Академический Документы
Профессиональный Документы
Культура Документы
OUTLOOK 2013
AUTOMATION & CONTROL
ENVIRONMENTAL
REFINING
GAS PROCESSING
PETROCHEMICALS
PETROLEUM TECHNOLOGY QUARTERLY
ptq
Q1 2013
cover and spine copy 14.indd 1 7/12/12 11:26:25
Are you looking to step up plant perform ance?
Topses next generation B R IM catalysts offer refiners the opportunity to increase
perform ance through an increase in catalyst activity.
U sing the original B R IM technology Topse has developed several new catalysts, resulting
in higher activity at low er lling densities.
The next generation B R IM catalysts display
- high dispersion
- high porosity
- high activity
W e look forw ard to stepping up your perform ance!
W W W .T O P S O E .C O M
Stepping up perform ance
next generation B R IM
technology
topsoe.indd 1 6/12/12 12:55:24
2013. 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 the publisher cannot be held responsible for any
statements, opinions or views or for any inaccuracies.
3 A 2020 vision of biofuels
Chris Cunningham
5 Outlook 2013
15 ptq&a
19 Refnery design in China
Sun Lili Sinopec Engineering Incorporation
31 Maximising ethane in liquids crackers
Muhammad Imran Technip Stone & Webster Process Technology
43 Tracking performance degradation in a debutaniser column
Ali Abdulal, Vinod Ramaseshan and Stanley Gustas III Saudi Aramco
Horia Orenstein, Mohammad Kurdi and Andrey Gaskov SAS Middle East
53 Operator training simulator for a multi-product fuels pipeline
Santosh Alladwar, Cordell Chapman, Billy Mayo, Surajit Dasgupta and
Javier Vazquez-esparragoza KBR
Daniel Vogt Krohne Oil and Gas
Keone Jackman Petroleum Company of Trinidad and Tobago Limited
63 Refnery operations gain from advances in digital automation
Tim Olsen Emerson Process Management
69 Controlling flm temperature in fred heaters
Jinyu Jiao, Yuriy Morayko, Morten Theilgaard and Michael Ho WorleyParsons Canada
73 High-capacity tray for debottlenecking a crude distillation unit
Kaushik Majumder Shell Projects & Technology
Giuseppe Mosca Sulzer Chemtech
Kent Mahon Refning NZ
79 Steam reformers for hydrogen and synthesis gas
Norm Pelletier and Goutam Shahani
Selas Fluid Processing Corporation, a division of Linde Engineering
85 VDU vacuum system revamp
Ahmet Bebek Turkish Petroleum Refneries Corporation
89 Reducing carbon emissions with antifoulants
India Nagi-hanspal, Mahesh Subramaniyam, Parag Shah and James Noland
Dorf Ketal Chemicals
95 Towards a zero (net) carbon refnery
Michael Stockle Foster Wheeler
101 Modular design of smaller-scale GTL plants
Kenneth Roberts Velocys
105 Starting up hydrotreating units with cracked feeds
Randy Alexander, Reinder Van Bruggen and Pierre Dufresne Eurecat US
111 Optimising a process scheme for platforming heaters
Adil Rehman, Suman Pachal, Shyam K Choudhary, Ugrasen Yadav and M K E Prasad
Technip KT India
119 Improved hydrogen yield in catalytic reforming
Roberto Amadei Chemical & Energy Development
127 Product News
Totals Central Germany refnery at Leuna, where Bilfnger Berger Industrial Services provides maintenance services.
Photo BIS Group
Q1 (J an, Feb, Mar) 2013
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
contents/ed com copy 17.indt 1 11/12/12 16:04:37
STRATEG IC
SO LU TIO N S
O PERATIN G
SO LU TIO N S
O RG A N ISATIO N A L
SO LU TIO N S
EN VIRO N M EN TA L
SO LU TIO N S
C A PITA L
SO LU TIO N S
SO FTW A RE
SO LU TIO N S
W hat is
Next Gen Per f ormance
SM
?
N extG en Perform ance is optim ising each and every opportunity. Its
addressing challenges before they becom e problem s and applying
solutions before others have fully considered them .
Its the result of tw o parties you and KBC w orking in a side-by-side
partnership, com bining our know ledge and experiences.
Why partner with KBC?
KBC is a leading independent consultancy to the energy sector
w orldw ide. W ith our proven m ethodologies and industry-leading
softw are, our experienced consultants consistently dem onstrate
innovative and creative thinking, fexibility, and a penchant for identi-
fying and im plem enting im provem ents that result in tangible value.
W e are dedicated to continually im proving our consultancy and
softw are solutions by being responsive to your requests, keeping
abreast of new technologies, and anticipating the needs of an
ever-changing industry.
W e m easure our success in term s of your success. Let us help
you achieve N extG en Perform ance.
Americas
+1 281 293 8200
Europe, Middle East , and Africa
+44 (0)1932 242424
Asia
+65 6735 5488
answers@kbcat.com w w w.kbcat.com
2011 KBC A dvanced Technologies plc. A ll rights reserved.
G iven our global experience in
w ide-ranging practice areas and
client sectors, KBC is uniquely
positioned to provide a fully
integrated response to your
m ost signifcant challenges.
kbc.indd 1 7/3/12 13:58:03
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 Q1 2013 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 Offce
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 18 No 1
Q1 (J an, Feb, Mar) 2013
A 2020 vision of
biofuels
I
t is not just another new year; it is a year closer to 2020, when the worlds
commitment to the Kyoto protocol will be put to the test. Although
governmental talking shops on global strategy continue to make little
progress, the steady march of the decade seems to be concentrating refning
minds on meeting carbon limits. And a big benefciary looks set to be the
biofuels industry.
During November, the US Senate voted to cancel restrictions in its annual
defence policy bill that would have severely limited the militarys efforts to
develop biofuels for its jet aircraft. The National Defense Authorization Act
usually attracts most attention for policy changes on matters such as deten-
tion of terrorists, but the 2013 version brought biofuels to the fore by amend-
ing a rule that would have barred the military from buying biofuels if they
cost more than petroleum-derived jet fuel.
There remained the need to reconcile the Senate vote with the version yet
to be approved by the House of Representatives, which included limits on
the military use of biofuels, before the bill was ready for presidential signa-
ture. However, the latest move by US lawmakers is a major boost for the
alternative fuels industry. Production of relatively small batches of biofuels
inevitably costs far more than conventional, petroleum-based production of
jet fuel. As long as the price comparison test remained, there would be lim-
ited opportunity for a signifcant contribution of biofuels to the military
requirement of distillates. Since military approval generally leads the estab-
lishment of developing technologies, this was signifcant progress.
The US Air Force has been testing small but expensive batches of alterna-
tive fuels in its aircraft to prove they can be reliably used once prices become
competitive with petroleum fuels. For its part, the US Navy has been the big-
ger spender on biofuels for both warships and aircraft as part of its green
feet programme. With a 2020 deadline in mind, the Navy has set that year
to meet a target of 50% biofuels in its non-nuclear fuel pool by that date.
While opponents of biofuels in the US argue that freeing the use of cur-
rently expensive fuels is poor timing during a period of cutbacks, supporters
say that home-grown fuels support the domestic farming industry and create
jobs while protecting the military from price shocks in the international
petroleum markets. Bioethanol is, of course, big business for Midwest corn
growers, but it is not clear which US farming communities will be meeting
the vastly increased demand for vegetable oils to make jet fuel or biodiesel.
There is not much open mention of climate change amidst that controversy,
but there is unreserved acknowledgement of the Kyoto protocol and its
regional interpretation in other developments in favour of renewables. Enis
plan for wholesale conversion of its Venice refnery to biofuels production
amidst plant closures in Europe is based on environmental targets set by the
European Union. The Venice plan contrasts with news elsewhere in Italy,
where several refneries have already closed or will close temporarily in 2013.
In India, a leading refner has been reviewing the options for introducing
bio-jet fuel production, and we aim to report in PTQ the outcome of those
deliberations later in the year. After its public relations disasters of a few
years ago, it appears that the approach of 2020 may be turning the tide in
favour of the biofuels industry.
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
contents/ed com copy 17.indt 2 11/12/12 14:54:39
Theres only one way to stay ahead of
the competition: always be evolving.
Meet Max Ovchinnikov:
Industry Game-Changer.
W hen youre leading the w orld in catalyst developm ent, your w ork is never done and thats just ne
by M ax O vchinnikov, C RITERIO N s Senior Research C hem ist and co-developer of such advanced
catalyst platform s as SEN TRY and C EN TERA
,
n
o
i
t
c
u
d
o
r
p
n
i
e
g
n
a
h
C
0
60
80
2
.
1
1
.
2
-
1
.
0
-
Year on year
2010
2009
1
.
5
9
.
1 2
.
1
-
0
.
5
0
.
2
7
.
0
-
9
.
2
8
.
4
1
.
9 7
.
0
7
.
0
0
.
0
8
.
0
5
.
0
4
.
0
2
.
4
6
9
.
6
6
4
.
6
2
4
.
4
2
4
.
9
2
.
9
Figure 1 Procurement ratio of imported crude oil by region
ADU RHT
VDU DCU
RFCC
C ru d e
o il
R e sid u e
H yd ro tre a te d
re sid u e
A R
VG O C G O
VR
C o k e
Figure 2 Process fow of DC-RHT combined process
sinopec.indd 2 11/12/12 16:42:44
According to preliminary esti-
mates, if a 1.2 million t/y DC unit
is installed in a 12.0 million t/y
renery that processes poor-quality
crude oil, the impurities content of
the crude oil can be even higher.
Sulphur content may be 8.8%
higher, the total content of nickel
and vanadium may be 18.1%
higher, and carbon residue content
may be 17.3% higher.
It is predicted, based on a pricing
system of $60 per barrel, that the
internal rate of return may increase
by 0.75% due to lower crude oil
costs and the inuence of compre-
hensive factors.
RHT-RFCC RICP
Heavy cycle oil (HCO) from the
RFCC unit is cycled to the RHT unit
to improve the properties of RHT
feed. HCO contains a large amount
of aromatic components that can
dissolve resin and asphaltene in
the residue, so as to improve RHT
reactions, reduce carbon deposits
on catalyst, prolong the service life
of catalyst, reduce gas yield, and
improve product distribution in the
RHT and RFCC units.
Both the light oil yield and liquid
yield of the RFCC unit in a typical
integrally combined process (RICP)
are 1.5-2 wt% higher than those of
a conventional process. Since less
slurry and coke is produced in the
RFCC unit, the energy consumption
of the unit is reduced.
According to preliminary
calculations, the overall energy
consumption of a RICP is 2 kg oe/t
of feed lower than that of a conven-
tional process.
S-RHT and IRCC-FR
Industrial application results also
show that due to the constraints of
operating conditions and catalyst in
the RHT unit, there is still a large
amount of non-cracking compo-
nents in the RFCC feed, which will
increase coke formation and gas
production during RFCC reactions.
The latest research shows that
coke formation and gas production
may be reduced and the light oil
yield of the RFCC unit may be
increased by reducing the operating
severity of the unit, increasing the
amount of cycle oil and routing the
www.eptq.com PTQ Q1 2013 21
cycle oil to a separately installed
RHT unit, where a hydro-saturation
reaction takes place in the presence
of a specic catalyst to deeply
improve the properties of RFCC
feed and convert non-cracking
components into cracking compo-
nents (see Figure 3).
The RFCC combined process
(IRCC-FR) is used to further
increase light oil yield. Preliminary
research results show that both
the light oil yield and liquid yield
of the RFCC unit increases by 6
wt%.
Further comprehensive research
is under way. The successful imple-
mentation of this research will
change traditional ideas of heavy
oil processing and provide more
exibility for poor-quality heavy oil
processing.
C ru d e
o il
ADU RHT
VDU HC HF
RFCC
Figure 3 Process ow of IRCC-FR
300
1000
800
200
100
m
p
p
,
d
e
e
f
e
h
t
n
i
V
+
i
N
0
Ebu||ated bed
S|urry bed
Net VR conversion,
50 60 70 80 30 40 90 100
Fixed bed
Figure 4 Adaptability of RHC technology
ADU
Gas oi|
HT
VDU
SDA
S|urry bed
RHC
FCC
CCR
Crude
oi|
Diese|
Aspha|t
Naphtha
Gas oi|
VR
Unconverted
oi| 5
Diese|
Gaso|ine
Gaso|ine
S<1000 ppm S<100 ppm
Figure 5 Process ow of the SDA-RHC-FCC combined process
sinopec.indd 3 10/12/12 13:01:12
SDA-RHC-FCC combined process
Hydrogen consumption for the
three processes is roughly the same,
but light oil yield in the slurry bed
process is remarkably higher than
that of the other two processes (see
Figures 4 and 5).
Since poor-quality heavy crude
oil is processed, deep-cut technol-
ogy and SDA technology are used
to treat heavier components, which
are difcult to process, in a slurry
bed hydrocracking unit, so as to
optimise the operating conditions
in the gas oil hydrotreating unit
and especially in the FCC unit.
Therefore, the renery-wide yield
of light oil in the slurry bed process
is the highest among the three
processes.
The problems of slurry bed tech-
nology concerning engineering
scale-up and operation stability
have been solved. It should be
commercialised at the appropriate
time so as to open up a new proc-
ess route for optimised processing
of poor-quality residue.
Cleaner production and clean
product strategy
It is reported that, by the end of 2010,
motor gasoline meeting National III
or National IV emission standards
accounted for over 96% of the
gasoline produced by Sinopec, and
the proportion of gasoline product
meeting National IV or higher
emission standards was only 12%.
Therefore, for the purposes of
gasoline quality upgrading, espe-
cially for the production of gasoline
meeting National IV or National V
emission standards, the sulphur
content of gasoline will decrease
from 150 ppm to 50 ppm or even
below 10 ppm.
The key is to solve the problem of
high sulphur content in FCC gaso-
line. Inuenced by chemical
equilibrium, the difculties of gaso-
line desulphurisation will further
increase, resulting in higher costs.
There are three ways to reduce
the sulphur content of gasoline
product: FCC feed pretreatment to
reduce the sulphur content of FCC
gasoline by improving the proper-
ties of FCC feed; desulphurisation
of FCC gasoline; and a proper
increase in the contribution of the
22 PTQ Q1 2013 www.eptq.com
Gas oil
FCC
RFCC
G a s
o il
R e sid u e
F C C d ie se l
F C C g a so lin e
S < 1 0 0 0 p p m
S < 1 0 0 -1 5 0 p p m
S < 3 0 0 0 p p m
Gas oil
HT
RHT
Figure 6 Process ow of front-end hydrogenation of gasoline
Feedstock Slurry bed Ebullated bed Fixed bed
hydrocracking hydrocracking hydrocracking
Kuwait atmospheric residue 650.00 650.00 650.00
Hydrogen 15.27 14.17 13.18
Total 665.27 664.17 663.18
Output of main products
LPG 55.82 31.21 70.54
Naphtha 78.48 72.91 57.71
Gasoline 125.75 61.48 148.12
Kerosene 75.96 75.96 64.27
Diesel 249.11 198.93 223.97
Unconverted oil 11.07 81.55
FCC slurry 0.00 12.01
Total of light oil products 529.30 409.29 494.07
Case analysis: comparison of several combined processes
Table 1
Item Conventional Continous liquid Ultra-low
hydrogenation phase HDS sulphur HDS
Scope of Straight-run diesel No limit >90% >90%
application Coker diesel or No limit <10% <10%
FCC diesel
Technical Flow characteristics Downow, Upow, continous Two reaction
characteristics conventional trickle liquid phase bed zones,
bed technology technology conventional
trickle bed
technology
Reaction zone 1 1 2
Recycle hydrogen Recycle hydrogen Cycle oil Recycle hydrogen
or cycle oil
Reaction Hydrogen partial 6.4 6.4 6.4
conditions pressure, MPa(g)
Hydrogen-to-oil ratio 300-500:1 1.5-2.0 300:1
(liquid phase
circulation)
Reaction 360-390 360-390 360-390/280
temperature, C
Space velocity, h
-1
1.5-2.0 1.5-2.0 3.0/6.0
Product Sulphur content, ppm >350/50 >50 >10
specication Emission standard National III/IV diesel National IV National V
Investment Investment, % Basis 10-15 lower Equivalent
and energy Energy Basis 15-20 lower 10 lower
consumption consumption, %
Comparison of new diesel hydrogenation technologies with conventional technology
Table 2
sinopec.indd 4 11/12/12 16:52:27
limiting factors love
limitless possibilities
Overcome limiting factors affecting ref nery capacity and
operating f exibility 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
Solutions
grace.indd 1 10/12/12 10:30:41
quality. Proper evaluation and
selection of fux oil is required to
ensure compatibility with the fuel
oil produced in the system.
Incompatible fux oil can lead to
precipitation of asphaltenes in the
system. Flux oil characterisation
should be done to confrm compati-
bility with the system. The
following tests are recommended:
Aromatic carbon
Unsubstituted aromatic carbon
Substituted aromatic carbon
Average chain length.
Most favourable fux oils are
heavy catalytic naphtha (HCN) and
heavy cycle oil (HCO). These are
produced in refneries. However,
these are not readily available, as
refners prefer to produce diesel.
The second best choice for fux oil
is light cycle oil (LCO) and it is
generally available from refneries.
There must be extensive changes
in the quench system in case of
limitations in fux oil availability.
Depending on how much ethane is
cracked, the quench oil tower may
need to be bypassed, with possible
modifcations to the quench water
tower to provide additional heat
transfer capability. Additional
equipment may need to be provided
to handle tar removal. A beneft of
bypassing the quench oil tower is
reduced pressure drop between
furnaces and the cracked gas
compressor. A higher cracked gas
compressor suction pressure can
therefore be employed.
Existing quench oil and quench
water towers may be limiting if
capacity increase is another objec-
tive in addition to feedstock
fexibility. Technips Ripple trays
have been used successfully for
numerous revamp applications for
quench oil and quench water
towers (see Figure 5).
Cracked gas system
Cracked gas compressor perform-
ance needs thorough evaluation.
With a higher ethane feed diet,
molecular weight reduces from ~30
to ~19 with an increase in volumet-
ric fow rate. One of the important
parameters to consider when shift-
ing from liquids cracking to
majority ethane cracking is poly-
tropic head. Polytropic head is
38 PTQ Q1 2013 www.eptq.com
inversely proportional to molecular
weight. As Equation 1 shows, a
lower molecular weight gas will
require a higher polytropic head for
the same compression ratio:
[1]
Where
H
P
Polytropic head, ft
MW Molecular weight
Z
AVG
Average compressibility
T1 Suction temperature, deg R
n Compression coeffcient
P1 Suction pressure, psia
P2 Discharge pressure, psia
Also, as Equation 2 shows,
compressor horsepower is directly
proportional to mass fow rate and
polytropic head:
[2]
Where
SHP Shaft horsepower
M Mass fow rate
H
P
Polytropic head
n
P
Polytropic effciency
1.02 2% gear losses
From the above equations, it is
clear that, for the same mass fow
rate and compression ratio, a lighter
cracked gas from majority ethane
cracking will require a higher shaft
horsepower as compared to heavier
cracked gas. However, as Figure 1
shows, majority ethane cracking
results in a higher ethylene yield
and therefore a lower mass fow of
cracked gas will be required to
produce the same quantity of ethyl-
ene as compared to liquids cracking.
An important variable to overcome
potentially higher horsepower
requirements is to increase frst-
stage suction pressure. As Equation
1 indicates, increased compressor
suction pressure will reduce the
compression ratio and polytropic
head requirements. On the negative
side, a higher frst-stage suction
pressure results in a higher radiant
coil outlet pressure. Since a higher
pressure at the radiant coil outlet
degrades the ethylene yield, increas-
ing the frst-stage suction pressure
tends to lower ethylene make. On
the positive side, the liquids crack-
ing ethylene yield is a stronger
function of the coil outlet pressure
as compared to gas cracking. With
a small sacrifce in ethylene yields,
gas cracking furnaces can therefore
be operated at a higher coil outlet
pressure. There will be a reduced
impact on coil outlet pressure in
case the quench oil tower is
bypassed.
A higher coil outlet pressure also
lowers the volumetric fow rate,
which is helpful in avoiding poten-
tial casing or suction nozzle
limitations. If feed fexibility is the
only target, with no increase in
ethylene production, it is sometimes
possible that a cracked gas compres-
sor driver designed for liquids
cracking furnace effuents can work
for cases with majority ethane
cracking effuents. In some cases,
especially where both feed fexibil-
ity and an increase in ethylene
production are targets, it may not
be possible to operate by just
increasing the frst-stage suction
pressure. Possible solutions for this
situation are:
Adding a booster stage
Adding a parallel stage with a
dedicated driver
Adding a new driver for one of
the stages to avoid limitations on
the existing driver
Adding wheels on stages
|
|
|
|
|
|
|
|
|
|
j
j
)
|
|
\
[
j
)
|
\
[
=
j
)
|
\
[
1
1
1545
1
1
2
1
n
n
AVG P
P
P
n
n
T
MW
Z H
02 . 1 ] 3300 ) [(
) (
n
H
P
P
m
SHP =
Figure 5 Quench oil and quench water
towers
technip s&w.indd 5 10/12/12 13:12:22
requiring higher polytropic head.
