You are on page 1of 60

gas

2014
PTQ supplement
gas cov copy 6.indd 1 06/03/2014 13:57
1
2
1
5
_
e
www.engineering-solutions.airliquide.com
Because our genes lead us to an irrepressible passion for new solutions.
The only thing longer than our company history
is the list of our innovations.
When it comes to providing innovative
plant and system solutions, Air Liquide
Global E&C Solutions is your partner of
choice – worldwide. No surprise, since
we innovate relentlessly in order to bring
sustainable and cost-effective solutions
to society, levera ging partnerships with
customers, suppliers, academics and
communities. This is not only demonstrated
by thousands of patents, but by pioneering
reference projects around the globe as
well. Yours could be the next one.
AIR2362022_AZ_Techno-DNA_7-8-2013_RZ.indd 1 21.05.13 11:26 gas air liquide.indd 1 04/03/2014 10:44
©2014. The entire content of this publication is protected by copyright full details of which are available from the publishers. All rights
reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means –
electronic, mechanical, photocopying, recording or otherwise – without the prior permission of the copyright owner.
The opinions and views expressed by the authors in this publication are not necessarily those of the editor or publisher and while every
care has been taken in the preparation of all material included in Petroleum Technology Quarterly and its supplements the publisher
cannot be held responsible for any statements, opinions or views or for any inaccuracies.
3 China’s sour gas
Chris Cunningham
5 Zero emissions in sulphur recovery
Michele Colozzi and Simona Cortese
KT-Kinetics Technology
Lucia Barbato
Processi Innovati

15 Eliminating the Claus furnace
Matt Thundyil, Sameer Pallavkar, Ramiro Vazquez and David Seeger
GTC Technology US, LLC
31 Zeolite based dryers in ethylene plants
Vassilios Zafirakis and Hans Hoefer
Grace Materials Technologies

49 Virtual commissioning of a gas handling system
Rainer Scheuring
Cologne University of Applied Sciences
Hans-Christian Haarmann-Kühn and Jürgen Essler
TGE Marine Gas Engineering
Torsten Felix
Flemming Automationstechnik
Michael Brodkorb
Honeywell Process Solutions

54 Protecting compressors with dynamic simulation
Nicholas Brownrigg
AspenTech

Woodside Petroleum’s Pluto LNG project in Western Australia. Photo: Woodside Petroleum
2014
www.eptq.com
gas
p
t
q
Y L R E T R A U Q Y G O L O N H C E T M U E L O R T E P
ed com copy.indd 1 12/03/2014 12:43
Lnglneerlng Soluuons for Cll, Cas, 8eñnlng & Chemlcal lndusLrles.
sales[bre.com www.bre.com 979-776-3220 uS 800-776-3220
ProMax is a powerful and versatile process simulation
package that is used worldwide to design and optimize
gas processing, refning and chemical facilities. Totally
integrated with Microsoft Visio
®
, Excel
®
and Word
®
,
ProMax is the best simulation tool for designing gas
plants and predicting performance under varying
process conditions.
Discover why engineers around the world depend on ProMax!
- Mode| any NGL, LNG, LÞG recovery and fracnonanon process
- kate heat exchangers |nc|ud|ng brazed a|um|num
- Þred|ct opnma| exchanger performance w|th acnve ranng dur|ng s|mu|anon
- Determ|ne hydrate formanon and CC
2
freeze-out temperatures us|ng best-|n-|ndustry pred|cnons
- S|ze two and three phase separators
- Use Lxce| to perform parametr|c stud|es and econom|c ana|yses
- Ca|cu|ate hydrate |nh|b|tor requ|rements us|ng unsurpassed MeCn]n
2
C]nC pred|cnons
- S|mu|ate an ennre process fac|||ty us|ng mu|np|e ßowsheet des|gn
- Mode| am|ne treanng, dehydranon, cryogen|c, and fracnonanon fac|||nes |n the same pro[ect
Process Focus: Gas Processing
ProMax
®
with TSWEET
®
& PROSIM
®
Process Simulation Software
bre.indd 1 8/3/12 10:09:31
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 driv-
ers 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 primar-
ily fatty acid methyl esters (FAME),
usually created by transesterifcation
between fats and methanol. Currently,
biodiesel is produced from various
vegetable and plant oils. First-genera-
tion 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 production in the US and
Europe in quantities that can produce
enough biodiesel to be used in a
commercial market with currently
applicable technologies.
First-generation feedstocks for
Gas 2014 3
Editor
Chris Cunningham
editor@petroleumtechnology.com
Production Editor
Rachel Storry
production@petroleumtechnology.com
Graphics Editor
Rob Fris
graphics@petroleumtechnology.com
Editorial
tel +44 844 5888 773
fax +44 844 5888 667
Business Development Director
Paul Mason
sales@petroleumtechnology.com

Advertising Sales
Bob Aldridge
sales@petroleumtechnology.com
Advertising Sales Office
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
ISSN 1362-363X

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.
China’s sour gas
C
hina has set about its plan to step up consumption of gas purposefully,
in a bid to cut levels of air pollution. For the time being, that means
accelerating imports as well as raising gas prices to encourage domestic
development.
The PRC’s gas imports are sourced as liquefed natural gas, chiefy from
Africa, and pipeline gas from central Asia. At the beginning of March, the
China Development Bank approved a $700 million loan to complete a section
of the Kazakhstan-China main gas pipeline, itself part of a vast network of
gas-carrying pipelines linking the southern former Soviet republics and
western China.
At the same time, domestic energy majors Sinopec and CNPC said that
they were likely to meet a state annual target of 6.5 billion m
3
of shale gas in
2015, rising to as much as 100 billion m
3
by the end of the decade. These
fgures compare with total production in 2013 of 200 million m
3
, so there is
some reliance on unproven reserves of shale gas.
China also has vast conventional reserves of natural gas, in particular in
the south western province of Sichuan and in Xinjiang to the west. So why is
domestic gas not already piped into the nation’s industrial centres rather
than from central Asia or via LNG terminals? For one thing, China’s onshore
gas reserves are classed as ‘very sour’ with a hydrogen sulphide content of,
typically, 15%. That is a high but not prohibitive barrier to gas processing.
Crucially, these are also high pressure felds. An accidental release of
untreated gas has the potential for disaster, as proved to be the case over a
decade ago when a blow-out poisoned the city of Chongqing.
Nonetheless, Sinopec has operated a safe production and processing site in
the region since 2010. The next big development is the Chuandongbei sour
gas project, run by Chevron and CNPC under a production sharing contract.
The project was originally due onstream in 2010, but has been beset by the
technical diffculties of recovering high-pressure sour gas and severe cost
over-runs. Chuandongbei is now expected to start up in 2015 and is slated to
reach daily production of over 500 000 cubic feet of pipeline quality gas.
Chevron is the only international oil major operating in China’s gas felds,
in conventional onshore felds, in shale deposits, and offshore. The compa-
ny’s total production of gas in China is reported at 20 000 b/d oil equivalent
in 2013. There is far to go for China’s natural gas industry.
Should China move ahead more purposefully with its sour gas processing,
there is a potential down-side for refnery costs worldwide. Throughout the
current century, China has been the world’s largest importer of sulphur. The
PRC’s fertilizer companies convert this to sulphuric acid for the manufacture
of ammonium phosphate which has transformed the nation’s level of self-
reliance in the production of crops.
Most of this sulphur for China originates in Canada’s sour gas processing
industry, as well as the Middle East’s oil refneries and gas production sites,
Kazakhstan’s Caspian developments and US West Coast refneries. It is a
market that provides a welcome outlet for — from the refners’ perspective
— a largely unwanted, low-value byproduct. The eventual disappearance of
a China market for sulphur will add to some refners’ disposal problems.
CHRIS CUNNINGHAM
p
t
q
Y L R E T R A U Q Y G O L O N H C E T M U E L O R T E P
ed com copy.indd 2 06/03/2014 12:16
Visit our webpage to learn more about all our products > www.jacobs.com/comprimo
Comprimo® Sulfur Solutions
Total Project Solutions
We are one of the world’s largest and most diverse providers of
professional technical services. Our global network includes more
than 200 ofces in more than 25 countries
From our gas treating & sulfur technology centers of expertise in
Leiden, The Netherlands and Calgary, Canada, we provide engineering,
procurement and construction services to our clients around the world
Gas Treating & Sulfur Technology
With more than 500 units built during the
last 40 years, we are a global leader in the
gas treating and sulfur recovery technology
gas jacobs.indd 1 04/03/2014 10:46
Zero emissions in sulphur recovery
E
nvironmental regulations
around the world nowadays
impose very strict SO
2
emis-
sions requirements. H
2
S faring is
no longer tolerated in industrial
complexes; best available technique
(BAT) has to be utilised to satisfy
such new requirements. The
achievement of a high level of relia-
bility and availability is also
mandatory for sulphur recovery
units (SRU) installed in fully inte-
grated industrial complexes where
production must be maximised.
The main targets of the SRU to be
installed in a new industrial
complex can be summarised as
follows:
• SO
2
emissions no higher than 150
mg/Nm
3
(50 ppmv)
• Ensure complete destruction of
impurities contained in sour gas
A process under development for sulphur recovery features an innovative
catalyst and a new scheme for the treatment of all sour gas feedstocks
MICHELE COLOZZI and SIMONA CORTESE KT-Kinetics Technology
LUCIA BARBATO Processi Innovati
feedstocks (NH
3
, HC, COS, CS
2
,
RSH, HCN, and so on)
• Maximum energy recovery
• Maximum availability and
reliability
• Minimum opex and capex.
The Modifed Claus section can
achieve a sulphur recovery eff-
ciency (SRE) of 92-96% when
provided with two reactors, and
94-98% when provided with three
reactors. Therefore, the addition of
a tail gas treatment (TGT) section is
always required to achieve a SRE
higher than 99.9%.
In the 1990s, Kinetics Technology
SpA (KT) introduced the proprie-
tary RAR (reduction, absorption,
recycle) technology which is recog-
nised as BAT and is specifcally
developed to minimise SO
2
emis-
sions with the aim of being in full
compliance with the most stringent
environmental regulations in force.
The RAR process is a reductive
amine based tail gas treatment
process, capable of achieving the
maximum mandated sulphur
recovery effciency (SRE>99.9+%).
A block diagram of the RAR
process is shown in Figure 1.
In the proprietary KT tail gas
treatment section, the Claus tail gas
is mixed with hydrogen rich gas
from battery limits, when available,
or it can be produced in a dedi-
cated reducing gas generator
(in-line heater). Tail gas is then sent
to the hydrogenation reactor, where
all components containing sulphur
are reduced and/or are hydrolysed
to H
2
S.
Tail gas is then sent to a quench
tower where, through direct contact
www.eptq.com Gas 2014 5
Sour water stripper
acid gas
Hydrogen make-up
Amine acid gas
Reduction
Absorption Recycle
Modified Claus
section
Liquid sulphur
degassing
section
Heating/cooling
section
Amine
regeneration
section
Incineration
section
Fuel gas
Flue gas
Sour
water
Liquid
sulphur
Claus
tail gas
Lean
amine
Rich amine
Acid gas
recycle
Super
enrichment
Off gas
recycle
TGT tail gas
Sweet tail gas
Vent gas
recycle
Undegassed
liquid sulphur
Figure 1 Block diagram of the RAR process
gas kt.indd 1 06/03/2014 12:21
6 Gas 2014 www.eptq.com
RAR process is shown in Figure 2.

Integrated sour gas selective
oxidative autothermal process
Environmental regulations will
become more and more stringent in
the near future, requiring SO
2
emis-
sions to the atmosphere to be much
lower than 150 mg/Nm
3
(lower
than 50 ppmv). Therefore, new
process schemes will be required in
order to be in compliance with the
strictest future regulations. For this
with circulating water, steam is
condensed and the gas is cooled
down. The H
2
S contained in the tail
gas is removed by the use of an
amine solution in the TGT absorber.
Sweet tail gas from the top of the
TGT absorber, containing a few
parts per million of H
2
S, is sent to
the thermal incinerator to be
converted to SO
2
before the fue
gases are released to the atmos-
phere. The amine solution is
normally regenerated in a dedicated
amine regeneration section,
included in the sulphur recovery
unit. The produced liquid sulphur
is treated in the liquid sulphur
degassing section where hydrogen
polysulphides and H
2
S are
completely removed. In order to
minimise SO
2
emissions to the
atmosphere, all of the gases released
during liquid sulphur degassing are
recycled back to the Claus section
and the TGT section. A typical
process fow scheme of the
SWS
acid gas
Acid gas
CA
Thermal
reactor
Sulphur condensers
Claus
stage I
RAR
WHB
Thermal
incinerator
WHB
Stack
Claus
stage II
Sulphur
Rich amine
Flue gas
H
2
CA
Fuel
gas
Acid gas recycle
Figure 2 Typical process flow scheme of the RAR process
Sour water stripper
acid gas
O
2
Reduction
Absorption Recycle
H
2
S cracking
section
(SOAP section)
Liquid sulphur
degassing
section
Heating/cooling
section
Amine
regeneration
section
Incineration
section
(Start-up only)
Fuel gas
Flue gas
to ATM
Hydrogen
rich gas
Sour
water
Liquid
sulphur
Claus
tail gas
NNF (for start-up only)
Lean
amine
Rich amine
Acid gas
recycle
Super
enrichment
Off gas
recycle
TGT tail gas
Vent gas
recycle
Undegassed
liquid sulphur
Amine acid gas
Figure 3 Block diagram of the new concept SRU
gas kt.indd 2 06/03/2014 12:21
Gas Processing Plant
Western USA
A World of Solutions
Visit www.CBI.com
SHALE GAS
PROCESSING
SOLUTIONS
Complete Capabilities Addressing
the Entire Natural Gas Value Chain
For more than 40 years, CB&I has excelled at creating
innovative solutions for the gas processing industry,
employing a deep portfolio of execution capabilities
and processing technologies. Drawing upon our
full-service technology plus EPFC expertise, CB&I has
installed more than 600 gas processing related projects.
From concept through commissioning across the entire
natural gas value chain, we deliver. Contact CB&I to
learn more about our modular gas processing plants
and superior gas technologies and how we can improve
your next gas processing project.
FIELD DEVELOPMENT
GAS CONDITIONING AND TREATING
SULFUR REMOVAL AND RECOVERY
FRACTIONATION
NITROGEN REJECTION
DEEP ETHANE AND LPG RECOVERY
LIQUID PRODUCT TREATING
UPGRADING OF LOW VALUE GAS
MODULARIZATION
cbi_ptq_gas_suppl_ad_apr_2014.indd 1 3/6/2014 1:37:20 PM
cbi.indd 1 07/03/2014 09:51
reason, KT is developing an inno-
vative process and a new concept
for the SRU.
The process scheme under devel-
opment is divided into two main
sections:
• H
2
S cracking section (sour gas
SOAP section)
• Tail gas treatment section (amine
based).
The core part of the innovative
SRU is the Sour Gas Selective &
Oxidative Autothermal Process
(Sour Gas SOAP) section which is
fully integrated with the RAR tail
gas treatment plant’s confguration
and process.
A block diagram of the novel
process is shown in Figure 3.
The new concept SRU developed
by KT is a simplifed scheme
equipped only with a catalyst-flled
reaction furnace package (catalytic
reaction furnace or CRF) followed
by one sulphur condenser, a liquid
sulphur degassing section and a
TGT section.
What is the advantage of such
a configuration?
The reason for such a minimal
confguration is strictly related to
the novel idea of operating the
Claus section in a different way:
• Use of a proprietary catalyst in
the reaction furnace to minimise
SO
2
production
8 Gas 2014 www.eptq.com
• Use of a specifc H
2
S/O
2
ratio in
order to maximise production of
hydrogen while minimising the
production of SO
2
• Use of oxygen instead of ambient
air.
The peculiarity of the process is
the possibility to install a catalyst
block directly in the reaction
furnace, in order to enhance the
selective conversion of H
2
S directly
to H
2
, S
x
and H
2
O instead of S
x
, SO
2

and H
2
O.
An interesting advantage of
the new concept SRU is the possi-
bility to achieve important capex
reductions, minimising the number
of items of equipment to
be installed in the plant (see
Figure 4).
Additionally opex can be reduced
with the process since a thermal
incinerator is virtually no longer
required in normal operations, the
hydrogen rich gas produced can be
utilised in a petroleum refnery,
and SO
2
and CO
2
emissions are
drastically reduced. In natural gas
production, where hydrogen users
are not available, the operating
parameters may be adjusted to
minimise production of hydrogen
rich gas.
Reaction mechanism
H
2
S cracking through partial oxida-
tion in the CRF using a proprietary
catalyst is based on the following
reactions:
H
2
S + 3/2 O
2
→ H
2
O + SO
2
(oxidation reaction) (1)
H
2
S + 1/2 O
2
→ 1/2 S
2
+ H
2
O
(partial oxidation reaction) (2)

2H
2
S + SO
2
↔ 2H
2
O + 3/n Sn
(Claus reaction) (3)
H
2
S ↔ H
2
+ 1/2 S
2
(cracking reaction) (4)
KT is developing the proprietary
catalyst in collaboration with the
University of Salerno (UNISA) in
Italy. The laboratory plant to test
the proprietary catalyst and the
novel process is located at the
University of Salerno. KT has
selected UNISA because it has the
best facilities in Europe on an
academic level as well as the high-
est competence in the feld of sour
gas heterogeneous catalysis. More
than 4000 hours of tests in the labo-
ratory plant at UNISA have been
performed. For feedback of the
results of laboratory tests, KT has
developed a thermodynamic and
kinetic model which fts with high
accuracy the results of the labora-
tory plant. The model utilises more
than 100 reactions and in this way
it approaches actual system
behaviour.
SWS
acid gas
Acid gas
CA
Catalytic
reactor
furnace
Sulphur condensers
Claus
stage I
RAR
H
2
rich gas
WHB
Thermal
incinerator
WHB
Stack
Claus
stage II
Sulphur
Acid gas recycle
Rich amine
Flue gas
H
2
CA
Fuel
gas
Figure 4 Process flow scheme for the new concept SRU
gas kt.indd 3 11/03/2014 17:01
www.eptq.com Gas 2014 9
H
2
; CO and H
2
S are the final prod-
ucts of reaction.
KT has evidence of such behav-
iour in all of the laboratory tests
performed.
New concept SRU for industrial
applications
Based on knowledge acquired from
the laboratory tests, KT has
and CS
2
in the gas phase at the
outlet of the catalytic reaction
furnace. The COS and CS
2
destruc-
tion mechanism is based on the
adsorption of hydrocarbons and
H
2
S over the catalyst surface. The
hydrocarbons and H
2
S then react
over the catalyst producing CS
2
.
The CS
2
produced reacts with SO
2
producing COS which reacts with
Figures 5 to 7 show the main
mechanisms of reaction involved in
the SOAP Process in the presence
of the catalyst.
H
2
S reaction mechanism
The heat for cracking reaction (4) is
made available by the oxidation
reaction (1 & 2) of H
2
S. The oxygen
provided to the system is only suffi-
cient to achieve the proper reaction
temperature to activate all the rele-
vant reactions over the catalyst.
A part of the H
2
S is oxidised to
SO
2
and H
2
O (1) and another part of
H
2
S is partially oxidised to S
x
and
H
2
O (2) producing the heat
required for the other reactions. At
the same time, part of the H
2
S
adsorbed over the catalyst reacts
with SO
2
according to the Claus
reaction (3), then the other part of
the H
2
S produces H
2
by cracking
reaction (4).
KT has evidence from laboratory
tests that, by utilising its proprie-
tary catalyst, the cracking reaction
(4), the partial oxidation reaction
(2) and the Claus reaction (3) are
favoured, while SO
2
formation is
negligible.
NH
3
reaction mechanism
One of the major merits of the
novel process is the possibility to
treat a very high content of ammo-
nia even without the presence of
H
2
S. NH
3
destruction follows the
same mechanism as H
2
S conver-
sion, therefore this reaction is also
conducted autothermally. A part of
the ammonia is oxidised with the
oxygen provided to the system to
produce N
2
and H
2
O, in turn
supplying the heat required for the
NH
3
cracking reaction producing
N
2
and H
2
.
This behaviour is very interesting
for the future treatment of extra
heavy crudes that will be increas-
ingly utilised in the coming years.
Hydrocarbons, COS and CS
2

reaction mechanism
A peculiarity of the catalyst is that
hydrocarbons adsorbed over the
catalyst will react preferably and
exclusively with H
2
S, therefore no
coke formation is expected. In this
way it is possible to avoid
completely the presence of COS
KT catalyst
O
2
(g)
Cracking
H
2
S

(g)
H
2
S

(ads)
H
2
S

(ads) H
2
(g) + S
2
(g)
SO
2
(g) + H
2
O

(g)
S
2
(g) + H
2
O

(g)
S
2
(g) + H
2
O

(g)
P
a
r
t
i
a
l

o
x
i
d
a
t
i
o
n
C
l
a
u
s

r
e
a
c
t
i
o
n
T
o
t
a
l

o
x
i
d
a
t
i
o
n
Figure 5 H
2
S reaction mechanism over proprietary catalyst
KT catalyst
O
2
(g)
Cracking
NH
3
(g)
NH
3
(ads)
H
2
(g) + N
2
(g)
N
2
(g) + H
2
O

(g)
T
o
t
a
l

o
x
i
d
a
t
i
o
n
Figure 6 NH
3
reaction mechanism over proprietary catalyst
KT catalyst
O
2
(g)
T
o
t
a
l
o
x
id
a
t
io
n
P
a
r
t
i
a
l

o
x
i
d
a
t
i
o
n
C
l
a
u
s

r
e
a
c
t
i
o
n
P
a
r
t
i
a
l

o
x
i
d
a
t
i
o
n
Cracking
O
2
(g)
H
2
(g)
S
2
(g)
CH
4
(g)
CH
4
(ads)
CH
4
(g)
H
2
S

(g)
H
2
S

(ads)
H
2
S

(ads)
CS
2
(ads)
COS

(ads)
H
2
(g) + S
2
(g)
SO
2
(g) + H
2
O

(g)
S
2
(g) + H
2
O

(g)
S
2
(g) + H
2
O

(g)
CO
2
(g) + H
2
O

(g)
CO(g) + H
2
S

(g)
CO(g) + H
2
(g)
H
2
S(ads) + CH
4
(ads)
T
o
t
a
l

o
x
i
d
a
t
i
o
n
Figure 7 Hydrocarbons, COS and CS
2
reaction mechanism over proprietary catalyst
gas kt.indd 4 06/03/2014 12:21
This is not a checkerwall. This is not a
choke ring.
This is a VectorWall
TM
and it will change
the way you look at your reaction furnace.
With unmatched mechanical stability,
the VectorWall
TM
will increase your furnace
capacity and eliminate ammonia carryover,
saving your company millions of dollars.
For more information check out our new video
at petrochemical.blaschceramics.com

