stages is sold to the
paper industry, so the process
delivers zero waste. The roaster
and its fltration system are the
subject of patents.
The hydrometallurgy process
The calcine that was produced
during the roasting process is
treated by hydrometallurgy in three
steps: frst calcine is leached to
produce a slurry containing tung-
sten and molybdenum, depending
on the composition of the catalysts
and their contaminants (which are
mainly phosphorus and arsenic). In
a step using decantation and fltra-
tion, a solid concentrate containing
alumina and silica, along with
oxides of nickel, is produced. This
material is reserved for the pyro-
metallurgical process.
The soluble fraction of the slurry
is purifed to remove phosphorus
and arsenic. These contaminants
are not recycled and are disposed
of as hazardous waste; they repre-
sent less than 1% of the starting
material. Tungsten and molybde-
num are present in the purifed
solution which is mixed with
reagents to precipitate calcium
tungstate or a mixture of calcium
molybdate and tungstate. The yield
of tungsten recovered in one shot is
more than 85%. These metallic
concentrates are calcined and sold
to steelmakers, more particularly to
companies looking for tungsten
units, for example the EuroW
company which produces tungsten
carbide and cemented carbides, or
Erasteel which produces steel and
tungsten alloys. The hydrometallur-
gical process is protected by a
patent.
The pyrometallurgy process
The leaching residues are melted at
a high temperature in a submerged
electrode arc furnace. Nickel alloys
are obtained from this operation;
they are sold to steelmakers while
the silico-aluminate slag produced
is used in roadworks, while dust
produced from the melting stage is
recycled.
Conclusion
The development of a recycling
process for NiW spent catalysts by
Eurecat, LElectrolyse and Valdi
was subsidised by the European
Union and was awarded Best of
the Best project status in 2009.
After several step developments in
the process, in 2013 Valdi success-
fully treated more than 800 tonnes
of spent NiW catalysts. New invest-
ments made in 2011 (a new roaster)
and 2013 (hydrometallurgy capac-
ity) enabled the company to reach
3000 tonnes capacity. With its pyro-
metallurgical and hydrometallurgi-
cal process, the company is able to
recycle a large range of spent mate-
rials, including NiW, NiMo,
NiCoMo and CoMo catalysts.
Jean-Pierre Dufour is the Technical and
Industrial Director of Valdi. He holds a PhD in
metallurgy and has worked for 20 years in the
metallurgical industries.
Sophie Comte is the Laboratory Manager with
Valdi. She started with the company seven
years ago as R&D project manager and holds a
PhD in environmental chemistry.