In case the quench oil tower is
bypassed to address quench area
issues, there will be a reduced
pressure drop between furnaces
and the cracked gas compressor.
For the same coil outlet pressure, it
will therefore be possible to
increase the compressor suction
pressure.
Another important consideration
for the cracked gas compressor is to
confrm that a suffcient operating
margin is kept between operating
point and surge limit. This operat-
ing margin tends to be narrowed
due to a higher operating speed
and polytropic head required for
lighter cracked gas operation.
Again, a combination of increased
suction pressure and a lower mass
fow required for lighter cracked
gas operation can help to stay away
from the surge limit. Compressor
performance curves need to be
studied carefully by use of a proc-
ess simulator to confrm compressor
performance for lighter cracked gas
effuent.
Chilling train and
expander/recompressor
As Figure 3 shows, switching from
naphtha cracking to majority ethane
cracking results in an increase in
hydrogen production from ~1 wt%
to ~4 wt% and a lower methane
production from ~13 wt% to
~5 wt%. Higher hydrogen produc-
tion may result in a higher
volumetric fow to the chilling train
(depending on composition) and
an increased pressure drop.
Replacement of some piping and
exchangers in the hydrogen system
may help to lower the pressure
drop. In the case of no modifca-
tions, this additional pressure drop
has to be compensated by addi-
tional discharge pressure from the
cracked gas compressor. This is not
always possible due to potential
limitations in cracked gas compres-
sors designed for operating with
heavier cracked gas.
Lower methane and C
3
+ produc-
tion may reduce the demethaniser
warmer level feeds, but can add a
burden to colder level feeds. If
hydrogen is being recovered in the
original liquids cracking scheme,
with no slip stream to the expander,
there will be an increase in the
hydrogen stream fow to the hydro-
gen recovery core exchanger, with
potential hydraulic and surface area
limitations. There will also be
reduced fow to the expander from
the demethaniser overhead due to a
lower methane content in the
cracked gas. A reduced fow to the
expander will lower the expander
effciency. The combined effect of
lower effciency and a reduced fow
is a reduction in refrigeration avail-
able to the chilling train from the
residue gas expander. The expander
feed system can potentially be
modifed to add load and recover
more refrigeration (see Figure 7).
Expander internals or the expander
itself may need to be replaced to
make it suitable for the new set of
operating fow rate and composi-
tion. Additional expander stages
are sometimes needed to process
the lighter molecular weight gas.
Ethylene losses from the demeth-
aniser overhead will also increase
due to the lower methane-to-
hydrogen ratio. Modifcations can
be made to the overhead system to
reduce ethylene losses. These modi-
fcations should be fexible enough
to allow reverting back to naphtha
or gas oil cracking, if required.
Another important consideration
is the percentage of hydrogen in the
residue gas. An ethane cracker
results in lower molecular weight
residue gas due to the increase in
hydrogen content. Since residue gas
is used as a regeneration gas for
dryers, there is the potential for an
increased pressure drop in the
regeneration gas system. It is,
however, possible to reduce the
regeneration gas fow and therefore
reduce the pressure drop due to the
improved specifc heat of lighter
regeneration gas.
Ethylene splitter, C
2
hydrogenation
and deethaniser systems
Shifting from naphtha or gas
oil cracking to majority ethane
cracking results in a higher ethane
per cent to ethylene splitter feed,
that is a lower purity feed to the
Figure 6 Ethylene unit Credit:Technip
R e -c o m p re sso r E x p a n d e r
Figure 7 Simplifed sketch of expander/recompressor
www.eptq.com PTQ Q1 2013 39
technip s&w.indd 6 11/12/12 16:49:42
ethylene splitter. This will require a
higher refux rate for the same
production capacity. If additional
capacity is another objective, the
required tower internal traffc will
be further increased. The existing
ethylene tower trays would need to
be checked for the extra capacity.
Possible solutions to overcome limi-
tations in tray hydraulics are the
replacement of existing trays with
high-capacity trays, or the addition
of another tower in series. The
deethaniser and C
2
hydrogenation
reactor systems also need to be
evaluated for hydraulic constraints
and any possible negative impact
due to the potential higher fow
rates.
Ethylene and propylene refrigeration
compressors
Depending on composition, availa-
ble credits and mass fow rate, a
higher ethylene refrigeration duty
may be required for the chilling
train during maximum ethane crack-
ing. An ethylene refrigeration
compressor revamp might be
required in this case. Scope of the
revamp can be extensive if the objec-
tive is to both increase plant capacity
and add feed fexibility. Overall, the
ethylene refrigeration compressor
may need a higher hydraulic capac-
ity and horsepower to enable
increased ethane cracking.
Propylene refrigeration load will
increase with two major users
during maximum ethane cracking.
First, as was explained above, the
ethylene splitter will require a
comparatively higher refux rate for
majority ethane cracking as
compared to naphtha or gas oil
cracking. For non-heat pumped
systems, a higher refux rate will
require a higher duty on the ethyl-
ene splitter condenser, which is the
largest user of the coldest propyl-
ene refrigerant. For low-pressure
heat pump systems integrated with
the ethylene compressor, there will
be an increased load on the heat
pumped ethylene compressor.
Second, the ethylene refrigeration
condenser duty will increase due to
a higher demand for ethylene
refrigeration. The ethylene refriger-
ation condenser is normally the
second largest user of the coldest
propylene refrigerant.
On the positive side, there will
be two important refrigeration
credits to propylene refrigeration
with increased ethane cracking.
These credits can be used to reduce
the load on the propylene refriger-
ation compressor. The frst credit
is the increased recycle ethane
production from ~5 wt% to
~35 wt%. Recycle ethane can
potentially be used for the coldest
level refrigeration credit. A higher
recycle ethane refrigeration credit
can help to lower signifcantly the
propylene refrigeration compressor
horsepower. An additional credit
is the fresh liquid ethane feed,
which can be fashed to furnace
feed pressure to achieve additional
refrigeration credit at the coldest
level. Ethane feed, however, may
need to be dried before fashing to
a lower pressure and temperature
to avoid the formation of hydrates.
Additional investment in dryers
and regeneration systems may
need to be considered. Higher
duties on the ethylene splitter and
ethylene refrigeration condensers
can be balanced by these additional
credits during periods of high
ethane cracking. The propylene
refrigeration compressor must be
checked on a case-by-case basis
before shifting to higher ethane
cracking.
Conclusion
While current economics for energy
and feedstock supply favour ethane
feed cracking, plant owners must
rigorously examine operational and
capacity constraints before transi-
tioning to maximum ethane feeds.
Detailed feasibility studies need to
be undertaken to evaluate the
impact of maximising ethane crack-
ing on plants originally designed
for naphtha or gas oil cracking. A
well-planned feasibility study can
help improve production margins.
Each ethylene plant is unique and
needs dedicated investigation for
suitability in feedstock transition.
References
1 American Chemistry Council, Shale Gas
and New Petrochemicals Investment: Benefts
for the Economy, Jobs, and US Manufacturing,
March 2011.
2 International Energy Agency, World Energy
Outlook 2011.
3 Bernard A, Pickett T M, Manek B M, Frye D K,
Flux oil stream import to quench system risk
and impacts, paper number 33e, AIChE 2011
Spring National Meeting, Chicago, Illinois, 14
March 2011.
Muhammad Imran is a Lead Process Engineer
for Technip Stone & Webster Process
Technology, Houston, Texas. He has 16 years
of experience with petrochemical, refnery and
gas processing projects, with the past seven
years focused on ethylene. He holds a bachelors
degree in chemical engineering from University
of Engineering and Technology Lahore and
a masters in chemical engineering from the
University of Houston.
Figure 8 Grassroots ethylene plant
40 PTQ Q1 2013 www.eptq.com
technip s&w.indd 7 10/12/12 13:13:00
roMax process slmulauon soware by 8ryan 8esearch & Lnglneerlng, lnc.
oqloeetloq 5olouoos fot tbe Oll, Cos, kefoloq & cbemlcol loJosttles.
sales[bre.com www.bre.com 979 776-3220 uS 800 776-3220
Process
Comparing Physical
Solvents for Acid Gas
Removal
Insight:
Physical solvents such as DEPG, NMP, Methanol, and Propylene Carbonate
are often used to treat sour gas. These physical solvents differ from chemical
solvents such as ethanolamines and hot potassium carbonate in a number of
ways. The regeneration of chemical solvents is achieved by the application of
heat whereas physical solvents can often be stripped of impurities by simply
reducing the pressure. Physical solvents tend to be favored over chemical
solvents when the concentration of acid gases or other impurities is very high
and the operating pressure is high. Unlike chemical solvents, physical solvents
are non-corrosive, requiring only carbon steel construction. Aphysical solvents
capac|ly lor aosoro|rg ac|d gases |rcreases s|gr|lcarl|y as lre lerperalure
decreases, resulting in reduced circulation rate and associated operating costs.
Typical Physical Solvent Process
DEPG (Dimethyl Ether of Polyethylene Glycol)
DEPG is a mixture of dimethyl ethers of polyethylene glycol.
Solvents containing DEPG are marketed by several companies including
Coastal Chemical Company (as Coastal AGR
PC and is by Fluor
Daniel, Inc. The light hydrocarbons in natural gas and hydrogen in synthesis
gas are less soluble in PC than in the other solvents. PC cannot be used for
selective H
2
S treating because it is unstable at the high temperature required
to completely strip H
2
S from the rich solvent. The FLUOR Solvent process is
generally limited to treating feed gases containing less than 20 ppmv; however,
|rproved slr|pp|rg W|lr red|ur pressure lasr gas |r a vacuur slr|pper a||oWs
treatment to 4 ppmv for gases containing up to 200 ppmv H
2
S. The operating
temperature for PC is limited to a minimum of 0F (-18C) and a maximum of
149F (65C).
Gas Solubilities in Physical Solvents
All of these physical solvents are more selective for acid gas than
for the main constituent of the gas. Relative solubilities of some selected gases
in solvents relative to carbon dioxide are presented in the following table.
The solubility of hydrocarbons in physical solvents increases with
the molecular weight of the hydrocarbon. Since heavy hydrocarbons tend
to accumulate in the solvent, physical solvent processes are generally not
economical for the treatment of hydrocarbon streams that contain a substantial
amount of pentane-plus unless a stripping column with a reboiler is used.
Gas Component
DEPG
at 25C
PC
at 25C
NMP
at 25C
MeOH
at
-25C
H
2
0.013 0.0078 0.0064 0.0054
Methane 0.066 0.038 0.072 0.051
Ethane 0.42 0.17 0.38 0.42
CO
2
1.0 1.0 1.0 1.0
Propane 1.01 0.51 1.07 2.35
n-Butane 2.37 1.75 3.48 -
COS 2.30 1.88 2.72 3.92
H
2
S 8.82 3.29 10.2 7.06
n-Hexane 11.0 13.5 42.7 -
Methyl Mercaptan 22.4 27.2 34.0 -
Choosing the Best Alternative
Adetailed analysis must be performed to determine the most economical
choice of solvent based on the product requirements. Feed gas composition, minor
components present, and limitations of the individual physical solvent processes are
all important factors in the selection process. Engineers can easily investigate the
ava||ao|e a|lerral|ves us|rg a ver|led process s|ru|alor sucr as ProVax
which has
oeer ver|led W|lr p|arl operal|rg dala.
For additional information about this topic, view the technical
article A Comparison of Physical Solvents for Acid Gas Removal at
http://www.bre.com/tabid/147/Default.aspx. For more information about ProMax,
contact Bryan Research & Engineering or visit www.bre.com.
bre.indd 1 7/3/12 10:45:54
For m ore inform ation, visit www.uop.com/uniex.
2012 H oneyw ell International, Inc. A ll rights reserved.
refinem ent redefined
UOP Unihox rosiduo-upgrading toonnology yiolds 25% moro
oloan uol, turning "bottom o tno barrol" into "top o tno lino."
Got tno most rom ovory barrol. Tno UOP Unihox Proooss oan doublo tno diosol yiold
o otnor rosiduo-upgrading toonnologios, turning tno "bottom o tno barrol" into moro
blaok on your bottom lino. Tno Unihox Proooss is a nign-oonvorsion, oommoroially
provon toonnology tnat prooossos low-quality rosiduo stroams liko vaouum rosiduo into
nign-quality distillato produots loading to rohnory margin inoroasos o up to 100%.
Simply put, witn tno Unihox Proooss rom UOP, you'll maximizo produotion and prohts
rom ovory barrol.
SPM-UOP-30 Uniflex Ad_Comp_v6.indd 1 3/8/12 2:35 PM
uop.indd 1 9/3/12 12:31:44
Tracking performance degradation in a
debutaniser column
C
omplex, capital-intensive
industries such as rening and
petrochemicals face a pressing
demand to minimise unplanned
downtime and achieve higher asset
uptime and improved Health Safety
& Environment (HSE) targets. To
improve the reliability and availabil-
ity of facilities and assets, Saudi
Aramco has been investigating the
application of SAS Predictive
Analytics and Data Mining in opera-
tions and maintenance. In this
regard, a pilot project was conducted
on the ability to deliver leading indi-
cators to the problem of performance
degradation in the Ras Tanura ren-
erys hydrocracking unit debutaniser
column. This article deals with this
A project applying statistical and data mining techniques to historical operating
data identied the root causes of a debutanisers poor performance
ALI ABDULAL, VINOD RAMASESHAN and STANLEY GUSTAS III Saudi Aramco
HORIA ORENSTEIN, MOHAMMAD KURDI and ANDREY GASKOV SAS Middle East
pilot project; understanding of the
use of a business analytical tool,
resulting in successfully achieving
the goals that deliver prediction of
performance degradation; the ag-
ging of leading indicators; and the
development of a dashboard for
operating staff, with recommenda-
tions for revision to the multivariable
control system (MVC) system to
alleviate the problem.
Hydrocracking unit
Saudi Aramcos Ras Tanura ren-
ery operates a 44 000 b/d
Unicracking process unit (hydroc-
racking unit) designed by UOP and
commissioned in 1999. This unit is
a two-stage maximum conversion
unit (>97%), where the primary
feed vacuum gas oil (SC-7 and SC-
8) is treated and cracked over three
reactors (C-100 and C-101 in the
rst stage, and C-200 in the second
stage) to naphtha and distillate
products.
Since the unit was originally
conceived to operate in both maxi-
mum distillate and naphtha mode,
the fractionation section is what in
hydrocracking parlance is termed a
debutaniser rst fraction scheme.
The reactor efuent rst comes to
the debutaniser, where lighter ends
are removed before being routed to
the main fractionation section.
Consequently, the debutaniser acts
both as a conventional stabiliser
www.eptq.com PTQ Q1 2013 43
Reaction section
Recycle compressor
Second stage cooler
First stage cooler
J 8 0 -D -1 3 0
Debutaniser Fractionator
Fractionation section
Feed/effluent
exchangers
VG O fe e d
E -3 1 5 E -3 2 4
J 8 0 -F -3 1 0
J 8 0 -F -3 2 0
J 8 0 -G -3 1 1
Figure 1 Ras Tanura hydrocracking unit
aramco.indd 1 10/12/12 13:18:43
44 PTQ Q1 2013 www.eptq.com
and as a hydrogen sulphide (H
2
S)
rejection column. To act as a H
2
S
rejection system, it is imperative
that the vapour-to-liquid ratio at
the bottom of this column is high
(>0.5 on mole basis). If it is lower,
H
2
S rejection is poor, leading to
corrosion in the downstream frac-
tionation column and pressure drop
in the second-stage reactor (C-200)
on account of particulate carry-over
with the recycled oil, over and
above possible leaks in the piping
on account of high temperature and
wet H
2
S corrosion.
Subsequently, since the feed to
the debutaniser essentially comes
from the reaction section, the
vapour-to-liquid ratio is strongly
inuenced by the product slate. Ras
Tanuras hydrocracking unit, since
start-up, has predominantly oper-
ated in distillate mode (>95% of the
time) and with catalyst systems
changed to low-zeolite catalyst in
the rst stage and amorphous cata-
lysts in the second (as opposed to
high-zeolite catalyst in the past).
The product slate has predomi-
nantly been heavier (a atter true
boiling point, TBP, curve), leading
to a higher than design requirement
in the reboiler duty (based on a
preferential once-through reboiler
[red heater] design), with no
improvement in the vaporisation
rate, leading to H
2
S slippage and
associated corrosion and reliability
problems. To mitigate this issue, a
multivariable controller (MVC)-
based control system for the
second-stage conversion levels was
developed by changing the C-200
inlet temperature and therefore the
weighted average bed temperature
(WABT). This was done because the
distillate selectivity per pass conver-
sion on the second stage is lower
when compared to the rst stage
for the same catalyst (the effect of
ammonia comes into the picture).
Consequently, the light end make
would increase and thereby main-
tain the vapour-to-liquid ratio in
the debutaniser. However, this
procedure leads to a debit in distil-
late make (the product of choice for
the renery).
A pilot project with SAS Institute,
using predictive analytics and data
mining techniques, was applied to
this problem, to ascertain if other
leading indicators could be tracked
to alleviate the situation and main-
tain the distillate make from the
unit. While the renery will conduct
other engineering modications to
the unit in the future, data mining
was considered to be a good
concept to deal with the current
situation, alleviate the problem, if
possible, and therefore optimise
any additional engineering solution.
Figure 1 shows a high-level over-
view of the hydrocracking unit in
question.
Technique and analytic procedure
The objectives of the project, called
the Predictive Performance
Degradation (PPD) project, can be
summarised as follows:
Perform data mining techniques
to understand inter-dependencies
among different debutaniser opera-
tional parameters that include:
Hot and cold feed tempera-
tures, ows and compositions
Internal reux
Tray 23 temperature: the
seventh tray from the top in the
rectication section
Reboiler duty and bottom
temperatures
Identify the cause and effect rela-
tionships among these operating
parameters and the following debu-
taniser performance-related issues:
Debutaniser instability
Light hydrocarbon slippage
and Reid vapour pressure
deviations
H
2
S slippage and corrosion
problems at the bottom of the debu-
taniser and the fractionator column
Proactively identify the critical
event and root causes that contrib-
ute the most to the above
performance issues
Generate proactive dashboards,
reports and other user interface
screens that provide information
about events, symptoms, causes
and effects of the performance
degradation of the debutaniser,
7 0
8 0
6 0
5 0
4 0
3 0
2 0
2 0 /4 /2 0 1 0 2 2 /4 /2 0 1 0 2 4 /4 /2 0 1 0
1 0
C A S E 1 0
R
e
f
l
u
x
f
r
o
m
G
-
3
1
0
A
/
B
f
l
o
w
Figure 2 Case 18, reux ow rate vs time
Analysis description SAS module used
H
2
S slippage event and root causes sequence analysis SAS Enterprise Guide
SAS Enterprise Miner
Visual discovery analysis for 24 H
2
S indirect indicators SAS/BASE
SAS/CONNECT
SAS JMP
Correlation and regression analysis for major operation parameters SAS/JMP
SAS/ETS
Non-linear regression analysis for cold-hot feed composition SAS/ETS
Neural network modelling for rst-stage efuent density SAS/BASE
SAS JMP
Visual discovery analysis for reux loss root causes SAS/BASE
SAS JMP
SAS Predictive Asset Maintenance modules used
Table 1
aramco.indd 2 12/12/12 10:52:14
thereby helping engineers and
operators to react in time and
minimise the debutanisers
instability.
The performed predictive analysis
is based on historical process infor-
mation data. It applies statistical
and data mining techniques to this
data to nd the factors and root
causes that affect the debutanisers
performance. Analytical capability
is the highlighted outcome of the
project. It has been determined that
there is a huge potential in benets
from the exploitation of historical
data. It has also been determined
that exploitation of such data in a
similar way has not been fully
applied in the past. Therefore, this
project was designed to utilise the
power of analytics to unleash the
potential of such historical data and
to make operations more stable and
protable.
Although the performance of the
debutaniser with regard to H
2
S slip-
page and its stability has shown
signicant improvement after the
MVC systems implementation, the
PPD project has shown that there is
still room for further improvements.
To facilitate the analysis, apart
from mechanical data, process and
control data were also analysed.
More than 300 tags for three years
data were processed (data taken
from before the implementation of
the MVC and also after), which
resulted in over 750 million records
being analysed and correlated to
nd the root cause of the debutan-
iser columns performance
degradation. The required modules
and components from the analytical
bundle SAS Predictive Asset
Maintenance were used. Table 1
lists the tools used.
