PTQ Gas 2011

gas
2011
PTQ supplement
gas cov copy 3.indd 1 9/3/11 13:04:31
h yd ro g e n -
syn th e sis g a s to c h e m ic a ls -
syn th e sis g a s to fu e ls -
a m m o n ia -
m e th a n o l -
D M E -
Haldor Topse offers you the technology, catalyst and experience for protable use of your natural gas feed
whether you are seeking a new plant or a revamp of an existing plant.
Through research-driven technologies, Haldor Topse offers impartial counselling on the selection of the best technology
and catalyst for your plant.
WWW. T O P S O E . C O M
Creating excellence from raw materials
you need a specialist
Add value to your natural gas
with Topse Technologies
www.ptqenquiry.com
for further information
haldor.indd 1 8/3/11 13:27:06
2011. The entire content of this publication is protected by copyright full details of which are available from the publishers. All rights
reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means
electronic, mechanical, photocopying, recording or otherwise without the prior permission of the copyright owner.
The opinions and views expressed by the authors in this publication are not necessarily those of the editor or publisher and while
every care has been taken in the preparation of all material included in Petroleum Technology Quarterly the publisher cannot be held
responsible for any statements, opinions or views or for any inaccuracies.
5 Amines in bad odour
ChrisCunningham
7 Enhanced sulphur recovery from lean acid gases
AngelaSlavensandJustinLamarBlack & Veatch
DjordjeLNikolicandTheoBrokShell Global Solutions International

19 Cryogenic monitoring
JerryWorsleySensornet
25 Glycol dehydration of high-acid gas streams
JohnCarrollGas Liquids Engineering
NathanHatcherandRalphWeilandOptimized Gas Treating
35 Controlling corrosion in amine treatment units
JayaRawat,PeddyVCRaoandNVChoudary
Bharat Petroleum Corporation Ltd
41 A promoter for selective H
2
S removal: part I
GeraldVorberg,RalfNotzandTorstenKatzBASF SE
WielandWacheandClausSchunkBayernoil Raffneriegesellschaft

48 Estimating gas hydrate inhibitor loss: a case study
AlirezaBahadoriCurtin University
ThePearlgas-to-liquidsproject,RasLaffan,Qatar. Photo: Shell
2011
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
contents/ed com copy 8.indt 1 9/3/11 14:40:37
P: 847.285.3850 | E: mgtsales@merichem.com | www.merichem.com
With a wide range of system capabilities,
Merichem Company offers both liquid
and solid media desulfurization
technologies to sweeten gas and air
streams containing virtually all levels
of hydrogen sulfide.
Our extensive experience includes
treatment of gasification syngas, natural
gas, amine acid gas, sour water stripper
gas, fuel gas, and Claus tail gas.
LO-CATs flexible technologies and
worldwide services are backed by over
three decades of reliability. J oin 200+
licensees in 29 countries on 6 continents
Problem: Sour Gas
Sweetest Solution: LO-CAT
H
2
SRemoval/Recovery
that are meeting their treating needs
by turning H
2
S into elemental sulfur
efficiently and effectively with LO-CAT
.
Learn how sweet it is, visit
www.merichem.com/LO-CAT
Merichem: A global provider of focused
hydrocarbon-treating technology,
chemical and service solutions.
MERICHEM COMPANY
Sweet Solutions.
C
M
Y
CM
MY
CY
CMY
K
MER-0308 LO-CAT Chemist PTQ Spread.pdf 1 1/7/11 3:55 PM
merichem.indd 2 8/3/11 13:29:37
P: 847.285.3850 | E: mgtsales@merichem.com | www.merichem.com
With a wide range of system capabilities,
Merichem Company offers both liquid
and solid media desulfurization
technologies to sweeten gas and air
streams containing virtually all levels
of hydrogen sulfide.
Our extensive experience includes
treatment of gasification syngas, natural
gas, amine acid gas, sour water stripper
gas, fuel gas, and Claus tail gas.
LO-CATs flexible technologies and
worldwide services are backed by over
three decades of reliability. J oin 200+
licensees in 29 countries on 6 continents
Problem: Sour Gas
Sweetest Solution: LO-CAT
H
2
SRemoval/Recovery
that are meeting their treating needs
by turning H
2
S into elemental sulfur
efficiently and effectively with LO-CAT
.
Learn how sweet it is, visit
www.merichem.com/LO-CAT
Merichem: A global provider of focused
hydrocarbon-treating technology,
chemical and service solutions.
MERICHEM COMPANY
Sweet Solutions.
C
M
Y
CM
MY
CY
CMY
K
MER-0308 LO-CAT Chemist PTQ Spread.pdf 1 1/7/11 3:55 PM
www.ptqenquiry.com
merichem.indd 3 8/3/11 13:29:51
Simplify
sulfur recovery
and cut your costs
Single source ISO 9001 technology and service provider
www.axens.net
Beijing +86 10 85 27 57 53 Houston +1 713 840 11 33
Moscow +7 495 933 65 73 Paris +33 1 47 14 25 14 Tokyo +81 335 854 985
Low-temperature tail-gas hydrogenation catalysts
that deliver superior and cleaner performance from simplified operations, and significantly
lower CO
2
emissions. Axens TG catalysts can work with steam reheating technologies to reduce
energy consumption. Operating costs and investments are also reduced. Its a winner every time.
www.ptqenquiry.com
axens.indd 1 10/12/09 12:59:39
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 2011 5
Editor
Chris Cunningham
editor@petroleumtechnology.com
Production Editor
Rachel Zamorski
production@petroleumtechnology.com
Graphics Editor
Rob Fris
graphics@petroleumtechnology.com
Editorial
tel +44 844 5888 773
fax +44 844 5888 667
Advertising Sales Manager
Paul Mason
sales@petroleumtechnology.com

Advertising Sales
Bob Aldridge
sales@petroleumtechnology.com
Advertising Sales Offce
tel +44 844 5888 771
fax +44 844 5888 662
Publisher
Nic Allen
publisher@petroleumtechnology.com
Circulation
Jacki Watts
circulation@petroleumtechnology.com
Crambeth Allen Publishing Ltd
Hopesay, Craven Arms SY7 8HD, UK
tel +44 844 5888 776
fax +44 844 5888 667
ISSN 1362-363X

Petroleum Technology Quarterly (USPS 0014-781)
is published quarterly plus annual Catalysis edition
by Crambeth Allen Publishing Ltd and is distributed
in the USA by SPP, 75 Aberdeen Rd, Emigsville, PA
17318. Periodicals postage paid at Emigsville PA.
Postmaster: send address changes to Petroleum
Technology Quarterly c/o PO
Box 437, Emigsville, PA 17318-0437
Back numbers available from the Publisher
at $30 per copy inc postage.