ww.eptq.com PTQ Q1 2014 145
responses that a high-fdelity model
with a non-complementary control
layer imported from the operating
plant. Use the high-fdelity control
emulation, but be tolerant of the
process model and accept that
some controller tuning is likely to
be benefcial.
OTSs will continue to become
more accurate, but this will not be
achieved through software devel-
opments alone. The more detailed
the simulation, the more data and
understanding that it holds, and,
consequently, the more engineering
effort required. As with any engi-
neering activity, the time and effort
should be assessed against the
value. The best OTS will not be the
best choice for everyone.
Martin Sneesby is an independent dynamic
simulation consultant with more than 20
years of experience in process simulation and
modelling, including many operator training
simulators and detailed engineering studies.
Email: martin.sneesby@apess.co.uk
system is complementary to a plant
or simulation. A low-fdelity
process model with a complemen-
tary (well tuned) control layer
might provide better overall
often need adjustment to work well
with even the most accurate simu-
lation models. Failure to allow for
controller tuning on the simulator
system may be detrimental to the
overall projects success.
Prognostications
High-fdelity process simulation
models are already very accurate, if
built well with the right data and
good engineering assumptions, and
offer a lot of value, particularly for
OTSs. However, not everything can
be modelled perfectly. Modelling
tools and principles can still be
improved. There is still room for
further development to reduce the
granularity of models (such as
smaller volumes, smaller time-
steps, more detailed unit operation
models), although there are dimin-
ishing returns.
The increase in fdelity of DCS
emulation that is achieved by using
real hardware and control confg-
urations is laudable, but it should
be remembered that a control
Green-eld projects share many of the
same characteristics as OTS projects
built for existing plants, but they tend
to be less technically challenging
and more schedule driven. This is
because there is not the same wealth
of detailed knowledge about the
operating plant and assumptions and
estimates become more acceptable.
Data are usually more readily available
and reasonableness will be judged by
less stringent criteria. The project team
generally recognises that value is more
closely linked to the delivery date
(to maximise training time) than it is
to accuracy. Indeed, the incremental
benet of a high-delity OTS for
training is small much can be
achieved with relatively simple, stable
and robust models.
WWW.ZWICK-ARMATUREN.DE
HIGH STANDARD VALVES
FOR NON-STANDARD CONDITIONS.
cat valdi.indd 3 12/03/2014 12:07
A Claus type sulphur recovery unit (SRU) without tail
gas treating is becoming rare. New emission regula-
tions, increasing the sulphur yield and higher sulphur
crudes are the main drivers for adding a tail gas treat-
ing unit (TGTU). In the past the most common
approach for 99.8% plus sulphur recovery was a SRU
followed by an amine based TGTU. In view of cost,
complexity and reliability, it may be worthwhile to
explore alternatives. One option is a combination of
two established processes, Jacobs Superclaus process
followed by a caustic scrubber. For an SRU upgrade of
an oil refnery in Taiwan, this combination was
preferred over an amine based TGTU solution.
Among the advantages of combining the Superclaus
process and a caustic scrubber are the lower installed
cost and higher reliability, according to Jacobs.
Investment costs are reduced, the process is less expen-
sive to operate and to maintain, requires a smaller
footprint and simplifes overall operation.
The overall system can achieve greater than 99.9%
sulphur removal at attractive capital and operating
costs. At least 99.0-99.4% of the H
2
S 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.
The residual SO
2
content in the fue gas is typically less
than 50 ppmv. Because of the nature of the caustic
scrubber process, an SO
2
content of less than 20 ppm is
achievable without excessive cost.
The Superclaus process consists of a thermal stage
followed by a minimum of three catalytic reaction
stages, with sulphur removed between stages by
condensers. The frst two or three reactors are flled
with standard Claus catalyst while the fnal reactor is
flled with selective oxidation catalyst. In the thermal
stage, the acid gas is burned with a substoichiometric
amount of controlled combustion air such that the tail
gas leaving the last Claus reactor contains typically
0.8-1.0 vol% H
2
S. The selective oxidation catalyst in the
fnal reactor oxidises the H
2
S to sulphur at an eff-
ciency of more than 85%. Depending on feed gas
conditions and unit layout, sulphur recovery up to
99.2% can be achieved.
The resulting tail gas is routed to an incinerator in
which all sulphur compounds are converted to SO
2
.
In the downstream caustic scrubber, an alkaline
scrubbing liquid is injected that quenches the incinera-
tor fue gas and removes SO
2
. In a downstream packed
bed the SO
2
removal is completed and the absorbed
SO
2
is removed as a sulphite salt. The clean (water
saturated) gas is then passed through mist removal
devices and released to atmosphere. The fnal SO
2
content is reduced to less than 50 ppmv.
The alkaline sulphite/sulphate solution reverses to
the vessel sump for recycle back to the inlet. In the
vessel sump, oxidation air is used to convert the sulph-
ites to sulphates that can be routed to waste water
treatment.
The combined technologies may be applied to both
new and existing plants. The overall sulphur removal
of over 99.9% can also be maintained in the case of
sub-optimal recovery upstream or bypassing of the
selective oxidation reactor stage. This makes it easier to
comply with emissions regulations. The reliability is
high, with less than 1% unscheduled shutdown time,
provided proper maintenance is applied.
Superclaus has a long catalyst lifetime, there are no
batch processes involved, and operation is simple and
continuous. In comparison to an amine based TGTU,
the footprint is small and the number of items of
equipment is limited.
For more information: Sander.Kobussen@Jacobs.com
A refnery wanted to validate ITWs Online Cleaning
and Improved Degassing/Decontamination technolo-
gies for future turnarounds/slowdowns and during the
run. Online Cleaning has been performed on the entire
CDU (which has two pre-heat trains) and lasted 24
hours. Subsequent to that, as the refner wanted to
further recover downtime (and reduce steam-out
requirements), ITW applied its Improved Degassing/
Decontamination technology, which lasted 24 hours. At
manways opening, the following results were achieved:
LEL = 0%; HC = 0 ppm; H
2
S = 0 ppm; benzene = 0
ppm; ammonia = 0 ppm; pyrophorics = nil.
Moreover, for the frst time in the units history, there
was no need for column wetting as pyrophorics were
completely eliminated. All the maintenance operations
were conducted under dry conditions.
Visual inspection of the bundles confrmed the
results of online cleaning. The refnery decided there-
fore not to extract some bundles during turnaround,
but to check their performance at unit start-up.
At start-up the results of online cleaning were
comparable to mechanical cleaning. The outlet
temperatures of the trains were the same and the foul-
ing factor of the H.E. was comparable to the
mechanically cleaned ones.
No waste was generated. All of the washing fuids
were sent to a slop oil tank and fully reprocessed at
Technology in Action