Or contact us: 1.518.436.1263
vectorwall@blaschceramics.com
Our VectorWall
TM
is saving SRU
reaction furnace operators
millions of dollars
Jeffrey J. Bolebruch Senior Market Manager
blasch.indd 1 11/03/2014 16:31
www.eptq.com Gas 2014 11
New concept SRU in gas production
The results of opex analysis for a
gas feld application comparing
conventional and new concept
SRUs are shown in Table 2 and
Figure 9.
According to the selected plant
confguration, capex reduction of up
process and in the TGT section.
Low pressure steam for amine
regeneration is imported from the
battery limit while the excess steam
is sent to a high pressure and/or
medium pressure steam header
according to relevant network
conditions.
performed process simulations of
the novel process with high accu-
racy, with the purpose of choosing
the best process scheme and the
best operating conditions. In this
regard it was possible to perform
several case studies for different
industrial applications including:
• Petroleum refning
• Oil and gas production
• Coal gasifcation
• Integrated gasifcation combined
cycle (IGCC) complex
• Chemical and petrochemical
complex.
In this article, we present the
results of our opex and capex anal-
ysis relevant to a petroleum
refnery and a gas feld.
New concept SRU in a petroleum
refinery
The results of opex analysis for a
petroleum refnery application,
comparing conventional and new
concept SRUs, are shown in Table 1
and Figure 8.
A conventional plant confgura-
tion consists of a Modifed Claus
section equipped with two Claus
reactors followed by an amine-based
reductive TGT section with a dedi-
cated amine regeneration system
and the relevant thermal incinera-
tion section. A liquid sulphur
degassing section is also foreseen.
The novel process is equipped
only with a CRF system and a
sulphur condenser followed by an
amine-based reductive TGT section
with a dedicated amine regenera-
tion system, and a liquid sulphur
degassing section with full recycle
of H
2
S to the Claus and TGT
sections. A thermal incinerator is
also foreseen but it is to be utilised
only during start-up and shutdown
operations (see Figure 4).
In plant confguration A, part of
the high pressure steam produced
is utilised as the heating medium in
the novel process and in the TGT
section, while the excess steam is
let down to low pressure steam
which is utilised as the heating
medium for amine regeneration.
With this proposed confguration,
steam export will not be provided.
Plant confguration B utilises the
produced high pressure steam as
the heating medium in the novel
40
80
70
100
90
60
50
30
20
10
0
−10
P
r
o
p
o
r
t
i
o
n
,

%
−20
L
P
S
M
P
S
H
y
d
r
o
g
e
n

r
i
c
h

g
a
s
C
O
2

e
m
i
s
s
i
o
n
s
F
u
e
l

g
a
s
E
l
e
c
t
r
i
c

p
o
w
e
r
Conventional SRU concept
Plant config. A of new SRU concept
Plant config. B of new SRU concept
Figure 8 Opex analysis for a SRU installed in a petroleum refinery

SRU conventional Plant configuration A Plant configuration B
concept new SRU concept new SRU concept
Low pressure steam import Yes Yes Yes
Medium pressure steam export Yes No Yes
Hydrogen rich gas No Yes Yes
CO
2
emissions Yes No No
Fuel gas Yes Yes Yes
Electrical power Yes Yes Yes
Profit index, %* 26 34 100
*Profit = (total cost of utilities)/(ton of sulphur produced)
Profit index = (100 x profit) / (profit of plant configuration with maximum. profit)
Opex analysis for a SRU installed in a petroleum refinery
Table 1
SRU conventional concept New SRU concept
Low pressure steam import Yes Yes
Medium pressure steam export Yes No
Hydrogen rich gas No No
CO
2
emissions Yes Yes
Fuel gas Yes Yes
Electrical power Yes Yes
Profit index, %* -100 -74
*Profit = (Total cost of utilities)/(ton of sulphur produced)
Profit index = (100 x profit)/(profit of plant configuration with maximum profit)
Opex analysis for a SRU installed in a gas field
Table 2
gas kt.indd 5 06/03/2014 12:21
12 Gas 2014 www.eptq.com
TGT is 99.9++%, much higher than
the 99.9+% of current BAT. The
reason for this difference is the fact
that in the novel process the quan-
tity of total sulphur (H
2
S+COS+CS
2
) in the top of the TGT
absorber is much lower than in the
conventional one
• The same fow scheme for the
SRU can be used in all industrial
applications.
Up to now only a few disadvan-
tages have been identifed:
• The H
2
S cracking section has a
lower sulphur recovery effciency
when compared with the conven-
tional one which has a reaction
furnace and two or three additional
Claus reactors
• In case of shutdown of the TGT
section, a higher quantity of SO
2
may be released to the atmosphere.
Nevertheless such a scenario is very
rare or near-impossible due to the
very high availability and reliability
of the TGT section, especially when
treating tail gas from the H
2
S crack-
ing section with zero SO
2
content
• The availability and reliability of
the proprietary catalyst needs to be
demonstrated in a pilot plant.
This article is based on a paper presented at the
ADIPEC 2013 Technical Conference, Abu Dhabi,
UAE, 10-13 November 2013.
Michele Colozzi is Business Development
and Technology Manager with KT-Kinetics
Technology SpA (KT) in Rome and is involved
in the commercialisation, design and
development of new processes, construction,
commissioning and start-up of sulphur and
gas treatment plants, oil and gas treatment
plants, hydrogen and synthesis gas generation
plants, and waste disposal plants. He holds a
chemical engineering degree from University
of Rome “La Sapienza”.
Simona Cortese is a Business Development
Coordinator with KT-Kinetics Technology.
She works as coordinator of the technical
team for the development of gas treatment
and sulphur recovery technical proposals for
PDP and FEED activities and as coordinator
of research and development team. She
graduated as a chemical engineer from the
University of Rome “La Sapienza”.
Lucia Barbato is a Process Engineer with
Processi Innovativi Srl, a process company
specialising in R&D and owned by KT-Kinetics
Technology. She is involved in the research
and development of projects for partial
oxidation of hydrocarbons and H
2
S, and
graduated cum laude as a chemical engineer
from the University of Salerno.
ppmv) which is not possible with
the current BAT
• There is complete destruction of
COS, CS
2
, hydrocarbons, RSH,
ammonia, HCN and other impuri-
ties in the catalytic reaction furnace
• The novel process can treat only
sour water stripper acid gases,
without any limitation on the
ammonia content in the sour gases;
this is not possible with the conven-
tional process
• A conventional plant requires
two or three Claus catalytic
converter systems (pre-heater, reac-
tor, sulphur condenser) to minimise
the SO
2
, COS and CS
2
content in the
tail gas to the TGT section. The
novel process requires only one
catalytic converter system since the
catalyst is installed directly in the
catalytic reaction furnace. Claus
catalytic converters are no longer
necessary due to the low content of
SO
2
in the process gas from the
CRF system
• The novel process will produce a
lower quantity of exhaust catalyst
compared to the conventional
process
• The reduced number of items of
equipment makes the investment
cost of the novel process cheaper
than the conventional one. In addi-
tion, the associated maintenance
costs will be reduced
• The major advantage of the novel
process is the higher overall SRE.
Indeed, the SRE of H
2
S cracking +
to 30% may be achieved due to the
reduced number of items of equip-
ment foreseen in the SOAP process.
The innovative plant confgura-
tion shows an important
peculiarity, the possibility to mini-
mise SO
2
emissions, since the total
sulphur (H
2
S+COS+CS
2
) contained
in the gas from the TGT absorber is
much lower than in the conven-
tional one, as explained in the
following formula:
[(STot)
Top TGT Absorber
]
Novel Process
<
1
[ (STot)
Top TGT
Absorber
]
Conventional Process
n

5<n<10
Utilising always the same plant
confguration — the new SRU
concept (SOAP+RAR) — it will be
possible to achieve SO
2
content in
gas released to the atmosphere
lower than 50 mg/Nm
3
(18 ppmv).
Conclusions
The sour gas SOAP is an environ-
mentally friendly solution capable
of valorising the sour gases which
are considered a problem in all
industrial complexes, producing
hydrogen as valuable product,
where required, with the purpose
of achieving zero emissions.
The main advantages of the novel
process are summarised below:
• In terms of SO
2
emissions, the
novel process is able to achieve
values lower than 50 mg/Nm
3
(18
60
100
90
80
70
50
40
30
20
10
P
r
o
p
o
r
t
i
o
n
,

%
0
L
P
S
M
P
S
H
y
d
r
o
g
e
n

r
i
c
h

g
a
s
C
O
2

e
m
i
s
s
i
o
n
s
F
u
e
l

g
a
s
E
l
e
c
t
r
i
c

p
o
w
e
r
Conventional SRU concept
New SRU concept
Figure 9 Opex analysis for a SRU installed in a gas field
gas kt.indd 6 11/03/2014 16:30
Copyright © 2013 Merichem Company
To win a race you need not only speed, but a timely start. Merichem Company has a proven track record
of providing hydrocarbon treating solutions that meet the stringent timelines of our customers. Get your
treating project off to a timely start by choosing Merichem Company as your technology licensor.
Midstream Solutions
Proven Results
Think Fast...Think Merichem
Midstream Solutions with Industry-Leading Speed to Market
Merichem Company’s portfolio of midstream treating technologies
includes our patented FIBER FILM
®
Contactor that offers a variety of
caustic, amine and acid treating processes. Merichem also offers
solutions for gas treating, including our patented LO-CAT
®
wet scrubbing,
liquid redox system that converts H2S to innocuous, elemental sulfur. In
addition to our large portfolio of licensed treating technologies,
Merichem offers industry leading spent caustic management services that
utilize Beneficial Reuse options to create a non-waste solution for your
spent caustics. For fast, effective treating solutions, choose Merichem.
www.merichem.com | 5455 Old Spanish Trail | Houston, Texas 77023 | 713.428.5000
®
C
M
Y
CM
MY
CY
CMY
K
Merichem Company-FY14-MID.pdf 1 2/24/2014 8:19:32 AM
merichem.indd 1 06/03/2014 12:04
cleaner, better shale gas & oil
At Zeeco, we understand the everyday realities of shale gas & oil.
From our field-proven, smokeless enclosed or open flares to thermal
oxidisers specifically designed to handle amine/BTEX off-gasses, we
have the solutions that keep your people and your facilities safe – and
keep you in compliance. Solutions engineered for your purposes,
available from standardised inventory, and backed up by the world
leader in combustion technology. Choose the company where the
experts answer the phone and the engineers understand your needs.
Zeeco: We work where you work.
Zeeco, Inc. 22151 E. 91st St., Broken Arrow, OK 74014 USA
+1-918-258-8551 sales@zeeco.com zeeco.com
©Zeeco, Inc. 2014
burners • flares • thermal oxidisers • aftermarket parts, service & solutions • rental equipment
meet the Zeeco team at the AFPM Reliability & Maintenance Conference, booth #731
Standard ZEECO
®
Vertical Thermal Oxidiser
for Gas Processing Applications
Zeeco_PTQ Gas_Mar.indd 1 3/10/14 4:51 PM
zeeco.indd 1 11/03/2014 09:10
Eliminating the Claus furnace
F
uel sources are becoming
increasingly sour and society is
increasing pressure on environ-
mental regulatory agencies to
improve emissions standards in a
desire for a cleaner environment.
As a result, producers and proces-
sors in the gas processing, refning,
petrochemical and other industries
have a growing focus on sulphur
removal technologies that not only
will improve the quality of their
product streams but will also
enable them to comply with the
more stringent environmental regu-
lations being imposed by
governmental agencies. Since the
price of sulphur is not suffciently
high for sulphur recovery processes
to be proftable, the value of a
sulphur recovery unit (SRU) is
defned by the lost opportunity cost
associated with downtime of the
associated hydrocarbon processing
unit as well as by its cost of opera-
tion. Then, for the producers and
processors, it is important to under-
stand the various sulphur removal
and recovery technologies available
so that the most reliable, lowest
cost processes can be selected and
used in their facilities.
Over the past decade, researchers
at Phillips 66 developed a catalytic
combustor that can be used to
replace the burner and thermal
reactor in Modifed Claus units and
improve the operations of the
SRU.
1-9
This new approach to the
Claus sulphur recovery process
was presented at the Laurance Reid
Gas Conditioning Conference in
2009.
10
Since that time, GTC
Technology has acquired the exclu-
sive licensing rights to the catalytic
combustion technology that
A novel approach to the conversion of hydrogen sulphide to elemental sulphur is
expected to extend the economic range for Claus plants
MATT THUNDYIL, SAMEER PALLAVKAR, RAMIRO VAZQUEZ and DAVID SEEGER
GTC Technology US, LLC
replaces the conventional Claus
burner and thermal reactor and
markets it under the name
GT-CataFlame. GTC markets the
complete process of GT-CataFlame
integrated with the downstream
Claus converters under the name
GT-Spoc (Sulphur Partial Oxidation
Catalysis). The ultimate design of
GT-Spoc is a single vertical vessel
that contains all the components of
GT-CataFlame, followed by the
Claus converters and sulphur
condensers.
Background: conventional Claus
The Claus process was patented by
Carl Friedrich Claus in 1883, and
introduced in 1936. The Modifed
Claus process is the most successful
commercial method for sulphur
recovery. In this framework, the
reaction of H
2
S with oxygen is
separated into two stages: (1) a
highly exothermic thermal stage
where approximately 50% of the
H
2
S is converted to elemental
sulphur and, of the remaining H
2
S,
a third of it is converted to SO
2;
and
(2) a moderately exothermic cata-
lytic stage where the remaining H
2
S
in the gas stream reacts with the
rest of the SO
2
to produce more
elemental sulphur. The reactions
are reversible, and conversions are
highly dependent on temperature,
sulphur content and moisture
content. To achieve sulphur conver-
sions greater than 60-70%, the
thermal stage is followed by
sulphur condensation and separa-
tion which is followed by reheating
upstream of a catalytic stage oper-
ated at temperatures higher than
the sulphur dew point. Additional
catalytic stages may be added to
increase the effciency of sulphur
removal.
The fame stability of the combus-
tion section is a critical parameter
in Claus operations. At a H
2
S
content of above 55%, the acid gas
can be sent directly to the furnace.
Between 30–55% H
2
S, the acid gas
or combustion air (or both) may
need to be preheated. At concentra-
tions below 30% H
2
S, the Claus unit
operates in a ‘split fow’ mode with
preheat and, as the H
2
S content
drops below 10%, fuel gas may
need to be added.
A two-stage Claus unit can
deliver 90–95% sulphur recovery
effciency, with a three-stage confg-
uration delivering 95–98% recovery.
The tail gas is generally sent to an
incinerator if 96–97% sulphur
recovery effciency is acceptable. If
sulphur recovery in the 99–99.5%
range is required, tail gas opera-
tions based on a continuation of the
Claus reaction under sub-dew point
is generally undertaken either on a
solid bed, or in the liquid phase. If
sulphur recovery effciencies of
99.9% are required, the sulphur in
the tail gas is generally converted
www.eptq.com Gas 2014 15
If sulphur recovery
efficiencies of 99.9%
are required, the
sulphur in the tail
gas is generally
converted to H
2
S by
hydrogenation and
hydrolysis
gas gtc (original).indd 1 06/03/2014 12:33
16 Gas 2014 www.eptq.com
near-equilibrium H
2
S conversion
and sulphur selectivity in one tenth
of the volume used by a conven-
tional Claus burner and reaction
furnace.
10,11
The catalyst also
contains components that eliminate
classic Claus catalyst deactivation
mechanisms of sulphur poisoning
and coke deposition of the catalyst
in the frst Claus converter during
normal sulphur recovery operation
and start-up/shutdown using fuel
gas.
A simplifed process fow diagram
for GT-CataFlame integrated with
the waste heat boiler (WHB), frst
Claus converter and sulphur
condensers is shown in Figure 1.
This diagram shows an expected
application where the catalytic
combustor is installed in place of a
poorly performing Claus furnace.
In GT-CataFlame the air and the
acid gas are preheated to approxi-
mately 220°C (428°F) before they
are mixed. The gases are blended in
a specially designed chamber to
thoroughly mix the two gases
upstream of the catalyst reactor
bed. The gas contacts the front face
of the reactor at a specifc actual
velocity, passes through the catalyst
bed in under one second, and
immediately contacts the WHB. In
the reactor, approximately 70% of
the H
2
S is converted to sulphur and
H
2
or H
2
O, 10% is converted to H
2
O
and SO
2
, and most of the hydrocar-
bons are converted to H
2
, CO and
H
2
O, and any ammonia is
converted to N
2
and H
2
. The gas
exiting the WHB may be sent to a
sulphur condenser where elemental
sulphur is removed; after the
condenser, the gas may be
re-heated to pass into the frst
Claus converter. Alternately, the
gas exiting the WHB may pass
directly into the frst Claus
converter and the previously
mentioned sulphur condenser and
re-heater may be eliminated; pass-
ing directly from the WHB into the
frst sulphur condenser the process
effectively ‘jumps’ the Gamson and
Elkins curve and 90% sulphur
recovery occurs after the frst Claus
converter.
12
The process that
includes sending the gas directly
from the WHB into the frst Claus
converter is part of the patents that
to H
2
S by hydrogenation and
hydrolysis. Then the H
2
S is
captured and recycled to the Claus
unit.
Modifed Claus units are chal-
lenging to operate reliably, and are
particularly prone to problems
during start-up and shutdown. For
example, the furnace is usually
started up and shut down using a
fuel gas stream as the fuel, rather
than the acid gas. When the feed
shifts from acid gas to predomi-
nantly fuel gas, with an associated
shift in oxidant ratio (air ratio),
higher temperatures and soot
formation often may result, among
other undesirable consequences.
The soot has a tendency to foul the
Claus catalyst downstream of the
furnace.
The operating cost associated
with sulphur removal for a Claus
unit with tail gas clean-up is in the
$100/t range when considering
utilities and maintenance costs. The
sulphur that is recovered is gener-
ally bright yellow and preferred in
the marketplace.
GT-Spoc: process description
GT-Spoc uses a patented, durable
catalyst in a ‘short-contact-time’
reactor, GT-CataFlame, to achieve
Reheat
Catalyst
bed
Condenser
Sulphur pit
Condenser
GT-CataFlame
WHB
Acid gas
Air
Boiler
feed water
HP steam
LP steam
Figure 1 Simplified process flow diagram of GT-CataFlame integrated with the first Claus
converter and sulphur condensers
Acid gas
Air
GT-CataFlame reactor
Equivalent conventional Claus furnace
Waste heat recovery
Waste heat recovery
Figure 2 A conventional Claus furnace and a GT-CataFlame designed for the same
application showing the relative size difference
gas gtc (original).indd 2 11/03/2014 19:05
www.eptq.com Gas 2014 17
chimney-type separation tray
where the liquid sulphur exits the
side of the unit and the gas passes
down through the chimney. After
this molten sulphur separation tray,
the remainder of the process is the
same as a conventional SRU except
that it is arranged in a vertical
configuration.
The process that includes remov-
ing the first sulphur condenser (the
one located after the WHB and
Stream Acid gas from amine unit Sour water stripper gas Combustion air
Mass flow, kg/h 4050 730 10 900
Flow rate, Nm
3
/h 2671 675 8500