Aminesinbad
odour
A
mine capture of acid gases and amine plants themselves are not
unexpected topics for discussion by industry experts in this issue of PTQ
Gas. However, recent news has focused not so much on amine technology
for dealing with acid gas components of natural gas as on its application to the
downstream capture of gas combustion products, meaning, of course, CO
2
. With
carbon issues looming over the hydrocarbons industry in all of its forms, the story
of the Mongstad gas power plant in Norway, which includes a fagship project to
demonstrate large-scale capture and storage (CCS) of CO
2
, is causing no end of
confusion for climate change policy makers in Europe.
DONG Energy and Statoil have been working at Mongstad to build a ther-
mal power plant, with the aim of integrating it into Statoils neighbouring oil
refnery, while DONG operates and owns the plant. It has been built to supply
Mongstad and a production platform in the North Sea with electricity and
heat, and to supply a gas treatment plant with electricity. So far so good.
One of the chief ideas behind Mongstad is to strengthen the refnerys posi-
tion as an effcient supplier of refned products. At maximum utilisation, the
power plant can deliver 280 MW of electricity and about 350 MW of heat. Now
it is up and running, it is no doubt meeting all of those targets. All of the CO
2

it currently produces is released to the atmosphere, which is fne at present
because full-scale CCS was not planned to begin until 2014.
But this is where the controversy begins. The CCS arm of the venture was
frst delayed to 2018 and most recently, on Statoils advice, the Norwegian
government has delayed it further. According to Norways Petroleum and
Energy Ministry, that advice centres not on the technical issues of making a
CCS project work but on whether amine technology can be used at all, on
grounds of health and safety. There is a need for greater knowledge of the
effects of using amine technology on health and the environment, says the
ministry. Now the government wants to look at alternative options for imple-
menting CCS.
While this has, not unexpectedly, put Statoil in bad odour with local environ-
mentalists, who were just about prepared to put up with another major
gas-fred power plant provided the CO
2
it produces could be effectively
contained, the ramifcations travel much further. The EU is preparing to imple-
ment a large-scale programme to demonstrate CCS. Following the European
tradition for these matters, individual countries will demonstrate their own
fagship projects. The most successful are likely to become standards for future
implementation among the 27 states of the EU.
That programme is scheduled to begin in 2015, which puts Mongstad out the
running. To make matters worse for the Norwegian government, it was being
relied on to supply a key demonstration of CCS in a major gas power plant,
which leaves a signifcant gap in the programme once it begins. A project
touted by the national government as a Moon landing when it was frst
announced looks to be frmly Earth-bound.
Why has there been further delay to the Mongstad project? Statoil is increas-
ingly worried, apparently, about the production of nitrosamines as a
component of the process of amine capture of CO
2
and the effects of these
substances on human health, which many view with more than a pinch of
scepticism.
CHRIS CUNNINGHAM
p
t
q
Y L R E T R A U Q Y G O L O N H C E T M U E L O R T E P
contents/ed com copy 8.indt 2 9/3/11 13:07:16
Bryan Research & Engineering, Inc.
P.O. Box 4747 Bryan, Texas USA 77805
979-776-5220 www.bre.com sales@bre.com
Selectingthe Best Solvent for Gas Treating
PROCESS INSIGHT
Selecting the best amine/solvent for gas treating is not a trivial task.
There are a number of amines available to remove contaminants such
as CO
2
, H
2
S and organic sulfur compounds from sour gas streams.
The most commonly used amines are methanolamine (MEA),
diethanolamine (DEA), and methyldiethanolamine (MDEA). Other
amines include diglycolamine
(DGA), diisopropanolamine (DIPA),

and triethanolamine (TEA). Mixtures of amines can also be used to
customize or optimize the acid gas recovery. Temperature, pressure,
sour gas composition, and purity requirements for the treated gas
must all be considered when choosing the most appropriate amine for
a given application.
Primary Amines
The primary amine MEA removes both CO
2
and H
2
S from sour gas
and is effective at low pressure. Depending on the conditions, MEA
can remove H
2
S to less than 4 ppmv while removing CO
2
to less than
100 ppmv. MEA systems generally require a reclaimer to remove
degraded products from circulation. Typical solution strength ranges
from 10 to 20 weight % with a maximum rich loading of 0.35 mole
acid gas/mole MEA. DGA
is another primary amine that removes

CO
2
, H
2
S, COS, and mercaptans. Typical solution strengths are 50-
60 weight %, which result in lower circulation rates and less energy
required for stripping as compared with MEA. DGA also requires
reclaiming to remove the degradation products.
Secondary Amines
The secondary amine DEA removes both CO
2
and H
2
S but generally
requires higher pressure than MEA to meet overhead specications.
Because DEA is a weaker amine than MEA, it requires less energy for
stripping. Typical solution strength ranges from 25 to 35 weight % with
a maximum rich loading of 0.35 mole/mole. DIPA is a secondary amine
that exhibits some selectivity for H
2
S although it is not as pronounced
as for tertiary amines. DIPA also removes COS. Solutions are low
in corrosion and require relatively low energy for regeneration. The
most common applications for DIPA are in the ADIP
and SULFINOL

processes.
Tertiary Amines
A tertiary amine such as MDEA is often used to selectively remove
H
2
S, especially for cases with a high CO
2
to H
2
S ratio in the sour gas.
One benet of selective absorption of H
2
S is a Claus feed rich in H
2
S.
MDEA can remove H
2
S to 4 ppm while maintaining 2% or less CO
2
in
the treated gas using relatively less energy for regeneration than that
for DEA. Higher weight percent amine and less CO
2
absorbed results
in lower circulation rates as well. Typical solution strengths are 40-50
weight % with a maximum rich loading of 0.55 mole/mole. Because
MDEA is not prone to degradation, corrosion is low and a reclaimer
is unnecessary. Operating pressure can range from atmospheric,
typical of tail gas treating units, to over 1,000 psia.