Since H
2
S slippage from the debu-
taniser column bottom is primarily
due to slumping of column oper-
ations, the analysis was started with
this in focus; however, measure-
ment of H
2
S, in particular transient
high values, in the laboratory and
in the oil stream proved to be a
great challenge. This resulted in
unreliable deterministic analysis of
H
2
S. To overcome this obstacle, 24
H
2
S slippage indirect indicators
were considered, based on opera-
tions and engineering opinion along
www.eptq.com PTQ Q1 2013 45
5 8
5 9
5 7
5 6
5 5
5 4
5 3
2 0 /4 /2 0 1 0 2 2 /4 /2 0 1 0 2 4 /4 /2 0 1 0
5 2
C A S E 1 8
2 7
2 8
2 6
2 5
2 4
2 3
2 2
2 0 /4 /2 0 1 0 2 2 /4 /2 0 1 0 2 4 /4 /2 0 1 0
2 1
C A S E 1 8
D
-
1
4
0
c
o
l
d
f
l
a
s
h
d
r
u
m
l
i
q
u
i
d
f
l
o
w
D
-
1
2
0
b
o
t
t
o
m
s
t
o
C
-
1
3
0
f
l
o
w
Figure 3 Case 18, cold and hot feed rates
4 8 0
4 6 0
4 4 0
4 2 0
4 0 0
3 8 0
2 0 /4 /2 0 1 0 2 2 /4 /2 0 1 0 2 4 /4 /2 0 1 0
3 6 0
C A S E 1 8
C
-
3
1
0
t
r
a
y
t
e
m
p
e
r
a
t
u
r
e
,
C
3 . 4
3 . 0
2 . 6
2 . 2
1 . 8
1 . 4
2 0 /4 /2 0 1 0 2 2 /4 /2 0 1 0 2 4 /4 /2 0 1 0
1 . 0
C A S E 1 8
R
e
b
o
i
l
e
r
r
e
t
u
r
n
t
o
C
-
3
1
0
p
r
e
s
s
u
r
e
d
i
f
f
e
r
e
n
c
e
Figure 4 Case 18 Tray 4 temperature and eccentric orice pressure drop
aramco.indd 3 10/12/12 13:19:04
debutaniser column; manipulated
variables and process disturbances
based on changes to feed quality;
and/or other disturbances to the
system.
Based on the inuence map,
multiple cases were analysed.
Figure 2 shows the trend of reux
ow changes for Case 18.
It is evident that initially the
reux rate is stable and then drops,
steadies itself, nally plunging to
15 000 b/d within 24 hours. When
this time period was superimposed
with changes in the hot and cold
feed rate from the hot and cold
ash drums to the debutaniser
column (the reactor efuents), it is
evident that the hot feed rate
increases when the cold rate corre-
spondingly goes down (see Figure
3). The colder feed is essential
liquid ashed from the hot separa-
tor/ash drum and, as such,
constitutes the lighter components
of the reactor efuent.
The reboiler duty is constant, but
jumps to a higher value as reux
suddenly drops. (The reboiler duty
is maintained constant, as the outlet
temperature is close to the equip-
ments maximum temperature)
There is a surge in tray 4s tempera-
ture and, despite the increase in
reboiler outlet temperature, the net
vaporisation has decreased on
account of a at TBP curve for
the reactor efuent. Figure 4 shows
the complete scenario.
It is important to note that the
reboiler outlet temperature is high,
but boil-up rates are low. This can
only happen if the debutaniser
bottoms liquid has a high density,
indicating a heavier hydrocarbon
composition.
To prove this hypothesis, the
liquid density at the inlet of the
reboiler (F-310) is required.
Unfortunately, there are no
sampling points in this area to
develop the corresponding distilla-
tion curve. Therefore, inferential
density was calculated, taking into
account the fractionators product
ow rates. The idea behind this
estimation is as follows. The liquid
composition coming into the F-310
reboiler is the same as the liquid
composition owing into the main
fractionation column. Nearly all
46 PTQ Q1 2013 www.eptq.com
with their hypotheses. Using SAS
Institute tools for statistical analysis
and data mining techniques, all of
these indicators and hypotheses
were carefully examined. The
discovery at this stage was that the
most signicant problem is loss of
reux ow in the debutaniser.
During reux losses following
signicant H
2
S slippage, leading
indicators were affected; namely, a
reduction in tray 23s temperature;
surges in the hot feed component
rate, with a corresponding dip in
the cold feed component rate;
surges in tray 4s (four trays above
the reboiler return line) tempera-
ture; a reduction in reboiler duty;
and a drop in the reboiler return
eccentric orice pressure drop,
along with a reduction in the
condenser outlet temperature.
Consequently, 22 cases of signi-
cant reux loss during a three-year
period from 2008-2010 were
analysed. Eleven of these were not
considered because of extremely
unstable debutaniser column opera-
tion. The other 11 cases of
signicant reux loss during the
period 2009-2010 were analysed in
greater detail.
Based on the above, a reux prob-
lem inuence map was developed,
which basically analyses the nd-
ings and shows inter-dependencies
between the various parameters:
control variables around the
0 . 8 2
0 . 8 3
0 . 8 1
1
7
/
0
4
/
2
0
1
0
0
8
:
0
0
2
1
/
0
4
/
2
0
1
0
0
:
0
0
2
3
/
0
4
/
2
0
1
0
1
7
:
0
0
2
6
/
0
4
/
2
0
1
0
0
9
:
0
0
2
9
/
0
4
/
2
0
1
0
0
:
0
0
D
e
n
s
i
t
y
a
v
e
r
a
g
e
0 . 8 0
C A S E 1 8
2 0
3 0
1 0
1
7
/
0
4
/
2
0
1
0
0
8
:
0
0
2
1
/
0
4
/
2
0
1
0
0
:
0
0
2
3
/
0
4
/
2
0
1
0
1
7
:
0
0
2
6
/
0
4
/
2
0
1
0
0
9
:
0
0
2
9
/
0
4
/
2
0
1
0
0
:
0
0
R
e
f
l
u
x
f
r
o
m
G
-
3
1
0
A
/
B
f
l
o
w
0
6 0
7 0
8 0
9 0
5 0
4 0
C A S E 1 8
Figure 5 Case 18, density and reux ow rates
High reboiler duty
but low vaporisation
and hence a sudden
reux drop
aramco.indd 4 11/12/12 17:05:45
combustion and environmental solutions. pure and simple.
Emissions regulations continue to tighten worldwide.
Stay ahead of the regulatory curve and protect your
bottom line by choosing innovative combustion solutions.
From our smokeless flares to our Next Generation Ultra-Low NO
x
GLSF Free-Jet burners, Zeeco pushes combustion technology
forward. Driven by the challenges our customers face, we design
flares, burners, and thermal oxidisers that simply work better and
last longer. Global offices, local expertise, and smart design.
Zeeco is combustion innovation... pure and simple.
burners flares thermal oxidisers aftermarket products & services
Zeeco, Inc. 22151 E 91st St., Broken Arrow, OK 74014 USA
+1-918-258-8551 sales@ zeeco.com zeeco.com
ZEECO
C
From i nl et t o out l et
2 4 0
With o u t ste a m in je c tio n
With ste a m in je c tio n
Figure 1 Effect of steam injection on flm temperatures
worley.indd 1 11/12/12 17:16:39
70 PTQ Q1 2013 www.eptq.com
gains radiant heat from the refrac-
tory. The portion of the tube facing
the burners has a higher local heat
fux, while the side facing the
refractory is much lower. For a
given fred heater with nominal two
diameter tube spacing and a very
uniform longitudinal heat fux
distribution, the local peak heat fux
(q
m
) is approximately 1.8 times the
average heat fux (q
a
) for single-
fred heating. In contrast, the
double-fred heater has radiant heat
on both sides of the tubes, which
greatly reduces the peak fux to
about 1.2 times the average heat
fux.
3
The correlations mentioned
above for single and double fred
can be simply represented in the
following equation:
q
m
= Xq
a
(1)
where X represents the time factor,
which is approximately equal to 1.8
and 1.2 for single and double fred,
respectively.
The local flm temperature can be
calculated by the following
equations:
4
T
f
= T
b
+ T
f
(2)
T
f
=
q
m
(
D
o
)
(3)
K
f
D
i
where T
f
and T
b
are flm tempera-
ture and oil bulk temperature,
respectively. T
f
is the flm temper-
ature rise and K
f
is a flm heat
transfer coeffcient.
From the equations, it can be seen
that it is the localised heat fux, not
the average heat fux, that directly
governs the flm temperature. For a
heater with a given average heat
fux, a double-fred heater has a
lower localised heat fux distribution
than a single-fred heater. A lower
localised heat fux reduces the flm
temperature at that location. Figure
3 shows a comparison of the flm
temperature between single and
double fring for the same heater
with the same average heat fux. The
results are from the simulation of a
vacuum heater with a design duty
of 50 MW. It can be seen that using
double fred can greatly reduce the
flm temperature of the radiant coil
for the heater.
Selecting the correct location and
amount of steam injection is critical.
It must be injected upstream of the
heater tubes with the highest flm
temperature, yet far enough down-
stream in the radiant section to
minimise incremental pressure drop
to ensure charge pump capacity is
not exceeded.
2
Steam injection can also change
the fow regime for two-phase fow.
The problem associated with slug
fow can be mitigated by steam
injection. However, it is not the
intention of this article to discuss
the effect of steam injection on the
fow regime because it is not related
to the topic. Caution is advised in
selecting a suitable steam condition
to match the process fuid condi-
tion, to make sure that no
condensation occurs after steam
injection into the process fuid.
Reducing tube size
Depending on the allowable pres-
sure drop, fred heater coils are
usually divided into multiple passes
to accommodate the total fow and
to meet pressure drop require-
ments. For a given fow rate and
the number of fow passes, oil mass
velocity increases with reducing
tube size. A higher oil mass veloc-
ity reduces the oil residence time
and increases the flm heat transfer
coeffcient, therefore lowering the
flm temperature. Figure 2 shows
the effect of tube size on flm
temperatures. The horizontal axis
represents the heater coil growth
from radiant inlet to radiant outlet.
The results come from the simula-
tion of a small heater with a design
duty of 15 MW. The heater is a
vertical cylindrical type with four-
pass fow. To demonstrate the effect
of tube size on the flm tempera-
ture, the radiant tube size is
changed from 6.625in to 4.5in, with
no change in the size of the convec-
tion tubes. More radiant tubes have
been added to the radiant section to
compensate for surface loss caused
by the reduction in tube size, to
keep the same heat fux. Case 1 is
for the heater with all tubes sized at
6.625in, while case 2 is for the same
heater with tube sizes changed
from 6.625in 4.5in, for the radiant
coils only. It can be seen that the
flm temperature in the radiant
section has been decreased because
of the reduction in tube size.
Reducing the tube size increases
the heater pressure drop, which
requires a much higher pump head
upstream of the heater. Since flm
temperature control is basically
intended to control the peak flm
temperature, it is advisable to
reduce tube sizes for tubes with
peak flm temperature only, to
minimise the increase in pressure
drop caused by reducing tube sizes.
Double fred vs single fred
A fred heater can be single or
double fred. The heat fux on the
tubes circumferential surface is not
uniform because of the shading of
radiant heat. The single-fred heater
receives radiant heat on one side of
the process tubes (directly from the
burner fame), while the other side
of the tubes, facing the heater wall,
4 1 0
4 2 0
4 0 0
3 9 0
3 8 0
3 7 0
3 6 0
F
i
l
m
t
e
m
p
e
r
a
t
u
r
e
,
C
3 5 0
4 . 5 in tu b e s
6 . 6 2 5 in tu b e s
From i nl et t o out l et
Figure 2 Effect of tube size on flm temperatures
worley.indd 2 11/12/12 17:16:50
Lowering the average heat fux
The frst step in designing a fred
heater is to set up the allowable
average radiant heat fux. For a
given heater, either single or double
fred, it is helpful to control the flm
temperature by lowering the aver-
age heat fux. From Equation 1, the
localised heat fux reduces with
lower average heat fux, no matter
whether it is single or double fred.
Oil flm temperature depends on
the heat fux and oil mass velocity.
Decreasing the heat fux reduces
the oil flm temperature at a fxed
mass velocity.
5
The average radiant section heat
fux is defned as the total radiant
section absorbed duty divided by
the total radiant section tube surface
area. For a given radiant duty of a
fred heater, the only way to lower
the average heat fux is to increase
the radiant sections surface area. It
may be reasonable to assume a rela-
tively low average heat fux to
design a fred heater with very tight
flm temperature control. However,
it is also noted that oil residence
time increases as the surface area
increases, which may partially
counteract the beneft of decreasing
the flm temperature by lowering
the average heat fux. A lower aver-
age heat fux means more capital
cost in the heater coils, which is
another drawback in controlling the
flm temperature by reducing the
average heat fux. Thus, the effect
on heater design of lowering the
average heat fux should be care-
fully evaluated before a decision
is made.
Other design considerations
There are other design considera-
tions that should not be neglected
when designing a fred heater with
better flm temperature control. The
radiant section heat fux at any
point in the heater is controlled by
the temperature difference between
the hot fue gas and oil in the tube.
The heat transfer rate increases with
the temperature difference between
the hot fue gas and the cold oil.
6
In
a vertical up-fred heater, it is not
rare to see that heat fux is low at
the foor and gradually increases
along the length of the fame. It
is highest at the point where
www.eptq.com PTQ Q1 2013 71
maximum combustion takes place
in the fame, then reduces at the top
of the fre box. Thus, real fred heat-
ers may encounter more or less
signifcant heat fux imbalances.
This heat fux imbalance can cause
high flm temperatures and high
rates of fouling formation. Efforts
should be made to optimise the
design parameters to minimise the
heat fux imbalance. These design
parameters include radiant section
height to width ratio, burner to
tube distance, number of burners,
fame shape and dimensions, and
radiant section tube layout.
7
Flame impingement can cause
extremely high localised heat fux,
which results in a higher flm
temperature and rapid coke forma-
tion. Flame impingement occurs
when a fame actually touches or
engulfs the tubes. Vinayagam has
discussed the causes of fame
impingement for a fred heater.
8
Some precautions need to be
considered in the design of fred
heaters to prevent fame impinge-
ment occurring; for instance, an
adequate fre box to contain the
fame, more and equally spaced
burners, the correct type of burners,
and improved distribution of
combustion air fow.
Summary
Film temperature control is critical
to the successful design of fred
heaters, especially for those heaters
employed in upgrading heavy feed-
stocks. This article has discussed
several ways to control the flm
temperature in the design of fred
heaters. These methods have been
discussed in detail and have proved
to be effective means of controlling
flm temperature.
References
1 Pelini R G, Heat fux and flm temperature
in fred thermal-fuid heaters, Chemical
Engineering, Dec 2008.
2 Hanson D, Martin M, Low capital revamp
increases vacuum gas oil yield, March, Oil &
Gas Journal, 2002.
3 Romero S, Delayed coker fred heater
design and operations, Rio Oil & Gas Expo and
Conference, 2010.
4 API 530, Calculation of heater-tube thickness
in petroleum refneries, 6th ed, Sept 2008.
5 Golden S W, Barletta T, Designing vacuum
units, Separations, Apr 2006.
6 Martin G R, Heat-fux imbalance in fred
heaters cause operating problems, Hydrocarbon
Processing, May 1998.
7 Nogay R, Prasad A, Better design method
for fred heaters, Hydrocarbon Processing, Nov
1985.
8 Vinayagam K, Minimizing fame
impingements in fred heaters, Chemical
Engineering, May 2007.
Jinyu Jiao is a Mechanical Engineer with
WorleyParsons in Calgary, Canada. He holds an
MS degree in mechanical engineering as well as
a PhD in chemical engineering.
Yuriy Morayko is a Mechanical Engineer with
WorleyParsons in Calgary, Canada. He holds a
BSc degree in mechanical engineering.
Morten Theilgaard is a Mechanical Department
Manager for WorleyParsons in Calgary, Canada.
He holds a BSc in mechanical engineering.
Michael Ho has eight years experience in the
oil and gas industry in an EPCM environment.
He holds masters and bachelors degrees in
mechanical engineering.
3 2 0
3 4 0
3 6 0
3 8 0
3 0 0
2 8 0
2 6 0
2 4 0
2 2 0
F
i
l
m
t
e
m
p
e
r
a
t
u
r
e
,
C
From i nl et t o out l et
2 0 0
S in g le fire d
D o u b le fire d
Figure 3 Comparison of the flm temperature between single and double fring
worley.indd 3 11/12/12 17:17:21
*ad copy 3.indt 1 10/12/12 16:40:47
High-capacity tray for debottlenecking a
crude distillation unit
T
he main fractionator of the
crude distillation unit (CDU-1)
in the Whangarei refnery of
The New Zealand Refning
Company (NZRC) was retroftted
with high-capacity internals to
increase the unit throughput from
8500 t/d to 13 000 t/d. Ultra-high-
capacity Shell ConSep* trays were
applied in the most capacity-
constrained HGO pumparound
(mid circulating refux) section of
the column, as no other frst-
generation, high-capacity tray was
found adequate to debottleneck this
section. By the application of these
trays, capex savings of the order of
$5.5-6 million were achieved
Use of ultra-high-capacity trays in the most constrained section of a CDU enabled
over a 50% increase in throughput at less cost than other debottlenecking options
KAUSHIK MAJUMDER Shell Projects & Technology GIUSEPPE MOSCA Sulzer Chemtech
KENT MAHON Refning NZ
compared to other conventional
debottlenecking options. This was
the frst application of ConSep trays
in a CDU main fractionator, and the
post-revamp test run established
realisation of the expected
performance.
Whangarei refnery targeted
expanding its refning capacity
through the Point Forward Project.
1
The project involved increasing the
throughput of the CDU-1 from 8500
t/d to 13 000 t/d, thereby increas-
ing the distillate component to
downstream processing and gener-
ating additional long residue to
replace imported long residue for
loading the vacuum distiller. Figure
1 shows a simplifed process fow
diagram of CDU-1.
Shell Global Solutions International
(SGSi) carried out the feasibility
study for the expansion of CDU-1.
Several options were studied to
debottleneck the main fractionator:
Replacement of the existing
column internals with high-capacity
internals including the ConSep tray
for the most capacity-constrained
HGO pumparound section
Installation of a new crude pre-
fractionator column to separate off
light naphtha and reduce the load
to the main fractionator. The capex
for this option was found to be $6
million higher than for option 1
www.eptq.com PTQ Q1 2013 73
Crude
desalter
Furnace-2
Furnace-1
Preheat
train-1
Preheat
train-2
Preheat
train-3
C ru d e
c h a rg e
Preflash
vessel
Main
fractionator
(C-150)
U n sta b ilise d
n a p h th a
K e ro se n e
H G O
L G O
L o n g re sid u e
S trip p in g
ste a m
S trip p in g
ste a m
S trip p in g
ste a m
Figure 1 Simplifed process fow diagram of CDU-1
sulzer.indd 1 10/12/12 13:57:21
74 PTQ Q1 2013 www.eptq.com
Installation of a new heavy end
column to recover the heavy gas oil
(HGO) dropped into the long resi-
due to ofoad the main fractionator.
The capex for this option was found
to be $5.5 million higher than for
option 1.
On the basis of a comparison of
the revamp options, NZRC decided
to proceed with the ConSep tray
alternative owing to this options
lowest capex and most favourable
economics.
Shell ConSep tray technology
The ConSep tray utilises the princi-
ple of de-entrainment by centrifugal
forces to remove the gravitational
limitation of jet ood. Separation of
the entrained liquid before entering
the next tray allows very high
vapour velocities to be achieved in
the column. The tray combines the
features of a contacting deck and a
separator deck in a single tray. The
basic features of the tray are shown
in Figure 2. The functioning of the
contacting deck, which in fact is a
normal tray, is limited by three
hydraulic mechanisms: jet ooding,
downcomer choking and down-
comer backup. The use of a
separator deck inuences all three
mechanisms:
2
The jet ooding limit is extended
as the entrained liquid is efciently
separated from the vapour to
prevent carry-over of liquid to the
tray above
The liquid entering the main
downcomer is largely coming from
the separator deck, where it is well
degassed. As a result, the down-
comer liquid handling capacity is
substantially increased
To eliminate downcomer backup
limitation, the separator deck is
designed with low-pressure drop
swirl tubes. The contacting deck is
also designed with a relatively high
open area.
Figure 3 shows the expected capac-
ity gain of the ConSep tray over
conventional trays and packing.