Composition, kg-moles/h
C
1
0.48 0 0
C
2
0.72 0 0
C
3
1.19 0 0
C
4
0.60 0 0
CO
2
1.19 0 0
H
2
O 2.86 6.00 5.4
H
2
S 112.21 12.00 0
Hexene 0 0.04 0
Phenol 0 0.02 0
N
2
0 0 295.6
O
2
0 0 78.5
NH
3
0 12.00 0
Conditions for comparison of GT-CataFlame and a Claus furnace
Table 1
HP steam
Boiler feed water
Boiler feed water
Boiler feed water
LP steam
Liquid sulphur
LP steam
Condensate
LP steam
Liquid sulphur
Tail gas
Preheated acid gas
Preheated air
GT-CataFlame catalyst
Claus catalyst
Claus catalyst
Figure 3 Ultimate vertical design of GT-Spoc with two Claus converters in series
are included in GT-Spoc. The gas
path after the first Claus converter
is the same as that of a conven-
tional Claus SRU.
A conventional Claus furnace and
a GT-CataFlame reactor designed
for the same acid gas application
are shown for comparison in Figure
2; the conditions used to design
both units are summarised in Table
1. In the figure, the furnace and
GT-CataFlame are drawn to the
same scale which shows the rela-
tively larger size requirement for
the furnace. This size difference
saves significantly on footprint and
the amount of refractory material
required. The GT-CataFlame refrac-
tory volume is 1/40
th
that of the
refractory volume required for the
furnace. This large reduction in
refractory is due in part because
the refractory lining the mixing
section is only for protection and
good engineering practice.
The ultimate design intended for
GT-Spoc is a vertical arrangement;
that is, GT-CataFlame is integrated
with the WHB and the downstream
Claus unit equipment, all combined
in a single vertical tower. A simpli-
fied diagram of the vertical
GT-Spoc with two Claus converter
stages is shown in Figure 3. The
acid gas and air are first pre-heated
as they were in the horizontal
arrangement described above. The
preheated gases are then combined
at the top of the unit in the mixing
chamber. The well-mixed gas then
travels downwards to the catalyst
where the H
2
S is converted to
sulphur. Next, the gas passes to the
first Claus converter without first
passing through a sulphur
condenser. In the vertical GT-Spoc
arrangement, the sulphur
condenser located between the
WHB and the first Claus converter
is removed. The WHB operating
temperature is adjusted so that the
gas out of the first Claus converter
remains above the sulphur dew
point temperature. The gas exits
the WHB and passes directly to the
first Claus converter bed. After the
Claus converter, the gas is cooled
in the first sulphur condenser and
molten sulphur forms as it does in
the conventional Claus SRU. The
molten sulphur collects on a
gas gtc (original).indd 3 06/03/2014 12:33
18 Gas 2014 www.eptq.com
• The least amount of unwanted
byproduct formation, primarily
COS and CS
2
• The destruction of ammonia
• Formation of H
2
and CO, rather
than any unwanted hydrocarbon
byproducts.
The results shown in Figure 4 indi-
cate the performance of a particular
catalyst formulation that was tested
for an extended period in the unit.
During this testing period, air-to-
H
2
S ratios and gas velocities were
being evaluated so, throughout the
duration of the testing, these param-
eters were varied to observe the
effects and to arrive at optimum
operating conditions. However,
even though conditions were
changed, H
2
S conversion and selec-
tivity towards forming elemental
sulphur remained high throughout
the period. At certain points, the
data refect operating conditions
that achieve better conversion and
sulphur yield so those operating
conditions will be used for commer-
cial unit design and operation.
Throughout the test period, the
range of gas composition was
approximately 80-82 mol% H
2
S,
11-12 mol% CO
2
, 5-6 mol% water,
and 1 mol% hydrocarbon content;
the remaining 1-2 mol% is nitrogen
and the typical analytical measure-
ment error.
The results indicate that over 80%
of the H
2
S was converted with
approximately 69-70% yield of
elemental sulphur; the remainder
was unreacted H
2
S and SO
2
, and
less than 1% formation of COS and
CS
2
combined. In addition there
was an average 7% yield of H
2
. This
performance was better than that of
a conventional Claus furnace where
50–70% of the inlet H
2
S is converted
and the elemental sulphur yield is
40–65% of the inlet sulphur.
Advantages
Based upon the results of the test-
ing that have been presented here
and previously,
10
several process
and economic advantages have
been identifed from having a short
contact time catalytic combustor
versus a fame reaction furnace.
Process advantages
• Reactants are pre-mixed prior to
showed signifcant improvement in
hydrogen and sulphur yields over
what was typically observed in
refnery Claus furnaces. Higher
sulphur and hydrogen yields
were believed to result from
direct oxidation and subsequent
splitting of H
2
S at high reaction
temperatures.
11
The results were
gathered with no hydrocarbons
in the inlet gas, therefore the
hydrogen must come from NH
3

and H
2
S.
Operating results
The researchers from Phillips 66
undertook a project to demonstrate
the catalytic combustion section of
GT-Spoc, GT-CataFlame. A pilot
unit installed at a refnery location
in the US operated intermittently
for more than a year. Testing was
performed to identify the operating
parameters and catalyst formula-
tion that resulted in:
• The highest H
2
S conversion to
sulphur
• Producing the necessary amount
of SO
2
for the downstream Claus
converters
upstream of the frst Claus
converter) is patented by Phillips
66 and licensed by GTC. This step
has the advantage of reducing the
number of pieces of equipment, the
overall SRU footprint, and reducing
the cost of the SRU; this confgura-
tion may be applied to a
conventional Claus SRU with or
without a GT-CataFlame as well as
in the vertical GT-Spoc SRU.
GT-Spoc: operating data
Phillips 66 conducted a large
number of tests on a catalytic
combustor that could be used to
replace the conventional Claus
furnace with the goal of providing
improved operation of Claus
SRUs. After extensive development,
the inventors arrived at a robust
catalyst formulation and a combus-
tor design that delivered results
which could be used to model
and design commercial facilities.
The summary results presented in
Table 2 are from tests using the
catalytic combustor compared to
measurements taken at the SRUs
of four refneries. The results
60
100
90
80
70
50
40
30
20
10
H
2

c
o
n
c
e
n
t
r
a
t
i
o
n
H
2
S

c
o
n
v
e
r
s
i
o
n
S
u
l
p
h
u
r

y
e
i
l
d
,

%
0
0 100 200 300 400 500 600 700 800 900
Elapsed time, hrs
Sulphur yield
H
2
S conversion
Outlet H
2
Figure 4 Results of H
2
S conversion, sulphur yield and hydrogen concentration in the
outlet gas for 900 hours of testing at various operating conditions in a GT-CataFlame unit
Unit GT-Spoc Refinery A Refinery B Refinery C Refinery D
NH
3
present? Yes Yes Yes Yes No
% S yield 74.0 63.4 48.1 54.9 68.2
% H
2
yield 7.0 6.3 3.5 5.4 4.2
Laboratory results for GT-Spoc catalyst compared with refinery Claus reaction
furnace sulphur and H
2
yields
Table 2
gas gtc (original).indd 4 06/03/2014 12:33
fast returns
A Century of Innovation
in the Oil and Gas Industry
1914 - 2014
Quickly maximize the value of natural gas assets with NGL recovery
solutions from UOP.
Get cash flow and profits sooner with hydrocarbon management solutions from
UOP, including the UOP Russell line of complete, modular, factory-built systems
that are designed to have you up and running six months quicker than other
options. The reduced on-site installation time, proven designs and pre-fabricated
solutions can deliver an earlier startup at a lower cost than stick-built facilities.
Or choose the UOP-Twister
®
Supersonic Gas Separation System, with a small
footprint and low maintenance that’s often ideal for remote and offshore facilities.
Or get maximum NGL/LPG recovery — an industry-leading 99+% — with solutions
from UOP and Ortloff Engineers, Ltd. In short, whether you need fast, compact,
efficient or all of the above, UOP has a solution for your NGL-recovery needs.
For more information about UOP Russell solutions, visit www.uop.com/uoprussell or
visit www.uop.com to learn about all of the UOP hydrocarbon management solutions.
© 2014 Honeywell International, Inc. All rights reserved.
UOP-Russell-Fast-Returns-Ad-Full-Page_UOP Ad 8.5x11 UOP6467 1/10/14 3:22 PM Page 1
uop.indd 1 04/03/2014 10:48
20 Gas 2014 www.eptq.com
re-heater that is located down-
stream of the WHB may also be
eliminated, so that the gas stream
exiting the WHB fows directly to
the frst Claus reactor stage.
COS and CS
2
reduction and
implications for Claus converter
bed design
The best way to minimise the
effects of hydrocarbon contamina-
tion of acid gases in any SRU is to
prevent hydrocarbon absorption in
the acid gas removal system (slug
and drop catching drums, aerosol
droplet removal, and lean amine/
sour gas differential temperature
control) and mitigate hydrocarbon
accumulation (rich fash drums
with skimming and gas removal,
carbon fltration, and regenerator
refux purge).
Despite these efforts, hydrocarbon
may get into the acid gas, leading to
the undesired byproducts COS and
CS
2
. Nominally, the byproducts
result from reacting after the fame
mixing zone where hot, oxygen-free
fame byproducts can mix with
ratios, making switching nearly
seamless
• COS and CS
2
formation are
signifcantly reduced
• The sulphur product contains
only 25-33% of the dissolved H
2
S/
H
2
S
x
of normal Claus sulphur.
Economic advantages
• Signifcant reduction in refractory
lining due to smaller size of
GT-CataFlame
• Small reactor volume and reduc-
tion in byproducts allows for
design changes reducing overall
unit footprint
• Design changes eliminate inter-
connecting piping and sulphur
rundown piping and equipment
• Because GT-CataFlame is smaller
than a typical furnace, there is less
refractory and catalyst mass to heat
up or cool down; this in turn
reduces the time required for shut-
down or start-up
• The Claus plant can be
constructed in a vertical orientation
where the unit is self-draining
• The frst sulphur condenser and
passing through the GT-CataFlame
reactor. The reaction is uniform
through the cross section of the
reactor eliminating the problems of
post-combustion mixing for
contaminant destruction
• Close-coupling of the catalyst
zone and waste heat boiler (WHB)
improves overall sulphur yield and
reduces air requirement due to
rapid reaction quenching
• No fames, fre-eyes, nor burner
management systems: a small
retractable burner or access to a
hydrogen containing fuel is all that
is needed to trigger the catalyst at
start-up
• Fuel gas oxidation for warm-up
takes place at 25% air stoichiometry,
eliminating oxygen breakthrough to
downstream Claus converter beds
during start-up and shutdown
• Low molecular weight hydrocar-
bons are converted to reduction tail
gas unit-friendly H
2
and CO by
catalytic partial oxidation. Soot
formation is virtually eliminated.
Air/fuel ratios for fuel gas oxida-
tion closely resemble acid gas/air
www.eptq.com PTQ Q1 2014 131
3 Respini, Jones, Spanu, Sesselego, Avoiding foul
play, Hydrocarbon Engineering, Nov 2006.
Matteo Virzi is Senior Technology Manager
with ISAB Priolo refinery in Italy. He is an
expert in distillation and thermal conversion
processes. With more than 20 years of
experience in technology, operations and
automation, he holds a degree in chemical
engineering from the University of Palermo.
Email: mvirzi@isab.com
Marco Respini is a Senior Technology Expert
with Baker Hughes Downstream Chemicals,
specialising in refinery and petrochemical
process improvements in fouling control.
He has 15 years of refining experience and
is currently involved in developing new
technologies for improving refinery conversion
processes. With extensive experience in
asphaltene related problems in oil production
and refining, he is an inventor of five US patents
and has published 10 technical papers and
seven conference papers on visbreakers and
heavy fuel oil stability problems. A graduate
of Milan University with a degree in industrial
chemistry, he has been a Research Fellow in
the field of organometallic catalysts and is a
registered professional chemist in Italy. He is
also a member of ACS and NACE.
Email: marco.respini@bakerhughes.com
which provided refnery personnel
with the right information to be
able to constantly keep the unit at
the best process severity for any
processed feed.
VisTec is a mark of Baker Hughes Incorporated.
References
1 Petralito G, Respini M, Achieving optimal
visbreaking severity, PTQ, Q1, 2010.
2 A Phase Separation Kinetic Model for Coke
Formation, Preprints ACS, Div. Pet Chem, 38,
428-433, 1993.
result in better handling of fouling,
and at the same time increase
HVGO directly (less gasoil recycled
as wash oil). Figure 17 shows the
trend of wash oil rate.
The decrease in wash oil rate
was continuously optimised. The
HVGO was analysed on a daily
basis with respect to the level of
contaminants, using the VCI tech-
nique that measures the coke
particles entrained within the
HVGO (thus not removed by wash
oil).
Also, the rate of antifoulant on
the wash oil was optimised and
increased when needed.
Conclusions
This article describes an example of
high conversion visbreaking with
increased run length when
compared to typical visbreaker run
lengths.
This successful result was made
possible by coupling Baker Hughes
VisTec anticoke/antifoulant treat-
ments with monitoring technology,
A controlled decrease
in wash oil can result
in better handling of
fouling, and at the
same time increase
HVGO directly
B
e
s
t
V
a
lv
e
s
s
in
c
e
1
8
6
7
baker hughes.indd 8 13/12/2013 11:38
gas gtc (original).indd 5 11/03/2014 19:05
www.eptq.com Gas 2014 21
Start-up and shutdown operational
improvements
Most of the time, starting up a Claus
plant is spent heating up the large
amount of refractory in the reaction
furnace chamber and catalyst in the
catalyst beds. The GT-CataFlame
reaction section is very compact and
in close proximity to the waste heat
boiler entrance, so there is a mini-
mum amount of refractory to be
dried and heated to operating
temperature.
The small amount of catalyst and
support materials in the reaction
section is easily heated up by a
small preheater such as a
Stackmatch, eliminating one of the
biggest problems in starting up or
recovering from a shutdown —
relighting the main burner.
Laboratory testing has shown that
hydrogen rich gases (~40% H
2
) and
air mixtures can be catalytically
ignited by the catalyst, potentially
bypassing the need for any type of
fred preheater.
Minimising the formation of
contaminants such as COS and CS
2

and operating the frst catalyst bed
directly after the WHB makes it
possible to reduce the amount of
Claus catalyst needed, speeding up
both the shutdown process
(sulphur removal or heat soak) and
the start-up process (warm-up).
Elimination of the sulphur
condenser following the WHB and
subsequent reheater eliminates
cool-down or warm-up of equip-
ment and interconnecting piping,
again speeding up the shutdown or
start-up process.
When heating up or cooling
down a Claus plant, fuel gas fring
at near stoichiometric ratios is
used, especially when Claus cata-
lyst is sulphur laden. Air-to-fuel
gas ratios for most natural gas
streams range from 9.5-10:1,
while air to acid gas ratios are
typically in the 1-2.5:1 range,
depending on acid gas concentra-
tion. This makes for diffcult
switching from one feed gas to
another. At stoichiometric ratios,
the adiabatic fame temperature
(~3500°F, 1925°C) for most fuel
gases is well above the temperature
limits of reaction furnace refractory
(~2800°F, 1535°C) and ferrules,
The GT-CataFlame demonstration
unit confrmed the results obtained
in the laboratory.
Turndown
Laboratory testing showed that
operations were stable down to
25% turndown when operating
only with acid gases. However, the
laboratory apparatus was capable
of running with any mixture
between pure natural gas and pure
H
2
S.
Demonstration plant data were
run at as low as 12% turndown
with minimal effect on perfor-
mance. Because of the ability to
operate with nearly any ratio of
fuel gas to acid gas, even lower
turndown rates can be achieved.
Air demand and related operational
improvements
While the overall recovery of
sulphur will fuctuate signifcantly
with air-to-acid gas ratios in both
the Claus and GT-Spoc processes,
the GT-CataFlame catalyst’s selec-
tivity for sulphur does not vary
signifcantly over a wide range of
air-to-H
2
S ratios.
The combination of pre-mixing
before reaction and close coupling
of the reactor to the WHB aids both
hydrogen formation reactions
(partial oxidation and dissociation)
and inhibits the main hydrogen
consumption reaction (recombina-
tion of H
2
and S). As fow rate
increases in the unit, improved heat
transfer due to higher Reynolds
number and thermal conductivity
of the gas (due to higher hydrogen
content) appear to aid in inhibiting
recombination. This translates to
more capacity with less overall
pressure drop.
unburned hydrocarbons, for
instance CH
4
and S
2
, in the acid
gases as in a typical 1/3-2/3 split-
fow furnace for leaner H
2
S content
gases.
14
Since the gases fowing to
the GT-CataFlame reactor are
pre-mixed and pre-heated, the
temperature is hotter than that of a
furnace (approximately 2200-2300°F,
1200-1260°C), and since the time
from reactor to WHB is short, plant
data show a signifcant reduction in
the amount of byproduct COS and
CS
2
when compared to a typical
Claus unit.
A study of 24 gas plants with a
broad range of inlet H
2
S concentra-
tions in the gas passing to the Claus
furnace showed that the COS and
CS
2
concentration in the gas exiting
the WHB was in the range 0.06-1.7
mol% for COS and 0.01–1.1 mol%
for CS
2.
13
At the optimised operat-
ing conditions for the GT-Spoc pilot
unit, 0.12–0.18 mol% COS and 0.36–
0.48 mol% CS
2
were measured in
the gas leaving the catalytic
combustor.
Also, since the formation of COS
and CS
2
is signifcantly decreased,
and the frst Claus converter condi-
tions will be hotter than a
conventional Claus SRU design, the
Claus converter following the reac-
tor can be much smaller, helping to
reduce the Claus catalyst bed size
and allowing for a more compact
footprint. In GT-Spoc, the frst Claus
converter may be hotter than in a
conventional Claus design if the gas
out of the WHB fows directly to the
Claus converter without being
cooled in a sulphur condenser.
Acid gas concentration requirement
Stable operation of the
GT-CataFlame catalyst was demon-
strated at acid gas concentrations as
low as 25% H
2
S using only
pre-heated air and pre-heated acid
gas. In addition, a wide range of
laboratory tests were conducted on
lean H
2
S streams containing CO
2
or
nitrogen diluents and typical light
hydrocarbon components. Sulphur
yields varied with H
2
S concentra-
tion, but the yields were still close
to equilibrium computations, unlike
lean acid gas Claus units that
produce little to no sulphur from
the thermal stage.
In general, the
more frequent the
shutdown and start-up
of a Claus unit, the
more problems can
be expected
gas gtc (original).indd 6 11/03/2014 19:06
22 Gas 2014 www.eptq.com
Capital cost comparison
An independent engineering frm
performed a cost study that
compared the capital cost of a
two-stage GT-Spoc unit to a
conventional two-stage Claus unit
with a furnace. The GT-Spoc SRU
was designed in the single, vertical
tower arrangement shown in Figure
3. The amine unit, the tail gas unit
(TGU) and the sulphur pit were
assumed to be similar for both
units and were not included in the
capital cost estimate. Therefore any
cost advantage of the effects of
reduced duty on the TGU and less
H
2
S and H
2
S
x
dissolved in the
sulphur are not included in the
comparison. The cost study was
prepared using the standard engi-
neering cost factors based on US
Gulf Coast prices.
Design basis
The composition of the acid gas
fowing to the GT-CataFlame or to
the Claus furnace is shown in Table
4. For purposes of presentation in
the table, the acid gas and sour
water stripper (SWS) off-gas are
added together and reported as the
composition of one stream, as they
would be in the case of the feed gas
to the GT-Spoc SRU. However, in
the case of a conventional Claus
stream, the SWS off-gas is added to
the burner and the acid gas is oper-
ated in split fow.
The fow rate of the combined
acid gas and SWS off-gas is 42.9
Nm
3
/h (3.8 MMscfd) and the
sulphur production rate is approxi-
mately 100 t/d.
Cost comparison results
The results of this cost estimate
indicated that the capital cost of a
GT-Spoc unit is approximately 20%
less than that of a conventional
Claus unit. A few major differences
between the two processes are the
reasons for the savings and they
are:
• The GT-CataFlame reactor is
approximately 20% of the volume
of an equivalent Claus furnace
• GT-CataFlame requires much less
refractory (as little as 1/40
th
the
amount required for a conventional
Claus furnace)
• The sulphur condenser located
Reduction of H
2
S and sulphanes in
produced sulphur
Experimental data from laboratory
tests demonstrate that the sulphur
produced in the GT-Spoc vertical
arrangement contains less dissolved
H
2
S and H
2
S
x
because all the
sulphur from both stages will only
be exposed to the H
2
S partial pres-
sure that would normally be found
at the second condenser of a
conventional Claus unit, which is
signifcantly lower than at the frst
condenser following the conven-
tional thermal stage. The results in
Table 3 show the difference in H
2
S
and H
2
S
x
content as the sulphur
from the sulphur condenser imme-
diately following GT-CataFlame
contains fve to six times more
dissolved H
2
S and H
2
S
x
as sulphur
from the condenser that follows
both GT-CataFlame and the frst
stage Claus converter.
requiring the incorporation of
steam or nitrogen diluent to hold
fame temperatures to a safe oper-
ating envelope.
Also, near-stoichiometric fames
have a tendency to form soot which
can rapidly plug catalyst beds
downstream of the fame.
GT-CataFlame catalyst can convert
fuel gas via partial oxidation (down
to as low as 25% of full stoichio-
metric) air, close to the ratios used
for acid gas operations. The catalyst
has components that resist the
formation of coke on the catalyst
surface which translates into less
production of soot.
Firing slightly over stoichiometric
can cause unwanted catalyst
exotherms, damaging both catalyst
and vessels. Since the catalyst
does not have to operate near
stoichiometric air-to-fuel ratios,
damaging exotherms on Claus cata-
lysts are eliminated.
In general, the more frequent the
shutdown and start-up of a Claus
unit, the more problems can be
expected. During the small demon-
stration plant’s long term catalyst
testing, the unit was shut down
and started up over 30 times.
Through all of the start-ups
and shutdowns, the GT-CataFlame
catalyst retained its original
activity.
Conditions:
Air/H
2
S 2.4
H
2
S flow 750 cm
3
/min After GT-CataFlame After GT-CataFlame and one Claus stage
% Sulphur yield 72.3 85.8
H
2
S in sulphur, ppmw (FTIR) 308.4 60.3
H
2
S
x
in sulphur, ppmw (FTIR) 340.1 54.2
Reduction of H
2
S and H
2
S
x
in sulphur passing through Claus catalyst
Table 3
Component Mole % Component Mole %
CO
2
0.20 nC
4
0.00
H
2
S 67.80 nC
5
0.02
C
1
0.11 nC
6
0.00
C
2
0.09 H
2
O 19.80
C
3
0.02 NH
3
11.94
iC
4
0.00 H
2
0.05
Acid gas composition used for a capital cost comparison of GT-Spoc and a
conventional Claus SRU
Table 4
The capital cost of
a GT-Spoc unit is
around 20% less than
that of a conventional
Claus unit
gas gtc (original).indd 7 11/03/2014 19:06
www.eptq.com Gas 2014 23
Enhanced Synthesis Gas Production.
6 US 7,138,101 Two Stage Catalytic Process
for Recovering Sulfur from an H
2
S Containing
Gas Stream.
7 US 7,226,572 Compact Sulfur Recovery
Plant and Process.
8 US 7,326,397 Catalytic Partial Oxidation
Process for Recovering Sulfur from an H
2
S
Containing Stream.
9 US 7,357,908 Apparatus and Catalytic
Partial Oxidation Process for Recovering Sulfur
from an H
2
S Containing Gas Stream.
10 Keller A, Is there “direct oxidation” of H
2
S
to sulfur?, Laurance Reid Gas Conditioning
Conference, Feb 2009.
11 Partial Oxidation of H
2
S to sulfur, a Claus
Alternative, British Sulphur Conference, Oct
2003.
12 Paskall H G, Sames J A, Sulphur Recovery,
13th Edition, 2010, 1-14.
13 Paskall H G, Sames J A, Sulphur Recovery,
13th Edition, 2010, 2-80.
14 Borsboom H, Clark P, Goar B G, New insights
into the Claus thermal Stage, chemistry and
temperatures, Laurance Reid Gas Conditioning
Conference, Feb 2009.
Matt Thundyil is Manager of Sulfur Business
for GTC Technology US, LLC, in Houston,
Texas. He has over 15 years’ experience in
the petrochemical and energy industry and
is the inventor on a number of patented
separations technologies. He holds a degree
in chemical engineering from the Indian
Institute of Technology (IIT), Madras, India, and
a PhD from The University of Texas, Austin, in
advanced separations technology for natural
gas processing.
Sameer Pallavkar is a Sulfur Process Engineer
with the Sulfur Business Group at GTC
Technology US, LLC’s Houston office. During
his PhD program at Lamar, he worked as
a research assistant in the Department of
Chemical Engineering, and was assigned to the
Microwave Reaction Engineering Laboratory
as a research candidate. He holds a degree in
chemical engineering from the University of
Mumbai, Bombay, and a Master’s from Lamar.
Ramiro Vazquez is a Licensing Manager,
Sulfur Business Unit, at GTC Technology US,
LLC’s Houston office, where he is responsible
for technology licensing and sales. He holds
a degree in chemical engineering from the
University of Nuevo Leon, Monterrey, Mexico,
and a Master’s in chemical engineering from
the University of Houston.
David Seeger, Senior Scientist for GTC
Technology US, LLC’s Sulfur Group in Houston,
holds a BA degree in chemistry from St.
Thomas University (St. Paul, MN) and MA and
PhD degrees in chemistry from the University
of Michigan (Ann Arbor, MI). He is the inventor
of a selenium removal process for FGD and has
led research efforts to develop and improve
processes for HAP, FGD and sulphur removal.
He is one of the founders of CrystaTech.
after the WHB and before the frst
Claus converter in a conventional
Claus SRU is eliminated in the
vertical GT-Spoc SRU design
• The frst reheater that follows the
frst sulphur condenser in a conven-
tional Claus SRU is also eliminated
as it is unnecessary in GT-Spoc
• Interconnecting piping is reduced
in GT-Spoc as the stages are close
coupled in a vertical arrangement
as compared the conventional
Claus SRU.
Summary
Conventional Modifed Claus tech-
nology with a free-fame thermal
section is well known to have a
number of operating challenges
associated with the reactions that
occur in the fame. An alternative,
proven approach is presented here,
where the free-fame thermal
section is replaced with a catalytic
combustor that utilises a durable
catalyst that enables sulphur partial
oxidation catalysis. This change at
the front end of a conventional
Claus process results in signifcant
process and economic advantages,
including improved sulphur yield,
reductions in COS and CS
2
forma-
tion, reduced impact of
hydrocarbons on air demand,
improved start-up and shutdown
operations, and reductions in the
Claus converter catalyst bed
sections. Overall, GT-CataFlame
allows upgrades of aging Claus
burners and enables a 20–30%
reduction in capital costs associated
with Claus plants, while reducing
the tail gas unit load due to its
higher conversion effciencies.
GT-SPOC and GT-CataFlame are marks of GTC
Technology US, LLC.
References
1 US 6,403,051 Recovery of Sulfur from H
2
S
and Concurrent Production of H
2
Using Short
Contact Time Reactors.
2 US 6,579,510 SPOX Enhanced Process for
Production of Synthesis Gas.
3 US 6,800,269 Short Contact Time Catalytic
Sulfur Recovery System for Removing H
2
S from
a Waste Gas Stream.
4 US 6,946,111 Short Contact Time Partial
Oxidation Process for Recovering Sulfur from
an H
2
S Containing Gas stream.
5 US 7,122,170 Catalysts for GT-SPOC™
www.eptq.com PTQ Q1 2014 121
of Heavy Hydrocarbonaceous Feedstock,
European Patent Application, EP 1754770A1,
Research Institute of Petroleum Industry.
8 Bahmani M, Sadighi S, Mohaddecy S R,
Mashayekhi M, Hydrocracker parametric
sensitivity study, PTQ, Q2 2009.
9 Sadighi S, Mohaddecy S R, Ghabouli O,
Rashidzadeh M, Optimisation of product yield
and coke formation in a RFCC unit, PTQ, Q2
2010.
Sepehr Sadighi is Project Manager, Catalysis
and Nanotechnology Division, Catalytic
Reaction Engineering Department, Research
Institute of Petroleum Industry (RIPI), Tehran,
Iran. He holds a PhD in chemical engineering
from Universiti Teknologi Malaysia.
Email: Sadighis @ripi.ir
Reza Seif Mohaddecy is Project Manager,
Catalysis and Nanotechnology Division,
Catalytic Reaction Engineering Department,
RIPI. He holds a MS in chemical engineering
from Sharif University of Technology.
Email: Seifsr @ripi.ir
Kamal Masoudian is Project Manager, Catalysis
and Nanotechnology Division, Catalyst
Characterization and Evaluation Department,
RIPI. He holds a BSc in chemical engineering.
Email: masoudiansk @ripi.ir
References
1 Bhaskar, M, Valavarasu G, Balarman K S,
Advantages of mild hydrocracking FCC feed
- a pilot plant study, Petroleum Science and
Technology, 21, 9 & 10, 1439-1451, 2003.
2 European patent application, EP
1754770A1, Process for Hydroconverting of
a Hydrocarbonaceous Feedstock, Research
Institute of Petroleum and NTI company, 2007.
3 Mohaddecy S R, Zahedi S, Sadighi S, Bonyad
H, Reactor modeling and simulation of
catalytic reforming, Petroleum & Coal, 48 (3),
28-35, 2006.
4 Lee R, Leunenberger E, Powell R, Optimizing
the cat feed hydrotreater/FCCU complex with
detailed simulation tools, Process Technology
Update, Desulphurization Process, Word
Refining, Jul/Aug 2001.
5 Dean R R, Mauleon J L, Combined Fluid
Catalytic Cracking and Hydrocracking Process,
US patent No: 4426276, 17 Jan 1984.
6 Talman J, Jonsgma B, Thamprajmachit B,
Cackett S, Wijk R, Synergistic integration of FCC
and hydroprocessing facilities for upgrading
bottom of the barrel, Asia Pacific Refining
Technology Conference, 2001, Kuala Lumper.
7 Kadiev K M, Mezhidov V K, Zarkesh J,
Masoudian S K, Process for Hydroconverting
60
50
40
30
20
10
D
i
e
s
e
l