Mixed Solvents
In certain situations, the solvent can be customized to optimize the
sweetening process. For example, adding a primary or secondary
amine to MDEA can increase the rate of CO
2
absorption without
compromising the advantages of MDEA. Another less obvious
application is adding MDEA to an existing DEA unit to increase the
effective weight % amine to absorb more acid gas without increasing
circulation rate or reboiler duty. Many plants utilize a mixture of amine
with physical solvents. SULFINOL is a licensed product from Shell Oil
Products that combines an amine with a physical solvent. Advantages
of this solvent are increased mercaptan pickup, lower regeneration
energy, and selectivity to H
2
S.
Choosing the Best Alternative
Given the wide variety of gas treating
options, a process simulator that can
accurately predict sweetening results is a
necessity when attempting to determine
the best option. ProMax
has been proven

to accurately predict results for numerous
process schemes. Additionally, ProMax
can utilize a scenario tool to perform
feasibility studies. The scenario tool may
be used to systematically vary selected
parameters in an effort to determine the
optimum operating conditions and the appropriate solvent. These
studies can determine rich loading, reboiler duty, acid gas content of
the sweet gas, amine losses, required circulation rate, type of amine
or physical solvent, weight percent of amine, and other parameters.
ProMax can model virtually any ow process or conguration
including multiple columns, liquid hydrocarbon treating, and split ow
processes. In addition, ProMax can accurately model caustic treating
applications as well as physical solvent sweetening with solvents such
as Coastal AGR
, methanol, and NMP. For more information about

ProMax and its ability to determine the appropriate solvent for a given
set of conditions, contact Bryan Research & Engineering.
www.ptqenquiry.com
bre.indd 1 8/3/11 13:36:38
Enhanced sulphur recovery from
lean acid gases
T
he exploitation of increasingly
diffcult natural gas reserves
has risen in recent years,
requiring the removal of a number
of harder to remove sulphur
species, as well as H
2
S, CO
2
and
H
2
O. In addition, maximum limits
for sulphur content in pipeline gas
continue to tighten. As a result, the
treating requirements for gas feld
development projects have signif-
cantly increased in complexity,
often requiring a combination of
process steps and units. As the
development of diffcult sour gas
felds is expected to further increase
in the future, strategic integration
of various gas-treating process units
is necessary to achieve an overall
optimised fow sheet with lowest
lifecycle cost. The ability to combine
technologies and process units in
the optimum confguration results
in a competitive advantage for
addressing challenges and opportu-
nities posed by undeveloped sour
gas felds.
In 2010, a study by Black &
Veatch compared several alterna-
tive Claus sulphur recovery unit
(SRU)/tail gas-treating unit (TGTU)
confgurations for achieving high
sulphur recovery and reliable oper-
ation from lean acid gas containing
high concentrations of carbonyl
sulphide (COS) and mercaptans.
The lean acid gas stream that was
used as the basis for comparison of
the process confguration test cases
had an H
2
S concentration of 25
mole% and an organic sulphur
concentration of 1 mole%. The
study noted that a lean acid gas
stream such as this could be
produced from sweetening natural
gas with a high COS/mercaptan
A preferred acid gas enrichment scheme is combined with two gas-treating
schemes to examine the lifecycle cost of a sour natural gas treatment facility
AngelA SlAvenS and JuSTin lAmAr Black & Veatch
DJorDJe l nikolic and Theo Brok Shell Global Solutions International
content and affrmed that, at such a
low H
2
S concentration, reliable
Claus unit operation can be
diffcult.
The 2010 study considered the
use of the acid gas enrichment
(AGE) process to increase acid gas
H
2
S concentration, as a means to
alleviate the problems associated
with the lean acid gas feed to the
Claus SRU/TGTU. It also addressed
the fact that the conventional AGE
processing scheme cannot achieve
high sulphur recovery when COS
and mercaptans are present, due to
the fact that selective treating
solvents commonly utilised for acid
gas enrichment do not absorb these
organic sulphur species and allow
them to slip to the incinerator.
Several acid gas enrichment
processing schemes were compared,
and it was concluded that when
acid gas COS/mercaptan levels are
high enough to reduce sulphur
recovery with conventional acid gas
enrichment below an acceptable
level, an alternative enrichment
design confguration should be
considered.
The purpose of the previous
study was to compare alternative
acid gas enrichment processing
schemes for a given lean acid gas
stream composition and fow rate,
but not to consider the impact that
the upstream acid gas removal unit
(AGRU) solvent selection can have
on the acid gas quality. However,
in most situations, it is important to
consider the overall sour gas treat-
ment fow scheme rather than
simply evaluating the acid gas
processing in isolation, as synergies
often exist between the various
processing units (AGRU/AGE/
SRU/TGTU) that provide opportu-
nities for optimisation of the overall
fowsheet.
This article compares three differ-
ent Sulfnol AGRU options for
development of the overall fow
scheme for treatment of a sour
natural gas stream containing
signifcant quantities of organic
sulphur (methyl and ethyl
mercaptan). Although COS was
an additional organic sulphur
compound considered in the 2010
study, it would typically not be
present in acid gas from the over-
head of a chemical amine solvent or
hybrid chemical/physical solvent
regenerator. This is primarily due
to the fact that absorbed COS is
hydrolysed to H
2
S and CO
2
in the
solvent regeneration step of the
AGRU. All Sulfnol solvents remove
COS to a large extent in the main
absorber of the AGRU. With
Sulfnol-X, deep removal of COS
can be achieved due to the presence
of piperazine, which enhances COS
hydrolysis. With Sulfnol-M, most
of the COS would be absorbed due
to the larger number of trays
www.eptq.com GAS 2011 7
it is important to
consider the overall
sour gas treatment
fow scheme
rather than simply
evaluating the acid gas
processing in isolation
b&v.indd 1 9/3/11 13:01:07
employed. Since Sulfnol is a hybrid
chemical/physical solvent, for the
reasons described above, COS
captured in the absorber will not be
present in the acid gas. Therefore, it
is anticipated that the small varia-
tions of COS in the sales gas and
8 GAS 2011 www.eptq.com
treated fash gas would not be a
differentiator between the three
cases studied in this article; hence,
COS has been omitted from the
analysis.
The resulting acid gas from each
AGRU confguration is processed
utilising some of the acid gas treat-
ment schemes presented in the 2010
article. The optimum confgura-
tions, in terms of relative Capex
and Opex, are presented.
Test cases
The sour natural gas stream used
for comparison of the various proc-
ess confguration test cases is shown
in Table 1. Sulphur content is 96.1
t/d. About 2% of the acid gas
sulphur is present as mercaptans;
therefore, recovery of mercaptan
sulphur is important if high sulphur
recovery is to be achieved.