3
The ow parameter () is dened
as:
=
L
P
v
V P
L
where
L
v
represents the liquid-to-
vapour mass ow ratio and
Pv
P
L
Swirl
tubes
Contacting
deck
Hi-fi
downcomer
Swirl deck
downcomer
Separator
deck
Figure 2 Schematic of ConSep tray
1 . 2
2 . 0
1 . 8
1 . 6
1 . 4
1 . 0
0 . 8
0 . 6
0 . 4
0 . 2
0 . 0 1 0 . 1 0 1 . 0 0
R
e
l
a
t
i
v
e
c
a
p
a
c
i
t
y
Fl ow paramet er, ()
0 . 0
C o n S e p tra y
S tru c tu ra l p a c k a g in g
C o n v. d o wn c o m e r tra y
H i-fi c a lm in g se c tio n tra y
Figure 3 Comparison of column internal capacities
Location Diameter Application Installed Capacity increase, %
UK 1.9 NGL debutaniser 1995 22
1
Australia 1.9 FCCU debutaniser 1996 30
2
Germany 2.2 HCU main fractionator 1999 50
Australia 1.7 NGL debutaniser 1999 15
1
Singapore 2.5 FCCU debutaniser 2000 20
1
Japan 2.1 FCC debutaniser 2006 10
Sweden 2.0 FCC debutaniser 2006 20
Sweden 1.0 C
3
/C
4
splitter 2007 50
USA 2.3 C
3
/C
4
splitter 2006 12
Canada 1.0 Depropaniser 2007 20
New Zealand 4.6 Crude distillation 2009 22
2
Singapore 1.8 PO drying column 2011 20
2
China 3.2 Ethylene fractionator 2011 Target 25%
3
Japan 2.5 FCC debutaniser 2013 (scheduled) Target 15%
3
1 Limited by reboiler capacity 2 Limited by feed to column 3 No data yet
Overview of ConSep tray applications
Table 1
sulzer.indd 2 11/12/12 18:05:10
represents the ratio of vapour-to-
liquid density.
Typically, the ConSep tray is capa-
ble of offering a 40-50% capacity
advantage over a wide range of frst-
generation, high-capacity trays. In
most revamps, the column retroft-
ted with the ConSep tray becomes
limited by other factors such as
availability of feed and/or
constraints on auxiliary equipment,
including reboiler, condenser,
pumps and so on, even before the
full potential of the ConSep tray is
realised. Table 1 shows some appli-
cations of this tray along with
benefts achieved and constraints
faced.
Modifcations of main fractionator
(C-150)
During the feasibility study, the
HGO pumparound section of the
column was found to be severely
limiting for the targeted throughput
of 13 000 t/d. This section was
already ftted with Shell Calming
Section (CS)* trays. Since the frst
generation of high-capacity inter-
nals was found inadequate to
debottleneck this section, ultra-
high-capacity ConSep trays were
selected. The trays were designed
to achieve 33% more capacity
compared to the CS tray.
The HGO pumparound section
consisted of three contacting trays
with a tray spacing of 500mm. A
one-for-one tray replacement with
ConSep trays was selected. Figure 4
shows a schematic drawing of the
HGO pumparound section ftted
with these trays. For the remaining
sections of the column, the follow-
ing internals were suggested:
Stripping section: Shell HiFi*
trays
Wash section: MellapakPlus
252Y** packing
All other sections: Shell CS trays.
As this was the frst application
of ConSep trays in this service, a
detailed study was carried out to
address the risks associated with
this application and the mitigations
were applied in the design. The
trays were manufactured by Sulzer
Chemtech. To ensure proper
performance of the trays in a rela-
tively new application, rigorous
quality control steps were followed
www.eptq.com PTQ Q1 2013 75
at the manufacturing site and a
detailed mock-up assembly of tray
components was carried out at the
refnery site prior to installation in
the column.
Post-revamp performance
From conception to implementa-
tion, the project took four years,
and the revamp was implemented
during a normal planned shutdown
in October 2009 without affecting
the units availability. The
performance test run of CDU-1 was
conducted in September 2010.
Crude throughput was maintained
at 13 000 t/d, although the crude
blend used during the test run was
marginally heavier than the one
considered for the revamp design.
H G O +
in te r re flu x
Tray 14
Tray 13
Tray 12
I n te r
re flu x
L G O
H G O
va p o u r
Figure 4 Schematic of HGO pumparound section of C-150
5 0 0
8 0 0
7 0 0
6 0 0
4 0 0
3 0 0
2 0 0
1 0 0
0
0 2 0 4 0 6 0 8 0 1 0 0
T
B
P
t
e
m
p
e
r
a
t
u
r
e
,
C
Wei ght , %
1 0 0
D e sig n b le n d
Te st ru n b le n d
Figure 5 Comparison of design and test run crude TBP
sulzer.indd 3 11/12/12 18:05:29
run conditions to evaluate hydrau-
lic loading of the ConSep trays. In
Table 3, the key performance
indicators for these trays operating
under test run conditions are
compared with the design condi-
tions. During the test run, the trays
were operating 10-15% lower than
design capacity, even at a design
intake of 13 000 t/d, due to:
Heavier feed than in the design
case
Less preheat recovery of the order
of 10-12C due to limitations in the
crude preheat train.
Based on a review of operating
experience since startup and the
76 PTQ Q1 2013 www.eptq.com
Comparison of the design crude
blend and crude blend used during
the test run is shown in Figure 5.
Key design conditions and test run
operating conditions of C-150 are
shown in Table 2. The product qual-
ity is compared in Figure 6 (a-e).
C-150 was simulated for the test
Parameters Design Test run Parameters Design Test run
Crude intake, T/D 13 000 13 077 Flash Zone press, barg 1.90 2.24
Naphtha, T/D 4405.0 3516.4 Feed temp, C 361.5 345.0
Kerosene, T/D 1498.8 2169.3 Top temp, C 180.8 175.7
Light gas oil, T/D 1517.0 1116.1 Kero draw temp, C 216.5 223.4
Heavy gas oil, T/D 2014.5 1731.9 LGO draw temp, C 244.2 250.1
Long residue, T/D 3573.3 4271.1 HGO draw temp, C 280.7 278.3
Strip steam, T/D 106.9 149.5 Flash zone temp, C 343.9 327.3
Top press, barg 1.65 1.95 Bottom temp, C 333.3 322.061
Key design and test run operating conditions of C-150
Table 2
3 0 0
4 0 0
2 0 0
6 0 0
7 0 0
5 0 0
1 0 0
0 2 0 4 0 6 0 8 0 1 0 0
T
e
m
p
e
r
a
t
u
r
e
,
C
Wt , %
0
Long resi due ASTM D86
1 5 0
4 0 0
1 0 0
5 0
3 0 0
3 5 0
2 5 0
2 0 0
0 2 0 4 0 6 0 8 0 1 0 0
T
e
m
p
e
r
a
t
u
r
e
,
C
LV, %
0
Heavy gas oi l ASTM D86
1 5 0
2 0 0
1 0 0
3 0 0
3 5 0
2 5 0
5 0
0 2 0 4 0 6 0 8 0 1 0 0
T
e
m
p
e
r
a
t
u
r
e
,
C
LV, %
0
Li ght gas oi l ASTM D86
D e sig n b le n d
Te st ru n b le n d
1 5 0
2 0 0
1 0 0
5 0
0 2 0 4 0 6 0 8 0 1 0 0
T
e
m
p
e
r
a
t
u
r
e
,
C
LV, %
0
Napht ha ASTM D86
3 0 0
2 0 0
1 0 0
2 5 0
1 5 0
5 0
0 2 0 4 0 6 0 8 0 1 0 0
T
e
m
p
e
r
a
t
u
r
e
,
C
LV, %
0
Kerosene ASTM D86
Figure 6(a-e) Comparison of design and test run product quality
sulzer.indd 4 10/12/12 13:57:54
test runs results, it could be
concluded that the revamp targets
for the CDU-1 main fractionator
(C-150) were achieved. No hydrau-
lic constraint was experienced in
achieving the design intake of
13 000 t/d and the required prod-
uct quality was achieved.
Conclusions
The performance of Shell ConSep
trays in the HGO pumparound
section of the CDU-1 main fractiona-
tor met the target of capacity
enhancement without any drawback
compared to the pre-revamp condi-
tions. During the test run, the trays
were operating at 10-15% lower than
the design capacity even at the
design intake of 13 000 t/d due to
heavier crude feed and lower feed
temperature. However, the built-in
capacity margin enabled stable oper-
ation for the trays at much above
the capacity limit of the frst genera-
tion of high-capacity trays.
The options to debottleneck
columns already equipped with the
frst generation of high-capacity
trays are limited. ConSep trays
provide an attractive solution for
www.eptq.com PTQ Q1 2013 77
such cases. In this revamp project,
use of only three of these trays in
the most capacity-constrained
section of the column made it possi-
ble to retroft the existing column
and made the capex option more
attractive over the other debottle-
necking options.
* Shell ConSep, Shell CS and Shell HiFi are Shell
trademarks. ** Mellapak Plus 252Y is a Sulzer
Chemtech trademark.
References
1 Refnery expansion means NZ more self
reliant, media release by NZRC, 16 July 2010.
2 Wilkinson P M, De Villiers W E, Mosca
G, Tonon L, Achieve challenging targets in
propylene yield using ultra system fractionation
trays, ERTC 2006.
3 De Villiers W E, Bravo J L, Wilkinson P M,
Summers D R, Further advances in light
hydrocarbon fractionation, PTQ Q3 2004.
Kaushik Majumder is Distillation Team Lead of
Shell Projects & Technology in Bangalore, India.
He holds a bachelors degree from Jadavpur
University, India, and a masters and doctorate
from Indian Institute of Technology, Delhi.
Email: Kaushik.Majumder@shell.com
Giuseppe Mosca is the Global Refnery
Technology Manager of Sulzer Chemtech.
He holds BS and MS degrees in chemical
engineering from the University La Sapienza
Rome, Italy.
Email: Giuseppe.Mosca@sulzer.com
Kent Mahon is a Process Engineer at Refning
NZ. He was the Senior Process Engineer and
Commissioning Process Engineer during the
Point Forward Project.
Email: Kent.Mahon@refningnz.com
Parameters Design Test run
Froth backup/CS height, % 68 60
Tray pressure drop, mbar 12.3 9.2
Tube food , % 73 60
Flow parameter 0.17 0.19
Overall column load factor, m/s 0.12 0.10
Flooding (CS tray), % 133 112
Key performance indicators for ConSep trays
Table 3
sulzer.indd 5 11/12/12 18:05:58
va|ue acc|ng cata|ysts, abscrbents, acc|t|ves anc
prccess technc|cgy fcr c|| refn|ng prccesses.
www.jmcata|ysts.com|rehner|es
UK
Te| +44 (0)1642 553601
USA
Te| +1 630 268 6300
Te| +1 732 223 4644
Naphtha HDS
D|ese| HDS
Hydrctreating cata|ysts
vacuum tcwer
Hydrcgen cata|ysts
FCC un|t
FCC additives
SOx NOx remcva|
l|ght c|efn prccuct|cn
Bcttcms ccnvers|cn
Veta|s traps
/ct|v|ty bccsters
CO cx|cat|cn
6as | |iquid
purifcaticn
S, Hg, C| remcva|
Cruce un|t
vCO HDS
Hycrcgen p|ant
Prccess
diagncstics
Untitled-1 1 31/07/2012 17:18
j matthey.indd 1 6/12/12 13:25:56
Steam reformers for hydrogen and
synthesis gas
T
he popular press and scien-
tic literature has been rife
with references to the discov-
ery of vast quantities of shale gas in
North America. This resource has
become economically recoverable
due to advances in horizontal drill-
ing and well fracturing or fracking.
The advent of shale gas has been
referred to as a game-changer,
which has the potential to make
North America energy independent
and the rening and petrochemical
sectors more cost competitive.
Many renery operations and
petrochemical processes require
large quantities of hydrogen (H
2
)
and synthesis gas (carbon monox-
ide [CO] and H
2
, also known as
syngas). The dominant method of
producing H
2
and syngas is steam
reforming of light and medium
hydrocarbons. This technology will
become even more popular in the
future with the advent of relatively
cheap natural gas.
Shale gas
The historic and future supply,
demand and pricing for US natural
gas in a recent Department of
Energy study is shown in Figure 1.
This graph shows total natural gas
production in the US, where shale
gas production will increase from 5
trillion cu ft in 2010 to 13.6 trillion
cu ft in 2035, representing an annual
increase of 4.1% per annum. This
growth in natural gas production
will make the US a net exporter in
the early part of next decade. The
price of natural gas is expected to
increase to ~$7.4/MMbtu (2010
dollars) in 2035 based on supply,
demand and incremental cost of
production. It is expected that natu-
Modern steam methane reformer design is a complex process that needs to be
carefully optimised in terms of capital and operating costs
NORM PELLETIER and GOUTAM SHAHANI
Selas Fluid Processing Corporation, a division of Linde Engineering
ral gas will be relatively cheap
compared to other fuels on an
energy equivalent basis. Based on
these projections, it is expected that
the dominant method of producing
H
2
and syngas is set to remain
steam methane reforming.
Petrochemical and rening
The rening and petrochemical
sectors consume large quantities of
H
2
, CO and syngas, which are
collectively known as HYCO.
Hydrogen is used for rening crude
oil, which is increasingly becoming
heavier and more sour. Also, ever
more stringent environmental regu-
lations will drive H
2
demand. In
addition, H
2
, CO and syngas are
used in various chemical synthesis
reactions. A summary of the major
industrial gases applications in
rening and petrochemicals is
shown in Table 1.
As noted previously, the domi-
nant method of producing H
2
, CO
and syngas is steam reforming of
natural gas. H
2
, CO and syngas
plants require a hydrocarbon feed-
stock such as natural gas or naphtha
and utilities such as water, electric-
ity and nitrogen. The outputs are
H
2
, CO, syngas and varying
amounts of steam. Given that these
plants are designed to last decades,
it is very important to understand
the current and future costs of feed-
stocks. This is essential to make the
appropriate trade-off between
capital cost and feedstock
consumption.
In particular, consumers of
tonnage quantities of industrial
gases have to make informed
investment decisions in an uncer-
tain economic environment. Plant
owners/operators need to examine
carefully their HYCO needs in
order to develop the most cost-
effective plant conguration for a
steam reformer. This can be done
by partnering with an experienced
engineering company that owns a
complete technology portfolio for
www.eptq.com PTQ Q1 2013 79
2 5
3 0
2 0
1 5
1 9 9 0 2 0 0 0 2 0 1 0 2 0 2 0 2 0 3 5
V
o
l
u
m
e
,
t
r
i
l
l
i
o
n
s
c
u
b
i
c
f
e
e
t
H isto ry P ro je c tio n s
1 0
2 0 1 0 2 0 3 5 1 9 9 0
0
5
1 0
H e n ry H u b sp o t m a rk e t
n a tu ra l g a s p ric e s
2 0 1 0 d o lla rs p e r m illio n B T U )
N e t e x p o rts, 2 0 3 5 = 5 %
N e t im p o rts, 2 0 1 0 = 1 1 %
C o n su m p tio n
D o m e stic p ro d u c tio n
Figure 1 Total US natural gas production, consumption and net imports
Source: Annual Energy Outlook 2013, Energy Information Administration, June 2012
selas.indd 1 10/12/12 14:05:47
80 PTQ Q1 2013 www.eptq.com
producing H
2
, CO and syngas.
Ideally, the engineering and manu-
facturing companies have to work
together as a single team to identify
the most effcient and economic
plant design. This includes an in-
depth assessment of capital and
operating costs, taking plant relia-
bility and process safety into
account in order to deliver an opti-
mal solution. It is important to
consider all the industrial gases as
well as steam and utility needs at a
given manufacturing complex, in
conjunction with available feedstock
and wastes to develop the best
possible long-term solution. All of
these industrial gas requirements
have to be examined over a long-
term horizon in a holistic manner.
Reformer
Steam reforming of hydrocarbon
feedstock to produce H
2
, CO and
syngas has been the dominant proc-
ess for producing H
2
, CO and
syngas for over 80 years. A simpli-
fed block fow diagram of the
steam methane reforming process
showing how the reformer fts into
the overall plant is depicted in
Figure 2.
Conversion of steam/hydrocar-
bon mixtures is carried out in
catalyst-flled tubes heated from the
outside for the production of
syngases rich in H
2
and CO.
Feedstock can be methane, ethane,
LPG or naphtha. The basic design
criteria for reformers are burner
design and confguration, heat
recovery from the hot fue gases
and furnace control. A variety of
design options with burners
arranged in the ceiling or at the
sidewalls of the furnace can be
considered. The blue-shaded steam
reforming block in Figure 2 repre-
sents the furnaces radiant section,
where the tubes receive most of the
input heat by radiation. In the
convection section, heat is recov-
ered in both the process streams
and steam system by convective
heat transfer. Some other considera-
tions are:
Desired operating temperature
and pressure
Amount of export steam
Feed fexibility for natural gas,
LPG and heavy naphtha
Burners for low NOx generation
and low noise generation
Selective catalytic reduction for
NOx control, if required
Flue gas heat recovery by preheat-
ing feed, water and generating
steam
Fully automatic reformer using
advanced control technology.
For large installations, the top-
fred design combines both process
and physical benefts to minimise
the total cost of ownership. The
top-fred design allows either
modularised or stick-built construc-
tion, taking the relative differences
in shop assembly versus feld
construction costs into account.
Furthermore, this confguration
results in a compact frebox, leading
to fewer burners relative to the
number of tubes. The smaller
surface area per unit volume when
compared with sidewall-fred
reformers minimises heat loss and
reduces construction time. The
main advantages of a top-fred
design are.
Easy access to burners
Simplifed combustion control
with fewer burners
Heat recovery fexibility (omega,
horizontal, vertical orientations
available)
Maximum reforming effciency
Highest heat fux due to co-
current fow.
Waste heat
recovery
Feed
pretreatment
Steam
reforming
CO-shift
conversion
Pressure
swing
absorption
F e e d
F u e l
D e m in e ra lise d
w a te r
H yd ro g e n
E x p o rt
ste a m
H
2
re c yc le
S te a m p ro c e ss
Ta ilg a s
Figure 2 Steam reforming
Carbon monoxide (CO) Hydrogen (H
2
)
Acetic acid Hydrotreating
Formic acid Hydrocracking
Polycarbonates Ammonia
Toluene di-isocyanate Aniline
1,4 Butanediol
Syngas (CO + H
2
) Cyclohexane
Methanol Caprolactum
Oxo-alcohols 2-Ethylhexanol
Hexamethylenediame (HMDA)
Hydrogen peroxide
Isononyl alcohol
Toluene diamine (TDI)
Industrial gases in refning and
petrochemicals
Table 1
The engineering
and manufacturing
companies have
to work as a single
team to identify the
most effcient and
economic plant design
selas.indd 2 10/12/12 14:06:03
3D TRASAR For
Crude Overhead Systems
Nalco Downstream Energy
Services is the global leader in
Refning Process Treatments.
For more than 80 years, we have
worked to build an unrivaled level
of experience and expertise with
over 1,200 employees operating
in more than 350 refneries and
400 petrochemical plants across
six continents.
3D TRASAR for Crude Overhead
Systems is a patent pending new
technology that combines an
automated analyzer, Pathfnder
n
a
i
r
o
d
a
u
c
E
a
y
a
M
h
s
o
o
r
o
S
o
p
a
N
e
t
n
e
i
r
O
6
1
y
e
r
e
M
n
a
r
d
u
B
s
a
R
b
i
r
a
h
G
s
a
R
z
o
o
r
w
o
N
e
n
e
c
o
E
h
e
i
d
e
u
o
S
a
n
a
r
f
a
a
Z
n
a
y
y
a
R
l
A
m
u
i
y
a
l
e
B
i
w
a
t
a
R
o
v
l
o
P
i
j
f
a
h
K
2
2
a
n
o
e
L
4
2
a
n
o
e
L
n
a
g
r
e
h
a
B
2
2
A
S
V
D
P
y
v
a
e
H
b
a
r
A
m
i
l
r
a
M
o
t
i
u
K
%
,
t
n
e
t
n
o
c
e
n
e
t
l
a
h
p
s
A
/
m
3
1 . 7 8
Figure 5 Specic fuel cost with and without antifoulant
dorf ketal.indd 3 10/12/12 14:19:23
www.eptq.com PTQ Q1 2013 93
References
1 Based on a natural gas price of about $6/MM
Btu for a typical 100 KBPSD renery that emits
1.2-1.5 MM t/yr of CO
2
.
2 Stockle M, Carter D, Jones L, Optimising
Renery CO
2
Emissions, Foster Wheeler
Technical Paper www.fwc.com/publications/
tech_papers/les/ERTC%20CO2%20paper%2
0Nov07.pdf
3 Brandt A R, Unnasch S, Energy intensity and
greenhouse gas emissions from California
thermal enhanced oil recovery, Energy & Fuels
2010: Keesom W, Unnasch S, Moretta J, Life cycle
assessment comparison of North American
and imported crudes. Technical report, Jacobs
Consultancy and Life Cycle Associates for
Alberta Energy Resources Institute, 2009.