Taiwanese tail gas treating with
Superclaus and caustic scrubbing

Online cleaning and decontamination of
a CDU
case studies copy 8.indd 1 11/03/2014 14:37
approaches for tail gas feeding and mixing have been
considered. The most reliable was executed and fnally
put into operation.
Two tangential feeds for tail gas with a swirl booster,
closely followed by a defector plate and homogenising
area were designed.
All of this led to a reduction in the required resi-
dence time from two seconds to less than 1.5 seconds.
A nominal 50 000 kg/hr of tail gas with less than 290
kJ/kg are oxidised at temperatures around 720C.
The combination of a HighTurb Burner with a Low
NOx approach and staged feeding of tail gas resulted
in a solution which meets the refners requirements:
Utilisation of amine off-gas as a booster fuel reduces
fuel costs and increases sulphuric acid production
Partial feed of tail gas to the burner reduces NOx
formation due to low temperature combustion
Staged feed of tail gas within the oxidiser reduces
the required combustion temperature with consequent
energy savings
Optimised feed and mixing devices inside the
oxidiser reduce signifcantly the size of the equipment
and consequently the capital cost.
CS developed, designed, delivered and commis-
sioned the complete package.
For more information: kraxner@comb-sol.com
As refneries, plants, production, and transport units
age, upgrading technologies is the desired alternative
to the capital expenditure required to replace entire
fare systems. Engineers from Saudi Aramco partnered
with Zeeco to develop their fare technology specif-
cally targeted to eliminate smoke in an effcient,
economical manner in areas without access to steam
and without the need for a new fare stack. The initial
design was conceived by Saudi Aramco through
advanced computational fuid dynamics (CFD) analy-
ses and feld tested in select southern area gas-oil
separation plant (GOSP) locations. Finalisation of the
design occurred after extensive full-scale testing at
Zeecos Research and Development Facility in Broken
Arrow, Oklahoma, under the observation of Saudi
Aramco personnel. The resulting HPAAS technology
utilises supersonic air injection in a manner that can be
easily adapted to retroft and upgrade outdated smok-
ing technologies.
In a typical low pressure fare system, a vast amount
of air is needed to prevent smoking in the fare. Due to
the large air fow required, the diameter of the air
delivery structure is larger than the gas riser itself to
reduce the velocity within the stack, minimise the pres-
sure drop within the system and reduce blower power
requirements.
By using much higher pressure air than in typical
designs, HPAAS fare tips utilise air in a manner
comparable to steam injection tips. However,
instead of injecting most of the air required for
smokeless faring, highly kinetic jet streams of
unit start-up, with no problems at all during
reprocessing.
For Improved Degassing/Decontamination, the
condensate was routed directly to the oily sewage. No
problems at all were recorded at the API separator/
waste water treatment plant.
Given the successful application, the refnery will
consider ITW Technologies applications as follows:
Avoid CDU slowdown to clean the preheat trains
(one day of online cleaning compared with about 25-30
days at half capacity for slowdown)
Regular application on the visbreaker to increase
unit run length
Rescheduling of future turnaround downtime by
taking into consideration both the reduced time for
safe entry and the reduced need for cleaning
exchangers.
For more information: mferrara@itwtechnologies.com
An improved generation of tail gas oxidisers from CS
Combustion Solutions features combustion chamber
components that create new possibilities for reftting,
optimising and debottlenecking of existing facilities
used for separating sulphuric compounds, says CS.
Tail gas cleaning after sulphur recovery units often
involves high demand for energy and the risk of high
NOx emissions. On the other hand, the output of
sulphuric acid is quite small.
For a refnery project in Russia, amine off-gas was
considered as a supporting fuel for the tail gas oxida-
tion process.
Both high content of sulphuric compounds and high
calorifc value were drivers for using amine gas. To
avoid a negative impact on acid quality from a high
nitrogen content a LowNOx concept was the preferred
choice of the refner.
Steam production was not considered as the main
driver for the combustion process since excess steam
was already produced at the site.
Over and above of all these requirements, the invest-
ment costs of such an oxidiser should not exceed the
price of standard solutions.
CSs approach to the combustion of tail gas starts
with partial feed of the gas directly to the high inten-
sity mixing burner that forms the basis of all of its
process burners.
Within the HighTurb burner, tail gas is partly used as
cooling agent to reduce fame temperature.
Consequently, production of thermal NOx is reduced.
In combination with staged feeding of tail gas within
the LowComb oxidiser, staged oxidation avoids
sudden temperature drop. This enables the user to
achieve the lowest possible emissions at the lowest
possible oxidiser temperatures with high reliability of
equipment.
Based on CFD combustion modelling, different