i
n
c
r
e
a
s
e
,

%
0
Simple series Series Parallel Residue
upgrading
Integration cases
39.89%
41.27%
37.53%
53.69%
Figure 7 Comparison to base case of growth in diesel production
HCR feed Ni+V, ppmwt Sulphur, wt% IBP, °C FBP, °C
Base 2.98 1.222 309 516.4
Simple series 2.98 1.222 309 516.4
Series 2.98 1.222 309 516.4
Parallel 3.18 1.323 285.5 516.3
Residue upgrading 3.18 1.323 285.5 516.3
Feed quality of HCR unit in base case and integration strategies
Table 6
RFCC feed Ni+V, ppmwt Sulphur, wt% IBP, °C FBP, °C
Base 14.723 1 264.6 592.6
Simple series 14.723 1 264.6 592.6
Series 14.615 1.006 264.6 599.5
Parallel 14.874 0.9667 267.3 592.6
Residue upgrading 13.734 1.088 267.3 592.1
Feed quality of RFCC unit in base case and integration strategies
Table 7
www.eptq.com Revamps 2013 63
processes might increase in the
future, and the fouling of units
might become a more serious
concern, potentially impacting a
large number of production days.
By regular application of online
cleaning, the unit can always be
operated under clean conditions. In
addition, the vast majority of
mechanical work, which takes up a
signifcant number of days during
unit shutdowns, can be replaced
with an online cleaning process that
requires no opening of equipment
or man entry, and can be carried out
in as little as 24 hours oil-to-oil.
Acknowledgement
Special thanks to all the Lotos team for the
support and valuable assistance provided
during the entire time that ITW Online
Cleaning has been applied on site.
Mariusz Hołowacz has been Crude Oil
Distillation Complex Manager with Grupa
Lotos in Gdansk, Poland, since 2008 and, since
2010, Manager of the new distillation unit 120.
Rafał Zaprawa was VDU and propane
deasphalting unit Shift Team Manager with
Grupa Lotos in Gdansk, Poland, from 2001-
2008, then Deputy Manager of the crude oil
distillation complex.
Marcello Ferrara is the Chairman of ITW. With
27 years’ experience in the petroleum business,
including oil exploration and production,
refining, petrochemicals, transportation,
and energy production, he holds a PhD in
industrial chemistry and international patents
for new processes and additive compositions
for environmental control and for improving
petroleum/petrochemical processes.
Email: mferrara@itwtechnologies.com
market considerations and by the
downtime required for a mechani-
cal cleaning turnaround. With a
downtime of 15-20 days, it is much
more economical to run the units
under non-optimised conditions
rather than lose production. This,
however, results in energy losses,
giveaway and capacity reduction,
which negatively impact unit
economics.
The introduction of online clean-
ing, by cleaning the unit in 24
hours on an oil-to-oil basis, allows
for the recovery of losses and the
operation of units under improved
and more reliable conditions.
Furthermore, turnarounds can be
avoided or rescheduled with
reduced downtime.
In the case of turnaround
improvement, an additional
reduction in downtime can be
achieved by applying ITW’s
improved degassing/decontamina-
tion technology.
Conclusion
The results of ITW Online Cleaning
have opened up new possibilities
for Lotos, whereby online cleaning
can be applied during a plant run,
to recover a unit’s performance and
improve the level of operational
excellence, and in preparation for a
turnaround, to reduce downtime
and the turnaround’s scope of work.
These options are increasingly
important, as the amount of oppor-
tunity crudes the company
Figure 7 HGO — crude exchanger
Heavy crude Oil
Atmospheric Distillation
Vacuum Distillation
Coker & Visbreaker Feed
Fluidized Catalytic Cracker
Bitumen






Non-intrusive flow
measurement
up to 400°C
Field-Proven at Refineries
Trouble free operation at
extreme pipe temperatures
No clogging, no abrasion,
no pressure losses
Installation and maintenance
without process interruption
Independent of fluid or pressure
Hazardous area approved





refinery.flexim.com
itw.indd 5 10/09/2013 17:06
ripi.indd 6 11/12/2013 20:18
gas gtc (original).indd 8 11/03/2014 19:06
RELIABLE SWISS QUALITY
API 618
Rod load up to 1'500 kN/335'000 Ibs
Power up to 31'000 kW/42'100 HP
FULL RANGE:
YOU GET MORE THAN JUST A
PROCESS GAS COMPRESSOR
Lubricated up to 1'000 bar, non-
lubricated up to 300 bar

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


Designed for easy maintenance
We are the competent partner
with the full range of services –
worldwide
→ www.recip.com/api618
YOUR BENEFIT:
LOWEST LIFE CYCLE COSTS
gas burckhardt.indd 1 04/03/2014 10:45
Predicting corrosion rates in amine and
sour water systems
C
orrosion is a ubiquitous prob-
lem in gas treating in the
petroleum and natural gas
industries, in syngas plants, in
processing unconventional gases
such as shale and coal seam gas,
and in numerous other treating
applications. The primary impuri-
ties removed in the treating process
are the acid gases carbon dioxide
and hydrogen sulphide. The corro-
sion of equipment and piping is an
inevitable consequence of removing
these very gases with amines, and
of handling sour water. There are
other corrosive impurities that
either enter in small amounts with
the gas, such as HCN and oxygen,
or that are produced in the amine
system itself, mostly heat stable
salts (HSS) derived from HCN.
Corrosion rates are affected by the
nature of the corrosive agent,
temperature, fuid velocity, the
presence of solids, and the metal-
lurgy involved.
1
To prevent
equipment failures, mitigate risk
and select optimal materials, one
must be able to predict corrosion
rates pertinent to the particular
processing conditions. This article
describes the underpinnings of a
chemistry-based predictive corro-
sion model built on both public and
much proprietary corrosion rate
data. The model includes depend-
ence on ionic solution composition
(speciation), fuid velocity, temper-
ature, HSSs, and metallurgy.
Corrosion in alkaline systems
Although the concepts presented
here apply equally to pH-neutral
and acidic systems, these systems
are not addressed because the
amount of corrosion data available
A chemistry-based predictive model predicts corrosion rates in specific
processing conditions
NATHAN A HATCHER, CLAYTON E JONES, G SIMON WEILAND and RALPH H WEILAND
Optimized Gas Treating, Inc.
for modelling is not as extensive.
The corrosive action of H
2
S is inher-
ently different from that of CO
2
in
that H
2
S can and does form a rela-
tively robust, protective iron
sulphide layer on the metal surface.
On the other hand, iron carbonate
forms a more fragile layer, so it
offers much less protection. There
are several tenets embedded in the
model:
• The corrosive agents are acids
• In and of itself, the amine (or
ammonia) is not corrosive
• The iron sulphide flm can
protect against further corrosion
• Iron carbonate also offers protec-
tion but to a lesser degree
• High fuid velocities physically
increase corrosion rates
• Higher temperature increases
corrosion rates
• Heat stable salts chemically exac-
erbate corrosion.
Although technically incorrect,
the industry continues to bandy
about such terms as ‘amine corro-
sion’ and ‘alkaline stress corrosion
cracking’ to describe corrosion that,
at the root level of chemistry, is
really caused by dissolved acid
gases in various forms. For exam-
ple, nearly 60 years ago,
Polderman
2
reported that 20 wt%
MEA without acid gas was actually
less corrosive to steel than pure
water. As far as the corrosive
agents themselves are concerned,
the important parameter is the
chemical activity of the dissolved
acid gas species responsible for
corrosion. The activity (vs concen-
tration) changes with the amine
type, amine concentration, acid gas
loadings, the concentrations and
identities of HSSs, and temperature.
This may make some amine
systems appear to be more suscep-
tible to corrosion than others;
however, the essential point is that
it is the activity of the corrosive
species that is of direct importance,
not the type of amine per se.
The chemical species of interest
are: bisulphide ion (HS

), free phys-
ically dissolved H
2
S, bicarbonate
ion (HCO
3

), and free physically
dissolved CO
2
, all of which are
oxidising agents. These species are
called protonic acids because they
can give up a hydrogen ion.
3

Sulphide (S
=
) and carbonate (CO
3
=
)
ions are also present; however, they
themselves are fnal reaction prod-
ucts and are unable to provide the
hydrogen ion necessary for the
oxidation of iron. Molecular hydro-
gen sulphide and carbon dioxide
react with iron only in the presence
of water. The fnal distribution of
molecular and ionic species is
found by solving the equations of
chemical reaction equilibria, atom
balances, and a charge balance. The
resulting set of species concentra-
tions is termed the solution’s
speciation.
www.eptq.com Gas 2014 25
Corrosion rates are
affected by the nature
of the corrosive agent,
temperature, fluid
velocity, the presence
of solids, and the
metallurgy involved
gas ogrt.indd 1 06/03/2014 13:37
26 Gas 2014 www.eptq.com
test, CO
2
was added to an amine
solution that already contained H
2
S.
The corrosion rate quickly rose and
in less than one hour a new higher
steady corrosion rate was estab-
lished. The precipitated layer of
iron carbonate is very fragile, so it
is unlikely to adhere as tenaciously
to metal surfaces as iron sulphide
does. Different abilities to adhere
are contained within the framework
of the present model.
Managing corrosion begins
primarily with protecting the iron
sulphide flm on the metal. Fluid
velocity creates sheer stress on
solid surfaces — the shear stress
can be enough to rip off the imper-
fect iron sulphide layer, and even a
low velocity seems very likely to
dislodge and remove iron carbonate
deposits. When this happens, fresh
iron is exposed and corrosion
increases in the region of high
shear stress. Subsequently, the
dislodged iron sulphide and iron
carbonate particles can act as scour-
ing agents and increase corrosion in
other areas of the plant via that
mechanism. For these reasons, the
gas treating industry has generally
adopted velocity limits for carbon
steel piping in amine service. Lean
amine velocities are typically held
below 7–10 ft/s while rich amine
velocities are commonly held below
3–5 ft/s.
Even a little fashing can greatly
increase the velocity in a pipe, and
the high velocity fow from the
passing biphase scours the iron
sulphide layer. This results in dras-
tically shorter piping life. Although
less accurately than for single-phase
fows, it is still possible to provide
reasonable estimates of corrosion
rates in two-phase fows, and this is
done in the present model.
Heat stable salts (HSS) affect
corrosion rates by altering solution
speciation, and they chemically
exacerbate corrosion by chelating
iron and destroying the passivating
iron sulphide layer:

FeS + 7 H
2
0 ↔ Fe(H
2
0)
6
+2
+ HS

+ OH

Fe(H
2
0)
6
+2
+

n HCOO

↔ + Fe(n HCOO

)
(2–n)
+
(6 – n)H
2
0
There are very few quantitative
measurements of the effect of HSSs
In their simplest stoichiometric
forms, the basic corrosion reactions
of dissolved H
2
S species with iron
are:

H
2
S(aq) + Fe(s) → FeS(s) + H
2
(g)
2HS
-
(aq) + Fe(s) → FeS(s) + H
2
(g) + S
=
(aq)

For CO
2
, the relevant reactions
are:

CO
2
(aq) + Fe(s) + H
2
O → FeCO
3
(s) + H
2
(g)
2HCO
3
-
(aq) + 2Fe(s) → 2FeCO
3
(s) + H
2
(g)
The oxidation reaction with
hydrogen sulphide is faster than the
reaction with bisulphide; however,
the alkalinity of the amine (and
ammonia) solutions means that the
dissolved H
2
S is predominantly in
the bisulphide form, with very little
remaining as free molecular hydro-
gen sulphide. This is also true of
dissolved carbon dioxide. The
concentrations of free H
2
S and CO
2

are pH dependent and pH is a func-
tion of amine strength, total
dissolved acid gas, temperature,
and to a lesser extent HSS concen-
trations. However, heat stable salt
species and their concentrations do
affect the speciation of the solution,
especially in lean solvents.
As discussed by Cummings et al
1
,
the sequence of physico-chemical
steps in the process of oxidising
iron consists of transporting the
acid from the bulk solution to the
metal surface, adsorption of the
acid onto the surface, reaction with
iron, and transport of reaction prod-
ucts back into the bulk solution. The
steps are similar to what occurs in
heterogeneous catalysis. The reac-
tion of H
2
S with the iron component
of various iron-based metallurgies
forms solid iron sulphide and
hydrogen gas, and as the reaction
proceeds, the surface of the iron is
changed to a mosaic of iron and
sulphide ions. The surface expands
by addition of sulphide, and the
liberation of hydrogen gas exacer-
bates the expansion. The surface
layer is somewhat porous, and it
adheres to the surface of the free
metal. The iron-H
2
S and iron-HS


reactions form reaction products in
completely different phases, as do
the reactions with carbon dioxide
and bicarbonate. Thus, because
reaction products are continually
removed from the reacting solution,
it follows from Le Chatelier’s princi-
ple that there is a strong
thermodynamic driving force
powering continued corrosion.
What limits the corrosion reac-
tions is primarily the amount of
bare, unreacted iron that the passi-
vating flm leaves available at the
metal surface. A secondary factor is
the concentrations of dissolved
reactant gases, H
2
S and CO
2
, which
are other parameters. Thus, H
2
S,
HS

, CO
2
and HCO
3

all react with
unprotected iron. To control the
concentration of dissolved gas in
their various forms, most practi-
tioners adopt a rich amine acid gas
loading, upper limit of <0.4 to 0.5
moles of total acid gas per mole of
molecular amine.
Passivation
In a new amine unit or in one that
has just been cleaned, during the
period immediately following intro-
duction of H
2
S, for example, the
corrosion rate climbs exponentially.
Measurements by Cummings et al.
3

using a bench-scale mini-amine
plant with relatively lean amine
indicated in some cases over 600
mpy corrosion rate during this
phase of operation. Eventually, the
corrosion rate levels off and begins
to fall as the iron sulphide layer
increasingly occludes contact
between H
2
S and iron (passivation).
Finally, once the iron sulphide layer
is established, the system settles
down to a nominal (and hopefully
low) residual corrosion rate. The
time over which the fnal steady
corrosion rate is attained is on the
order of hours or even days. During
these three phases of passivation,
the solution frst becomes nearly
black, and then it changes to dark
green, lighter green, and fnally
slightly amber. Chelated iron
(which results from complexation)
scatters light and turns otherwise
contaminant-free solution to amber.
The coloration depends on the size
of the iron sulphide particles. The
results of experiments on the time
dependence of iron passivation by
carbon dioxide are sparse. In one
gas ogrt.indd 2 06/03/2014 13:37
Linde AG
Engineering Division, Dr.-Carl-von-Linde-Strasse 6–14, 82049 Pullach, Germany
Phone +49.89.7445-0, Fax +49.89.7445-4908, naturalgas@linde-le.com, www.linde-engineering.com
With Linde, customers rely on long time expertise in the supply
of individually designed LNG plants. To deliver a high level of
quality at low costs and in shorter time, Linde has translated
the idea of product standardization into the LNG business,
and developed the StarLNG™ standardisation concept for a
capacity range between 100 and 1,500 tons per day.
The idea of StarLNG is to standardise and optimise a small-to-
mid scale LNG plant based on a wide set of process variations.
This “Process Toolbox” is designed to cover about 90 percent
of real-life parameters.
With this innovation, customers benefit from a scalable,
flexible LNG plant design including many alternatives, based
on pre-engineered documents and a 3D CAD-model for a fully
modularized plant. StarLNG combines the highest safety
standards, reliability, and efficiency, together with
competitive CAPEX and OPEX and shortest delivery time.
StarLNG™
The leading small-to-mid scale LNG plant.
gas linde.indd 1 04/03/2014 10:42
on corrosion and, although data on
certain HSSs are included in the
current model, more data over a
broader range of species and condi-
tions would be very welcome.
Model
Our mechanistic, chemistry-based
corrosion rate model was
28 Gas 2014 www.eptq.com
developed on the basis of the
sequence of steps outlined earlier,
namely, diffusion of the corrosive
agent to the surface, adsorption,
reaction with iron, and either depo-
sition or diffusion of products away
from the surface. The parameters in
the model have been regressed to a
large number of corrosion
measurements taken in well-con-
trolled bench scale fow
apparatuses and static equipment
using both H
2
S and CO
2
rich amine
systems and sour water. Although
the details of the quantitative
model are proprietary to Optimized
Gas Treating, Inc., the corrosion
rate in the model is a function of
the important factors:
Corrosion rate = f(a
H
2
S
,a
HS
_, a
CO
2
, a
HCO
3
_
, v, T)