India Nagi-Hanspal is Lead Renery Engineer,
Technical Services with Dorf Ketal Chemicals,
Mumbai, India. She holds a MEng degree in
chemical engineering from Imperial College
London. Email: indianagi@dorfketal.com
Mahesh Subramaniyam is Director of Research
& Development with Dorf Ketal Chemicals. He
7 . 9 6
7 . 9 8
7 . 9 4
7 . 9 2
7 . 9 0
7 . 8 8
7 . 8 6
N o a n tifo u la n t With a n tifo u la n t
S
p
e
c
i
f
i
c
f
u
e
l
c
o
n
s
u
m
p
t
i
o
n
,
k
g
/
m
3
7 . 8 4
Figure 6 Specic fuel consumption with and without antifoulant
4 1 , 0 0 0 in
fu e l sa vin g s
5 9 , 0 0 0 in
c a rb o n c re d its
Figure 7 Financial benet of antifoulant
treatment
holds a PhD in chemistry from Indian Institute
of Technology, Mumbai.
Email: drmaheshs@dorfketal.co.in
Parag Shah works in Global Renery Technical
Services with Dorf Ketal Chemicals in software
development for desalter adequacy testing and
monitoring fouling in preheat exchanger trains.
He holds a BEng in chemical engineering from
Mumbai University.
Email: paragshah@dorfketal.com
James Noland is Senior Director of the Process
Chemicals Division of Dorf Ketal USA, LLC. He
holds a BEng in chemical engineering from
Mississippi State University, USA.
Email: james.noland@dorfketal.com
dorf ketal.indd 4 12/12/12 10:55:27
www.metso.com/oilandgas
Focusing on your safety
W e of er peace of m ind w hen it com es to your process safety and productivity.
M etso is a single source for all your process control needs, from autom ation system s
to intelligent f ow control solutions that are proven in the toughest conditions.
O ur 30,000 professionals based in over 50 countries deliver sustainability and
prof tability to custom ers w orldw ide.
metso.indd 1 11/9/12 11:42:58
Towards a zero (net) carbon refnery
T
he well-to-wheel emissions
for petrol and diesel can be
split into three main parts: oil
production and transportation, oil
processing and transportation, and
emissions from the consuming
engines. In order to reduce emis-
sions and meet targets for carbon
Review of options for refners to reduce the carbon emissions from their refneries.
The technologies already exist to develop a refnery that has zero net emissions
MICHAEL STOCKLE
Foster Wheeler
reduction, it is likely that all three
parts of the chain will need to make
some contribution.
While the processing of oil in the
refnery contributes around
only 5-10% of the total well-to-
wheel emissions, the opportunities
for reducing these are signifcant
due to the nature of the processes
used, the fuels available and the
fact that these are large-scale fxed
location processes making applica-
tions such as carbon capture and
storage (CCS) more practical than
they would be on individual
vehicles.
www.eptq.com PTQ Q1 2013 95
H
2 S F G
VDU
FCC
FCC Nap
HDT
Visibreaker
Amine
treatment
Fuel gas
system
Hydrogen
plant
SRU
L P G
L C O
N a p h th a
L C O /D C O L C O /D C O
C ru d e
H
2
L P G
H
2
H
2
S F G
H e a vy
N a p h th a
L ig h t
N a p h th a
S F G
H
2
S F G
S F G H
2
S F G WN
F G
L P G
CDU
LPG
treater
Naphtha
HDT
Naphtha
splitter
Reformate
splitter
Kero
HDT
CCR
ISOM
Diesel
HDT
N a p h th a
K e ro se n e
D ie se l
VG O
Va c u u m re sid u a ls
H
2
S
G a s o il
D ie se l
K e ro se n e
R e sid u e
A tm o sp h e ric
re sid u a ls
L P G
G a so lin e
H S F O
N a p h th a
J e t
D ie se l
S u lp h u r
H
2
H
2
R e fo rm a te
I so -M
C
4
s
H e a vy
re fo rm a te
R e fo rm a te
Tre a te d g a so lin e
A tm o sp h e ric re sid u a ls
F G
L P G
S F G
Figure 1 Refnery confguration
fw.indd 1 12/12/12 12:13:12
96 PTQ Q1 2013 www.eptq.com
This article will look at the main
sources of carbon emissions from a
refnery and consider a range of
options for reducing the carbon
impact of the refnery, starting with
fuel substitution and energy eff-
ciency, then moving on to how
technologies such as CCS and
renewable power generation could
be integrated into the refnery.
The article will look at the impact
of making these changes on refnery
economics and show one possible
solution that, at a suffciently high
carbon price, could see a zero net
carbon refnery (that is, one where
no fossil fuel-derived carbon is
emitted).
Sources of CO
2
in the refnery
Sources of carbon dioxide (CO
2
) in
the refnery can be categorised into
four main groups:
Fuel to process units
Steam and power production
Hydrogen production
FCC coke (where applicable).
The emissions of an individual
refnery depend on a number of
factors. The confguration of the
refnery is a key factor, as complex
upgrading refneries produce more
CO
2
than simple hydroskimming
confgurations, but they tend to
produce more of the fuels society
demands (gasoline and diesel) per
barrel of oil processed. The fuels
used in the refnery also have an
impact, as do the crudes processed.
Heavy, sour crudes require more
energy to process than do light,
sweet crudes.
In order to quantify the relative
sizes of the emissions, we can look
at a typical refnery confguration
and the CO
2
emitted from it. In this
article, we have considered an FCC-
based European refnery processing
150 000 b/d of Ekofsk Crude that
generates all power and steam on-
site from refnery fuel gas and fuel
oil. Figure 1 shows the main confg-
uration of this refnery. The
emissions from this refnery are
shown in Figure 2.
Options for reducing refnery carbon
emissions
There are a number of ways of
reducing the carbon emissions of a
refnery, ranging from relatively
simple, low-cost options to complex,
capital-cost-intensive options. These
options can be grouped into a
number of areas and we will
consider each area in turn.
Energy effciency
The frst area of focus is energy eff-
ciency. The other options we pursue
will tend to be more expensive than
the current energy sources in the
refnery (if they were cheaper, they
would already be used) and so
making the most of the energy we
do use will become even more
important.
Even refneries that consider
themselves good performers in
terms of energy effciency can do
more, and this is illustrated by a
recent study Foster Wheeler
completed for a top-quartile refn-
ery in northern Europe. This study
resulted in operational improve-
ment and investments being
identifed that could save around
10% of the fuel used in the refnery,
and the potential projects had an
estimated payback of less than 18
months. In our example, we will
assume our base case 150 000 b/d
refnery has already achieved a
high level of effciency and
5 0 0 0
4 0 0 0
3 0 0 0
2 0 0 0
1 0 0 0
T
P
D
o
f
C
O
2
Base sources
0
U tilitie s
H P U
Visb re a k e r
F C C
D H T
N a p h th a
C D U
S R U
5 0 0 0
4 0 0 0
3 0 0 0
2 0 0 0
1 0 0 0
T
P
D
o
f
C
O
2
Base Nat . gas f i ri ng
0
U tilitie s
H P U S R U
Visb re a k e r
F C C
D H T
N a p h th a
C D U
U tilitie s
H P U
S R U
Visb re a k e r
F C C
D H T
N a p h th a
C D U
Figure 2 Base case carbon emissions
Figure 3 Comparison of impact of natural gas fring
Fuel Calorifc value, kJ/kg CO
2
tonnes/tonne CO
2
tonnes/FOE* tonne
*Fuel oil equivalent
Methane 50 000 2.75 2.20
Ethane 47 500 2.93 2.47
Propane 46 300 3.00 2.59
Ethylene 47 200 3.14 2.67
Fuel oil 40 000 3.21 3.21
Hydrogen 119 900 0 0
CO
2
emissions for typical fuels
Table 1
Hydrocarbon fuels
contain very different
levels of carbon and
result in signifcant
carbon emissions
for the same level of
energy requirement
fw.indd 2 12/12/12 12:13:23
will focus on other options to
reduce the emissions from the
refnery.
Fuel substitution
Hydrocarbon fuels contain very
different levels of carbon and result
in signifcant carbon emissions for
the same level of energy require-
ment. Table 1 compares the impact
of different fuel sources on carbon
emissions.
Figure 3 shows the total refnery
CO
2
emissions from our 150 000
b/d refnery, assuming all heat and
power is generated on-site,
compared with likely emissions if
natural gas was imported into the
refnery for power generation.
We can see that moving from
fuel oil fring to natural gas fring
reduces emissions from the refn-
ery by about 13%. Of course, the
economics of this option are highly
dependent on the relative prices of
natural gas and fuel oil. In the US,
where the differential between
natural gas and crude price is
wide, natural gas fring will be
increasingly attractive. In Europe,
the incentive will be lower, but
will still generally favour natural
gas fring where available. Figure 4
shows the natural gas price
required to break even on energy
cost versus fuel oil at varying
carbon prices.
Green heat and power
In our example 150 000 b/d refn-
ery, we have assumed that all of
the refnery's power requirements
are generated on-site. One way to
reduce the refnery's emissions
would be to generate this power
from low-carbon sources, either on-
site or by purchasing green power
over the fence.
If we replace all of the on-site
power generation with imported
low-carbon power, emissions drop,
as shown in Figure 5.
This option reduces on-site emis-
sions by 28%, but is dependent on
a reliable supply of green power. If
power was required to be gener-
ated on-site, options such as wind
and solar could be evaluated, but
again the dependability of these
sources may be an issue, and
investment is likely to be required
www.eptq.com PTQ Q1 2013 97
3 0
2 5
2 0
1 5
1 0
5
5 0 6 0 7 0 8 0 9 0 1 0 0 1 1 0 1 2 0
N
a
t
u
r
a
l
g
a
s
p
r
i
c
e
,
$
/
M
M
B
T
U
Fuel oi l pri ce, $/ bbl
0
0
$ /to n n e C O
2
5 0
2 5
5 0 0 0
4 0 0 0
3 0 0 0
2 0 0 0
1 0 0 0
T
P
D
o
f
C
O
2
Base Low carbon
el ect ri ci t y purchase
0
U tilitie s
H P U S R U
Visb re a k e r
F C C
D H T
N a p h th a
C D U
U tilitie s
Visb re a k e r
F C C
D H T
N a p h th a
C D U
H P U S R U
3 , 0 0 0 , 0 0 0
2 , 5 0 0 , 0 0 0
2 , 0 0 0 , 0 0 0
1 , 5 0 0 , 0 0 0
1 , 0 0 0 , 0 0 0
5 0 0 , 0 0 0
B a se F u ll p o st-c o m b u stio n
c a p tu re
P re -c o m b u stio n
c a p tu re e x c e p t F C C )
T
P
A
C
O
2
0
E m itte d
C a p tu re d
in backup systems (which will not
offer a carbon saving) or energy
storage.
If we combined natural gas fring
and low-carbon power purchase,
we can achieve an emissions reduc-
tion by almost 40% from the
original base case. To make further
reductions in emissions, other, less
conventional options need to be
investigated. These could include:
Renewable sources of heat, such
as heat from concentrated solar
power (CSP) or electrical heating
from green power
Carbon capture
Figure 4 Natural gas breakeven price at varying carbon price
Figure 5 Impact on carbon emissions of electrical power purchase
Figure 6 Impact of carbon capture
fw.indd 3 12/12/12 12:13:35
costs of this are high and signifcant
levels of power are required.
Therefore, this is not considered
further here.
A number of biomass-based fuels
could potentially be used in the
refnery, with everything from
wood chips to algae being potential
fuels for the utilities boilers. For the
process units, converting to solid
fuels would be more challenging:
so, in order to keep the assump-
tions simple, we will look at an
example where some of the
purchased natural gas is replaced
with bio-derived methane. This
could be produced from anaerobic
digestion or by methanation of bio-
derived syngas.
The production of the biogas uses
energy and, because this is not all
carbon-free, there are still some
emissions associated with using this
gas. In our example where we
combine power import, natural gas
import and full post-combustion
capture, our refnery still emits
around 140 000 t/y of CO
2
(about
8% of the original emissions). To
get to a zero net carbon refnery,
this amount would need to be
equivalent to the bio-derived
content of the biogas.
Figure 7 shows the emissions on
the zero net refnery. Table 2 shows
the level of biogas fring required at
different levels of carbon reduction
achieved by the biogas.
From Table 2, we can see that at
high rates of replacement of natural
gas with biogas, it is possible to
actually achieve negative overall
emissions from the refnery and, in
effect, decarbonise some of the
product, although the impact on
products is small, with even the
most optimistic case reducing emis-
sions from the products by about
only 2.5%.
Conclusions
We have seen that a number of
options exist for refners to reduce
the carbon emissions from their
refneries. The technologies already
exist to develop a refnery that has
zero net emissions. The challenge in
achieving this is not technical, but
commercial, with refners requiring
suffcient incentive and capital to
invest, and suffcient protection
98 PTQ Q1 2013 www.eptq.com
Use of biomass-derived or other
low-net carbon fuels.
There are several potential
renewable sources of heat that
could be used in a refnery. Green
power could be used in electrical
heaters, and CSP or geothermal
power could be used as sources of
heat to generate steam. These
options have not been considered
in detail in this article, but could
offer opportunities for refners
with easy access to abundant and
cost-effective sources of these types
of energy.
Carbon capture
There are a number of ways that
carbon capture can be integrated
into the refnery. All three of the
commonly proposed capture
schemes (pre-combustion, post-
combustion and oxyfuel capture)
can be used in the refnery, with all
three schemes technically viable in
some of the refnery processes.
Carbon capture can typically
recover up to 90% of the carbon in
the fuel and can enable a refnery to
get much closer to a zero carbon
target. We will look at two options.
The frst one sees us change all of
the fred heaters in the refnery to
fre on hydrogen and capture the
CO
2
produced in making hydrogen
(pre-combustion capture). For the
second, post-combustion capture
can be added on all sources of
carbon from the refnery.
Carbon capture can deliver a
large reduction in emissions, but it
still cannot achieve 100% carbon
emission-free processing in the
refnery. Even combining capture
with the other options considered,
using natural gas as fuel, importing
power and implementing CCS,
around 8% of the carbon emitted in
the base case is still emitted. Of
course, consideration also needs to
be given to where the CO
2
is stored,
but this is outside the scope of this
article.
To achieve a net emission of no
fossil-derived carbon, we need
alternative fuel sources that are not
fossil-fuel based. The obvious
choice here is biomass-based fuels.
The carbon from these sources has
been captured from the atmosphere
and will be captured again as plants
grow, absorbing the emissions. An
alternative option could be to use
hydrogen produced from electroly-
sis using green electricity, but the
2 0 0 , 0 0 0
4 0 0 , 0 0 0
6 0 0 , 0 0 0
8 0 0 , 0 0 0
1 , 0 0 0 , 0 0 0
1 , 2 0 0 , 0 0 0
1 , 4 0 0 , 0 0 0
C a rb o n
e m issio n s
T
P
A
C
O
2
0
C a p tu re d
B io -d e rive d
Figure 7 Carbon emissions from the zero
net carbon refnery
Net emissions (TPA) % natural gas from biomass
when using biogas 10% 20% 50% 70% 100%
Biogas emissions as 0% 68 830 886 -202 944 -338 830 -542 660
% of natural gas 20% 82 418 28 064 -135 000 -243 710 -406 774
50% 102 801 68 830 -33 085 -101 029 -202 944
80% 123 184 109 595 68 830 41 652 886
Impact of bio-derived gas and carbon capture on emissions
Table 2
Carbon capture can
typically recover up
to 90% of the carbon
in the fuel and can
enable a refnery to
get much closer to a
zero carbon target
fw.indd 4 12/12/12 12:13:49
against the impacts that lower
carbon emissions could have on
their competitive position in their
domestic and export markets.
Current high oil prices mean that
the returns on energy efciency are
good and most reneries are likely
to nd some attractive investments
in this area. Where natural gas is
available, changing from fuel oil
ring to natural gas ring is also
likely to be attractive. Steps beyond
these options are generally less
attractive nancially unless there is
a sufciently high carbon price or
appropriate level of government
subsidy.
Previous Foster Wheeler studies
have shown a carbon price of
around $60/t CO
2
is required to
justify the investment in carbon
capture. The higher cost of
biomass-derived fuels means that
similar or higher levels of subsidy/
carbon tax are required to
make this option economically
attractive.
www.eptq.com PTQ Q1 2013 99
This results in a hierarchy of
actions for any rener looking to
reduce carbon emissions:
Maximise design efciency and
operate efciently
Minimise the carbon in the fossils
fuels you use
Capture the carbon you cannot
avoid and/or move to low-carbon
fuels and power.
The right solution will, of course,
depend upon the specic circum-
stances and business objectives of a
particular renery, and the full
range of options need to be studied
in order to develop a robust techni-
cal solution and optimised
investment plan.
Mike Stockle is Chief Engineer of Rening
Technology with Foster Wheeler and currently
works in the Business Solutions Group in
Reading, UK. During his time at Foster Wheeler,
he has worked on a number of rening projects,
ranging from a grassroots renery conguration
studies and FEEDs, through major renery
revamps, to a number of small studies. He is an
experienced LP modeller and has undertaken
a number of feasibility and front-end studies
looking at the impacts of changing markets
and legislation on reneries across the globe.
He holds an MEng in chemical engineering
from Nottingham University and is a Chartered
Engineer and a Fellow of the IChemE.
Email: michael_stockle@fwuk.fwc.com
Current high oil prices
mean that the returns
on energy efciency
are good and most
reneries are likely to
nd some attractive
investments in
this area
Tripl e eccentric
Butterfl y Val ves
Check Val ves
Doubl e Bl ock and Bl eed
ESD Val ves
www.zwick-armaturen.de
fw.indd 5 12/12/12 12:13:59
C O M P R E S S O R S T U R B I N E S G L O B A L S E R V I C E
EBARA CORPORATION
w w w .elliott-turbo.com
Cust omer:
Ethylene plant, Saudi Arabia.
Chal l enge:
O verhaul and repair four critical steam turbines in 12 days.
Resul t :
Elliott coordinated six m onths of pre-planning
and com pleted the project four days early.
They turned to Elliott
for planning and execution.
The custom er turned to Elliott because they understood that planning is everything
and that few com panies can m atch Elliott in turnaround planning and execution.
W ho w ill you turn to?
The world turns to Elliott.
elliott.indd 1 1/6/12 21:23:46
Modular design of smaller-scale GTL
plants
M
odularisation is a construc-
tion method that involves
designing and building
refnery and gas processing facili-
ties as series of unitised (or
modularised) process and utility
systems. The units are designed
and fabricated in a controlled shop
environment, then shipped to the
project site, where they are installed
on foundations and connected
together. Compared to traditional
feld construction methods, where
all materials are delivered to the
project site and then built on loca-
tion, modularisation is proving to
be an effective solution for achiev-
ing fast-track project schedules, and
for building refning and gas
processing facilities in remote
locations. For example, Ventech
Engineers International LLC
(Ventech), based in Pasadena,
Texas, is an engineering, procure-
ment and construction company
that specialises in the design and
fabrication of modularised facilities.
A project in Siberia illustrates
the capabilities of modularisation
when dealing with remote and
challenging environments. Ventech
designed, fabricated and shipped a
2000 b/d crude processing plant to
the remote Siberian town of Urai in
just six months after contract sign-
ing. Urais temperatures drop as
low as -45C (-50F), so the modules
were designed and fabricated to
accommodate these conditions. In
particular, the crude process
module was completely enclosed
and heated. Ventech personnel,
assisted by local labour, installed
the modules at the project site and
started refnery operations four
weeks after the modules delivery.
Modularisation is opening up opportunities for GTL plants to monetise small-scale
natural gas deposits both on- and offshore
KenneTh RobeRTs
Velocys
Ventech has applied modular
design and construction most
recently in the Kurdistan region of
northern Iraq. A large-scale expan-
sion to a refnery in Kurdistan
started during 2010, with Ventech
providing a 20 000 b/d modular-
ised crude processing plant that
was added to the existing 20 000
b/d refnery. Made up of
26 modules, the new crude unit
was designed and fabricated at
Ventechs Pasadena fabrication site.
The modules were shipped to the
port of Mersin, Turkey, then trans-
ported by truck to the project site
in Kurdistan, where the refnery
was erected in 16 weeks.
The refnery has continued to
expand in a modular fashion; once
completed in 2013, total capacity
will be over 185 000 b/d and it
will remain the sole producer of
unleaded gasoline in Iraq. In all,
Ventech has provided modularised
crude distillation units, naphtha
hydrotreaters, catalytic reformers,
isomerisation units, demercaptani-
sation systems, gas plants and
supporting utilities to this three-
phase project. Modularisation has
made the difference in this
geographically and politically chal-
lenging location, enabling expansion
work to be accomplished quickly.