The treated gas specifcations
used for the comparison test cases
are shown in Table 2. The specifca-
tions considered are typical for
pipeline-quality natural gas. Table 3
and Table 4 describe the three
process confguration test cases
compared in this article, each of
which is illustrated in Figures 13.
The Case 1 process confguration
is shown in Figure 1. Sour natural
gas is processed in a Sulfnol-X
AGRU, producing a lean acid gas
stream containing H
2
S and
mercaptans. Sulfnol-X technology
employs piperazine accelerator in
addition to MDEA, sulpholane and
water. This solvent formulation
meets both low H
2
S and CO
2
speci-
fcations, as well as low
specifcations of trace sulphur
species (such as mercaptans, COS
and organic sulphides). The
Sulfnol-X unit has a standard
confguration consisting of a main
absorber, hydrocarbon fash vessel
with associated absorber and a
regeneration section. The hydrocar-
bon fash minimises co-absorbed
hydrocarbons in the acid gas. The
lean acid gas stream from the
AGRU regenerator is processed in a
conventional two-bed Claus SRU.
SRU tail gas fows to an MDEA-
based Shell Claus Offgas Treatment
(SCOT)1 TGTU, where residual
sulphur species in the SRU tail gas
are converted to H
2
S, which is
subsequently removed in the down-
stream TGTU absorber. The TGTU
regenerator regenerates the rich
MDEA solvent, and the overhead
from this tower is routed to the
front of the SRU for the recovery of
sulphur in this stream.
Table 3
Table 4
Component Sour natural gas, lbmoles/hr Sour natural gas, mole fraction
H
2
O 10.63 0.126
CO
2
784.69 9.300
H
2
S 270.00 3.200
N
2
421.88 5.000
C1 6618.37 78.440
C
2
210.94 2.500
C
3
101.26 1.200
iC
4
4.21 0.050
C
4
4.21 0.050
iC
5
1.26 0.015
nC
5
1.26 0.015
C
6
3.37 0.040
CH
4
S 4.06 0.048
C
2
H
6
S 1.34 0.016
Total 8437.47 100.000
Temperature, F 95
Pressure, psia 870
Sour natural gas feed stream used for comparison of test cases
Table 1
Gas stream Specifcation Value
Sales gas H
2
S, ppmv <5
Total sulphur, ppmv <20
CO
2
, mole% <1.0
Fuel gas H
2
S, ppmv <100
Tail gas ex-TGTU absorber H
2
S, ppmv <250
Treated gas specifcations
Table 2
Test case Description
1 Sulfnol-X main gas treating, conventional SRU/SCOT1 using MDEA
2 Sulfnol-X main gas treating, SRU/SCOT using Sulfnol-M acid gas enrichment
and tail gas treating, routing of enrichment absorber overhead to SCOT
3 Sulfnol-M main gas treating with integrated enrichment step and tail gas
absorber, conventional SRU/SCOT using Sulfnol-M, routing of enrichment
absorber overhead to SCOT
Test cases for comparison
Table 3
Test case AGRU solvent AGE solvent TGTU solvent AGE overhead routing
1 Sulfnol-X MDEA
2 Sulfnol-X Sulfnol-M Sulfnol-M To TGTU, per Case 4 from 2010 article
3 Sulfnol-M Sulfnol-M Sulfnol-M To TGTU, per Case 4 from 2010 article
Test case process confgurations
Table 4
b&v.indd 2 9/3/11 14:41:41
increase acid gas strength for
improved Claus operation.
The Case 2 process confguration
is shown in Figure 2. Sour natural
gas is processed in a Sulfnol-X acid
gas removal unit similar to that in
Case 1. Acid gas from the Sulfnol-
X unit is subsequently processed in
an enrichment absorber, which
absorbs H
2
S from the acid gas using
Sulfnol-M solvent. Sulfnol-M tech-
nology employs MDEA, sulpholane
and water, without piperazine, as is
the case with Sulfnol-X. This results
in AGE and TGTU absorbers, which
remove H
2
S and mercaptans while
minimising CO
2
absorption.
Reduced CO
2
solubility is desirable
in the AGE and TGTU absorbers to
prevent excessive CO
2
fow through
the SRU and TGTU, which increases
their size. AGE absorber overhead
gas, containing most of the CO
2
in
the acid gas as well as other compo-
nents not absorbed by the
Sulfnol-M solvent, fows to the
TGTU hydrogenation reactor rather
than to the incinerator, similar to
Cases 4A, 4B and 4C presented in
the 2010 article. Routing the AGE
overhead to the hydrogenation
reactor allows conversion of the
Figure 1 Case 1 process confguration: Sulfnol-X gas treating with conventional SRU/SCOT
Tre a te d
n a tu ra l
g a s
To fu e l
S o u r
n a tu ra l
g a s
A ir
Main
absorber
Flash
drum
Regenerator
Thermal
reactor
Inline
burner/mixer
Quench
column
WHB
Claus converters
A ir
S u lp h u r S u lp h u r S u lp h u r
A ir
F u e l
Incinerator
F u e l
Hydrogenation
reactor
TGTU
absorber
Regenerator
Case 1 represents a conventional
base case to compare with Cases 2
and 3. The Case 1 AGRU produces
a lean acid gas containing nomi-
nally 25 mole% H
2
S and 0.5 mole%
mercaptans, an acid gas stream
similar to that used for comparison
of the test cases explored in the
2010 article, with the absence of
COS. As explained in the 2010 arti-
cle, acid gas at or below this
concentration makes sustainable
Claus performance diffcult due to
a low thermal reactor temperature
and acid gas fame instability. Cases
2 and 3 utilise enrichment to
b&v.indd 3 9/3/11 16:22:19
Tre a te d
n a tu ra l
g a s
To fu e l
S o u r
n a tu ra l
g a s
A ir
Main
absorber
Flash
drum
Regenerator
Thermal
reactor
Inline
burner/mixer
Quench
column
WHB
Claus converters
A ir
A ir
F u e l
Incinerator
F u e l
Hydrogenation
reactor
Inline
burner/mixer
A ir
F u e l
Hydrogenation
reactor
AGE
absorber
TGTU
absorber
Regenerator
Figure 2 Case 2 process confguration: Sulfnol-X gas treating with SRU using enrichment, enrichment absorber overhead routed to SCOT
mercaptans in this stream to H
2
S,
for absorption in the downstream
TGTU absorber, which also employs
Sulfnol-M solvent. Concentrated
acid gas fows from the Sulfnol-M
regenerator to the SRU/TGTU for
sulphur recovery.