Here: a
H
2
S
= activity of dissolved H
2
S
a
HS
_ = activity of bisulphide ion in
solution
a
CO
2
= activity of dissolved CO
2
a
HCO
3
_
= activity of bicarbonate ion in
solution
T = absolute temperature
v = superfcial fuid velocity in
pipe.
The model contains many of the
elements that comprise the tenets
discussed earlier and in a form
consistent with the various steps
occurring during corrosion, includ-
ing adsorption isotherms and the
temperature dependence of both
reaction kinetics and turbulent
diffusion. The model even includes
corrosion by water alone as a limit-
ing case in highly dilute solutions.
Data for corrosion by H
2
S were
correlated over the range of condi-
tions shown in Table 1. Although
scatter in the data is sizeable, if for
no other reason than that repeata-
ble corrosion rate measurements
are hard to make, especially by
weighing coupons, Figure 1 shows
that this model fts the measure-
ments over nearly three orders of
magnitude. Data for corrosion by
CO
2
were taken in both a static cell
and in a fow apparatus. Parameter
ranges and a parity plot comparing
the model with measurements of
corrosion in CO
2
-only amine
systems are shown in Table 2 and
Figure 2, respectively.
In correlating the model,
Arrhenius kinetics was found to ft
the data well. This is in agreement
with the proposal of Cummings et
al.
3
These authors assumed a power
law kinetics expression in terms of
molecular and ionic concentrations;
however, we found a better ft
using activities rather than concen-
100
1000
10
1 M
o
d
e
l

c
o
r
r
o
s
i
o
n

r
a
t
e
,

M
P
Y
0.1
0 1 10 100 1000
Measured corrosion rate, MPY
Rich amine
Sour water
Figure 1 Parity plot of measured vs model corrosion rates in rich amine and sour water
systems containing H
2
S only
100
1000
10
1 M
o
d
e
l

c
o
r
r
o
s
i
o
n

r
a
t
e
,

M
P
Y
0.1
0 1 10 100 1000
Measured corrosion rate, MPY
Figure 2 Parity plot of measured vs model corrosion rates in amines containing CO
2
only
gas ogrt.indd 3 12/03/2014 12:58
www.eptq.com Gas 2014 29
3 Cummings A L, Veatch F C, Keller A E,
Corrosion and corrosion control methods in
amine systems containing H
2
S, NACE Corrosion
1997 Conference, Paper No, 341.
Nathan A Hatcher joined Optimized Gas
Treating, Buda, Texas, as Vice-President,
Technology Development, in 2009. He holds a
BS in chemical engineering from the University
of Kansas and is currently a member of the
Amine Best Practices Group.
Email: nate.hatcher@ogtrt.com
Clayton E Jones joined Optimized Gas Treating,
Inc as a Software Development Engineer in
2012. He holds a BS in chemical engineering
from McNeese State University and a MS in
chemical engineering from the University of
New Mexico.
G Simon Weiland recently joined Optimized
Gas Treating as a Software Development
Engineer. He will receive a BS in chemical
engineering from the University of Oklahoma
in May 2014.
Ralph H Weiland is a co-founder of Optimized
Gas Treating with offices in Clarita, OK,
Houston, TX and Buda, TX. He holds BASc,
MASc and PhD degrees in chemical engineering
from the University of Toronto.
Email: ralph.weiland@ogtrt.com
underpinning lends it not only to
confdent interpolation, but also to
reliable extrapolation to conditions
well outside measured ranges.
References
1 Cummings A L, Waite S W, Nelsen D K,
Corrosion and corrosion enhancers in amine
systems, Brimstone Sulfur Conference, Banff,
Alberta, April 2005.
2 Polderman L D, Dillon C P, Steel A B, Oil &
Gas J., 16 May 1955, 180.
trations. Note that neither amine
strength, nor amine type, nor
ammonia content appears explic-
itly, although these parameters
have signifcant effects on the activ-
ity of the corrosive species in
solution because speciation
depends on the amine type, its
concentration, and the temperature.
Corrosion of other metallurgies is
included in the model’s capabilities,
as is the corrosion of common pipe
fttings. The model’s mechanistic
Parameter Range
Lab velocity 0–25 m/s
Temperature 55–120°C
Sour water 1–30 wt% NH
4
HS equivalent
3.5–10 bar H
2
S partial pressure
Amines MEA (18–30 mass %)
DEA (30 mass %)
MDEA (45 mass %)
0.1–0.8 mole H
2
S/mole amine
Calculated pH 6.0–9.5
Parameter ranges for corrosion by H
2
S
Table 1
Parameter Range
Lab velocity 0–13.5 m/s
Temperature 20–160°C
Amines MEA (6–30 mass %)
DEA (10–40 mass %)
AMP (9–35 mass %)
MDEA (35 mass %)
0.0–0.4 mole CO
2
/mole amine
Calculated pH 8.4–11.36
Parameter ranges for corrosion by CO
2
Table 2
www.eptq.com PTQ Q1 2014 91
Ivelina Shishkova is R&D Department
Manager with Lukoil Neftohim Burgas. She
holds a MS in organic chemistry engineering
and a PhD in petroleum refining from Sofia
Chemical and Technological and Metallurgical
University, and has authored more than 20
technical papers.
Email: Shishkova.Ivelina.K@neftochim.bg
Rosen Dinkov is the Quality Manager in the
Process Engineering department of Lukoil
Neftohim Burgas. His research interests include
crude oil characterisation, bio/conventional
fuels blends characterisation and modelling of
refinery distillation processes. He holds a MS
in organic chemistry engineering from Burgas
University and a PhD in the technology of
fossil and synthetic fuels from the University
of Chemical Technology and Metallurgy, Sofia.
Email: Dinkov.Rosen.K@neftochim.bg
Vladimir Jegorov is the Sales Development
Manager for Grace in the CIS region. Prior to
joining Grace, he was an FCC process engineer
at the Mazheikiai refinery in Lithuania.
Petko Petkov is a full professor and rector
of the Burgas University ‘Assen Zlatarov’. He
teaches in the social science department in
the field of oil refining and lubricants, and has
authored more than 180 scientific papers and
five books.
Email: PST_Petkov@abv.bg
higher FCC C
4
yield led to higher
production of alkylate, which
resulted in the production of 2%
more premium grade gasoline.
References
1 Watanbe K, Nagai K, Aratani N, Saka Y,
Chiyoda N, Mizutani H, Techniques for octane
enhancement in FCC gasoline, 20th Annual
Saudi-Japan Symposium, Dhahran, December
2010. 17. Montgomery J A, Guide to Fluid
Catalytic Cracking, Part 1, 1993.
Ivan Chavdarov is a Chemical Engineer in
the Process Engineering department of Lukoil
Neftohim Burgas, Bulgaria. His activities are
focused on guiding the operation of the units
of the FCC complex, troubleshooting support
and optimisation of the performance of the
FCC complex.
Email: Chavdarov.Ivan.S@neftochim.bg
Dicho Stratiev is Chief Process Engineer with
Lukoil Neftohim Burgas. He holds a MS in
organic chemistry engineering, and a PhD and
a DSc in petroleum refining from the Burgas
University ‘Assen Zlatarov’. He has authored
more than 130 papers.
Email: Stratiev.Dicho@neftochim.bg
Regular A-92
Premium A-95
Super A-98
RVP = 60kPa
1%
28%
71%
RVP = 50kPa
1%
26%
73%
Figure 3 Effect of changing the RVP on
refinery gasoline grades produced during
the Resolution catalyst period
MINIDIS ADXpert
Grabner Instruments Messtechnik GmbH | A-1220 Vienna, Austria | Dr. Otto-Neurath-Gasse 1
Phone +43-1-282 16 27-0 | Fax +43-1-280 73 34 | info.grabner-instruments@ametek.at
QUICK, ACCURATE, PORTABLE
DISTILLATION
Portable, True Atmospheric D7344 Distillation •
Excellent D86, ISO 3405, IP 123 Correlation •
Fully Automated Measuring and Cleaning •
6 mL of Sample •
No Glassware, no Open Flame •
Distillation of Gasoline, Jet Fuel, Diesel, Solvents, •
Unknown Samples, Raffinates and Biodiesel
Specified for Use by US Pipelines •
ADXpert half page 1.indd 1 06.12.2013 11:46:37
grace.indd 4 11/12/2013 12:49
gas ogrt.indd 4 11/03/2014 12:32
enersul.indd 1 10/03/2014 11:01
Zeolite based dryers in ethylene plants
S
team cracking represents more
than 95% of ethylene produc-
tion. This article focuses on
product stream drying duties in
ethylene production based on steam
cracking of paraffnic hydrocarbons.
Steam cracker feeds can be sepa-
rated into two categories:
• Natural gas liquids (NGLs):
ethane, propane and butane
• Heavier hydrocarbons such as
naphtha and gasoils.
North America (>5%) and the
A review of the role and nature of adsorbent dryers in the production of ethylene
by steam cracking
VASSILIOS ZAFIRAKIS and HANS HOEFER
Grace Materials Technologies
Middle East (>80%) use gaseous
feedstock to produce ethylene,
whereas in Europe (>80%), in Asia-
Pacifc (>81%) and in Latin America
(>65%) naphtha and gasoil feeds
dominate. On the world scale,
ethylene feedstocks are as follows:
• Naphtha 52%
• Gasoil 6%
• Ethane 28%
• Propane 8%
• Butane 3%
• Other 3%
Process sections of an ethylene
plant
In general, ethylene production has
four major sections:
1. Cracking heater section including
transfer line heat exchangers
(TLE)
2. Gasoline fractionator and quench
water tower
3. Compression section with both
acid gas removal and a main
drying unit
4. Chilling train and separation
www.eptq.com Gas 2014 31
Compression
Compression
Methanising
Deethanising
Depropanising
C
2
H
2

hydrogenation
C
3
H
4

hydrogenation
Demethanising
Acid gas
removal
Drying
PSA
Cracking
C
2
-cut
drying
C
3
-cut
drying
H
2
drying
Water
quench
Chilling
Ethylene
fractionation
Propene
fractionation
Feedstock
liquid
High
purity
hydrogen
Medium
purity
hydrogen
Methane
Ethylene
Propene
TLE and oil
quenching
Primary
fractionation
Pyrolysis
fuel oil
Debutanising
C
4
-cut
Pyrolysis
gasoline
Figure 1 Simplified process flow diagram for ethylene production by liquid cracking with a front-end demethaniser. In green boxes:
zeolite based adsorbers.
gas grace.indd 1 11/03/2014 09:30
32 Gas 2014 www.eptq.com
550-650°C. The lowest temperature
refers to the start of the cracking
cycle (SOR), while the highest
temperature represents the maxi-
mum temperature, where the coil
and the TLE need to be cleaned
(EOR) and the heater has to be shut
down.
Gasoline fractionator and quench
water tower
One of the major differences
between liquid and gas crackers is
the use of a gasoline fractionator.
Liquid crackers require gasoline or
primary fractionators. This is not
necessary for crackers using ethane
or propane as feedstock (see Figure
2).
Furthermore:
• For liquid feedstock downstream
of the TLEs, the required cooling to
temperatures of 200-230°C cannot
be performed in standard shell and
tube heat exchangers; condensation
of a signifcant portion of liquids in
the cracked gas leads to excessive
fouling by carbon deposits.
section for cracked gas separation
into the desired products.
A variety of commercial process
routes is available. Whereas crack-
ing heaters, quench towers and
compression sections are similarly
confgured (sections 1-3 above), for
separation (section 4) three main
process routes are established. They
are characterised by the frst sepa-
ration tower and the position of the
unit for hydrogenation of C
2
acety-
lenes in the cracked gas. These
three process sequences are:
1. Demethaniser frst with tail-end
hydrogenation
2. Deethaniser frst with front end
hydrogenation
3. Depropaniser frst with front end
hydrogenation.

Cracking heater section and TLEs
The ethylene industry is served by
a limited number of licensors and
contractors with their own ethylene
process technologies. They have
designed and built most of the
world’s ethylene plants. In the
process route shown in Figures 1, 2,
3 and 4 hydrocarbon feedstock
(ethane, propane, butane, naphtha
and gasoil) is fed into a pyrolysis
heater (furnace) and preheated to
temperatures of 500-650°C. Steam
enters the heater coil and the
hydrocarbon/steam mixture is
further heated to temperatures of
750-890°C. Within this range, satu-
rated hydrocarbon molecules crack
to form olefns such as ethylene,
propylene and butene, and diole-
fns, as well as other hydrocarbon
molecules such as methane,
benzene, toluene, and hydrogen
and other products. After this
pyrolysis reaction the produced gas
from the heater coils is rapidly
cooled in tubular heat exchangers
(transfer line exchangers, TLE)
against boiler feed water, produc-
ing high pressure steam (HPS). In
the case of gaseous feeds, the gas
leaves the TLE at a temperature of
300-400°C. For liquid feeds such as
light to heavy naphtha, tempera-
tures are 420-450°C, and for gasoil
Compression
Compression
Methanising
Deethanising
Depropanising
C
2
H
2

hydrogenation
C
3
H
4

hydrogenation
Demethanising
TLE
Acid gas
removal
Drying
Drying
PSA
Cracking
C
2
-cut
drying
C
3
-cut
drying
H
2
drying
Water
quench
Chilling
Ethylene
fractionation
Propene
fractionation
Feedstock
gas
Gasoline
High
purity
hydrogen
Medium
purity
hydrogen
Methane
Ethylene
Propene
C
4
+
hydrocarbons
Figure 2 Simplified process flow diagram for ethylene production by gas cracking with a front-end demethaniser. In green boxes: zeolite
based adsorbers
gas grace.indd 2 11/03/2014 09:43
Therefore, the cracked gas is further
cooled by injection of oil (oil
quench) by means of devices called
quench fttings. Oil quenching is
performed downstream of the TLEs
after each heater, or in the
combined cracked gas line (transfer
line) from all heaters. The mixed
two-phase fow of cracked gas and
quench oil is separated in the
bottom section of the primary frac-
tionator. In this tower the cracked
gas is separated from the fuel oil,
which leaves the tower from the
bottom. The overhead is directed to
the quench water tower
• For gaseous feedstock, cracked
gas from the TLEs is further cooled
to about 200°C against feed or
boiler feed water (BFW) in the
so-called secondary TLE, before
entering the quench water tower
where the cracked gas is further
cooled.
In a liquids cracking plant, this
tower essentially works as a partial
condenser to the upstream oil frac-
tionators, which condense both the
www.eptq.com Gas 2014 33
steam and a signifcant part of the
gasoline components. Part of the
gasoline is used as refux for the
upstream oil fractionator. The
remaining gasoline leaves the unit
after stripping in a side tower for
stabilisation.
The cracked gas leaves the top of
the quench tower at a temperature
of 35-40°C, slightly above atmos-
pheric pressure and enters the
compression section of the plant.
Compressor section, acid gas
removal and drying of cracked gas
The cracked gas is compressed in
four to fve stages to pressures of
35-37 bar. Usually, compression of
the charge gas from gaseous feed-
stock cracking takes place in
centrifugal compressors employing
four stages, while fve compressor
stages are required for compression
of cracked gas from the cracking of
liquid feedstocks such as naphtha
or gasoil. Consequently, in plants
processing gaseous feedstock a
higher compression ratio is
employed, yielding a higher
discharge temperature compared to
liquid based plants. The number of
stages depends primarily on gas
composition and the highest
temperature allowed for the
discharge of the different compres-
sion stages. Such a temperature is
typically below 100°C in order to
avoid excessive equipment fouling
through polymerisation of diolefns
and other precursors in the cracked
gas. As the pressure increases at
the different compressor stages,
water and hydrocarbons are
condensed in the downstream cool-
ers and separated from the gas in
the interstage separators.
Water is returned to the quench
water tower while the hydrocar-
bons from the frst three stages are
recycled to the gasoline stripper in
the upstream hot section of the
plant. The acid gas removal system
is usually positioned after the third
or fourth compression stage.
Cracked gas leaving the acid gas
removal unit has a remaining acid
Compression
Compression
Methanising
Deethanising
Debutanising Depropanising
C
2
H
2

hydrogenation
C
3
H
4

hydrogenation
Chilling and
demethanising
TLE and oil
quenching
Acid gas
removal
Primary
fractionation
Drying
Drying
PSA
Cracking
C
2
-cut
drying
C
3
-cut
drying
H
2
drying
Water
quench
Chilling
Ethylene
fractionation
Propene
fractionation
Feedstock
liquid
Pyrolysis
fuel oil
High
purity
hydrogen
Medium
purity
hydrogen
Methane
Ethylene
Propene
C
4
-cut
Pyrolysis
gasoline
Figure 3 Simplified process flow diagram for ethylene production by liquid cracking with a front-end deethaniser. In green boxes: zeolite
based adsorbers
gas grace.indd 3 11/03/2014 09:44
34 Gas 2014 www.eptq.com
purity is withdrawn from the
lowest temperature separator; this
can be further purifed in a down-
stream system. The condensates
produced over several stages are
fed to the demethaniser (see Figure
1). In recent plants, mixed refriger-
ants are used instead of cascade
refrigeration systems.
The process diagrams in Figures 3
and 4 show simplifed process
routes for liquid cracking plants
using a separation sequence of
deethaniser frst (see Figure 3) or
depropaniser frst (see Figure 4).
2,5
The separation section
The layout of the separation section
depends on the selected separation
sequence, the feedstock and the
targeted products. The following
represents the demethaniser frst
concept (see Figure 1):
• Demethaniser
• Deethaniser, acetylene converter,
acetylene converter dryer
• Ethylene fractionator (splitter)
ment. Therefore the condensate
always contains traces of free
water. This needs to be considered
in the design of downstream
condensate dryers. Complete
removal of water is mandatory for
the charged gas or liquid streams to
be cooled to temperatures below
15°C in the downstream chilling
train and the separation section, to
avoid the plugging of lines and
equipment due to formation of
hydrates and ice. For the removal
of moisture from these streams, the
most current technology uses solid
alumino-silicates (zeolitic molecular
sieves) which will be discussed
later. Previously, dual systems
(glycol followed by alumina) were
employed.
After drying, the cracked gas is
partially condensed in several
stages of a cascade refrigeration
system (propylene, ethylene, meth-
ane), resulting in cold product
streams at temperatures of -135°C
to -145°C. Hydrogen of 95% mole
gas content of about 1 ppmv.
Depending on the technology
selected downstream of the acid
gas removal system, one or two
additional compression stages may
be employed. Additional C
2
- and C
3

rich condensate is produced in
these compression stages. This is
fed into the stripper, where the C
2
and lighter fractions are forced
through the tower top back to a
compression section operating at
lower pressure. The bottoms of this
tower are sent to the downstream
depropaniser. If the operating
temperature in the depropaniser is
lower than 15°C, the bottom stream
has to be dried to avoid the forma-
tion of hydrates.
The cracked gas is saturated with
water before compression as well
as after each intercooler stage and
down-stream of the acid gas
removal system. The condensate is
also saturated with water, while
excess water separation depends on
the design of the separator equip-
Compression
Compression
Methanising
Deethanising Depropaniser
C
2
H
2
/ C
3
H
4

hydrogenation
C
3
H
4

hydrogenation
Chilling and
demethanising
Acid gas
removal
PSA
Cracking
C
2
-cut
drying
C
3
-cut
drying
H
2
drying
Water
quench
Ethylene
fractionation
Propene
fractionation
Feedstock
liquid
High
purity
hydrogen
Medium
purity
hydrogen
Methane
Ethylene
Propene
TLE and oil
quenching
Primary
fractionation
Pyrolysis
fuel oil
Debutanising
C
4
-cut
Pyrolysis
gasoline
Drying
Chilling
Figure 4 Simplified process flow diagram for ethylene production by liquid cracking with a front-end depropaniser. In green boxes:
zeolite based adsorbers
gas grace.indd 4 11/03/2014 09:44
www.eptq.com Gas 2014 35
provider may minimise by chemi-
cal modification of the selected
adsorbent.
Kinetics or, better, mass transfer
kinetics, is the overall term related
to intra-particle mass transfer char-
acteristics causing breakthrough
behaviour as well as lost bed mass.
Lost bed mass is that part of the
adsorber bed which is still active
even though the breakthrough has
just started. Fast kinetics provides a
sharp breakthrough curve with a
small lost bed mass, while slow
to be adsorbed (the adsorbate). In
practice, one has to distinguish
between equilibrium capacity,
given through isotherms, and the
typically lower overall dynamic
capacity, which is characterised by
specific operating conditions.
Selectivity is the ratio of the
amount of purposely adsorbed
component to that of all others in a
fluid to be processed. In reality,
selectivity is never 100% for the
targeted adsorbate due to co-
adsorption, which the adsorbent
• Depropaniser, methyl acetylene
and propadiene (MAPD) converter,
MAPD converter dryers
• Propylene fractionator (splitter)
• Debutaniser.
The purpose of the demethaniser
is the separation of methane and
other lights (CO, H
2
) from C
2
+
components which are sent to the
deethaniser. This tower separates
C
2
components (acetylene, ethylene,
ethane) from heavier hydrocarbons.
C
2
hydrocarbons leave the deetha-
niser as the overhead product,
while the C
3
and heavier compo-
nents, the bottoms, are directed to
the depropaniser. The overhead of
the deethaniser is heated and
hydrogenated in order to convert
the acetylenes into ethylene and
ethane. Because traces of water
may be present during this hydro-
genation, drying of the C
2
hydrocarbons is necessary before
they enter the ethylene fractionator,
typically operating at -10°C to
-35°C and 16-18 bar. The ethylene
product is the overhead of the
ethylene fractionator. The bottom
product of the ethylene fractionator
is ethane, which is recycled back to
the heaters for cracking to extinc-
tion. In the depropaniser, C
3
+
hydrocarbons are separated into
two fractions. The C
3
fraction as
overhead and the heavier C
4
frac-
tion leaving the tower as bottom
product are sent to the debutaniser.
Due to the very low C
3
+ fraction in
ethane based crackers, C
3
and C
4