Design concepts for modularisation
In modular plant design, the differ-
ent process units are contained in
separate modules. For ease of ship-
ment to any location, Ventech
fabricates its modules to a standard
size of 13.5ft wide x 12ft high x 40ft
long. Each of the modules and
its associated equipment has a
reference code, which helps
www.eptq.com PTQ Q1 2013 101
Figure 1 Fischer-Tropsch module process unit Courtesy of Ventech
velocys.indd 1 10/12/12 14:27:48
102 PTQ Q1 2013 www.eptq.com
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
Distributed GTL plants, using
smaller Fisher-Tropsch (FT) reactors
and located near the natural gas
source, can be designed for use in
both offshore and remote on-shore
locations. The FT reactors devel-
oped by Velocys, the US-based
member of the Oxford Catalysts
Group, employ microchannel tech-
nology. This is a developing feld of
chemical processing that intensifes
chemical reactions and reduces the
dimensions of the reactor systems
by increasing the heat removal
capability to enable reactions to
occur at rates 10-1000 times faster
than in conventional systems. This
makes it possible to use more active
FT catalysts.
Velocys reactors are the only FT
reactors to use a highly active FT
catalyst developed by Oxford
Catalysts to accelerate FT reactions
by a factor of 10-15 times compared
to conventional reactors. Individual
reactors are designed to produce
125 b/d of FT product, and desired
plant capacity is reached by linking
together, or numbering up, multiple
reactors to meet the availability of
gas feed. For example, by linking
up 40 individual 125 b/d reactors,
a capacity of 5000 b/d can be
achieved with a natural gas feed
rate of approximately 50 million cu
ft/day. As a result, the distributed
GTL plant technology developed by
Velocys has the fexibility to scale
production to match the available
resource. Plants of this type can
also be used as a fexible and
economical way to expand the
capacity at existing petroleum refn-
eries when an economic gas supply
is available.
Velocys is currently working with
Haldor Topse and Ventech to
design a modular GTL plant using
its microchannel FT reactor (see
Figure 2). The modular GTL plant
will be comprised of standard-sized
13.5 x 12 x 40ft (4.1 x 3.65 x 12m)
modules that will be built in
Ventechs fabrication facility. In
these GTL plant designs, autother-
mal reforming (ATR) reactors from
Haldor Topse will be used to
produce the syngas feedstock for
the Velocys FT reactors. Haldor
Topse is also providing the hydro-
processing unit that takes raw
www.eptq.com PTQ Q1 2013 103
products from the FT reactor and
produces fnished fuels.
First commercial plant
The frst company to select the
modular GTL plant design is
Calumet Specialty Products Partners
L.P., an independent producer of
speciality hydrocarbon products in
North America. Calumet plans to
use the modular GTL plant, which
has a nominal capacity of 1000 bpd,
in the expansion of its Specialty
Products facility at Karns City,
Pennsylvania. The very pure,
paraffnic hydrocarbons produced
by the GTL plant will be used as
feedstock for some of Calumets
1500 products, including food-grade
waxes, solvents and lubricants.
Calumet was attracted to the idea
of modular GTL technology because
it offers the ability to use readily
available, low-cost natural gas to
reduce overall costs, increase the
security of feedstock supply and
improve product quality.
The GTL plants modules will be
fabricated in Ventechs Texas fabri-
cation facility, then transported by
truck to the Karns City site for
installation and integration with
Calumets existing refnery. Velocys
engineers are working with Ventech
and Haldor Topse on the plant
design, which is due be completed
by late 2012. Fabrication is expected
to begin during the frst half of 2013
and the modular GTL plant would
come on-stream during the second
half of 2014.
More projects
Meanwhile, both Velocys and
Ventech expect that the Calumet
project will be just the frst of many
small-scale modular GTL plants
announced, several of which are
already undergoing preliminary
design. Velocys continues to receive
numerous inquiries about the
microchannel technology, not only
for distributed GTL but also for
biomass-to-liquids (BTL) applica-
tions. For example, Solena has
recently chosen the Velocys FT
process for its GreenSky London
Project, which will produce renew-
able jet fuel in the UK.
Given the low natural gas prices
in the US, and the small and often
remote nature of many gas deposits
around the world, interest in modu-
lar distributed GTL plants
incorporating microchannel FT reac-
tors is expected to grow. Once the
Calumet GTL plant comes on
stream, Velocys, Ventech and Haldor
Topse are confdent that such
plants will prove to be the best way
forward for monetising stranded
and undervalued natural gas.
Kenneth Roberts is Senior Business
Development Director with Velocys, Houston,
Texas. With over 35 years experience in the
energy industry, he supports engineering
and business development activities in
collaboration with Ventech and Haldor Topse
for the modular GTL design. He holds a BS in
mechanical engineering and a MBA in strategic
planning from the University of Texas at
Austin.
H ig h h e a t flu x
1 0 tim e s h ig h e r th a n
c o n ve n tio n a l re a c to rs
S yn g a s
F T p ro d u c ts
0 . 2 5 to 1 0 m m
0 . 2 5 to 1 0 m m
Figure 2 Microchannel FT reactor internals Courtesy of Oxford Catalysts Group
velocys.indd 3 10/12/12 14:28:05
Commerci al l y Proven Route for the Producti on of Gasol i ne through
Methanol Synthesi s and Conversi on
ExxonMobil Research and Engineering Companys Methanol to Gasoline (MTG) Technology, was commercially operated in
New Zealand for over 10 years. Now we offer the second generation MTG technology which provides:
8ulphur-free low oenzene gasoline
A product full] compatiole with refner] gasoline
8imple, reliaole fxed oed reactor design
So Advanc ed,
Yet So Si mpl e.
3225 Gallows Road, Fairfax, virginia 22037-0001, USA
+1-703-846-2568 fax +1-703-846-3872
www.exxonmobil.com/refningtechnologies tsl.email@exxonmobil.com
exxon.indd 1 10/12/12 11:14:52
Starting up hydrotreating units with
cracked feeds
R
efners can realise signifcant
operating margins by process-
ing cracked feedstocks such as
coker naphtha and light cycle oil
(LCO). Unfortunately, after catalyst
change-outs on hydrotreating
(HDS) units, catalyst manufacturers
recommend that the refner only
process straight-run feeds for at
least the frst three days of the cata-
lyst cycle. This delay in processing
cracked feeds has several potential
costs including:
Lost profts from processing
cracked and heavy feeds
Storage costs for a three-day
supply of straight-run feed
Storage costs for three days of
cracked feed that must be held
aside for processing later
The purchase of additional
straight-run feed if a three-day
supply is not available.
Problem with cracked feeds
Cracked feeds such as coker naph-
tha contain signifcant quantities of
highly reactive olefns and diole-
fns, while cracked gas oils also
contain polynuclear aromatics. If
these components are exposed to
the hyperactive catalytic sites that
exist on freshly sulphided cata-
lysts, coke and gums will form that
readily deposit on the catalyst
surface. Such deposits tend to
obstruct the catalysts pores and
inhibit access to the active sites,
leading to a permanent loss of
catalyst activity and a reduction in
cycle length.
To avoid the adverse effects of
cracked feeds at startup, catalyst
manufacturers typically recommend
a break-in period of at least three
days, feeding only straight-run
A catalyst sulphiding process enables the introduction of cracked or heavy feeds
to HDS units without the recommended three-day delay following changeover
RANDY ALEXANDER, REiNDER VAN BRuggEN and PiERRE DufRESNE
Eurecat US
product to the unit. During this
three-day break-in period, a small
amount of soft coke will form on
the catalyst surface, gradually
tempering the hyperactivity of the
catalyst. Once catalyst activity has
stabilised, cracked feeds can be
introduced to the unit without the
risk of premature coke and gum
formation.
Eurecat has developed an ex-situ
treatment that allows the introduc-
tion of cracked or heavy feeds
without the recommended three-
day delay. This process, called
Totsucat CFP, combines the
Totsucat activation process with an
additional step to protect the cata-
lyst from the harmful effects of
cracked feed, ensuring long-term
reactor performance irrespective of
when cracked feeds are introduced
to the unit.
This is a carefully controlled
procedure that gently moderates
the hyperactivity that occurs imme-
diately after catalyst sulphiding.
Catalysts treated with this are
fully sulphided and activated prior
to loading. The activity will mimic a
catalyst that has already been condi-
tioned with straight-run feed for
three days. Cracked and heavy feed-
stocks can be gradually introduced
during the frst hours of a startup
without causing permanent damage
to the catalyst. Coke and gum
deposits will be minimised, since the
catalyst activity has been carefully
stabilised following activation.
Recent unit data also show that
this treatment can signifcantly
reduce the deactivation rate of a
hydrotreating unit, allowing the
refner to extend their cycle by as
much as 50%.
Totsucat catalyst activation process
Unlike with other pre-sulphurisa-
tion processes, Totsucat-treated
catalysts are delivered to the HDS
unit totally activated and ready to
work immediately. The uncertainty
of incomplete activation is elimi-
nated and startup time is greatly
reduced.
The advantages of the process
include:
Load-and-go reactor startups. The
catalyst is fully sulphided and acti-
vated and does not require any
additional sulphiding agents or
complex activation procedures
No sulphiding exotherms since
the sulphiding step is completed in
Eurecats equipment
No need for additional hydrogen
at startup
Minimal sour water formation
during startup
The hazards and odours associ-
ated with sulphiding chemicals are
avoided since spiking agents are no
longer needed at startup
No need to measure H
2
S levels in
the recycle gas
www.eptq.com PTQ Q1 2013 105
Catalyst
manufacturers
recommend a break-
in period of at least
three days, feeding
only straight-run
product to the unit
eurecat.indd 1 10/12/12 14:32:54
106 PTQ Q1 2013 www.eptq.com
Negligible amounts of H
2
S are
released during reactor heat-up,
protecting sulphur-sensitive units
downstream and avoiding the need
to are undesirable byproducts
Circulating high concentrations of
wet H
2
S during in-situ sulphiding
can cause severe corrosion prob-
lems. Totsucat eliminates this issue
since only low concentrations of
dry H
2
S will be recirculating
through the system.
Starting a unit with Totsucat-
treated catalysts is similar to a
restart after an emergency shut-
down. Once the reactor is loaded,
it only requires heating up to the
start-of-run temperature. Time-
consuming drying steps, holding
steps and tests to determine H
2
S
concentrations are not required.
Startup time is reduced to a few
hours, while startup risks are mini-
mised. The catalyst bed will
achieve peak performance since
sulphiding and activation are
controlled at very precise
conditions.
The treatment is particularly
useful for:
Critical path units where lost
production time is costly
Units with temperature limits,
where effective sulphiding is dif-
cult, if not impossible, to achieve
Reneries with limited hydrogen
supplies
HDS units that run in parallel or
series, where only one unit is taken
down while the other is still
operating
Units that cannot tolerate H
2
S
breakthrough to downstream units
Units that utilise gas-phase star-
tup, increasing the risk of
catastrophic exotherms during in-
situ sulphiding
Units where the use of sulphiding
chemicals is undesirable or
uneconomical.
Case study 1
A diesel hydrotreater (DHT) located
on the US East Coast was loaded
with Totsucat CFP-activated cata-
lyst in the summer of 2011. This
moderate-pressure unit (800 psi
with a hydrogen partial pressure of
6 8 0
7 2 0
7 1 0
7 0 0
6 9 0
6 7 0
6 6 0
6 5 0
6 4 0
6 3 0
0 2 0 4 0 6 0 8 0 1 0 0 1 2 0 1 4 0 1 6 0 1 8 0 2 0 0 2 2 0 2 4 0 2 6 0 2 8 0
T
e
m
p
e
r
a
t
u
r
e
,
F
6 2 0
C yc le 2 5 n o rm a lise d
C yc le 2 5 a c tu a l
C yc le 2 4 n o rm a lise d
y = 0 . 1 6 4 2 x +6 3 4 . 5 1
R
2
= 0 . 4 6 8 2
Figure 1 Reactor temperature (WABT) after 90 days on oil
7 2 0
7 0 0
6 8 0
6 6 0
6 4 0
6 2 0
0 3 0 6 0 9 0 1 2 0 1 5 0 1 8 0 2 1 0
T
e
m
p
e
r
a
t
u
r
e
,
F
Days on st ream
6 0 0
A c tu a l c yc le
N o rm a lise d c u rre n t c yc le T-re q .
N o rm a lise d p re vio u s c yc le T-re q .
D e a c tiva tio n = 3 . 7 F /m o n th
S O R WA B T = 6 4 5 F
P re vio u s d e a c tiva tio n = 8 . 9 F /m o n th
Figure 2 Reactor temperature (WABT) after 180 days on oil
7 2 0
7 6 0
7 4 0
7 0 0
6 8 0
6 6 0
6 4 0
6 2 0
0 3 0 6 0 9 0 1 2 0 1 5 0 1 8 0 2 1 0 2 4 0 2 7 0 3 0 0 3 3 0
T
e
m
p
e
r
a
t
u
r
e
,
F
Days on st ream
6 0 0
A c tu a l c yc le
N o rm a lise d c u rre n t c yc le T-re q .
N o rm a lise d p re vio u s c yc le T-re q .
C u rre n t c yc le d e a c tiva tio n = 4 . 2 F /m o n th
S O R WA B T = 6 4 5 F
P re vio u s c yc le d e a c tiva tio n = 7 . 8 F /m o n th
Figure 3 Reactor temperature (WABT) after 330 days on oil
Circulating high
concentrations
of wet H
2
S during
in-situ sulphiding
can cause severe
corrosion problems
eurecat.indd 2 10/12/12 14:33:07
A B B Lifecycle S ervices. N o one else know s your
system s like w e do.
W ith over 30 years of autom ation system heritage, no one else know s your A B B
control system s better than w e do! A nd w ith our outstanding lifecycle policies and
program s, system expertise, evolution options, value added services and determ ined
com m itm ent to our custom ers, w e can help you evolve to the latest in control system
technology w ith S ystem 800xA . S o, if you have an IN FI90, M O D 300, A dvant, D C I, or
other control system , w e are here to m ake sure your control system w ill continue to
m eet your business needs now and in the future. For m ore inform ation:
w w w. a b b . c o m /c o n tro ls y s te m s
PTQ_Sweden_Evol_A4.indd 1 08/09/2011 12:30 abb auto.indd 1 10/12/12 11:07:28
550-600 psi) processes up to 46 000
b/d of a mix of 35-40% straight-run
feed and 60-65% cracked stocks,
including 30% LCO. The reactor
holds 106 000 lb (48 tons) of CoMo
108 PTQ Q1 2013 www.eptq.com
catalyst. Due to the severity of the
feeds being processed and the size
of the reactor, the rener typically
achieves cycle lengths of less than
12 months before the end-of-run
(EOR) temperature is reached, forc-
ing a unit shutdown and a catalyst
change.
The following data compare the
reactor weighted average bed
temperature (WABT) for two cycles:
one utilising Totsucat D-treated
catalyst (Cycle 24) and another
where Totsucat CFP was used to
activate and protect the catalyst
(Cycle 25). Figure 1 shows the
WABT for the unit after 90 days.
Cycle 24 was considered a success-
ful run since it was one of the
longest cycles achieved by this unit.
But, in Cycle 25, LCO was intro-
duced in signicant quantities on
Day 2. After 90 days on oil, Cycle
25 exhibited a lower deactivation
rate than Cycle 24.
Figure 2 conrms that the deacti-
vation rate of the catalyst remains
much lower with Cycle 25 at
180 days on oil. The rate was only
4F (2.2C) per month versus 9F
(5C) per month for the previous
cycle.
Figure 3 shows that the trend
continued past the normal EOR for
this unit. Unit run data show that
the lower deactivation rate with
Cycle 25 continued until EOR. This
allowed the rener to extend the
cycle to 420 days, which is approxi-
mately 60% longer than anticipated
for this catalyst ll.
In comparing the two cycles, the
rener explained that Cycle 25 was
actually more severe than Cycle 24.
For Cycle 25, an additional
325 000 barrels of LCO had been
accumulated offsite during the
turnaround and required treatment.
Therefore, in addition to the normal
amount of LCO processed, an addi-
tional 3000-5000 b/d was fed to the
reactor for the rst 90 days of Cycle
25. The rener also explained that
this unit had averaged a deactiva-
tion rate of 20F (11C) per month
over the last 10 cycles.
In the past, the catalyst manufac-
turer recommended that a gentle
and lengthy startup should be
employed for this unit in order to
avoid premature activity loss.
In this case, Totsucat CFP has
proven that a run can be signi-
cantly extended without the
need for a gentle startup, allowing
the rener to gain signicant
6 0 0
7 0 0
6 5 0
5 5 0
5 0 0
0 5 0 1 0 0 1 5 0 2 0 0 2 5 0 3 0 0 3 5 0
W
A
B
T
,
F
Days on oi l
4 5 0
To tsu c a t C F P R u n 2
To tsu c a t C F P R u n 1
I n situ with b re a k in
Figure 4 Reactor temperature (WABT)
1 2 0
1 4 0
1 0 0
8 0
6 0
4 0
2 0
0 5 0 1 0 0 1 5 0 2 0 0 2 5 0 3 0 0 3 5 0
T
e
m
p
e
r
a
t
u
r
e
,
F
Days on oi l
0
I n le t to b o tto m T To tsu c a t C F P R u n 2
I n le t to b o tto m T To tsu c a t C F P R u n 1
I n le t to b o tto m T in situ b re a k in
Figure 5 Inlet to bottom T
120
140
100
80
60
40
20
0 50 100 150 200 250 300 350
0
0
b
/
d
,
d
e
e
f
a
h
t
h
p
a
n
r
e
k
o
C
Days on oil
0
Coker naphtha - Totsucat CFP Run 2
Coker naphtha - Totsucat CFP Run 1
Coker naphtha - in situ break in
Figure 6 Coker naphtha feed rate
eurecat.indd 3 12/12/12 10:59:06
exhibited a greatly reduced deacti-
vation rate, extending the unit cycle
length by 60%. And in a coker
naphtha unit, it helped extend unit
cycle length by over 50%.
Randy Alexander is the Director of Sales and
Marketing for Eurecat US, Houston, Texas. He
has over 25 years experience in the areas of
hydrotreating and hydrogenation catalysts, is
responsible for sales and marketing of catalyst
sulphiding and regeneration services and for
directing new product development, and holds
a bachelors degree in chemical engineering
from the University of Texas and a MBA from
the University of Houston.
Email: ralexander@eurecat.com.
Reinder van Bruggen joined Eurecat France in
2003 as Area Sales Manager and is currently
Totsucat Product Manager. He holds an MSc in
chemical engineering from Twente University
of Technology, The Netherlands.
Pierre Dufresne is Vice President, Research and
Development, for Eurecat SA. He holds a degree
from the Chemistry School of Mulhouse and a
PhD from the University of Lille. He has co-
authored around 70 patents.
Email: p.dufresne@eurecat.fr
www.eptq.com PTQ Q1 2013 109
margins by processing much more
LCO.
Case study 2
A naphtha hydrotreater (NHT)
located in a Gulf Coast renery is
now on its eighth cycle using
Totsucat CFP to pre-activate the
catalyst. The unit processes 35 000
b/d of feed consisting of 25-50%
coker naphtha. Due to limited stor-
age capacity, the rener struggled
to delay the introduction of cracked
feeds and typically brought them in
prior to the recommended three-
day break-in period.
Figure 4 shows the reactor
temperature for a typical cycle as
compared to two cycles utilising
Totsucat CFP ex-situ activation.
The rst run, dened as in-situ
with break-in, utilised a catalyst
that was sulphided in-situ and
then fed only virgin feed for three
days. For the other two cycles, the
reactor was loaded with treated
catalysts and progressively fed
coker naphtha during the rst day
of the cycle. Both of these latter
runs showed steady performance.
Figure 5 denotes the temperature
difference across the reactor during
each cycle. On average, Totsucat
CFP provided a temperature gradi-
ent across the bed similar to the
in-situ cycle.
Figure 6 shows that the amount
of coker naphtha fed to the unit
during each run was comparable.
The reners experience shows
that using Totsucat CFP pre-
activated catalysts can extend run
length without the need to delay
cracked feeds for three days at the
beginning of the cycle.
Conclusion
The addition of CFP to the Totsucat
process enables the rener to begin
introducing cracked stocks to their
hydrotreater on the rst day of a
unit startup without the need for
the three-day break-in period utilis-
ing only straight-run feed. In a
moderate-pressure diesel HDS unit,
catalyst treated with this process
Taking the Guesswork Out of CO
2
Removal
ProTreat Simulation Software
Get Real
Dont Guess!
SYNGAS, LNG AND CARBON CAPTURE
ProTreat accurately predicts column performance.
100% Mass-Transfer-Rate based no ideal stages.
Grassroots? Revamp? Designing for Packing?
Using MDEA with piperazine? HotPot?
Build a virtual plant with the industrys
most advanced gas treating simulator.
www.ogt rt .com +1 281 970 2700 info@ogt rt .com
Learn how state-of-the-art simulation science can beneft you.
OGT-0117 PTQ Half Horizontal Ad.indd 1 11/13/12 10:40 PM
eurecat.indd 4 12/12/12 10:58:27
Don t mi ss out
on t he i ndust r y s
most i mpor t ant
event of 2013
Marriott Rivercenter
San Antonio, Texas
March 17 19, 2013
The AFPM Annual
Meeting brings
together the rening
industrys most
important players
each year. Dont miss
this opportunity to be
a part of the action.