Routing of an enrichment
absorber overhead stream to a
TGTU was patented by Shell in
1982. This feature is crucial to
achieving high sulphur recovery
effciency when acid gas for enrich-
ment contains sulphur species other
than H
2
S. With conventional acid
gas enrichment, these other sulphur
species are not absorbed using
selective treating solvents and
therefore fow unrecovered to the
incinerator. Routing the enrichment
absorber overhead to the TGTU
allows conversion of these other
sulphur species to H
2
S, which can
be recovered in the TGTU absorber.
In order to avoid premature ageing
of the hydrogenation catalyst due
to the presence of heavier hydrocar-
bons, two parallel hydrogenation
catalyst beds are employed, one for
the main Claus tail gas and one for
b&v.indd 4 9/3/11 16:22:29
risk of contamination/premature
ageing to the AGE overhead reac-
tor, which is the smaller of the two.
If the need for premature change-
out of the catalyst in this reactor
arose, it would not result in signif-
cant losses in overall sulphur
recovery effciency.
the AGE overhead gas. This is a
deliberate deviation from the 2010
study, which considered only a
single hydrogenation reactor for all
cases, which routed AGE overhead
gas through the TGTU. The reason
for the inclusion of parallel hydro-
genation reactors is to restrict the
Tre a te d
n a tu ra l
g a s
To fu e l
S o u r
n a tu ra l
g a s
A ir
Main
absorber
Flash
drum
Hot flash drum
Thermal
reactor
Quench
column
WHB
Claus converters
A ir
F u e l
Incinerator
TGTU
absorber
Regenerator
Inline
burner/mixer
A ir
F u e l
Hydrogenation
reactor
Inline
burner/mixer
A ir
F u e l
Hydrogenation
reactor
Enrichment
absorber
It should be noted that the
Sulfnol-M solvent circulation
system is integrated between the
AGE and TGTU absorbers similar
to the Case 4C confguration in the
2010 study; 100% of the semi-rich
solvent from the TGTU absorber is
cascaded to the AGE absorber. Case
Figure 3 Case 3 PFD: Sulfnol-M gas treating with integrated enrichment step and tail gas treating, enrichment absorber overhead
routed to SRU/SCOT
b&v.indd 5 9/3/11 16:22:38
AGRU
(Sulfinol-X)
AGRU
regenerator
(Sulfinol-X)
SRU
Sour
natura|
gas
275.40
Treated
natura|
gas
0.04
HP ash
gas
0.08
Rich
275.87
Lean
0.59
Acid gas
275.28
TGTU
(MDEA)
SRU
tai| gas
24.42
Recyc|e acid gas
23.82
Incinerator
TGTU
tai| gas
0.59
Stack
gas
0.72
Su|phur
product
274.68
Su|phur
pit vent
0.05
Figure 4 Sulphur balance for Case 1: Sulfnol-X gas treating with conventional SRU/SCOT (sulphur fows in lbmoles/hr)
AGRU
(Su|no|-X}
AGRU
regenerator
(Su|no|-X}
SRU
AGE
(Su|no|-M}
Amine
regenerator
(Su|no|-M}
Sour
natura|
gas
275.40
Treated
natura|
gas
0.04
HP ash
gas
0.08
AGE absorber overhead
7.69
Rich
275.87
Lean
0.59
Acid gas
275.28
TGTU
(Su|no|-M}
SRU
tai| gas
16.32
Incinerator
TGTU
tai| gas
0.56
Stack
gas
0.69
Su|phur
product
274.72
Su|phur
pit vent
0.05
Acid gas
291.04
Rich
291.85
Lean
0.82
Semi-
rich
24.27
Figure 5 Sulphur balance for Case 2: Sulfnol-X gas treating with SRU using Sulfnol-M enrichment, AGE overhead routed to SCOT
(sulphur fows in lbmoles/hr)
b&v.indd 6 9/3/11 13:01:55
Sulfnol-M solvent from all three
absorbers. Concentrated acid gas
fows from the Sulfnol-M regenera-
tor to the SRU/TGTU for sulphur
recovery.
Compared with Case 2, Case 3
has one solvent system instead of
two, which eliminates one regenera-
tor and reduces the equipment
count in the solvent storage and
drain systems, as well as reducing
operational complexity. The formu-
lation of Sulfnol-M allows this
solvent to be used both for the
selective and non-selective removal
of H
2
S in the presence of CO
2
,
depending on the process condi-
tions. While Sulfnol-M cannot
reach the LNG-like deep CO
2
removal specifcation that can be
achieved with Sulfnol-X, the speci-
fcation of <1 mole% CO
2
selected
for this article can be easily
achieved.
Comparison of sulphur recovery
effciency
The overall sulphur recovery for
each of the test cases is given in
Table 5. The H
2
S removal and
sulphur recovery steps in each case
have been modelled on a consistent
basis. As a result, differences
between cases refect their particu-
lar process characteristics, without
infuence from the inconsistencies
associated with variations in param-
eters such as H
2
S removal effciency,
4C was the optimal case from the
2010 article, hence this confgura-
tion was selected without further
analysis of Case 4A or 4B confgu-
rations in this evaluation.
The Case 3 process confguration,
which integrates acid gas enrich-
ment within the AGRU, is shown in
Figure 3. Sour natural gas is proc-
essed in a Sulfnol-M AGRU and
the resulting acid gas is processed
in a conventional two-bed Claus
SRU. The Sulfnol-M AGRU
contains an enrichment step, which
consists of an additional hot fash
vessel and an enrichment absorber.