Improved aring in the Middle East

High efciency tail gas oxidation in
Russia
inexpensive. With all of the simplifcations to the
equipment, turnaround installations have been
achieved in as little as two days providing signifcant
down-time savings to the user.
For more information: Clayton_Francis@zeeco.com
The Habshan Sulphur Granulation Plant, located about
16km south east of the existing Habshan Gas Complex,
will be the worlds largest sulphur forming plant when
measured by capacity. The plant also includes exten-
sive storage, blocking and rail car loading capabilities.
In early 2011, Dodsal Engineering and Construction,
the UAE based contractor, awarded Enersul, located in
Calgary, Alberta, a contract for 12 of Enersuls GXM1
sulphur granulation units as well as block pouring
towers and other equipment.
Sulphur granulation is a size enlargement process.
Liquid sulphur is sprayed onto small sulphur seeds
undersized granules and cooled to solidifcation.
With repeated application of liquid sulphur, the seed
increases in volume and weight. As the granule is
enlarged, each coating of liquid sulphur is fully and
structurally bonded to the layer beneath. This ensures
a spherical granule, entirely dry and completely free of
voids.
The process occurs in a rotating drum. Undersized
granules, which serve as seeds, are fed into the granu-
lating drum and are lifted by fights attached to the
inside of the rotating drum. The result is a continuous
falling curtain of sulphur granules. Liquid sulphur is
sprayed onto the seed curtain by a series of spray
nozzles located along the length of the drum. The cool-
ing process is achieved by the evaporation of water,
which is sprayed throughout the length of the drum
from the process water system.
As such, the GX granulation process produces a
spherical formed product, completely void of fat
surfaces, resulting in lower friability and fugitive dust
and has been the industry benchmark for over 30
years. The GXM1 is Enersuls high capacity sulphur
granulation unit, which can produce 1500 t/d with
higher throughput being achieved through multiple
process lines. The 12 Habshan units will produce 15
000 t/d of premium formed sulphur.
The Habshan project featured a staged delivery
schedule with the frst four granulation units to be
delivered in 56 weeks, the second four in 62 weeks,
and the fnal third in 68 weeks. Enersul met all deliv-
ery and project milestones and the plant was being
commissioned in March 2014.
Enersuls technology for the sulphur industry
includes sulphur forming technologies (GX and
WetPrill), formed and liquid sulphur storage and lead-
out, sulphur block pouring and block reclamation, H
2
S
degassing (HySpec), and liquid sulphur cooling.
For more information: pwories@enersul.com
air inspirate ambient air into the combustion zone.
Less than one tenth of the air fow is required
compared to low pressure technologies. Additionally,
since this minute air volume is injected at high pres-
sures, the delivery of air occurs in a much smaller
pipe, often of only 2 or 3 inches nominal diameter. The
weight and concerns of a large diameter air ducting
stack are eliminated completely.
The patented, proprietary design and layout of the
HPAAS injection tips are crucial to the technology.
Supersonic air injection nozzles maximise the eff-
ciency of the air injected through the system. The jet of
air produced creates a turbulent wake, inducing ambi-
ent air into the combustion zone and multiplying the
effect of the relatively tiny fow of compressed air so
that a signifcant portion of the air required for
complete combustion is available at the fare tip exit.
The correct placement and mixing set up by these tips
is also paramount to the success of the design. When
the combination of induced and injected air reaches
the fare tip exit, the velocity, angle of approach, and
volume have been optimised for interaction with the
gases and shaping of the fame. The nozzle placement
is refned to ensure the induced air is also fully mixed
with the waste gas stream and complete, smokeless
combustion occurs. Unlike steam and axial airfow, the
HPAAS can run momentarily without air, with no
immediate mechanical degradation because the injec-
tion nozzles are located below the fame suffciently far
to allow brief upsets to the air supply without instan-
taneous harm.
A specialised windshield is integral to the induction
of ambient air. By enhancing the upward draft around
the perimeter of the fare tip, the momentum of the air
within the windshield overcomes the tendency for
fame pull-down on the downwind side of the fare. In
effect, it shepherds the fame in an upward column
and prevents the impingement on the fare tip barrel
common to utility fares, protecting the equipment and
adding operational life to the tip. In fares without this
advantage, prolonged operation commonly damages
pilot and thermocouples beyond use as well as crack-
ing or buckling the barrel.
The greatest distinction between HPAAS and other
smokeless technologies is the ability to easily retroft
the technology to upgrade existing fares. Previously,
integrating an air-assist technology was mostly
achieved by replacing the fare entirely. In the Middle
East or other installations where suffcient steam was
not available, new smokeless operation was also
unobtainable.
With the advent of a small-bore air utility line and a
HPAAS fare tip of comparable size to the old tip,
existing fare stacks can be reused with no structural
modifcations. The tips are confgured to be direct-
bolted replacements to their predecessors; the only
welding required is simple utility bracketing along the
fare stack. If the compressor is located near the fare
stack, horizontal piping and supports are minimised.
Power, piping and controls to the compressor, buffer-
ing tank and control valve are simplistic and