The 2013 program has something
for everyone from process
and operations engineers to
senior management. Keynote
sessions feature top government
ofcials and corporate CEOs.
Breakout sessions cover nearly
every facet of rening technology,
as well as public policy
issues impacting the industry.
Dont miss the networking
opportunities at our receptions
and afliate- hosted events.
Be there and be sure to ex tend
your reach within the industry.
Register at www.afpm.org.
afpm.org
EXTEND
YOUR REACH
ATTEND THE 2013
AFPM ANNUAL MEETING
afpm.indd 1 6/12/12 12:54:27
Optimising a process scheme for
platforming heaters
C
oncern over safety issues,
especially in industries such
as rening, lead operating
personnel to be more and more
reluctant to operate their plants
near design or permissible limits,
thereby avoiding the risk of mishap
or industrial accident. On the other
hand, they are under pressure to
increase their prot margins.
CCR platforming heaters in an
operating oil renery have been in
operation for the last 30 years or so.
Due to the existing process scheme,
a thermal imbalance in all four of
the CCR platforming units heaters
had been observed in terms of
permissible limits for tube metal
temperatures and uneven bridge-
wall temperatures. The operator
decided to engage a consultant to
analyse the problem and to come
up with an optimised process
scheme, thereby avoiding operation
of the furnaces near the design
limits of the process coils.
The study was carried out based
on various document and data
provided for the existing system in
the form of DCS output to obtain
full details of the process.
Methodology of study
The feed property grid was gener-
ated using PRO II and a
thermodynamic method (Grayson
Streed). A reasonable process-side
fouling factor in the convection
zone has been considered, which is
quite likely to happen over a long
period of operation. The CCR plat-
forming heaters were studied using
commercial software (FRNC-5PC
Version 4.18 Mod 7.6) with the
following assumptions and data for
modelling: radiant heat loss of 2.5%;
Thermal imbalance resulting from an existing process scheme for CCR platforming
furnaces is addressed by process schemes for distributing heat load more uniformly
ADIL REHMAN, SUMAN PACHAL, SHYAM K CHOUDHARY, UGRASEN YADAV and M K E PRASAD
Technip KT India
ambient air temperature of 17C
and relative humidity of 95%, as
per the meteorological data
provided by the client; maintaining
the oxygen percentage at the outlet
of the convection section by keep-
ing the quantity of air the same as
in the existing operating case in
order to account for the ingress of
air, which might have taken place
over a long period of operation.
The algorithm followed for the
study is shown in Figure 1.
Denition of uid
Fluid in this case is dened and
identied as process feed entering
the reactor. Thus, process uid enter-
ing Reactor 1 is called Fluid 1,
process feed entering Reactor 2 is
Fluid 2 and process feed entering
Reactor 3 is Fluid 3. This denition
of uid is shown in Figure 2.
Comparison of existing scheme with
proposed scheme
The existing scheme according to
DCS outputs is summarised in the
process ow diagram shown in
Figure 3. The proposed schemes
results obtained from FRNC have
been summarised in the form of
www.eptq.com PTQ Q1 2013 111
S te p 1
E va lu a tio n a n d c h e c k in g th e
c o m p le te n e ss o f in p u t d a ta
p ro vid e d b y th e c lie n t
S te p 2
U tilise th e c o lle c te d d a ta to
d e ve lo p a b a se c a se
sim u la tio n m o d e l
S te p 3
R e vie w th e re su lts o f th e
sim u la tio n m o d e l fo r th e
a c tu a l c a se
S te p 4
D e ve lo p sim u la tio n m o d e ls
fo r p ro p o se d o p tio n s
S te p 5
R e vie w th e re su lts o f d iffe re n t
sim u la tio n m o d e ls fo r
p ro p o se d o p tio n s
S te p 6
C o m p a riso n o f va rio u s
o p tio n s a n d se le c tio n o f m o st
o p tim ise d o p tio n
S te p 7
P re p a ra tio n o f a stu d y re p o rt
a n d issu e th e re p o rt fo r c lie n t
re vie w a n d c o m m e n ts if a n y
Figure 1 Algorithm followed for the study
P a c k in o x
F lu id 1
F lu id 2 F lu id 3
C o n ve c tio n +
5 0 1 A a n d 5 0 1 B
R e a c to r 1
R e a c to r 2
R e a c to r 3
F u rn a c e 5 0 2
F u rn a c e 5 0 3
Figure 2 Denition of uid
technip.indd 1 10/12/12 14:37:10
112 PTQ Q1 2013 www.eptq.com
process fow diagrams shown in
Figures 4, 5 and 6 for proposed
Options I, II and III, respectively.
Results and discussion
Mechanical aspects
Option I
This scheme is identical to the exist-
ing scheme with regard to the
process fow pass arrangement
inside the heaters. Thus, modifca-
tions inside the heater are minimal
with respect to the existing scheme.
However, the north-side intercon-
necting process header between
501A and 501B needs to be cut in
order to connect the process outlet
of 501A to the inlet of Reactor 1.
Similarly, the outlet of Reactor 2
needs to be connected to the inlet of
501`B. Also, the outlet of 502, which
goes to Reactor 2 in the existing
scheme, needs to be dismantled and
rerouted to the inlet of 503.
Ultimately, this scheme would
require a lot of external mechanical
modifcations and rerouting of hot
piping, cutting of large-sized process
headers and so on, and thus would
require a long shutdown time.
Option II
In this scheme, the interconnecting
north-side header between 501A
and 501B needs to be blinded/
dismantled and the outlet of 501B,
which goes to the inlet of Reactor
1 in the existing scheme, needs to
be routed to Reactor 3. Similarly,
the outlet of Reactor 2, which goes
to the inlet of 503 in the existing
scheme, needs to be rerouted to
the inlet of 501B, and the outlet of
502, which goes to Reactor 2 in the
existing scheme, needs to be
connected to the inlet of 503. The
interconnecting north-side header
between the inlet and the outlet of
502 needs to be dismantled/
blinded. This option would require
changes inside the heater process
fow pass arrangement (single pass
for 501A and 501B in Option II,
against double pass for 501A and
501B in the existing scheme). This
scheme would require a lot of
external rerouting of heater outlet
and inlet piping to and from the
reactors and thus would direction-
ally require a longer shutdown
time compared with Option I.
Option III
This scheme is similar to the exist-
ing scheme with regard to the
process fow arrangement inside
the heaters. The interconnecting
north-side header between 501A
and 501B need not be cut.
However, the Packinox outlet,
which goes to the inlet of 501A in
the existing scheme, now needs to
be routed to the inlet of 502. The
Reactor 1 outlet, which goes to the
inlet of 502 in the existing scheme,
needs to be routed to the inlet of
501A. Similarly, the outlet of 502,
which goes to the inlet of Reactor
2 in the existing scheme, needs to
be routed to the inlet of Reactor 1.
Apart from external rerouting,
there is no cutting/dismantling of
interconnecting headers and no
change in heater internal pass fow
arrangements, with respect to the
existing scheme, in Option III.
Thus, Option III would direction-
ally require minimum mechanical
modifcation and rerouting
compared with Options I and II,
and consequently would require
the least shutdown time.
A b s d u ty 3 . 4 8 M W
Convection
section
Platformer heater
B-501A
D o u b le -p a ss d e sig n )
A b s d u ty 0 . 1 2 M W
M a x . T M T 4 9 9 C
B rid g e wa ll te m p . 4 7 6 C
Platformer heater
B-502
D o u b le -p a ss d e sig n )
A b s d u ty 9 . 1 M W
M a x . T M T 5 7 5 C
B rid g e wa ll te m p . 8 8 9 C
Platformer heater
B-503
S in g le -p a ss d e sig n )
A b s d u ty 3 . 0 5 M W
M a x . T M T 5 6 0 C
B rid g e wa ll te m p . 8 7 4 C
Platformer heater
B-501B
D o u b le -p a ss d e sig n )
A b s d u ty 2 . 1 1 M W
M a x . T M T 5 3 2 C
B rid g e wa ll te m p . 6 4 9 C
R e c yc le g a s
3 2 . 2 t/h 8 0 t/h
M ix stre a m 8 0 t/h +3 2 . 2 t/h )
Packinox
C-516
Reactor 1
D-501
Reactor 3
D-503
Reactor 2
D-502
4 7 1 C
3 1 B a rg
5 1 5 C
3 0 . 3 3 B a rg
4 7 9 C
2 9 . 0 5 B a rg
4 3 2 C
2 9 . 6 3 B a rg
5 1 0 C
2 9 . 5 5 B a rg
4 9 8 C
3 0 . 4 6 B a rg
5 0 6 C
2 8 . 9 7 B a rg
4 9 7 C
3 0 . 5 9 B a rg
5 0 1 . 4 C
3 0 . 5 9 B a rg
Figure 3 Existing process scheme
technip.indd 2 10/12/12 14:37:21
www.eptq.com PTQ Q1 2013 113
A b s d u ty 2 . 0 7 M W
Convection
section
Platformer heater
B-501A
D o u b le -p a ss d e sig n )
A b s d u ty 3 . 1 M W
M a x . T M T 5 4 0 C
B rid g e wa ll te m p . 7 0 7 C
Platformer heater
B-502
D o u b le -p a ss d e sig n )
A b s d u ty 7 . 1 4 M W
M a x . T M T 5 4 5 C
B rid g e wa ll te m p . 8 1 4 C
Platformer heater
B-503
S in g le -p a ss d e sig n )
A b s d u ty 1 . 9 6 M W
M a x . T M T 5 4 4 C
B rid g e wa ll te m p . 7 6 1 C
Platformer heater
B-501B
D o u b le -p a ss d e sig n )
A b s d u ty 3 . 0 5 M W
M a x . T M T 5 3 1 C
B rid g e wa ll te m p . 6 9 8 C
R e c yc le g a s
3 2 . 2 t/h
8 0 t/h
M ix stre a m 8 0 t/h +3 2 . 2 t/h )
Packinox
C-516
Reactor 1
D-501
Reactor 3
D-503
Reactor 2
D-502
4 7 1 C
3 1 B a rg
5 1 5 C
3 0 . 4 6 B a rg
4 7 9 C
2 9 . 0 5 B a rg
4 3 2 C
2 9 . 7 6 B a rg
4 9 4 . 3 C
2 9 . 6 9 B a rg
5 1 0 C
2 9 . 6 1 B a rg
4 7 9 C
2 9 . 1 1 B a rg
5 0 6 C
2 8 . 9 7 B a rg
4 9 0 . 1 C
3 0 . 5 9 B a rg
4 9 0 . 1 C
3 0 . 5 9 B a rg
Figure 4 Proposed process scheme (Option I)
A b s d u ty 2 . 0 7 M W
Convection
section
Platformer heater
B-501A
S in g le -p a ss d e sig n )
A b s d u ty 3 . 1 1 M W
M a x . T M T 5 5 2 C
B rid g e wa ll te m p . 7 0 7 C
Platformer heater
B-502
D o u b le -p a ss d e sig n )
A b s d u ty 7 . 1 4 M W
M a x . T M T 5 4 5 C
B rid g e wa ll te m p . 8 1 4 C
Platformer heater
B-503
S in g le -p a ss d e sig n )
A b s d u ty 1 . 9 6 M W
M a x . T M T 5 4 4 C
B rid g e wa ll te m p . 7 6 1 C
Platformer heater
B-501B
S in g le -p a ss d e sig n )
A b s d u ty 3 . 0 5 M W
M a x . T M T 5 4 2 C
B rid g e wa ll te m p . 6 9 8 C
R e c yc le g a s
3 2 . 2 t/h
M ix stre a m 8 0 t/h +3 2 . 2 t/h )
Packinox
C-516
Reactor 1
D-501
Reactor 3
D-503
Reactor 2
D-502
4 7 1 C
3 1 B a rg
5 1 5 C
3 0 . 5 7 B a rg
4 7 9 C
2 9 . 0 5 B a rg
4 3 2 C
2 9 . 8 7 B a rg
4 9 4 . 3 C
2 9 . 7 9 B a rg
5 1 0 C
2 9 . 7 1 B a rg
4 7 9 C
2 9 . 2 1 B a rg
5 0 6 C
2 9 . 1 9 B a rg
4 9 0 . 1 C
3 0 . 5 9 B a rg
4 9 0 . 1 C
3 0 . 5 9 B a rg
8 0 t/h
Figure 5 Proposed process scheme (Option II)
technip.indd 3 10/12/12 14:37:32
Maximum tube skin temperature
The maximum tube skin tempera-
ture encountered in Options I and
II is 545C, well within the permis-
sible tube metal temperature of
559.8C. A comparison of maximum
tube metal temperature is shown in
Figure 7 for various proposed proc-
ess schemes compared with that
encountered in the existing process
scheme.
Bridgewall temperature
Heating for Reactor 1: comparing
bridgewall temperature by fuid
number, we observe that, for Fluid
1, a maximum bridgewall tempera-
ture of 707C is encountered in
Options I and II, while Option III
results in a bridgewall temperature
of 652C
Heating for Reactor 2: comparing
bridgewall temperature by fuid
number, we observe that, for Fluid
2, a maximum bridgewall tempera-
ture of 814C is encountered in
Options I and II, while Option III
results in a maximum bridgewall
temperature of 784C
Heating for Reactor 3: comparing
114 PTQ Q1 2013 www.eptq.com
bridgewall temperature by fuid
number, we observe that, for Fluid
3, a maximum bridgewall tempera-
ture of 871C is encountered in
Option III, while Options I and II
result in a bridgewall temperature
of 698C. A comparison of bridge-
wall temperatures is shown in
Figure 8 for the various proposed
process schemes compared with that
encountered in the existing scheme.
Imbalance in bridgewall
temperature
Imbalance in the bridgewall temper-
ature has been observed in the actual
scheme, with hot fue gases at 631C
for 501A to 822C for 503. This, as
such, is not good for intermediate
tube supports in the convection
section and would result in uneven
thermal expansion problems.
The problem of imbalance in the
bridgewall temperature has been
addressed in Options I and II,
where bridgewall temperatures are
almost in the same range, from
707C for 501A to 761C for 503,
thereby reducing uneven stresses
on heater hardware and minimising
fue gas maldistribution and associ-
ated hot fue gas mixing problems.
The problem of imbalance in the
bridgewall temperature has also
been addressed in Option III, where
bridgewall temperatures are almost
in the same range, from 765C for
501A to 871C for 503, thereby
reducing uneven stresses on heater
hardware and minimising fue gas
maldistribution and associated hot
fue gas mixing problems.
Suitability of furnace hardware
The existing case bridgewall
temperature for 501A is 631C. It
has been found that the bridgewall
temperature for 501A is 765C,
which is the maximum for Option
III, against 707C for Options I and
II. However, a bridgewall tempera-
ture of 765C is well below the
permissible limit of 950C. Thus,
the furnace hardware of 501A is
suitable for all options.
The existing case bridgewall
temperature for 501B is 669C. It
has been found that the bridgewall
temperature for 501B is a maximum
784C for Option III, against 698C
A b s d u ty 2 . 2 M W
Convection
section
Platformer heater
B-501A
D o u b le -p a ss d e sig n )
A b s d u ty 4 . 5 5 M W
M a x . T M T 5 0 9 C
B rid g e wa ll te m p . 7 6 5 C
Platformer heater
B-502
D o u b le -p a ss d e sig n )
A b s d u ty 2 . 9 8 M W
M a x . T M T 5 3 6 C
B rid g e wa ll te m p . 6 5 2 C
Platformer heater
B-503
S in g le -p a ss d e sig n )
A b s d u ty 3 . 0 6 M W
M a x . T M T 5 5 9 . 8 C
B rid g e wa ll te m p . 8 7 1 C
Platformer heater
B-501B
D o u b le -p a ss d e sig n )
A b s d u ty 4 . 5 6 M W
M a x . T M T 5 4 7 C
B rid g e wa ll te m p . 7 8 4 C
R e c yc le g a s
3 2 . 2 t/h
M ix stre a m 8 0 t/h +3 2 . 2 t/h )
Packinox
C-516
Reactor 1
D-501
Reactor 3
D-503
Reactor 2
D-502
4 7 1 C
3 1 B a rg
5 1 0 C
2 9 . 5 7 B a rg
4 7 9 C
2 9 . 0 7 B a rg
5 1 5 C
3 0 . 5 3 B a rg
4 7 1 C
2 9 . 7 B a rg
5 0 6 C
2 9 B a rg
4 9 1 . 1 C
3 0 . 5 9 B a rg
8 0 t/h
Figure 6 Proposed process scheme (Option III)
technip.indd 4 10/12/12 14:37:42
Beach Rotana Hotel, Abu Dhabi
March 24-28
2013
Organised by
Officially
Supported by
F
O
R
P
R
O
G
R
A
M
M
E
S
&
R
E
G
I
S
T
R
A
T
I
O
N
p
l
e
a
s
e
v
i
s
i
t
w
w
w
.
s
o
g
a
t
.
o
r
g
www.sogat.org
March 24
Workshops
Process Optimisation (Day 1)
Filtration and Separation (Day 1)
Acid Gas Injection
H2S Safety Detection and Survival Solutions
March 25
Workshops
Process Optimisation (Day 2)
Filtration and Separation (Day 2)
Reducing Liquid Flaring and Oil Wastage
during Sour Well Testing and Intervention
Claus andTail Gas Treatment
March 26-28
9
th
International
SOGAT Conference
March 25-28
SOGAT Exhibition
Official
Media Partner
Shaila DSa Dome Exhibitions PO Box 52641 Abu Dhabi, UAE Tel: +971 2 674 4040 Fax: +971 2 672 1217
C
M
Y
CM
MY
CY
CMY
K
SOGAT2013Ad1 BLEED216x303mm for PTQ.pdf 1 12/6/2012 4:23:00 PM
sogat.indd 1 6/12/12 13:30:14
for Options I and II. However, a
bridgewall temperature of 784C is
well below the permissible limit of
950C. Thus, the furnace hardware
of 501B is suitable for all options.
The existing case bridgewall
temperature for 502 is 893C. It has
been found that the bridgewall
temperature for 502 is a maximum
116 PTQ Q1 2013 www.eptq.com
814C for Options I and II, against
652C for Option III. Option III
brings about relief to the heater 502
by bringing down the bridgewall
temperature to the lowest among
all three options and is well below
the permissible limit of 950C. Thus,
the furnace hardware of 502 is suit-
able for all options.
The existing case bridgewall
temperature for 503 is 822C. It has
been found that the bridgewall
temperature for 503 is a maximum
871C for Option III, against 761C
for Options I and II. However, a
bridgewall temperature of 871C is
well below the permissible limit of
950C. Thus, the furnace hardware
of 503 is suitable for all options.
Burner fexibility and utilisation
It has been found that, for the exist-
ing scheme, the burner for 501A is
operating at a very low load (27.6%
of design load). The total available
facility of 10 burners has not been
utilised and only one burner is put
into operation. Thus, burner fame
stability, fuel gas pressure control,
optimum excess air requirement
and control may be some issues of
concern.
In Options I and II, the burner
needs to be operated at around 62%
of the burner design load, which is
quite good from the point of view
of operations and burner utilisation,
while there is better utilisation of
the 501A burner in Option III,
where the burner load is 96.5% of
design load capacity.
It has been found that for the
existing scheme, the burner for
501B is operating at almost half
load (48.6% of the burner design
load). In all, 9 out of 10 burners are
put to use.
In Options I and II, the burner
needs to be operated at around 61%
of the burner design load, which is
quite good from the point of view
of operations and burner utilisation,
while there is better utilisation of
the 501B burner in Option III,
where the burner load is 98.5% of
design load capacity.
For the existing scheme, burner
502 is operating at around 88% of
design load, which is a good level
of utilisation. All 15 out of 15 burn-
ers are put to use.