In this enrichment section, rich
solvent is heated and fashed at
reduced pressure. The CO
2
-rich
fash vapour fows to an enrichment
absorber, where lean Sulfnol-M
solvent absorbs H
2
S and some
mercaptans. The enrichment
absorber overhead fows to a sepa-
rate hydrogenation section of the
TGTU, for conversion of mercaptans
to H
2
S, similar to the Case 2 fow-
sheet. Also similar to Case 2, the
Sulfnol-M solvent circulation
system is integrated between the
AGE and TGTU absorbers, as per
the Case 4C confguration in the
2010 study. Case 3 employs a single
regenerator for regeneration of the
AGRU
(Sulfinol-M)
AGRU
regenerator
(Sulfinol-M)
SRU
Sour
natura|
gas
275.40
Treated natura| gas
0.04
HP ash
gas
0.03
Rich
276.35
Lean
1.02
Acid gas
291.84
TGTU
(Sulfinol-M)
SRU
tai| gas
17.04
Incinerator
TGTU
tai| gas
0.53
Stack
gas
0.61
Su|phur
product
274.80
Su|phur
pit vent
0.05
Lean
1.80
Rich
24.63
Enrichment absorber overhead
6.32
Figure 6 Sulphur balance for Case 3: Sulfnol-M gas treating with integrated enrichment and tail gas treating, AGE overhead routed to
SCOT (sulphur fows in lbmoles/hr)
Table 5
Case 1 2 3
HP fash gas to incinerator, lbmol/hr 0.083 0.080 0.031
TGTU H
2
S to incinerator, lbmol/hr 0.425 0.410 0.391
TGTU COS to incinerator, lbmol/hr 0.163 0.152 0.143
Pit sweep sulphur to incinerator, lbmol/hr 0.045 0.045 0.045
Total sulphur to incinerator, lbmol/hr 0.716 0.687 0.610
Incinerator stack SO
2
, ppmv dry and air-free 266 269 249
Total sulphur in treated natural gas, lbmol/hr 0.043 0.043 0.044
Overall sulphur recovery, % 99.72% 99.73% 99.76%
Sulphur recovery of the sulphur complex, % 99.77% 99.78% 99.78%
Comparison of sulphur recovery effciency for test cases
b&v.indd 7 9/3/11 13:02:04
reboiler duty or circulation rate.
Overall sulphur recovery reported
in Table 5 considers the sulphur
balance envelope around the entire
gas treatment facility (AGRU/
AGE/SRU/TGTU) and includes
sulphur lost in the treated natural
gas stream, while sulphur recovery
of the sulphur complex fgures
reported in Table 5 considers only
the sulphur balance around the
sulphur production facilities (AGE/
SRU/TGTU). It is important to note
that the 2010 study reported
sulphur recovery effciency fgures
for the sulphur recovery complex
only (ie, the second set of fgures in
the table) and not for the overall
gas treatment facility. Furthermore,
the sulphur complex recovery eff-
ciency fgures reported in the last
row of Table 5 should not be used
in direct comparison with the
recovery effciency fgures reported
in 2010 for assessment of solvent
performance. This is because the
2010 study considered a lower
TGTU absorber overhead H
2
S speci-
fcation than that in the Sulfnol
schemes of this study (approxi-
mately 180 ppmv versus 250 ppmv,
respectively).
All three cases are similar in
sulphur recovery effciency (see
Table 5). The sulphur balance
for each case is illustrated in
Figures 46.
Comparison of solvent systems
Solvent circulation rates and other
solvent system design data for each
of the test cases are shown in
Table 6.
Solvent fow rate to the main
absorber is about the same in all
cases, with only a slightly higher
requirement for the Sulfnol-M case
(Case 3). In Cases 2 and 3, lean
Sulfnol-M solvent fows to the
TGTU absorber with subsequent
cascading of the semi-rich TGTU
absorber bottoms to the AGE
absorber. Cases 2 and 3 have the
same lean solvent fows to the
TGTU absorbers and the same
semi-rich solvent fows to the AGE
absorbers. Thus, Case 2 has essen-
tially the same total solvent fow as
Case 3, with the only difference
being a minor incremental increase
in fow for the Sulfnol-M main
absorber in Case 3. Sulfnol-M
requires more steam for regenera-
tion than Sulfnol-X; therefore, Case
3 has a higher solvent regeneration
steam consumption rate than Case
2. Case 1, because it does not
perform acid gas enrichment, has
the lowest total solvent circulation
and steam consumption rates.
Comparison of SRU/TGTU train size
Most SRU and TGTU equipment
sizes are primarily determined by
the process gas volumetric fow
rate. SRU, TGTU and incinerator
volumetric gas rates for each test
case are given in Table 7.
Case 1 requires the largest SRU/
TGTU train size due to the fact that
it processes SRU feed gas with the
lowest H
2
S concentration; fow rates
are similar to those for Case 1 from
the 2010 study. Cases 2 and 3
signifcantly reduce the SRU size
because these alternatives minimise
CO
2
fow through the SRU by slip-
ping CO
2
from the enrichment step
with the use of Sulfnol-M solvent.
Since the enrichment absorber over-
head is routed to the TGTU in
Cases 2 and 3, the TGTU and incin-
erator sizes are not substantially
smaller than the Case 1 size.
Capital and operating
cost comparison
Table 8 shows estimated capital and
operating costs for each case. Rough
Case 1 2 3
Main absorber
Lean solvent, gpm 1236 1236 1265
TGTU absorber
Lean solvent, gpm 325 1015 1012
Inlet H
2
S, mole% 1.35 1.31 1.35
Inlet CO
2
, mole% 48.43 50.74 49.31
Overhead H
2
S, ppmv 248 249 248
CO
2
slip, % 93.6 88.9 88.2
AGE absorber
Lean solvent, gpm - - -
Inlet H
2
S, mole% - 24.10 19.55
Inlet CO
2
, mole% - 69.62 74.90
Overhead H
2
S, ppmv - 6526 6562
CO
2
slip, % - 88.1 88.7
Main treating regenerator
Acid gas H
2
S, mole% 24.20 24.10 53.15
Total lean solvent, gpm 1284 1284 2338
Regeneration steam, lb/hr 55 888 55 888 129 387
AGE/TGTU regenerator
Acid gas H
2
S, mole% 28.46 57.22 -
Total lean solvent, gpm 325 1015 -
Regeneration steam, lb/hr 19 221 53 949 -
Total solvent circulation, gpm 1609 2299 2338
Total regeneration steam, lb/hr 75 109 109 837 129 387
Solvent system comparison for test cases
Table 6
Case 1 2 3
SRU tail gas, lbmol/hr 1870 1063 1108
TGTU feed gas, lbmol/hr 1870 1798 1727
Incinerator process gas feed, lbmol/hr 1713 1648 1576
SRU/SCOT train size comparison for test cases
Table 7
Most SRU and TGTU
equipment sizes are
primarily determined
by the process gas
volumetric fow rate
b&v.indd 8 9/3/11 14:42:33
1
1
2
2
_
e
Build on our technologies.
Call us, we inform you: +49 (0) 69 58 08-0
www.lurgi.com
A member of the Air Li qui de Group
Model f or our t echnol ogi es.
Lurgi is the worldwide leading partner when clean conversion is postulated. We command
sustainable processes which allow us to make better use of oil resources or biomass than
ever before.