separators are not present.
Zeolitic adsorbents in
ethylene plants
Any adsorbent has basically seven
features, six which qualify such
material for specific adsorption
duties, namely:
• Capacity
• Selectivity
• Kinetics
• Process compatibility
• Regenerability
• Mechanical stability (crush
strength, attrition)
• Cost.
Adsorption capacity is the
amount of adsorbate related to
adsorbent mass, which is a function
of operating temperature and the
partial pressure of the component
SweetSulf
TM
Sprex
®
AdvaSulf
TM
AdvAmine
TM
COSWEET
TM
HySWEET
®
A UNIQUE TASTE OF SWEET FOR YOUR GAS
50 YEARS OF EXPERIENCE IN GAS SWEETENING AND
SULPHUR RECOVERY PROCESSES
With its unique and complete proprietary technologies portfolio,
PROSERNAT offers optimized solutions to bring on specs any
type of gas contaminated with CO
2
, H
2
S, COS and organic
sulphur species, while producing sulphur with the most stringent
emissions standards.
www.prosernat.com
gas grace.indd 5 11/03/2014 12:33
36 Gas 2014 www.eptq.com
increasing load, and measuring the
required crushing force. The results
vary greatly, depending on the
samples’ geometries and shapes.
Comparative measurements with
this method are meaningful only
for samples with the same geomet-
rical shape and the same size
distribution for beads, and the
same D and L/D ratio for extru-
dates. Direct comparison of beads
and extrudates is less useful due to
different types of force applied to
the samples, for instance point vs
linear or even area-shaped. A more
realistic comparison divides the
measured force by the length of the
extrudate (N/mm), as recom-
mended by ASTM.
Bulk crush strength is commonly
measured by placing a representa-
tive sample of beads or extrudates
in a defned cylinder, applying a
force with a piston, and measuring
the amount of fnes through a 420µ
mesh. Bulk crush strength is
expressed as the pressure (MPa =
N/mm
2
) applied to generate 1%
fnes. This method best represents
the real conditions in a fxed bed
adsorber and is increasingly used.
Attrition and abrasion tests give
information about the expected
behaviour of an adsorbent during
processes such as packaging, trans-
portation, loading, as well as during
operation in the fxed bed. Bulk
crush strength together with attri-
tion tests are of great importance.
They simulate dust formation by
attrition due to mechanical parti-
cle-particle interaction in a fxed
bed. Adsorbents with high bulk
crush strength and low attrition and
abrasion can be produced by apply-
ing sophisticated beading or
pelletising procedures during the
adsorbent manufacturing process.
When comparing materials from
different suppliers, it is important
to ensure that the measuring
method is the same, otherwise the
results are misleading.
The cost of the adsorbent
comprises its purchase and unit
operating costs. Selectivity, kinetics,
compatibility and regenerability
directly impact the performance
and lifetime summarised in the
productivity and fnancial feasibil-
ity of any gas purifcation process.
with a boiling point range of
120-400°C. The adsorbent provider
has to reduce the concentration of
such active sites, or deactivate them
by masking with chemical
additives.
Regenerability ensures adsorption
and desorption in sequential cycles
with uniform performance. The
adsorbate must be reversibly
physisorbed. Irreversible side reac-
tions and the accumulation of such
reaction products in the pore
system of the zeolite, as well as the
deterioration of the working bed
capacity, are often described as
adsorber bed ageing. Reducing
working bed capacity goes with
increased operating costs due to
shorter cycle times, higher amounts
of regeneration gas and increased
pressure drop, fnally resulting in a
changeout.
Another issue is the mechanical
stability of the adsorbent in the
fxed bed. Commercial adsorbents
need to have adequate mechanical
strength so that losses during
handling and processing are mini-
mised. Three measured parameters
— single piece crush strength
(ASTM D-4179, ASTM D-6175),
bulk crush strength (ASTM 7084-
04) and attrition/abrasion (ASTM
4058) — describe expected mechan-
ical behaviour. Single piece crush
strength is measured by placing
one adsorbent body, bead or extru-
date between two plates, applying
kinetics leads to an extended curve
and a large lost bed mass. The
effect of an extended breakthrough
curve can be counteracted either by
using more adsorbent or by accept-
ing reduced throughput per given
adsorption cycle. A third option is
the use of smaller adsorbent bodies
offering a larger overall adsorbent
bed surface. As a rule of thumb,
one can say that the mass transfer
zone of 1.6 mm to 3 mm beads is
just a third that of 2.5 mm to 5 mm
beads. The smaller beads, however,
lead to a higher pressure drop.
Pressure drop and mass transfer
zone length can be optimised by
using a split bed. In typical bed
designs, smaller beads make up the
mass transfer zone while the larger
beads form the equilibrium zone.
Process compatibility describes a
number of effects during the
adsorption/desorption process
which may reduce the lifetime of
the adsorber bed. In an ethylene
plant, the charge gas contains very
reactive components such as acety-
lenes, olefns, diolefns and styrene.
They lead to oligomerisation/
polymerisation and fnally to
coking during regeneration. The
carbon rich mass from heavier
hydrocarbons formed at catalyti-
cally active sites of the desiccant
body is also known as ‘green oil’.
This represents the most critical
compatibility issue. Green oil is a
mixture of mainly C
4-20
oligomers
4 A type Zeolite 3 A type Zeolite
Atoms not shown to scale
Ion
exchange
Na
+
Na
+
Na
+
Na
+
Na
+
K
+
K
+
K
+
Aluminium
Silicon
Oxygen
Cations
Na
+
K
+
Figure 5 Schematic representation of the Na
+
/K
+
ion exchange of an A-type zeolite struc-
ture with one sodalith cage and eight truncated octahedra at its corners
gas grace.indd 6 11/03/2014 17:09
MAN Diesel & Turbo has unique compression solutions in its portfolio for the complete range of CO
2
, N
2
,
propylene and vapour related applications, with single machines or complete train solutions. Others talk
about CO
2
compression. We have the credentials! Over 200 RG Gas compressor units are in operation
or on order. We are world`s number one in CO
2
high pressure applications, thanks to sophisticated test-
ing facilities and proven track records in the field. With an RG Gas compressor from MAN Diesel & Turbo
you will gain an exceptionally flexible and optimized solution to maintain your business economic.
Find out more at www.mandieselturbo.com
Marine Engines & Systems Power Plants Turbomachinery After Sales
Actions speak louder than words
High pressure CO
2
gas compression
Typ 02 - Actions speak louder than words (RG Gas) - 297x210.indd 1 2014-01-27 11:59:29
man.indd 1 04/03/2014 10:47
FLEXSORB
®

Leader in H
2
S Removal
Sour Gas Treating Problems Solved
✓ Absorbs the H
2
S, Rejects the CO
2
✓ Cost-Effective for Grassroots and Retrofits
✓ Removes H
2
S to Less Than 10 ppm
✓ Vast Commercial Experience (100 + units)
✓ Enables Simple, Low Cost Retrofits
3225 Gallows Road, Fairfax, Virginia 22037-0001, USA
www.exxonmobil.com/refiningtechnologies • +1-703-846-2568 • fax +1-703-846-3872 • tsl.email@exxonmobil.com
exxon.indd 1 07/03/2014 09:52
www.eptq.com Gas 2014 39
molecules are positioned during the
adsorbing phase, relying on strong
physical forces called physisorp-
tion; this is fully reversible, in
contrast to chemisorption where
chemical reactions are involved.
plants, because it reduces product
losses through low co-adsorption.
The 3A zeolite has a channel
system with precisely dimensioned
openings and negatively charged
cages. In these cages, adsorbed
The cost however, which may be
generated by a production loss due
to a malfunction of the adsorption
unit, is far higher than the cost of a
changeout of the unit. Reported
cases comprise the selection of
non-process-compatible adsorbents
and/or a design which is not
adequate for the process unit.
Therefore, during design and
changeout of an adsorber unit,
consultation with experienced
adsorbent suppliers is inevitable for
ensuring reliable unit operation.
Industrially relevant adsorber
zeolites are artifcially crystallised
alkaline and alkaline earth alumi-
no-silicates, which have the general
formula (M’, M’’)•[(AlO
2
)
x
•(SiO
2
)
y
]
•zH
2
O. Such zeolites are crystalline
solids consisting of tetrahedral SiO
4

and A
l
O
4
building blocks, forming
microporous structures. Each
aluminum tetrahedron carries a
negative charge, which has to be
neutralised with a metal ion (M’,
M’’) such as Na
+
, K
+
and/or Ca
2+
.
Per unit cell, the SiO
4
and A
l
O
4

tetrahedra form eight truncated
octahedra, which build one ‘soda-
lith cage’, and are stacked in a cube
resulting in an A-type zeolite.
The 3A type zeolite, the work-
horse of ethylene plant dryers, is
produced via Na
+
-> K
+
ion
exchange, starting from an 4A type
zeolite with a pore opening of
about 4Å (0.4 nm), whereas the
larger K
+
ions narrow the pore
openings by about 1Å (0.1 nm, see
Figure 5).
The impact of Na
+
/K
+
ion
exchange on the dimension of
A-type zeolite pore orifces, and the
subsequent change of equilibrium
adsorption capacities for various
adsorbates, is shown in Figure 6.
The adsorption of all adsorbates is
reduced when K
+
concentration
increases. Reduction of water
capacity, however, is much less
pronounced than for any other
adsorbate shown. In ethylene
drying, an A-type zeolite with a K
+
exchange of 50% already shows
effcient moisture pick-up features
and, in parallel, low adsorption
capability for ethylene. This is
important for the application of 3A
zeolite in drying ethylene rich
streams, as occurs in ethylene
Carbon dioxide at 700 torr, 25ºC
Ethylene at 700 torr, 25ºC
Ethane at 700 torr, 25ºC Methanol at 4.0 torr, 25ºC
Water at 4.5 torr, 25ºC
Oxygen at 700 torr, –183ºC
0.15
0.25
0.30
0.20
0.10
0.05
g
a
b
s
o
r
b
a
t
e
/
g
a
b
s
o
r
b
e
n
t
0
0 0.1 0.2 0.3 0.4 50. 0.6 0.7 0.8 0.9 1.0
Exchange of K,
K
+
/ (K
+
+Na
+
)
Figure 6 Impact of Na
+
/K
+
ion exchange on equilibrium adsorption capacities measured
on powders
7
Figure 7a Ball-and-stick representation of the structure Figure 7b Microporous channels
with openings of a 3A zeolite, consisting of eight sodalith cages. 3Å for moisture
transportation Figure 7c Sodalith cages with diameter >0.8 nm
Figure 7d 3A zeolite pore structure comprising moisture adsorbance channels and cages
for moisture transportation and storage
a
b
c d
gas grace.indd 7 11/03/2014 09:44
40 Gas 2014 www.eptq.com
tions are 15-20 wt%. This reduces
the nominal equilibrium adsorption
capacity by approximately the same
percentage because clay based
binders have limited to no adsorp-
tion properties.
The binder is fundamentally
important to both the mechanical
stability and the catalytic activity of
the molecular sieve body. Figure 9
shows the principal dependence of
the amount of used binder on the
crush strength of the zeolite bead
and its corresponding water equi-
librium capacity. With increasing
binder concentration the crush
strength also increases but, at the
same time, the equilibrium water
adsorption capacity deteriorates.
Catalytic activity in the adsorbent
body has to be minimised.
Consequently, for ethylene plant
adsorbents reactive metal-free clay
binders have to be selected.
Beside the degree of K
+
exchange
and the type and amount of binder,
the fring conditions for beads in
the adsorbent manufacturer’s
calciner play an important role in
the bead’s performance, in terms of
mechanical stability and adsorption
properties. The two key process
parameters are calcination tempera-
ture and residence time. These
parameters also impact co-adsorp-
tion behaviour towards ethylene
and eventually to polymerisation
and coking.
In summary, the selected optimal
binder ensures a number of impor-
tant characteristics for the fnal
adsorber bead, namely:
• The right meso- and macroscopic
transportation pore confguration
• A mechanically stable bead with
low attrition
• Minimised catalytic activity when
exposed to the reactive species of
hydrocarbon feedstock.
The uniform crystalline structure
of zeolitic molecular sieves, in
combination with a secondary
transportation pore system,
provides predictable and reliable
adsorption properties.
Typically, zeolitic adsorber beads
are commercially available in bead
sizes ranging from 1 mm to 5mm in
diameter. Depending on the adsor-
bent duty, bead size is selected by
trading adsorption kinetics with
molecules with the right dynamic
dimensions may enter the pores,
whereas larger molecules will be
rejected. Figure 8 depicts a symbolic
3A zeolite structure with water
molecules adsorbed from a wet
ethylene stream.
After crystallisation resulting in a
particle size of approximately 10µ,
the individual A-type zeolite crys-
tals, however, are far too small for
use in fxed bed adsorbers.
Consequently, they are mixed with
a clay binder for moulding into
macroscopic beads. Around binder
zeolite contact points the pores of
the zeolite will be at least partly
blocked. Typically binder propor-
The interconnecting channels allow
the adsorbate molecules to travel
into the cages (adsorption) or leave
them (desorption). Both channels
and cages build up the 3A zeolite
pore system shown in Figures 7a to
7d, according to the University of
Princeton’s ZEOMICS library.
8
The red spheres represent oxygen
atoms, the sticks are oxygen-alumi-
num and oxygen-silicon bonds. For
a better view of the Na
+
ions, the K
+

ions and the truncated octahedra
are omitted.
The 3A zeolite framework of
symmetrically orientated tetrahedra
and octahedra and its pore struc-
ture is stereo-selective; smaller
15
25
20
10
5
W
a
t
e
r

a
d
s
o
r
p
t
i
o
n

c
a
p
a
c
i
t
y
,

w
t
%
0
0 5 10 15 20 25 30
Amount of binder, wt%
Water adsorption capacity
Crush strength (arbitrary units)
Figure 9 Principal dependence of binder amount on crush strength and equilibrium water
capacity
9
Figure 8 Idealised stereo-selective moisture adsorption on a 3A zeolite
gas grace.indd 8 11/03/2014 09:44
Lewis
®
Pumps comprehensive
product line includes:
Sulphuric Acid Pumps
Molten Sulphur Pumps
Axial Flow Phosphoric
Acid Pumps
Molten Salt Pumps
Sulphuric Acid Valves
• Butterfly Valves
• Gate Valves
• Globe Valves
Lewmet
®
Specialty Items
• Dilution Pipes
• Orifice Plates
• Strainers
• Thimbles
• Vortex Breakers
To learn more about Lewis
®
Pumps
or Weir Minerals products, visit:
www.weirminerals.com/Lewis
Weir Minerals Lewis Pumps is known internationally in the
sulphur, sulphuric acid and phosphoric acid industries with
equipment installed in more than 120 countries worldwide.
Founded in 1891 as Chas. S. Lewis & Co., Inc., today Weir
Minerals Lewis Pumps is the recognized world leader for
pumps and valves in these difficult pumping applications.
LEWIS
®
PUMPS
Vertical Chemical Pumps
The world standard for vertical pumps and services to
the sulphur chemicals industry
Weir Minerals Lewis Pumps
8625 Grant Rd.
St. Louis, MO 63123
T: +1 314 843-4437
F: +1 314 843-7964
Email: sales.lewis@weirminerals.com
Copyright © 2013, EnviroTech Pumpsystems, Inc. All rights reserved. LEWMET, LEWIS, and LEWIS PUMPS are registered trademarks of Envirotech Pumpsystems, Inc.; WEIR is a registered trademark of Weir Engineering Services Ltd.
weir .indd 1 06/03/2014 12:06
AMACS 24-7 Emergency Service Available at (281) 716 - 1179
© 2013 Amacs Process Towers Internals. All Rights Reserved. Member of Fractionation Research
www.amacs.com
CARTRIDGE
TRAYS
DUAL FLOW GALLERY
DISTRIBUTOR
VALVE BUBBLE CAP
TRAYS FOR EVERY PROCESS
For gas processing, refining and a wide range
of process applications. AMACS offers a wide
selection of alloys, sizes and configurations
to meet any mass transfer requirement. As a
leader in the research and design of trays and
tower internals, AMACS can design to your
specifications or improve your process with
our latest technologies.
• Valve (numerous options)
• Sieve or perforated
• Bubble cap
• Cartridge trays
• Dual Flow
• Gallery Distributor Tray
For all your process internal requirements
make one stop- AMACS.
SIEVE OR
PERFORATED
TRAYS FOR EVERY PROCESS
For gas processing, refining and a wide
range of process applications. AMACS
offers a wide selection of alloys, sizes
and configurations to meet any mass
transfer requirement. As a leader in the
research and design of trays and tower
internals, AMACS can design to your
specifi cations or improve your process
with our latest technologies.
• Valve (numerous options)
• Sieve or perforated
• Bubble cap
• Cartridge trays
• Dual Flow
• Gallery Distributor Tray
For all your process internal
requirements make one stop- AMACS.
AMACS 24-7 Emergency Service Available at (281) 716 - 1179
© 2013 Amacs Process Towers Internals. All Rights Reserved.
www.amacs.com
amacs gas.indd 1 11/03/2014 09:18
www.eptq.com Gas 2014 43
for adsorbate molecules is the driv-
ing force. In depressurisation mode,
the pressure inside the adsorber
vessel will be lowered by a factor >
10. At low pressures the affnity of
the zeolite pores for the adsorbate
molecules is signifcantly reduced,
resulting in adsorbate release from
the adsorber bed. When the sorbent
material has been suffciently freed
of adsorbate, the adsorber is ready
for the next cycle.
In order to allow continuous
TSA/PSA gas conditioning, two or
more adsorber beds in the gas
treatment unit are in adsorption,
regeneration or standby modes.
Dryers in ethylene plants
Complete removal of the water
contained in cracked gas is manda-
tory in order to avoid plugging of
downstream cryogenic sections due
to freezing or formation of
hydrates. Depending on the
cracked feedstock and the confgu-
ration of the ethylene plant the
following dryers may be in opera-
tion (the frst three are mandatory
for all ethylene plants):
1. The charge gas dryer or main
ethylene dryer, located after the
acid gas removal unit
2. The condensate dryer for condi-
tioning the gasoline stream from
the condensate stripper
3. The acetylene converter or
secondary dryer downstream of the
C
2
acetylene converter
4. The MAPD converter dryer
downstream of the MAPD
converter
its accepted maximum impurities
level, the adsorber has to be
regenerated.
Two regeneration methods are
used: thermo-swing adsorption/
desorption (TSA) and pressure
swing adsorption/desorption
(PSA). They rely on the fact that,
apart from the physic-chemical
nature of the adsorbate, adsorption
and desorption are pressure,
temperature and time controlled. In
TSA units, temperature is the
process variable; in PSA units pres-
sure is the variable.
In an ethylene plant, with the
exception of the Hydrogen PSA
purifer, all other adsorption units,
including the hydrogen dryer
downstream of the methanator, are
TSA types. In the TSA case, thermal
regeneration is accomplished by
passing hot product gas through
the bed, heating it up to a tempera-
ture signifcantly above the
temperature during adsorption.
The hot adsorber bed decreases its
affnity for adsorbates and releases
them into the gas fow now acting
as purge fuid. After completion,
the adsorber bed is cooled down,
regaining its adsorption capacity.
Then the unit is ready for the next
cycle. Typical regeneration gas
temperatures are 240°C for 3A
molecular sieves and 250-280°C for
4A molecular sieves.
In contrast, pressure swing regen-
eration takes place at the same
temperature as adsorption. In the
adsorption mode at elevated pres-
sures the affnity of zeolite pores
pressure drop and mechanical
stability. As a rough guide, the
smaller the beads the better the
kinetics, but the higher the pressure
drop the lower the single piece
crush strength, and vice versa.
Commercial adsorber beads have a
comparatively narrow bead size
range in order to have as close as
possible to mono-disperse particle
distribution characteristics targeting
adsorber bed porosities of 30% to
35%. Depending on the supplier´s
philosophy, zeolitic adsorber extru-
dates or pellets with L/D ratios
from nearly 1 to 5 are in use.
Extrudates with higher L/D ratios
may lead to a lower space flling
and a lower bulk density. They
tend to have favourable mass trans-
fer properties, however, sacrifcing
adsorption capacity and jeopardis-
ing the adsorber bed’s mechanical
stability. The pressure drop for
beads is more favourable than for
extrudates (see Figure 10), even in
the typical gas velocity range
between 5 to 12 m/min.
Even more importantly, plant
operators occasionally report signif-
icantly increased dust formation
when loading and unloading the
adsorber vessel, as a result of extru-
date breakage and split-offs.
Adsorber zeolites have a strong
affnity for polar and polarisable
molecules, whereas the higher the
degree of polarity, the higher the
selective adsorption power. Since
water is the most polar molecule
known, it is the preferred adsorb-
ate. In a moisture-containing
hydrocarbon stream, next to water
heavier hydrocarbons will be
picked up frst.
Depending on the water/hydro-
carbon ratio, for some time the
zeolite adsorber bed will simultane-
ously pick up water and some
hydrocarbons. At the end of the
adsorption cycle, the part of the
adsorber bed next to the inlet is
enriched with adsorbed water,
whereas the middle part of the bed
contains a higher concentration of
hydrocarbons. The outlet part of
the bed carries the lightest hydro-
carbons. This effect is given by the
higher polarity of water over heav-
ier and lighter hydrocarbons.
Before the product gas has reached
12
20
18
16
14
10
8
6
4
2
P
r
e
s
s
u
r
e

d
r
o
p
,

m
b
a
r
/
m
0
0 2 4 6 8 10 12 14 16 18 20
Superficial velocity, m/min
2.5-5.0 mm beads
3.7 mm extrudates
Figure 10 Pressure drop development of beads vs extrudates
gas grace.indd 9 11/03/2014 09:44
44 Gas 2014 www.eptq.com
stage at lower operating pressure
(see Figure 3). This is often used
when the ‘depropaniser first’ with
front end hydrogenation route has
been chosen.
The compressor section of an
ethylene plant is prone to fouling,
given the nature of the cracked gas,
which may contain varying concen-
trations of hydrogen, paraffins and
aromatics as well as reactive
compounds such as acetylenes,
olefins and diolefins. Under certain
conditions, compounds such as
butadiene, styrene, isoprene and
vinyl acetylene may form polymers.
Consequently, charged gas dryers,
which are embedded in this section,
are also impacted by this fouling
tendency. Such fouling is the result
of oligomerisation, polymerisation
and condensation reactions, given
by:
• Free radical polymerisation
reactions
• Diels-Alder polymerisation of
active Dienes (Cyclopentadiene,
Methyl cyclopentadiene, and so on).
The radical polymerisation reac-
tions are caused by heat alone, but
hydrocarbons into the regeneration
off-gas due to absorption and
co-adsorption outperforming the
dual system of glycol based absorb-
ers followed by alumina adsorbers
• More economical to regenerate.
Two design options with differ-
ent operating pressures are used,
depending on the selected process
route, which also define the posi-
tions of the dryers in the ethylene
plant:
1. Dryers operated at high pressure
(26-37 bar, 13-15°C)
2. Dryers operated at low pressure
(12-25 bar, 12-14°C).
The dryer operating temperature
is usually set at 2-3°C above the
hydrate formation temperature. It
has to be kept in mind that a
temperature rise by 1°C results in a
6.5% increase in water load to be
removed by the dryer. Therefore
careful temperature control is
advised.
Typically, the dryers are posi-
tioned after the last compression
stage, as in option 1 (see Figures 1
and 2). In option 2, the dryers are
positioned after the next-to-last
5. The hydrogen dryer downstream
of the methanator for upgrading
the hydrogen stream from the cold
box and or the PSA unit for hydro-
gen of high purity.
The charge gas dryers or main
cracked gas dryers
The main cracked gas dryer is posi-
tioned in the compressor section
downstream of the acid removal
unit (caustic and water wash
tower). Complete removal of water
to a level below <1 ppmv is manda-
tory. A specially developed 3A
molecular sieve is state-of-the-art
technology for operating such
dryers. It has almost completely
replaced the dual system of glycol
absorbers followed by alumina
adsorbers. The main technical and
commercial reasons for replacing
earlier drying concepts with zeolitic
molecular sieves are:
• Achieving gas specifications <0.1
ppmv water content
• Achieving significantly lower
‘green oil’ formation
• Achieving longer on-stream time
• Significantly lower losses of
We make technology work.
www.hoerbiger.com
Like musicians set the tone,
HOERBIGER sets standards in
the Oil and Gas Industry.
Image_ad_KT_2014.indd 1 06.03.2014 15:46:22
gas grace.indd 10 11/03/2014 09:45
www.eptq.com Gas 2014 45
shows a typical aging curve and
the corresponding pressure drop
development of a naphtha based
ethylene plant’s charged gas dryer.
are enhanced by peroxides. Also,
the Diels-Alder condensation reac-
tion contributes to fouling due to
formation of heavy hydrocarbons
condensing on the metal surfaces
and gradually dehydrogenating.
Both polymerisation reactions are
catalysed in the presence of iron.
The temperature is a key variable
for such reactions.
10,11
Grace and others encountered the
following experiences regarding the
operation of charge gas dryers:
• All gas cracking units show a
high fouling rate in the compressor
section, whereas plants based on
liquid cracking report low to
moderate fouling.
• Lower pressure charge gas dryers
have to cope with a higher water
load (see Table 1) and are more
prone to fouling compared to those
plants with the dryers set up at
higher pressures (32-37 bar vs 12-22
bar) after the last compression
section. This fouling is caused by
non-condensed and non-separated
heavier fouling precursors.
Regarding the frst experience, it
is known that gas-based plants,
typically operating four compressor
stages, generate higher discharge
gas temperatures because of the
higher compression ratios per
stage. This effect is further
enhanced by higher concentration
of hydrogen and its higher isotropic
exponent k=C
p
/C
v
:


T
s
, P
S
= Suction temperature,
pressure
T
D
, P
D
= Discharge temperature,
pressure
By contrast, higher content of
aromatics in the cracked gas from
liquid based crackers helps to
dissolve and wash out polymers
generated by the heat of the
compression.
Undesired charged gas-adsorber
bed interactions sooner or later lead
to deterioration of the adsorber
bed. Consequently, adsorber bed
capacities will be reduced and the
pressure drop will increase; both
are caused by the formation of
polymerisates and coke. Figure 11
Dryer operating pressure, bar 35 15
Operating temperature, °C 15 15
Water content of charge gas, mg/STDm
3
450 840
Charged gas water content as a function of operating pressure
Table 1

47% 50%
3%
>4.5 years
2.5-4.5 years
<1.5 years
18%
50%
32%
Cycle times too short
Pressure-drop issues
Vessel inspections
11%
81%
8%
Kept employed dessicant
Changed technology for
general improvement
Change of targeted contaminants
80
100
90
70
60
B
e
d

d
e
t
e
r
i
o
r
a
t
i
o
n

d
u
e

t
o

a
g
e
i
n
g
,

%
0.19
0.21
0.17
0.15
0.18
0.20
0.16
P
r
e
s
s
u
r
e

d
r
o
p
,

b
a
r
50
0 6 12 18 24 30 36 42 48 54 60
Months of operation
Pressure drop
Bed deterioration
Figure 11 Ethylene plant charged gas dryer bed aging and pressure drop increase
Figure 12a Desiccant life time in ethylene
plants
Figure 12b Reasons for changeout
Figure 12c Consequences of the changeout
a b
c
It is run at high pressure over an
operational plant period of 60
months, equalling about 650 regen-
eration cycles. After 650 cycles, the
adsorption bed capacity has been
reduced to nearly 70% and the
pressure drop over the bed has
increased from 0.15 bar to 0.21 bar.
According to an AIChE survey,
12
operators from gas (60%) and
liquid (40%) processing ethylene
plants reported the observations
shown in Figures 12a to 12c.
The design philosophy for a
gas grace.indd 11 11/03/2014 09:45
18-20 bar for the gas phase, and
16-28 bar for the liquid phase, with
temperatures typically in the range
35-50°C. The best desiccant for this
drying step, like the second drying
step, is a 3A type zeolite with espe-
cially low polymerisation and
coking characteristics. Such a desic-
cant can be used for gas and liquid
phase operations. The regeneration
gas and the regeneration tempera-
ture are the same as for the above
charged gas dryer.
The condensate dryer
Depending on the technology
chosen, hydrocarbon condensates
generated at the fourth compres-
sion stage have to be separated in
the knockout drum upstream of the
charge gas dryer. These condensate
dryers operate at pressures in the
range 30-35 bar, at temperatures of
12-15°C. Typically, two dryers are
required: one in service, the other
in regeneration or standby mode.
The state-of-the-art adsorbent is a
special 3A type zeolite. The hydro-
carbon condensates contain
oligomer and/or polymer precur-
sors, which are formed in the
upstream compression section.
Consequently, the selected adsor-
bent should exhibit little tendency
to coking and polymerisation.
Depending on cycle times, which
are 24-48 hours, the lifetime may
reach five to six years. The conden-
sate is saturated with water,
depending on composition and
operating conditions. The quality of
the upstream separator must ensure
that practically no traces of free
water reach the dryers. If free water
is expected always to be present in
the condensate, a layer of macropo-
rous silica gel is installed in the bed
inlet to guard the zeolitic adsorber
bed against free water. The mois-
ture concentration in the outlet is
specified to be below 1 ppmv. This
is to avoid the formation of
hydrates in downstream piping and
equipment. The regeneration gas
and the regeneration temperature
are the same as for the dryers.
The hydrogen dryer
In addition to methane, hydrogen
off-gas from the ethylene plant’s
cold box contains carbon monoxide
46 Gas 2014 www.eptq.com
any freezing or gas hydrate forma-
tion in the ethylene splitter. The
secondary dryer runs in a one-
vessel mode with a service time of
20 days prior to regeneration.
Consequently, during regeneration
the secondary dryer step is
bypassed. The desiccant of choice
for the secondary dryer is a type
3A zeolite with especially low
polymerisation and coking charac-
teristics. Such desiccant should
exhibit low co-adsorption for acety-
lene and olefins. Depending on the
employed technology and the
selected separation sequence of the
first tower, the operation conditions
are: pressures of 20-30 bar and
temperatures of -30-30°C. The
regeneration gas and the regenera-
tion temperature are the same as
for the charged gas dryer.
The MAPD converter dryer
The MAPD converter dryer is posi-
tioned between the methyl
acetylene and propandiene hydro-
genation unit (MAPD) and the
ethylene splitter to remove any
moisture formed during the hydra-
tion of the C
3
cut. This dryer is
needed to avoid any water content
in the final product, which is espe-
cially important for naphtha based
crackers. Typically, two vessels are
installed, one in service, the other in
regeneration mode, for both liquid
and gas phase operations, depend-
ing on the chosen hydrogenation
technology. The moisture content of
the inlet fluid is typically specified
at <100 ppmv. For gas and liquid
phases, the dryer outlet fluid mois-
ture contents are specified to be
below 1 ppmv. The corresponding
operating pressures are in the range
charged gas dryer depends on the
contractor:
• Main/guard bed configuration
(Lummus, Technip Stone &
Webster): the moisture analyser is
installed between beds. The guard
bed height is typically sized for six
to eight hours of additional adsorp-
tion time, allowing operators to
take action in case breakthrough
conditions occur. An additional
function of the such guard bed is to
adsorb trace contaminants such as
ammonia and methanol
• The single bed (Linde) adsorbent
section below the moisture analyser
functions as a guard bed. In this
design, one support grid has been
omitted and in the same available
volume more adsorbent material
can be installed. Furthermore, this
concept allows an easier loading of
the adsorbent material. Both
options are shown in Figure 13:
• Typical arrangements: two or
three vessels in total, one or two in
adsorption, one in regeneration
mode
• Flow: downwards during adsorp-
tion, upwards during regeneration
• Regeneration gas: methane frac-
tion from demethaniser overhead,
also utilised for all subsequent
dryers. Typical regeneration gas
inlet temperature range is 210-230°C
• Adsorption time typically 48-64
hours.
Typical lifetimes of low pressure
operated units are less than five
years, while high pressure units
may achieve more than five years
of operation.
The secondary dryer
The secondary dryer is positioned
between the acetylene converter
and the ethylene splitter. Its
purpose is to remove moisture
formed in the acetylene converter
during the selective hydrogenation
of C
2
acetylenes. This is caused by
the presence of free oxygen from
decomposed oxygenates and/or
oxygen ingress into the ethylene
unit. A typical specified moisture
load for the inlet gas is below 20
ppmv. Licensors offer designs with
a limit of 50-100 ppmv of moisture
as a safety margin. The targeted
moisture concentration of the outlet
gas is below 1 ppmv, thus avoiding
x = 0.8 to 1.0 mm
x x
Figure 13 Typical design layouts for
charged gas dryers. Left: main bed/guard
bed. Right: single bed
gas grace.indd 12 11/03/2014 09:45
www.eptq.com
are zeolitic compounds with very
promising adsorption characteris-
tics. Unfortunately, these materials
are still either diffcult to produce,
or thermomechanically or chemi-
cally not stable enough to serve as
adsorbents in ethylene plants.
The shale gas boom leading to
the erection of megacrackers is
expected to rely on established
adsorption technology. Specially
developed 3A type zeolites will
continue to play a dominant role in
this industrial application.
References
1 True W R, Oil & Gas Journal, July 2013, 90-
103.
2 Falqi F H, The Miracle Petrochemicals,
Olefins Industry: An In-Depth Look at Steam
Crackers, 7-8, Universal-Publishers, Boca Raton,
Florida, USA, 2009.
3 Kniel L, et al, Ethylene Keystone to the
Petrochemical Industry, 1980, 78-107, Marcel
Dekker Inc. New York, NY.
4 Kapur S, Cracking Technology for the
Production of Ethylene, Chapter 6.1, Handbook
of Petrochemicals Production Processes,
McGraw-Hill Professional, 2004.
5 Zimmermann H, et al, Ethylene, Ullmanns
Enzylopädie der Technischen Chemie, 2007,
34-47, Wiley-VCH Verlag GmbH & Co. KGaA,
Weinheim.
6 Knaebel K S, Adsorbent Selection,
Adsorption Research Inc, www.adsorption.
com/publications/AdsorbentSel1B.pdf (June
2004), 2003.
7 Breck D W, et al, J. Am. Chem. Soc., Vol. 78,
No. 23, Dec 1956, 5967.
8 http: //helios. princeton. edu/zeomics/
cgi-bin/view_structure.pl?src=iza&id=LTA&
token=e8279dfc54f813ef8f88b9baad8 bcc72
9 Rode E J, et al, Molecular sieve adsorbent
properties and their impact on cracked gas
drying, AIChE Spring National Meeting 1997,
Houston, TX.
10 De Haan S, et al, Overview of Cracked Gas
Compressor Fouling: Theories and Practices,
Presented at: 2001 Spring National Meeting,
23-26 April 2001, Houston, TX.
11 Snider S, Olefin Plant cracker Gas
Compressor Fouling, PTQ Gas, 2007, 36-40.
12 O’Brien B, et al, Cracked Gas Drier Survey,
Spring National Meeting, 21-25 March 2010,
San Antonio, TX.
Hans Hoefer, a materials scientist by
education, worked for almost 27 years for WR
Grace in the Adsorbents Business as Manager
for Research & Development, Technical
Customer Service and Market Development.
Vassilios Zafirakis is Manager of Technical
Customer Services Adsorbents with WR
Grace, Worms Germany. He holds a Master’s
in process engineering from the University of
Stuttgart, Germany.
(CO). Since this acts as an inhibitor
for palladium and silver hydro-
genation catalysts, untreated
hydrogen off-gas cannot be used
for hydrogenation (acetylene
converter, MAPD converter). For
this reason, the CO has to be
removed. There are basically two
established options (refer to process
fow diagrams):
1. A methanation reactor for low
grade hydrogen
2. PSA for high grade hydrogen
purifcation.
In the frst case, CO in the hydro-
gen rich stream is catalytically
converted to methane:
CO + 3H
2
-> CH
4
+ H
2
O
H
2
O is a poison for the hydro-
genation catalyst and has to be
removed. A TSA dryer equipped
with a 3A or a 4A type zeolite has a
typical cycle time of 24 hours. The
operating pressure is 6-30 bar at
15-30°C. The regeneration tempera-
ture is 240°C. The regeneration gas
is the demethaniser overhead. If the
feed is free of olefnic and acety-
lenic compounds, the lifetime may
be between fve and 10 years.
In the second case, the hydrogen
rich stream is purifed by a three-
layer PSA unit consisting of: 5A
type zeolite to remove nitrogen and
CO; activated carbon to remove
methane and CO
2
; and activated
alumina to eliminate H
2
O and NH
3
.
The adsorption pressure is 15-30
bar. The desorption pressure is
typically lowered by factors of
10-13 but is always kept slightly
above atmospheric pressure. The
operating temperature is 15-35°C.
The high grade hydrogen has a
typical purity of 99.5 vol% with
concentrations of CO <10 ppmv,
CH
4
<10 ppmv, and N
2
<10 ppmv.
Outlook
The adsorption technology
discussed here has its roots in the
early 1950s. Over the years, a lot of
progress has been made in the
adsorbents’ performance and
economical feasibility. Although
nowadays there are about 160 differ-
ent zeolite structures, none of those
could replace the type 3 A zeolite
for ethylene drying. Certainly there
Our knockout drum needs
a mist eliminator!
Can we add one without
overhauling the vessel?
IT’S A MORE COMMON problem than
you might think. A vertical knockout
drum removes free liquid from a
certain gas stream. But at the time the
plant was built, a mist eliminator was
not considered necessary. Now mist
is carrying over and causing trouble
downstream. There is no manway, so
adding a conventional mist eliminator
would require cutting the vessel open.
Instead, if the flanged exit nozzle
is at least six inches in diameter, you
can simply hang a patented MistFix
®

mist eliminator inside, securing it
between the flanges.
Even when there is a manway,
MistFix avoids entry, welding, and
recertification in retrofits. For new
and old vessels, it drastically cuts
installation cost and downtime.
Phone: 713-434-0934 • Fax: 713-433-6201
amacs@amacs.com
24-hr Emergency Service: 281-716-1179
MistFix
®
insertion mist eliminator
Visit our new website at
www.amacs.com
gas grace.indd 13 11/03/2014 09:45
Organised by CRU Events
3-6 November 2014, Marriott Rive Gauche, Paris, France
Simultaneous English/Russian interpretation/Синхронный перевод на английский/русский
See you in Paris to celebrate
our 30th Anniversary!
Register today at www.sulphurconference.com
Organised by CRU Events
Operator/
Producer
50%
discount
Early bird
rate just
€1405
Oil | Gas | Fertilizers | Metallurgy | Industrial uses
Sulphur 2014
The world’s leading industry event
2014 Call for papers
Sponsored by:
Official publications:
Case studies demonstrating operational experience in:
- Process and efficiency improvements
- Advances in environmental efficiency and strategies
for compliance
- Equipment improvements and optimisations
- Operational problems and solutions
- Best practice health and safety initiatives
- Application of improved technology or catalysts
- Restoration, revamping or maintenance
Technology innovations and improvements:
- New technology developments, which, if not proved
so far, offer the realistic prospect of economic
advantage in plant operations
- Technology innovations that improve environmental or
health and safety regulation compliance
Interested in presenting a paper in Paris? Submit your abstract by 14 April 2014 online at www.sulphurabstracts.com
or visit www.sulphurconference.com for further details
Sulphur is an excellent place to see the latest developments in both sulphur & sulphuric acid with
great presentations as well as a great place to meet both new and existing equipment suppliers.
Research Engineer, Haldor Topsøe A/S


48_sulphur.indd 1 06/03/2014 13:49
Virtual commissioning of a gas
handling system
G
as cargo such as liquid natu-
ral gas, ethane, ethylene and
propylene is transported in
liquefed form under cooling
temperatures and pressures.
Typically, pressures are low, which
leads to very low temperatures.
Although the cargo tanks are well
insulated, gas continuously evapo-
rates from the liquid cargo. This
gas is called boil-off gas and has to
be reliquefed during maritime
transport in the reliquefaction
system.
The reliquefaction system by TGE
Marine Gas Engineering for fve
new vessels of Navigator Gas ship-
ping company was designed with
UniSim Design. First, a stationary
simulation model was used for the
basic design. Later, a dynamic
simulation model was developed
for the design of the automation
and control system.
There are many differences
between automation systems that
are installed on a vessel and auto-
mation systems of a production
plant. For instance, it is much more
diffcult to provide support
and maintenance on a vessel
than on land. As a conse-
quence, sea based automation
systems have to be designed,
verifed and tested very care-
fully. The requirements in
this feld are to some extent
similar to the requirements in
the automobile and aviation
industries.
In order to develop auto-
mation systems that are safe,
reliable and error free, auto-
mobile and aviation
industries rely on hardware
in the loop simulation. With
Gas reliquefaction units were designed by means of stationary and dynamic
process simulation models
RAINER SCHEURING Cologne University of Applied Sciences
HANS-CHRISTIAN HAARMANN-KÜHN TGE and JÜRGEN ESSLER TGE Marine Gas Engineering
TORSTEN FELIX Flemming Automationstechnik MICHAEL BRODKORB Honeywell Process Solutions
this technology, the automation
system can be verifed within a
simulation based test environment.
TGE has decided to use hardware
in the loop simulation in the design
of the automation system of the
new cargo handling system, too.
This allows not only the develop-
ment of an error-free automation
system but also a substantial reduc-
tion in commissioning time.
Reliquefaction system
About 50% of the existing feet of
liquefed ethylene (LEG) carriers
uses TGE designed cascade relique-
faction units to cool down cargo
and maintain tank pressure.
Liquefed ethylene is usually trans-
ported using semi-pressurised,
fully refrigerated vessels that are
equipped with IMO type C tanks
(see Figure 1) and cascade refrigera-
tion systems.
The basic technology of the
cascade refrigerant systems have
remained unchanged for years. The
cargo tanks are cooled by an open
cooling system compressing the
BOG in a two-stage process and
condensing it against a refrigerant
(see Figure 2). Two crank, oil-free
piston compressors usually serve as
cargo compressors. The refrigerant
(propylene or R404A in place of
R22) is compressed by oil injected
screw compressors and condensed
against seawater.
To date, the biggest ethylene
carriers in operation are 22 000 m³
vessels built by Jiangnan Shipyard
for Navigator Gas. A new genera-
tion of these vessels with 21 000 m³
capacity is under construction and
the frst of fve vessels are to be
delivered in 2014.
The vessels are designed to carry
21 different LPG, chemical gas and
chemical cargoes. For 15 of them
(including ethylene, propylene,
ammonia, VCM and butadiene),
refrigeration is provided. The
vessels are capable of transporting
up to three different grades, two of
which may be cooled at the same
time. Operating conditions and
modes vary widely depending on
the specifc cargo.
A challenge in the design
of the reliquefaction units
was to improve perfor-
mance and operability in
order to meet increased
requirements for loading
rates and cooling down
times. In addition, the
same reliability and
robustness of the units,
which have operated
successfully for 14 years,
had to be maintained.
As an example, cooling
down capacity for ethylene
in one-grade operation
is 50% higher for the
www.eptq.com Gas 2014 49
Figure 1 5500 m³ bi-lobe tank under construction at Jiangnan
heavy industry site in Shanghai

gas honeywell.indd 1 11/03/2014 11:51
50 Gas 2014 www.eptq.com
Aspen Hysys Dynamics are based
on so-called frst principle process
modelling engines that allow realis-
tic modelling of the transient
behaviour of processes typically
found in the oil and gas and chemi-
cal industries. In order to create a
process model, the user selects
readily available components and
thermodynamic packages to defne
physical properties and phase
equilibria of the system and then
creates a fowsheet by adding and
linking generic unit operation
models (such as pipes, vessels,
pumps and distillation columns)
and control equipment (valves,
PIDs, and so on). The resulting
model can be initialised to a specifc
initial condition and run through
different predefned scenarios as
part of a dynamic simulation study.
Dynamic simulation studies are a
standard tool in the process indus-
tries for analysing and optimising
transient process behaviour.
Application examples for operabil-
ity or safety studies include
dynamic fare load estimation in
new systems. In order to meet
all requirements, compressors
of higher capacity and new
modes of operation have been
introduced.
Dynamic plant simulation and
design of automation system
Dynamic Process Simulators like
Honeywell UniSim Design,
Invensys Dynsym, or AspenTech
Cargo liquid
separator
LPG
condensor
Refrigerant
liquid
separator
Refrigerant
compressor
Refrigerant
condensor
Refrigerant
collecting
vessel
Refrigerant
economiser
Cargo vapour
compressor
Cargo
condensor
Cargo
receiver
Cargo
tank
Cargo
vapour
Out
Sea
water
In
Out
Sea
water
In
Figure 2 Reliquefaction system
C
M
Y
CM
MY
CY
CMY
K
OGT_124_ProTreat_Half_Page_Horizontal_PTQ_Gas_Supplement_2014.pdf 1 3/4/14 8:29 AM
gas honeywell.indd 2 11/03/2014 17:13
www.eptq.com Gas 2014 51
Automation system
(Siemens S7)
Simple signal simulation
(Simit)
Comprehensive plant simulation
(UniSim)
Figure 4 HIL simulation of TGE’s gas carrier
cargo handling system
Cargo
condensor
Refrigerant
compressor
Refrigerant condensor
Refrigerant
economiser
Refrigerant
collecting
vessel
Figure 3 Dynamic simulation model of reliquefaction system
refneries
1
or onshore gas felds,
2

and compressor studies.
3
In order to design a control
concept for the reliquefaction
system, a dynamic simulation
model was developed (see Figure
3).
This model served as the basis for
the design of a new control concept
that makes it possible to operate
the reliquefaction system with all
cargoes and all operation modes
within a single control structure. It
is TGE’s frst fully automated
cascade refrigerant cycle for LEG
carriers. In addition, the reliquefac-
tion system is kept in a stable and
stationary operation point inde-
pendent of any load on the system.
In order to ensure that the relique-
faction system operates in a
satisfactory manner in all operating
conditions, special attention was
paid not only to effciency but also
to robustness of the solution.
Hardware-in-the-loop simulation
Hardware-in-the-loop (HIL) simu-
lation is a technique that is used in
the product development cycle in
which real components interact
with simulated components.
4
Early
applications date back to the 1970s.
5