In proposed Options I and II, the
burner needs to be operated at
around 64% of the burner design
load, which is quite a good load
from the point of view of opera-
tions and burner utilisation, while
the level of utilisation of the 502
burner in Option III is 23% of
burner design load capacity. Since
5 7 0
6 1 0
5 9 0
5 5 0
5 3 0
5 1 0
4 9 0
4 7 0
E x istin g sc h e m e
F R N C o u tp u t)
F u rn a c e 5 0 3
sin g le p a ss)
F u rn a c e 5 0 1 B
d o u b le p a ss)
F u rn a c e 5 0 1 B
sin g le p a ss)
F u rn a c e 5 0 3
sin g le p a ss)
5 6 0
5 3 1
5 4 2
5 5 9 . 8
P ro p o se d
sc h e m e I
P ro p o se d
sc h e m e I I
P ro p o se d
sc h e m e I I I
T
u
b
e
m
e
t
a
l
t
e
m
p
e
r
a
t
u
r
e
,
C
Case
4 5 0
5 7 0
6 1 0
5 9 0
5 5 0
5 3 0
5 1 0
4 9 0
4 7 0
E x istin g sc h e m e
F R N C o u tp u t)
F u rn a c e 5 0 2
d o u b le p a ss)
F u rn a c e 5 0 1 A /B
d o u b le p a ss)
5 0 2 d o u b le p a ss)
5 0 3 sin g le p a ss)
5 0 2 d o u b le p a ss)
5 0 3 sin g le p a ss)
P ro p o se d
sc h e m e I
P ro p o se d
sc h e m e I I
P ro p o se d
sc h e m e I I I
T
u
b
e
m
e
t
a
l
t
e
m
p
e
r
a
t
u
r
e
,
C
Case
4 5 0
5 7 5
5 4 5 5 4 5
5 4 7
5 7 0
6 1 0
5 9 0
5 5 0
5 3 0
5 1 0
4 9 0
4 7 0
E x istin g sc h e m e
F R N C o u tp u t)
C o n ve c . +5 0 1 A /B
d o u b le p a ss)
5 3 2
5 4 0
5 5 2
5 3 6
C o n ve c . +5 0 1 A
d o u b le p a ss)
C o n ve c . +5 0 1 A
sin g le p a ss)
C o n ve c . +5 0 2
d o u b le p a ss)
P ro p o se d
sc h e m e I
P ro p o se d
sc h e m e I I
P ro p o se d
sc h e m e I I I
T
u
b
e
m
e
t
a
l
t
e
m
p
e
r
a
t
u
r
e
,
C
Case
4 5 0
Figure 7 Maximum tube metal temperatures for three process schemes
technip.indd 5 10/12/12 14:37:53
there are 15 burners in 502, a few
burners may be put out of opera-
tion in such a way that uniformity
of heat distribution is maintained
(maybe by using an alternate burner
as far as possible) and keeping the
heat load of the burner above the
turndown limit.
It has been found that, for the
existing scheme, the burner for 503
is operating at 94.5% of burner
design load. In the existing scheme,
fve out of fve burners are put to
use. In proposed Options I and II,
the burner needs to be operated at
around half load (50% of burner
design load). There is better utilisa-
tion of the 503 burner in Option III,
where the burner load is 87% of
burner design load capacity, which
is quite good from operation and
burner utilisation point of view.
Conclusion
Considering mechanical aspects,
including time required for carrying
out the modifcations, and process
aspects, such as maximum tube
metal temperature, thermal imbal-
ance and burner fexibility,
improvements would be best
addressed by carrying out modifca-
tions to the existing process scheme
recommended in proposed Option
III to achieve the desired process
optimisation within the least time.
Acknowledgement
Special thanks to Mr S Pachal for his
contribution to carry out the study and analysis
of the problem.
Adil Rehman is Principal Engineer, Syngas-
Thermal-Oil and Gas and Special Projects,
Process & Technology Department with
Technip KT India Ltd, New Delhi. He holds a
masters degree in petrochemical engineering
from Aligarh Muslim University, Aligarh, India.
Email: arehman@technip.com
Suman Pachal is Senior Engineer, Syngas-
Thermal-Oil and Gas and Special Projects,
Process and Technology Department with
Technip KT India Ltd. He holds a bachelors
degree in chemical engineering from University
of Calcutta, Kolkata, India.
Email: spachal@technip.com
Shyam K Choudhary is Principal Engineer,
Syngas-Thermal-Oil and Gas and Special
Projects, Process & Technology Department
with Technip KT India Ltd. He holds a bachelors
degree in chemical engineering from BIT, Sindri,
India. Email: skchoudhary@technip.com
Ugrasen Yadav is Deputy General Manager,
Refnery & Petrochemicals, Process &
Technology Department with Technip KT India
Ltd. He holds a masters degree in chemical
engineering from HBTI, Kanpur, India.
Email: uyadav@technip.com
M K E Prasad is Head of the Process and
Technology Department with Technip KT
India Ltd. He holds a BTech degree in chemical
engineering from Osmania University,
Hyderabad, India.
Email: mkeprasad@technip.com
www.eptq.com PTQ Q1 2013 117
7 0 0
8 0 0
7 5 0
9 0 0
8 5 0
6 5 0
6 0 0
5 5 0
5 0 0
4 5 0
E x istin g sc h e m e
F R N C o u tp u t)
7 6 1 7 6 1
P ro p o se d
sc h e m e I
P ro p o se d
sc h e m e I I
P ro p o se d
sc h e m e I I I
B
r
i
d
g
e
w
a
l
l
t
e
m
p
e
r
a
t
u
r
e
,
C
Case
4 0 0
F u rn a c e 5 0 2
d o u b le p a ss)
F u rn a c e 5 0 1 A /B
d o u b le p a ss)
5 0 2 d o u b le p a ss)
5 0 3 sin g le p a ss)
5 0 2 d o u b le p a ss)
5 0 3 sin g le p a ss)
8 8 9
7 8 4
7 0 0
8 0 0
7 5 0
9 0 0
8 5 0
6 5 0
6 0 0
5 5 0
5 0 0
4 5 0
E x istin g sc h e m e
F R N C o u tp u t)
C o n ve c . +5 0 1 A /B
d o u b le p a ss)
6 4 9
7 0 7 7 0 7
6 5 2
C o n ve c . +5 0 1 A
d o u b le p a ss)
C o n ve c . +5 0 1 A
sin g le p a ss)
C o n ve c . +5 0 2
d o u b le p a ss)
P ro p o se d
sc h e m e I
P ro p o se d
sc h e m e I I
P ro p o se d
sc h e m e I I I
B
r
i
d
g
e
w
a
l
l
t
e
m
p
e
r
a
t
u
r
e
,
C
Case
4 0 0
7 0 0
8 0 0
7 5 0
9 0 0
8 5 0
6 5 0
6 0 0
5 5 0
5 0 0
4 5 0
E x istin g sc h e m e
F R N C o u tp u t)
6 9 8 6 9 8
P ro p o se d
sc h e m e I
P ro p o se d
sc h e m e I I
P ro p o se d
sc h e m e I I I
B
r
i
d
g
e
w
a
l
l
t
e
m
p
e
r
a
t
u
r
e
,
C
Case
4 0 0
8 7 4 8 7 1
F u rn a c e 5 0 3
sin g le p a ss)
F u rn a c e 5 0 1 B
d o u b le p a ss)
F u rn a c e 5 0 1 B
sin g le p a ss)
F u rn a c e 5 0 3
sin g le p a ss)
Figure 8 Comparison of bridge wall temperatures for three process schemes
technip.indd 6 12/12/12 11:01:05
www.wraconferences.com/europeanfuels | +44 (0) 20 7384 7700 | Hannah.Wharrier@clarionevents.com
5 -7 March 2013
NH Grand Hotel Krasnapolsky
Amsterdam, T e Netherlands
www.wraconferences.com/europeanfuels
Find out how to capitalize on
the emerging market for biofuels
for transportation with case studies
from Shell and the EFOA
Engage with end-users from
across the transport fuels industry,
including Boeing, KLM, Daimler
and Renault
Gain insight into how the whole
fuels ecosystem is evolving
Hear how partnerships and
new operational efciencies
can help you with renery survival
and efciency focus sessions
Network with 200 senior
executives from across the
European fuels value chain
and benchmark your
operating strategies
Malcolm McDowell,
Head of Unit, DG Energy,
European Commission
Chris Beddoes,
Secretary General,
European Petroleum
Industry Association
Marco Schiavetti,
Director - Supply & Trading,
Saras
Atanas Ivanov,
Deputy Chief
Process Engineer,
Lukoil
Arthur Reijnhart,
General Manager
- Alternative Energy Fuel
Development Strategy, Shell
Aksel Skjervheim,
Head of Fuel Markets,
Gasnor
Ignaas Caryn,
Director Innovation and
Corporate Venturing and
Biofuels Strategy, KLM
Marek Herra, Production
Director, Grupa Lotos
3DAY EVENT
5-7 March 2013
GALA DINNER
5 March
Excellent forum:
engaged, competent
and honest participants
contributed to positioning
some of the bricks of this
very complex puzzle.
Ron Timmers,
Head of Industry Segment,
Linde AG
Very informative conference,
will denitely recommend it to
my colleagues.
Faisa Mohammed
Al Haddad,
Senior Analyst
- Projects & Special Studies,
Kuwait Petroleum
Corporation
Sponsored by: Media Partners: Supporting Associations:
THE EUROPEAN
FUELS CONFERENCE
14
th
Annual Meeting
UNITING THE EUROPEAN REFINING AND TRANSPORTATION
INDUSTRY IN NEW FUEL SOLUTIONS
KEY SPEAKERS:
KEY REASONS TO ATTEND:
efc.indd 1 10/12/12 16:42:25
Improved hydrogen yield in catalytic
reforming
T
he advanced fuel technol-
ogy is a patented invention
(US Patent 6207040, European
Patent 0914405) and states that the
higher-boiling C
6
hydrocarbons,
including naphthenes, benzene and
hexane, are more proftably sent to
the light tops isomerisation process
rather than to catalytic reforming.
The critical full-range naphtha-split-
ting mode to be constructed in the
refnery (see Figure 1) requires
negligible investment.
The economic gains of advanced
fuel technology
1
include:
A gain of more than 4% on the
recombined isomerate-reformate
blend volume yield per 100 tonnes
of isomerisation-reforming feed.
2
(This is mainly explained by the
good C
6
isomerate octane and, in
particular, by the even better C
6
blending isomerate octane, plus the
C
6
isomerate weight yield being
much higher and the C
6
isomerate
density being much lower than the
reformate weight yield and
density.)
A gain of around three octane
number points on the recombined
isomerate-reformate blend (result-
ing from 5-6 more points of the
reformate RON
1,3
acquired thanks
to the better reformability quality of
the reforming feed), exchangeable
for an additional gasoline yield.
These results are obtained by
managing the full-range naphtha
splitting unit in such a way as to
determine the two ranges of concen-
trations (0-4 vol% C
7
hydrocarbons
in the tops, 0-0.5 vol% C
6
hydrocar-
bons in the bottom, see Figure 1),
which defne the splitting
mode and are the necessary and
suffcient condition for performing
A process step that sends higher-boiling C
6
hydrocarbons to light tops isomerisation
delivers an increase in hydrogen production from naphtha catalytic reforming
ROBERTO AMADEI
Chemical & Energy Development
the advanced fuel technologys
functions.
Effcient separation by distillation
in the full-range naphtha splitting
unit can be achieved according to
different strategies, depending on
choices the refner has to make
between highly effcient equipment,
which is more capital intensive, and
highly effcient operation, which
consumes more energy. The devel-
opment in no way limits freedom
of choice in a compromise between
these options.
In practice, separation effciency
can be improved by modifying the
equipment (for instance, number
and type of distillation stages, inter-
nals type) and/or the refux ratio.
There is also the option to deter-
mine an optimum compromise
between improvements in equip-
ment and refux ratio.
Hydrogen
A further important feature of
advanced fuel technology is the
gain it offers in catalytic reforming
hydrogen production/availability.
The gain in production of catalytic
reforming hydrogen (net of
increased use in isomerisation) is
estimated in the range of 28% to
over 48% based on conventional
hydrogen production in catalytic
reforming. When removing all or
nearly all of the C
6
molecules (shift-
ing them to the isomerisation
process) from the catalytic reform-
ing feed, it has been demonstrated
in the refnery that the yield of cata-
lytic reforming hydrogen increases
signifcantly. This is in agreement
with the theory that the reforming
feed quality, after C
6
removal,
becomes much more favourable to
reformings dehydrogenation reac-
tions with hydrogen production
and much less favourable to reform-
ings hydrocracking reactions with
hydrogen consumption.
We will frst analyse the basis of
the technology before examining
the results of the refnery runs.
Reformate octane number
First, we can consider the aromati-
sation catalytic reforming reactions:
1 naphthene 1 aromatic + 3 H
2
- 48-55 Mcal/
kmol (million calories per thousand moles) (1)
1 paraffn 1 aromatic + 4 H
2
- 60-65 Mcal/
kmol (2)
It is well known that the delta
octane number [C
8
+ aromatics - C
8
+
(naphthenes + paraffns)] is much
higher (around four times) than for
[C
6
aromatics - C
6
naphthenes]. C
6
paraffns cannot be considered
because the C
6
paraffns that pass
through catalytic reforming do not,
for practical purposes, increase their
octane number. In fact, C
6
paraffns
either crack or pass through uncon-
verted (although an exception is
www.eptq.com PTQ Q1 2013 119
Isomerisation
unit
Catalytic
reforming
unit
N a p h th a
L ig h t to p s
0 -4 vo l% C
7
)
H e a vy n a p h th a
0 -0 . 5 vo l% C
6
)
Splitting
unit
Figure 1 The advanced fuel fow diagram
ced.indd 1 11/12/12 14:58:40
120 PTQ Q1 2013 www.eptq.com
made for a 10% maximum quota of
normal hexane with a RON value
of about 26).
The heat absorbed in catalytic
reforming by a one-molecule
aromatisation reaction, generating
one aromatic molecule, is roughly
speaking about the same, irrespec-
tive of the number of carbon atoms.
This means that, with about the
same heat absorption, C
8
+ has a
delta octane from the aromatisation
of one molecule about four times
higher than for a C
6
molecule.
Regarding one molecule, with
approximately the same heat
absorption, C
8
+ has a delta octane
four times higher than C
6
s delta
octane for a gasoline quantity on
average about 40% higher (because
the relevant gasoline weight is
exactly proportional to the molecu-
lar weight, while the relevant
gasoline volume is approximately
proportional).
The hydrocarbons C
6
, C
7
, C
8
, C
9
and so on compete with one
another in order to make use of the
available heat. When C
6
is present,
it provides a modest octane increase
by making use of a portion of the
heat subtracted from the availabil-
ity of C
7
+. When C
6
is present, in
order to increase the available heat,
the only option is to increase the
catalytic bed temperature.
When C
6
is removed, the refor-
mate octane increases automatically
due to the strong octane upgrading
available for C
7
+, which is simply a
result of the absence of C
6
. The
presence of C
6
appears to be very
harmful to catalytic reforming
performance.
The highest boiling hydrocarbon
among all of the C
6
s is cyclohexane,
which in reforming has a fair
conversion rate (although lower
than the C
7
+ naphthenes), but a
relatively modest potential for
octane increase. As a consequence,
it is important to transfer as much
C
6
as possible from the reforming
feed to the isomerisation feed. The
more C
6
transferred, the better the
reforming performance and the
refnery economics.
The theory applies to any crude
oil and to any refnery confgura-
tion. This theory is in full agreement
with refnery experimental results.
1
The frst conclusion is that the
removal of C
6
hydrocarbons
increases signifcantly the reformate
octane number process variables
being equal and the catalytic bed
temperature, and consequently the
catalyst life, being equal. In order
to keep the reformate octane
number constant, it is necessary to
decrease signifcantly the catalytic
bed temperature.
Catalytic reforming hydrogen
production
Catalytic bed temperature decrease
When C
6
is removed from the cata-
lytic reforming process, the
reformate octane number increases
signifcantly, which means that a
large decrease in the catalytic bed
temperature is necessary to obtain a
determined reformate octane
number.
We will now discuss hydrocrack-
ing reactions on the premise that
the categories of hydrocracking
reactions include the dealkylation
reactions. Hydrocracking reactions
are rather slow and very sensitive
to temperature, hydrocracking rates
decreasing rapidly as temperature
decreases. Thus, as soon as the C
6
s
are removed from the catalytic
reforming process, hydrocracking
decreases sharply. As a result, the
following main consequences arise,
all favouring a net increase in
hydrogen production during cata-
lytic reforming:
A quantity of hydrogen molecules
is not consumed and is therefore
saved, the proportion of saved
hydrogen molecules being one
molecule for each hydrocarbon
molecule not cracked. In other
words, catalytic reforming hydrogen
production increases by one mole-
cule for each molecule not cracked
The decrease in the contribution
of paraffns hydrocracking to refor-
mate octane causes a compensating
increase in aromatisation, implying
a further increase in hydrogen
production
The decrease in paraffns and
naphthenes hydrocracking favours
dehydrogenation reactions and
produces hydrogen
By means of dealkylation, the
heavier aromatics are converted
into light aromatics such as toluene
pile caps to spread the ground-
bearing loads. The super lift crane
was scheduled to perform numer-
ous super lifts during a 38-day
outage period. It would be lifting a
total weight of more than 9.4
million lb (4250 tonnes).
Due to the risks associated with
lifts, let alone the risks associated
with super lifts, a detailed manual
was produced and a third-party
company was employed to help
provide a cold set of eyes not only
to the lift plans but also to the erec-
tion and disassembly of the crane.
Additionally, each crane foundation
was tested to 125% of expected load
and checked for settlement.
Other yard activities included:
Assembly of regenerator cyclones
and top head
Assembly of reactor cyclones and
top head
Final assembly of middle section
of regenerator
Final assembly of lower section of
regenerator with upper combustor
riser
Final assembly of lower reactor
and stripper vessel
Final assembly of lower section of
combustor riser
Final assembly of regenerator
internal riser.
Replacement of the FCC unit
cooler bundle had presented an
initial problem. But that was solved
when the development of a new
method to replace the catalyst
cooler bundle was added to the
work scope. The original plan was
to use a 1000-tonne crane ftted
with nearly 400ft (122m) of boom
due to the height and weight of the
bundle. But now that the unit had a
new confguration, access was
prohibited. To solve the problem,
the project team developed an inno-
vative approach for bundle
replacement utilising two 100-tonne
chain hoists. The new procedure
was both safer and more cost effec-
tive than the previously scheduled
method.
The project faced a delay in the
delivery of a large expansion joint
assembly that would have been
held up due to Canadian travel
restrictions. The team opted to
complete the expansion joints on
site, which allowed the three
smaller sections to be shipped
instead of the overall assembly that
weighed approximately 150 000 lb
(68 tonnes) and was 15ft (4.5m)
wide, 15ft (4.5m) tall and over 40ft
(12.2m) long. Further complicating
the fabrication were the tolerances
on the joint that included a 7-inch
cold offset and bellow welding that
had to be completed under feld
conditions.
Many areas were completed as
far as possible during the pre-
turnaround construction phase. The
project was completed within
schedule, on budget and with a
safety record of zero recordables.
Over $3 billion in upgrades have
made the plant one of the most eff-
cient in the US. Valero St Charles is
now producing a variety of quality
fuels, making extra energy available
to the area grid and employment to
many workers.
Some of the projects fnal statis-
tics included:
Over 200 000 man hours no
recordables
Under 0.5% weld reject rate
AltairStricklands portion of the
turnaround was completed in 46
scheduled days
4.5% under budget
Over 20 million lb (9000 tonnes)
of steel removed/replaced
Over 33 super lifts during TAR
with Deep Souths 36000
Over 40 major lifts during TAR
with CC2800
Over 15 major lifts during TAR
with 800T
Over 20 major lifts during TAR
with CC2400
Over 150 critical lifts (over
50 000 lb [23-tonne] lifts) overall.
whitney Strick|and ls vl ChleI MurkeLlng
OIIcer wlLh AlLulrSLrlckund und
AlLulrSLrlcklund Holdlngs, LLC. He munuges
boLh domesLlc und lnLernuLlonul sules Ior
AlLulrSLrlcklund und domesLlc sules Ior
AlLulrSLrlcklund Holdlngs. He uLLended
Lhe UnlverslLy oI 1exus uL AusLln und klce
UnlverslLy ln HousLon, 1exus, und ls MCSL
quullIed.
/ma||. wstt|c||aaJa|ta|tstt|c||aaJ.com
Chad 8eckemeyer ls currenLly u lro|ecL
Munuger Ior AlLulrSLrlcklund und hus been
wlLh Lhe compuny slnce 1998.
www.eptq.com kVAMS 2012 25
assess achievement of emission
reductions required under the
uated Linde in regard to how
accurately its products measured
, nitrogen oxide and carbon
dioxide. Lindes US production site,
at Alpha New Jersey, passed the
audit with zero failures and, in
2011, the company was granted
continued approval to produce EPA
protocol gas standards both at this
site in the US and at an additional
As in other parts of the world, the
European Union is introducing
sions from light road vehicles,
gen particulates and oxides. Since
2009, it has been mandatory to have
ultra-low-sulphur petrol and diesel;
that is, fuels containing less than 10
Analysis of sulphur compounds in
fuels has therefore not only become
a critical requirement, but so has the
need to measure for lower and lower
levels of sulphur compounds. It is
now vital to be able to detect
extremely low levels of sulphur,
down to parts per billion, and a
range of instrumentation techniques
and detectors are required to meet
cal methods used for the
determination of sulphur species in
fuel samples is GC separation
Heavy crude 0il
Atmospheric 0istillation
Vacuum 0istillation
0oker & Visbreaker Feed
Fluidized 0atalytic 0racker
Bitumen
Non-nlrusve low
measuremenl up lo /000
FeLd-Proven al Relneres
Trouble Iree operation at
extreme pipe temperatures
No clogging, no pressure losses
nstallation and maintenance
without process interruption
ndependent oI uid or pressure
Hazardous area approved