With our technologies we can produce synthesis gases, hydrogen or carbon monoxide: for
downstream conversion to petrochemicals. Based on resources like natural gas, coal and tar
sand we produce synthesis gas which we convert into low-pollutant fuels.
Enhanced sustainability: from biomass which does not compete with the food chain, we
can recover ultra-pure fuels burning at a low pollutant emission rate which are excellently
suited for reducing the carbon footprint. You see, we are in our element when it comes to
sustainable technologies.
www.ptqenquiry.com
15_lurgi.indd 1 8/3/11 13:42:28
order of magnitude capital costs are
provided on a US Gulf Coast basis
and include engineering, procure-
ment and installation of all process
and utility equipment; storage,
metering and loading facilities;
plus, construction indirects, spares,
engineering and commissioning.
Items such as owners costs, land
costs, escalation and contingency
are excluded. Operating costs are
discounted for each year of plant
life and are added to capital cost to
form each cases net present cost.
The conventional SRU/TGTU
arrangement paired with Sulfnol-X
treating for Case 1 serves as a
benchmark for comparison between
the other cases, similar to the way
in which the Case 1 conventional
SRU/TGTU arrangement of 2010
served as a benchmark for the
enrichment cases studied at that
time. The Sulfnol-X design of Case
1 represents an optimum approach
if considering only the AGRU and
not the impact that the resulting
lean acid gas has on the down-
stream SRU. However, as
ascertained in the 2010 study, Claus
SRU operational diffculty caused
by low acid gas H
2
S concentration
Case 1 2 3
Fuel
Consumption, MMBtu/hr 32.3 26.8 30.6
Yearly cost 1.7 1.4 1.6
Power
Consumption, kW 1314 1348 1490
600 psig saturated steam
Export, lb/hr 19 940 17 481 17 301
Yearly credit 1.4 1.2 1.2
50 psig saturated steam
Import, lb/hr 58 964 99 383 118 769
Maintenance and non-utility operating costs
Yearly cost (10% TIC) 8.7 9.2 9.4
Cost summary
Total yearly operation and maintenance cost 13.2 16.0 17.6
Total installed cost 86.6 92.2 93.9
Net present cost 171.9 195.6 207.6
Notes: Costs are reported in millions of US dollars.
1. Costs were prepared using the following information:
* $6.00/MMBtu LHV fuel
* $0.064/kWh power
* $6.96/1000 lb 50 psig sat. steam
* $8.31/1000 lb 600 psig sat. steam
* 355 operating days per year
* 15% annual rate of return
* 25-year plant life
Costcomparisonfortestcases
makes the acid gas enrichment
achieved in the other cases desira-
ble, in spite of their larger amine
units with higher energy
consumption.
The Case 2 Sulfnol-X design is
essentially the same as Case 1;
however, in Case 2, the lean acid
gas from the AGRU is enriched to
improve Claus SRU operating
performance, while also giving the
beneft of reduced SRU size. The
AGE/SRU/TGTU for Case 2 is
similar to Case 4C from the 2010
study, but with Sulfnol-M solvent
Table8
instead of MDEA. In the 2010 study,
the AGE/SRU/TGTU confguration
of Case 4C resulted in slightly
lower capital cost than the conven-
tional SRU/TGTU confguration of
Case 1, which did not include
enrichment. Here, Case 2 is slightly
higher in capital cost than Case 1.
The difference between the results
of the 2010 study and the results
reported here is explained primarily
by the fact that the hydrogenation
step was split into two reactors
instead of combined into a single
reactor, as in the 2010 Case 4
confgurations.
The important comparison is
between Cases 2 and 3, both of
which produce acid gas with a H
2
S
concentration high enough to ensure
stable Claus SRU performance. In
both of these cases, the SRU, TGTU
and incinerator portions are similar
in size and cost; therefore, the
difference in capital cost is in
the AGRU/AGE/TGTU treating
sections. Although Case 3, with its
single Sulfnol-M solvent system,
requires fewer equipment items in
these sections, some of them are
larger and therefore slightly more
expensive than the equivalent equip-
ment items in Case 2. Sulfnol-M
requires more regeneration energy
than Sulfnol-X, making the regener-
ation system larger. Additionally,
Sulfnol-M requires a larger absorber
in the AGRU.
Despite the fact that this analysis
shows a slight preference for Case 2
over Case 3, the capital cost differ-
ential is small and quite possibly
within the margin of error of the
estimates. When operating costs are
considered, there is a greater depar-
ture between the net present cost of
the two options, with Case 2 still
emerging as the preferred confgu-
ration. However, the small
difference in net present cost
suggests that a detailed analysis
between these two options should
be carried out when selecting the
preferred scheme for a specifc sour
natural gas-treating application.
Summaryandconclusions
A 2010 analysis by Black & Veatch
acknowledged that acid gas enrich-
ment should be considered when
upstream AGRU confgurations
Alternativeacidgas
enrichmentschemes
mustbeconsideredto
achievehighsulphur
recoveryeffciency
whenorganicsulphur
compoundsare
present
b&v.indd 9 9/3/11 13:02:27
produce lean acid gas with H
2
S
concentrations lower than that
required to achieve reliable Claus
SRU operation (<25 mol%), and
concluded that alternative acid gas
enrichment schemes must be
considered to achieve high sulphur
recovery effciency when organic
sulphur compounds are present.
The preferred enrichment scheme
from the 2010 study incorporates
recycle of the unabsorbed enrich-
ment absorber overhead organic
sulphur to the TGTU for
high recovery effciency, and
employs maximum solvent integra-
tion between the AGE and TGTU
absorbers for minimum capital and
operating cost. The preferred
enrichment scheme from the
2010 study has been selected in
combination with two different
Sulfnol gas-treating schemes to
examine the lifecycle cost of the
entire sour natural gas treatment
facility.
In the current analysis, Case 2,
Sulfnol-X main gas treating with
Sulfnol-M acid gas enrichment and
tail gas treating, results in a slightly
lower lifecycle cost than Case 3,
which is a fully integrated Sulfnol-
M approach for main gas treating,
acid gas enrichment and tail gas
treating. Although the Case 2 life-
cycle cost is lower, minimal cost
difference between the two schemes
suggests that a detailed analysis
should be carried out to determine
the preferred option for a specifc
sour natural gas treating applica-
tion. Additionally, it is important
not to overlook the somewhat
intangible operational and mainte-
nance benefts associated with
operating a single solvent system in
Case 3, versus two separate systems
in Case 2, when selecting the best
confguration to meet a sour gas
processors needs.