Today, HIL simulation has become
an integral component in the devel-
opment process of electronic control
units especially in the automotive
and aviation industries.
4,6
In the
basic structure of a HIL simulator,
the control unit is a real component
and the plant is simulated.
Due to the fact that a gas carrier
cargo handling system is installed
on a seagoing vessel, activities such
as troubleshooting and maintenance
are diffcult to handle. As a result,
the automation system of a gas
carrier cargo handling system has
to be designed, verifed and tested
very carefully. The aim is to have
an error-free system. Figure 4
presents the HIL simulation struc-
ture of TGE’s gas carrier cargo
handling system.
The automation system of TGE’s
cargo handling system is based on
a Siemens S7 programmable logic
controller (PLC).
In order to test the automation
system, two simulators were used:
Simit and UniSim. Simit can be
smoothly integrated into Siemens
automation hardware and software
infrastructures and supplements
Dynamic simulation
studies are a
standard tool in the
process industries
for analysing and
optimising transient
process behaviour
gas honeywell.indd 3 11/03/2014 17:13
52 Gas 2014 www.eptq.com
behaviour of this system was exam-
ined for a representative subset of
cargos and all load scenarios.
In a fnal test session, human
machine interface, start-up behav-
iour, normal operation and several
fault scenarios were verifed with
experienced engineers from the
shipping company.
As a result of these inquiries,
several logic faults and control
algorithm specifc implementation
errors have been eliminated in front
of the tests at the shipyard and the
trial run.
Conclusion
TGE Marine Gas Engineering has
developed a new, more fexible,
and more powerful gas carrier
cargo handling system. This system
is able to handle a high number of
different cargos such as LPG, gases
and chemicals. For the majority of
these cargoes, refrigeration is
provided. This function is
performed by reliquefaction units
that were designed by means
of stationary and dynamic
UniSim Design process simulation
models.
Elimination of errors and debug-
ging of automation software is one
of the main tasks during the
commissioning phase of a cargo
handling system. Due to the fact
that commissioning time is very
limited, the complex logics of the
automation system and the
dynamic behaviour of the relique-
faction units cannot be verifed in
detail during this phase. Therefore,
TGE decided to carry out a detailed
HIL simulation study of the gas
carrier cargo handling system at an
earlier design stage.
In the HIL simulation study the
automation system was extensively
tested and evaluated. Simit was
used for the verifcation of the
complex logics. The S7 code
components, which control the
central reliquefaction system, were
tested by connecting them to a
dynamic UniSim Design process
simulation model and carrying out
a virtual commissioning. As a
result, several logic faults and
control algorithm specifc imple-
mentation errors have been
eliminated.
in the shipyard. Subsequently, a
trial run with one cargo and dura-
tion of about one week takes place.
It is neither possible to verify the
complex logics of the automation
system in detail nor to analyse and
optimise the dynamic behaviour of
the reliquefaction system for
different cargos and load scenar-
ios. As a consequence, TGE
decided to carry out a detailed
HIL simulation study of the gas
carrier cargo handling system at
an earlier design stage.
Simit provides a test environment
for the S7 based automation system.
A strong point is its ability to test
the logic controls of the S7 PLC.
Therefore Simit was used to test
and verify the logic controls of the
cargo system.
UniSim was employed for a
detailed analysis of the S7 code
components that control the central
reliquefaction system. The dynamic
the automation interface hardware
(Profbus DP, Profnet IO).
Communication between the auto-
mation system and Simit is carried
out as in the real system at the feld
bus level.
7
Simit does not provide a frst
principles modelling engine.
Consequently, a dynamic UniSim
model was used for testing the reli-
quefaction system. In these tests,
Simit established communications
between UniSim and the S7 PLC.
The interface between UniSim and
Simit was developed using UniSim
Design OLE Automation client-
server technology.
Virtual commissioning of
automation system
Elimination of errors and debug-
ging of automation software is one
of the main tasks during a system’s
commissioning phase. In the case of
machine tools, a study from 1997
provides the data for Figure 5.
8
The
duration of commissioning takes
approximately 15-25% of the
entire project. About 90% of this
period is required for commission-
ing the automation system, of
which up to 70% are needed for the
elimination of software errors (see
Figure 5).
Commissioning engineers for
gas carrier cargo handling systems
still experience the same diffcul-
ties. In addition, commissioning of
these systems has to deal with a
core problem: commissioning time
is very limited. Some preliminary
testing without cargo can be done
Entire project
duration
Duration of
commissioning
Duration for
automation
system
15-25%
Up to 90% for
automation
system
Up to 70% for
software errors
Figure 5 Fraction of commissioning and project duration for software error elimination
Elimination of errors
and debugging of
automation software
is one of the main
tasks during the
commissioning phase
of a cargo handling
system
gas honeywell.indd 4 11/03/2014 11:52
www.eptq.com Gas 2014 53
from Stuttgart University, Germany.
Email: rainer.scheuring@fh-koeln.de
Hans-Christian Haarmann-Kühn is the Head
of the Engineering Department at TGE Marine
Gas Engineering GmbH, Bonn, Germany. He
has a degree in process engineering and a
PhD from Technical University RWTH Aachen,
Germany. Email: hans-christian.haarmann-
kuehn@tge-marine.com
Jürgen Essler is a Project Engineer at TGE
Marine Gas Engineering GmbH in Bonn,
Germany. He has a degree in mechanical
engineering and a PhD from Technische
Universität Dresden.
Email: juergen.essler@tge-marine.com
Torsten Felix is certified PLC engineer
and Project Manager at Flemming
Automationstechnik GmbH.
Email: t.felix@flemming.de
Michael Brodkorb is Advanced Solutions
Consultant at Honeywell Process Solutions in
Tarragona, Spain. He has a degree in chemical
engineering from University of Dortmund,
Germany, and a PhD from Bradford University,
UK. Email: michael.brodkorb@honeywell.com
4 Wältermann P, Hardware-in-the-Loop: the
technology for testing electronic controls
in automotive engineering. 6. Paderborner
Workshop: Entwurf mechatronischer Systeme,
Heinz Nixdorf Institut, Paderborn, 2.-
3.4.2009,. www.dspace.de/shared/data/pdf/
paper/HIL_Overview_Waeltermann_03_E.pdf.
5 Waite W F, Simulation – the past as prologue
…probably, ITEA Journal, Mar/Apr 2002, 7-10.
6 Ledin J A, Hardware-in-the-Loop Simulation.
Embedded Systems Programming, Feb 1999,
42–60.
7 Siemens: SIMIT V7 - open simulation
platform for virtual commissioning, www.
industry.siemens.com/verticals/global/en/
chemical-industries/chemical-industries-
products-systems/Documents/SI MI T_
Brochure_EN.PDF.
8 VDW-Bericht: Abteilungsübergreifende
Projektierung komplexer Maschinen
und Anlagen. Verein Deutscher
Werkzeugmaschinenhersteller, Aachen, 1997.
Rainer Scheuring is Professor of Automation
Technology and Control Theory at Cologne
University of Applied Sciences. He has a
degree in technical cybernetics and a PhD
With the HIL simulation, virtual
commissioning of the automation
system has been realised. This will
signifcantly contribute to a safe,
reliable and error-free gas carrier
cargo handling system.
UniSim is a registered trademark of
Honeywell International Inc. DYNSYM is a
registered trademark of Invensys Inc. Aspen
Hysys Dynamics is a registered trademark of
Aspen Technology Inc. SIMIT is a registered
trademark of Siemens AG
References
1 Gruber D, Leipnitz D-U, Sethuraman P,
Alos M A, Nogues J M, Brodkorb M, Are there
alternatives to an expensive overhaul of a
bottlenecked flare system?, PTQ, 93-95, Q1
2010.
2 Panigrahy P, Balmer J, Alos M A, Brodkorb M,
Marshall B, Dynamics break the bottleneck,
Hydrocarbon Engineering, 93–96, Sept 2011.
2 Nugues J M, Brodkorb M, Feliu J A, How
can dynamic process simulation be used for
centrifugal compressor systems, Hydrocarbon
Engineering, 92–98, Aug 2012.
4
th
OPPORTUNITY CRUDES CONFERENCE
Downstream Meets Upstream: Right Crudes to Make Products in Demand
WESTIN OAKS HOTEL | HOUSTON, TX (US)| SEP. 14-16, 2014
This international conference addresses the difficulties and benefits of processing aromatic opportunity crudes and paraffinic
light oil alone and together as well as acidic or high TAN crudes.
REGISTER TODAY at opportunitycrudes.com/houston2014 to take advantage of the early bird discount.
We have assembled a diverse team of technology experts and business strategists to share their experiences and insights into many
urgent issues facing refiners worldwide:
 Accelerated US shale/tight oil output, influx of NG condensate, expanding Canadian oilsand output, and increased processing of heavy oil
outside the US
 Immiscibility concerns when blending aromatic heavy crudes and paraffinic tight oils into so-called "dumbbell" crudes
 Corrosion, fouling, and asphaltene deposition in railcars, pipelines, terminals and refinery units
 Uncertainty over FCC and coker operations due to a shift in crude types, a decline in gasoline demand, and a naphtha “surplus” from tight oil processing
 Seeking advanced residue and tight oil processing technologies to produce more middle distillates and petrochemicals amid shrinking fuel oil demand
Our speakers come from over 25 companies, including Albemarle, Baker Hughes, BASF, Cameron, Canadian Energy Research Institute,
Center for Strategic and International Studies (CSIS), Ecopetrol, EnSys Energy, Haverly Systems, Intertek, Invensys Spiral, Johnson
Matthey-INTERCAT, KBC Advanced Technologies, Nalco Champion, Schneider Electric, Soluble Solutions, Turner Mason, and others.
Don't miss out on opportunities to exchange ideas, share experiences, and network with professionals and executives working in upstream,
midstream, and downstream of the opportunity crudes and tight oil business.
Host
HYDROCARBON PUBLISHING COMPANY
Translating Knowledge into Profitability
SM

Sponsors Media Promoztional Partners
gas honeywell.indd 5 11/03/2014 11:52
Protecting compressors with dynamic
simulation
I
ntegrated dynamic simulation of
gas processing and petroleum
refning processes is vital in the
prevention of catastrophic equip-
ment occurrences. The protection of
compressors from mechanical fail-
ure is essential to maximise
operating time and ensure safer
operations.
This article highlights how
dynamic engineering software
successfully meets tighter product
specifcations through improved
understanding of plant operability
issues and ensures faster, safer
plant start-ups to help avoid
unplanned downtime. The results
deliver maximum plant availability
and productivity.
Better understanding leads to
better plant control
Understanding the process is the
frst step in the design of an effec-
tive control strategy. To prescribe a
solution, the process control engi-
neer must analyse all variables,
process dynamics and unit interac-
tions. First principle models (steady
state and dynamics) provide a
better understanding of the process
dynamics and its interactions,
enabling engineers to evaluate and
tune strategies before implementa-
tion. Using dynamic simulation
early in the design phase can help
identify important operability and
control issues and infuence the
design accordingly.
Compressors used to increase the
pressure of a gas are necessary in
many process designs. They are
highly specifc pieces of equipment
and often custom designed and,
therefore, expensive to purchase.
Compressors operate with fast
Dynamic simulation supports reliable operation of compressor installations to
deliver maximum plant availability and productivity
NICHOLAS BROWNRIGG
AspenTech
dynamics, meaning that a small
disturbance can disrupt a compres-
sor away from steady state
performance very quickly.
When steady state performance is
interrupted, a compressor can expe-
rience a phenomenon known as
surge. This can lead to malfunction-
ing or, worse, induce major
compressor damage requiring
equipment replacement. In addition,
a compressor surge can harm indi-
viduals working near compressors
or release noxious gases into the
surrounding atmosphere.
Using dynamic simulation can
help to predict compressors’ devia-
tion in behaviour from steady state
that could cause compressor surge.
Dynamic control schemes can also
be modelled to limit steady state
deviations if a disturbance is expe-
rienced in the plant.
The frst step in any dynamic
simulation is obtaining a steady
state model. A steady state
compressor model shows undis-
turbed compressor operation or, in
other words, how the compressor
would ideally operate if uninter-
rupted, given the process
parameters. To generate the steady
state compressor model in process
simulation, the inlet streams of the
compressor model must be defned,
as well as one of three compressor
specifcations: the change in pres-
sure across the compressor, the
compressor’s pressure ratio or the
energy supplied to the compressor.
Model based operations improve
decision support and safety
Compressors are vital to gas
processing and similar operations,
so it is important that models
include a built-in response to
potential disturbances, such as a
compressor surge. Dynamic
compressor models can optimise
design and operational perfor-
mance to achieve the following:
• Analyse a range of process
disturbance scenarios
• Add a proper control scheme to
appropriately respond to process
disturbances
• Perform process safety studies
• Finely tune process parameters to
achieve improved production results.
Process engineers use
AspenTech’s aspenONE solutions
to achieve model based operations
for decision support and safety. In
particular, Aspen Hysys Dynamics
extends Aspen Hysys steady state
models into dynamic process
models, enabling design and verif-
cation of process control schemes,
safety studies, relief valve sizing
and rating, failure analysis and
development of start-up, shutdown
and operating mode changes.
Engineers can create steady state
and dynamic models all within the
same simulation environment,
which saves signifcant time and
effort (see Figure 1).
Extensive numerical and graphical
results are generated for dynamic
models in Aspen Hysys Dynamics,
which allows for in-depth analysis
to be performed on each completed
simulation (see Figure 2).
Many compressor disturbances
can be explored, including decreas-
ing or increasing the feed rate to
the compressor and modifying the
composition or physical properties
of the compressor feed stream,
among others. Control schemes and
parameters can also be modifed,
54 Gas 2014 www.eptq.com
aspentech.indd 1 12/03/2014 12:49
www.eptq.com Gas 2014 55
behaviour, many production
processes in the oil and gas, energy
and chemicals industries will expe-
rience significant improvements in
control and safety. Knowing the
dynamic response of a compressor
is important when designing or
operating a process, and engineer-
ing dynamic software is a tool that
allows users to perform dynamic
simulations to obtain this informa-
tion. Engineering modelling
software provides multiple outlets
to view and tabulate the dynamic
response for equipment and
controllers, including strip charts,
tables and compressor curves.
When Hysys and Hysys Dynamics
are properly utilised conjointly,
compressor surge can be avoided.
Therefore, through better under-
standing of plant behaviour,
companies will save equipment
costs.
Verifying safe operation of a
plant revamp
Petronas, a Malaysian oil and gas
company, is one of the world’s
largest and most profitable compa-
nies. Petronas operates a liquefied
natural gas (LNG) plant that under-
went a revamp. While designing
the revamp, Petronas turned to
Aspen Hysys Dynamics to verify
that the new operating conditions
and control schemes, as well as
yielding more or less robust
compressor control, depending on
the user’s requirement. It is a
powerful tool that can be utilised to
ensure safe and proper functioning
of a process’s compressors.
Engineers can broaden their abil-
ity to design safer, more operable
plants, without over-designing, by
learning to do the following:
• Use data visualisation to help
optimise performance
• Implement and test control
schemes
• Schedule events to study safety
scenarios, start-up and more
• Switch steady state models to
dynamic mode for greater accuracy
and flexibility.
Reported benefits using dynamic
modelling can generate $15 million
savings through improved and
faster start-up procedures. Also, the
avoidance of over-designing relief
systems can achieve $10 million in
capital cost savings, improved
safety through better operational
procedures, better control system
design and proper relief-valve
sizing delivers enormous benefits.
Operators can achieve better design
decisions through detailed analysis
of the trade-offs between process
operability and process integration.
Safety and controllability mitigates
equipment malfunction
By understanding dynamic plant
9
3
2
6
5
N2
Q-23KA00
7 23L1005-03
4
1
8
Q-100
VLV-102
V-100
23FE1015
PIC-100
dP
scrubber
dP and vol
23Lxxxx
Bath tube
strainer
23KA001
23KA001:
surge controller
E-100
VLV-101
VLV-100
TEE-101
MIX-100
Molar flow kgmole/h
kPA
kPA
%
1900
1.073
192.7
100.00
Pressure drop
Feed pressure
Percentage open
23FE1015
* Actual feed volume flow
Compressor speed rpm
kW
ACT_m
3
/h
kPA
6285
3725
26159.5
183.6
Power
Capacity *
Feed pressure
kPA 1220 Product pressure
23KA001
Figure 1 Steady state model in Aspen Hysys
Figure 2 Simulation of a compressor installation in Aspen Hysys Dynamics
TRF-1
aspentech.indd 2 12/03/2014 12:49
56 Gas 2014 www.eptq.com
start-up and shutdown procedures,
would not lead to compressor
surge in one of the plant’s many
compressor units.
1
In Hysys Dynamics, models were
constructed that represented actual
equipment operation, piping
arrangements and potential control
schemes. Of note, a realistic emula-
tion of a controller used by
Petronas for anti-surge control in
compressors was replicated. In
addition, vendor predicted and as
tested operating curves for the axial
and centrifugal compressors in the
plant were input into the package
for enhanced prediction of
compressor behaviour.
During revamp studies, a variety
of new operating conditions and
failure scenarios for the plant was
considered. The use of dynamic
simulation allowed for the testing
of the operating conditions to
ensure that they would not cause
disruptions with equipment perfor-
mance and, in particular,
compressor surge. Failure scenario
analysis enabled Petronas to view
the operating limits of the compres-
sors along with specifc
circumstances that would directly
lead to surge in order to better
avoid them. Failure scenario analy-
sis also helped to confrm that relief
valves were adequately sized to
handle peak relief loads.
While performing dynamic simu-
lation, it was elsewhere discovered
that a propane compressor did, in
fact, experience surge upon the
closure of suction valves on a
chiller and also during a trip of a
cryogenic heat exchanger. This risk
was quickly mitigated since it was
uncovered early in the design
process, before implementation of
equipment. In all, over 45 different
scenarios, along with the start-up
and shutdown conditions of the
plant, were simulated and analysed
in an iterative process utilising the
event scheduler feature inside
Hysys Dynamics. Using the pack-
age to carry out dynamic
simulation for the revamp of the
plant ultimately lent a better insight
into the dynamic performances that
new plant designs would exhibit.
By knowing where potential
dangers in the process lay, Petronas
was able to better size and operate
equipment in order to avoid the
occurrence of a catastrophic event.
Operability analysis for emergency
system responses
Wood Group Production Services
Network (Wood Group PSN) is a
consulting group that assists oil
and gas companies in attaining
better performance and production,
from process development to oper-
ational diagnostics. Due to the
complex nature of safety systems,
performing safety system analysis
is a task that Wood Group PSN
often executes.
Safety systems are often designed
more conservatively than required
in order to protect against worst
case scenarios. However, the worst
case scenario calculation is a poor
representation of actual process
circumstances and it does not offer
any insights regarding how a
process behaves on its way to
reaching the worst case. Safety
systems’ performance and sizing
can be vastly improved through
dynamic simulation to ascertain
answers into how processes actu-
ally respond to disturbance. For
this reason, Wood Group PSN was
tasked with creating dynamic simu-
lations to model the response to
steady state disturbance of emer-
gency systems present in a plant
with the goal of ultimately enhanc-
ing the safety of the process. To
accomplish this, the company chose
to use Hysys Dynamics.
A comprehensive model of the
plant’s safety system was devel-
oped to investigate equipment
failures as conditions moved away
from the steady state and to evalu-
ate alternative process
confgurations towards bettering
safety system effectiveness.
Compressors were heavily featured
in many of the dynamic models
created, and the performance and
effciency curves for the compres-
sors, as well as the inertia and
friction profles for each compres-
sor, were included to better tailor
the model to ft the actual process.
Studying failure scenarios
allowed Wood Group PSN to view
which circumstances initiated
compressor surge and to what
degree. Accordingly, the safety
margins of the compressor units
were increased until confrmed to
protect against the occurrence of
surge. Using dynamic simulation
enabled the safety margins to be
increased up to the critical level of
protection based on predicted
process responses and eliminated
the need for over-design to provide
a blanket protection.
A colour scheme was imple-
mented on the dynamic simulations
to easily identify when material
became trapped and caused pres-
sure build-up when malfunctions
occurred. Identifying where block-
ages in operation take place helped
to prevent surge, since fow disrup-
tions is one of the main causes for
breakdown. Additionally, using
dynamic simulation eliminated the
need for the installation of a hot
gas bypass that served to prevent
compressor malfunctioning. This
reduction in design saved over $3.2
million in capital expenditure for
the plant. The requirement of the
hot gas bypass is an example of the
over-design that can materialise
when worst case scenario analysis
is used for the creation of safety
systems and how the use of
dynamic simulation can help
determine where unnecessary
components of the safety systems
exist.
The models created in Hysys
Dynamics were also able to serve
as a basis for operator training and
on-going development in the plant.
With an improved knowledge of
how the plant runs, operators were
better at steering the process to
avoid perturbations from steady
state and how to best remedy situa-
tions that have left steady state
operation.
References
1 AspenTech Global Conference, 2010.
2 AspenTech User Conference, 2009.
Nicholas Brownrigg is Dynamics Product
Manager, AspenTech.
aspentech.indd 3 12/03/2014 12:49
2014 Speakers include:
Boost your refining profitability through operational
excellence and increased product yields
8th Annual
Global Refining Summit
31 March-2 April 2014
Hesperia Tower, Barcelona, Spain
Save £300
*

when you book your place by 31 March 2014
refiningsummit.com/PTQ
*The offer is for end-users only and is not available to equipment manufacturers and solution providers.
enquire@wtgevents.com | +44 (0) 20 7202 7690 | refiningsummit.com
Lars Thorstholm
Manager Strategy
Statoil ASA
Giacomo Rispoli
Executive Vice President
eni S.p.A.
Rustem Gimaletdinov
Head of Department, Technical &
Engineering Excellence, Global Refining,
Petrochemicals & Gas Processing
JSC LUKOIL
Rob Overtoom
Technology Manager Pearl GTL, Qatar
Shell
Marcus Lippold
Vice President, Business Strategy
and Regulatory and Public Affairs
MOL Group
Mohamed Saad Al-Kubaisi
Technical Consultant
Qatar Petroleum Refinery
global summit.indd 1 04/03/2014 11:08
Your objectives
in focus
Make the most of today’s and tomorrow’s
challenges with leading-edge solutions from Axens
- Clean and alternative fuel technologies
- Petrochemicals
- Energy efficiency
- High performance catalysts & adsorbents
- Revamps
Single source technology and service provider
ISO 9001 – ISO 14001 – OHSAS 18001
www.axens.net
210/297.qxd:A4 13/08/10 12:12 Page 1
www.ptqenquiry.com
for further information
axens.indd 1 13/9/10 13:27:18