Worlds largest capacity sulphur
forming plant
21st World Petroleum Congress 116
Aerzener 111
Air Liquide Global E&C Solutions IFC
Albemarle Catalysts Company IBC
AMACS Process Tower Internals 54 & 57
Axens OBC
BASF Corporation, Catalysts Division 9
Bilnger Water Technologies 49
Bryan Research & Engineering 95
Burckhardt Compression 17
CB&I 7
ContiTech Schlauch 41
Criterion Catalyst & Technologies 4
Delta Valve 26
Dunn Heat Exchangers 91
Emerson Process Management 105
Enersul 35
Everlasting Valve Company 112
ExxonMobil Research and Engineering Company 19
Four Quest Energy 92
GE Water & Process Technologies 45
Grabner Instruments 121
Heurty Petrochem 58
Hoerbiger Kompressortechnik Holding 23
ITW Technologies 88 & 123
John Zink Hamworthy Combustion 74
Johnson Matthey Process Technologies 69
KBC Advanced Technologies 2
Koch-Glitsch 53
Kurita 28
Merichem Company 14
Metso Automation 80
OHL Gutermuth Industrial Valves 38
Onis 102
Optimised Gas Treating 65
Process Consulting Services 10 & 12
Rentech Boiler Systems 32
Sabin Metal Corporation 47
Sandvik Materials Technology 86
Spectro Analytical Instruments 79
Streamlight 31
Sulzer Chemtech 50
ThyssenKrupp Industrial Solutions 66
UOP 24, 61 & 76
Weir Minerals Lewis Pumps 99
Wood Group Mustang 42
Yokogawa Europe 106
Zeeco 70
Zwick Armaturen 124
Alphabetical list of advertisers
For more information on these advertisers, go to www.ptqenquiry.com
ad index copy 3.indd 1 11/03/2014 14:38
want results? demand ACTION
.
If raising the alky feed and distillate from your FCC unit is high on the agenda, you demand more than a
conventional catalyst you demand ACTION
. With a 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 success with the undisputed leader demand ACTION.
REFINERY CATALYST SOLUTIONS
CONTACT AN ALBEMARLE REPRESENTATIVE AT (281) 480-4747 TO DISCUSS WAYS YOUR CATALYST
OPERATIONS CAN BENEFIT FROM THE ADVANTAGES OF ACTION TECHNOLOGY.
View our article on ACTION in the latest issue of Catalyst Courier: www.albemarle.com/CatalystCourier82
albemarle.indd 1 07/03/2014 12:59
Stimulate the heart of
your hydroprocessing unit
Single source lSO 9001 lSO 14001 OHSAS 18001
www.axens.net
Impulse
TM
, the catalyst technology that combines
the stability you recognize with the activity you need
Imp_PTQ.indd 1 04/12/2012 14:07:08
axens.indd 1 6/12/12 13:17:48

PTQ Q2 2014 PDF

Enviado por

Dados do documento

Descrição original:

Título original

Direitos autorais

Formatos disponíveis

Compartilhar este documento

Compartilhar ou incorporar documento

Opções de compartilhamento

Você considera este documento útil?

Este conteúdo é inapropriado?

Direitos autorais:

Formatos disponíveis

PTQ Q2 2014 PDF

Enviado por

Direitos autorais:

Formatos disponíveis

SPECIAL FEATURES

FLOOD AND SPOT

l&T.indd 7 10/03/2014 13:30

SHAPED SUPPORT GRID (SSG)

INLET DIFFUSER BASKET

BILFINGER WATER TECHNOLOGIES

Engineered to endure harsh conditions.

technology is protected by patent in the U.S.A.; other patents pending.

Supersonic Gas Separation System, with a small

C and BORIS EELJ

c is Rijeka Renery Sector Director

GLSF Free-Jet Burner

VALVE COMPANY, INC.

Valve Manufacturers Since 1906

. With a continuing record of success, Albemarles ACTION

Você também pode gostar