Sulfnol and SCOT are trademarks of Shell.

Further reading
1 McNamara H J, Schilk J A, US Patent No.
4 356 161, Washington, DC, US Patent and
Trademark Offce, 1982.
2 Slavens A, Lamar J, ODell S, Francoviglia
L, Enhanced sulfur recovery from lean acid
gases containing COS and mercaptans, 2010
Laurance Reid Gas Conditioning Conference.
3 Nikolic D L, Wijntje R, Hanamant Rao P P,
Do you have hard to handle gases? Consider
using this second-generation hybrid solvent
for treating, Hydrocarbon Processing, Jul 2009,
6771.
Angela Slavens is Vice President and Sulphur
Technology Manager with Black & Veatch,
Overland Park, Kansas, USA.
Email: slavensaf@bv.com
Justin Lamar is Sulphur Section Lead with
Black & Veatch. Email: lamarja@bv.com
Djordje Nikolic is Licensing Manager for
Upstream Gas Processing Gas & Liquids
Treating with Shell Global Solutions
International, Amsterdam, The Netherlands.
Email: Djordje.Nikolic@Shell.com
Theo Brok is a Senior Process Design
Engineer and Subject Matter Expert, Gas &
Liquids Treating, with Shell Global Solutions
International.
Email: Theo.Brok@Shell.com
What Goes Around...
Remove Sol i ds and Hydrocarbons f rom
your sol vent wi t h MPRs Hi gh Capaci t y,
Easi l y Regenerat ed Medi a.
Available Worldwide
The MPR SSX and HCX processes use our unique and proprietary media, which
can easily be regenerated after exhaustion. They complement each other in way that
allows them to surpass the capacity of standard cartridges and activated carbon
lters, while bringing outstanding longevity.
SSX and HCX Solve The Following Common Problems:
Plugging of heat exchangers, piping, towers
and instruments
Poor hydrocarbon separation from
amine solutions
High Amine losses
Unstable Unit operations
Accelerated corrosion rates due
to erosion
Deactivation of Claus Catalyst in the SRU
SSX and HCX processes work by removi ng sol i ds t o
submi cron si ze part i cl es, whi l e al so removi ng hydrocarbons
(i ncl udi ng di esel , gas oi l s, paraf ns, BTEX and vari ous pet rol eum
i nt ermedi at es) f rom ami ne sol ut i ons.
Eliminating cartridge lters and activated carbon means eliminating lter replacement,
disposal and labor
SSX and HCX processes are available exclusively from MPR Services, Inc..
www.ptqenquiry.com
b&v.indd 10 9/3/11 13:02:38
Consuttlng Englneerlng Constructlon Operotlon I www.bv.com
Sweet
Sour
When lt comes to lnnovotlve process deslgn ond brood
experlence, Etock & veotch ls the provlder of cholce.
You con count on us to understond your speclfc buslness
drlvers ond detlver unblosed, toltored gos treotlng ond
processlng sotutlons for your unlque needs. Whether lt`s
one of our lnnovotlve, proprletory technotogles or o host of
other technotogles, took to Etock & veotch for best-votue
ond lntegroted sotutlons for ony condltlons.
We`re bultdlng o wortd of dlfference. Together.
Gas processing expertise in a variety of avors.
11_GOC_PTQ's Gas Supplement_AD.indd 1 2/21/11 2:41 PM
www.ptqenquiry.com
b&v.indd 1 8/3/11 14:17:05
Cryogenic monitoring
E
nvironmental and safety issues
are becoming ever more
important in the liquefed
natural gas (LNG) process industry.
As plants, pipelines and storage
facilities become larger in scale and
more complex, maximising safety
and minimising risk requires an
absolute understanding.
It is estimated that more than 400
liquefaction, receiving and storage
facilities are either operating, under
construction or planned today. This
effectively equates to more than 500
storage tanks and well in excess of
1000 km of pipeline that requires
monitoring.
Recent years have seen signifcant
developments in the technologies
available to effectively monitor an
ever-increasing number of facilities
and associated infrastructures. With
conventional technology, there tradi-
tionally has been a gap between
what is thought to be occurring
throughout an installation and what
is actually happening. However,
technologies are now available that
can overcome the limitations of
traditional measurement methods to
close that monitoring gap.
Fibre optic distributed tempera-
ture sensing (DTS) is one such
technology that has come to the
fore as a realistic alternative moni-
toring solution. This method of
monitoring has several key opera-
tional advantages over conventional
real-time data (RTD) sensors. One
example of the benefts of DTS over
traditional monitoring methods is
that conventional thermal point
sensors are not able to provide full
coverage of a pipe or tank and
therefore do not have the capability
to pinpoint the precise location of
Fibre optic distributed temperature sensing has several advantages over
conventional real-time data sensors in cryogenic leak detection at LNG terminals
JErry WorsLEy
Sensornet
any cool-down or leak event. By
contrast, using fbre optic DTS, the
location of such an incident can be
identifed almost instantly and
remedial action taken. In fbre optic
DTS, the fbre is the sensor and the
measurement is totally non-
intrusive. No welds are required to
the pipe or vessel and the fbre can
be easily strapped to the outside of
the pipe with no interruption to the
operational process. The fbre
provides its own communications
path and no additional tie cables
are required. Most importantly,
DTS using optical fbres enables the
capture of data every metre at near
real-time speeds. It can detect and
report cryogenic leaks to the emer-
gency shutdown (ESD) system
within 15 seconds. This enables the
ESD system to register and close
valves within 45 seconds.
A DTS system from Sensornet
provides a cryogenic monitoring
solution that can instantly locate a
temperature event anywhere in a
cryogenic process plant such as a
LNG terminal. The DTS system can
be used in digital leak detection
and thermal profling applications,
including LNG, liquefed petroleum
gas (LPG), ethylene and CO
2
load-
ing and unloading pipelines,
storage tanks, spill channels and
impounding basins, as well as the
cool-down monitoring of pipes and
storage tanks.
From an operational perspective,
the benefts of a digital cryogenic
temperature-monitoring system can
be summarised as follows:
Leaks can be detected rapidly and
mitigating action taken immediately,
thus minimising the risk to opera-
tions personnel and equipment
Such a system can lead to
improved plant operation, allowing
the monitoring system maintenance
schedule to be aligned to other
safety integrity systems in the plant
The sensing element is an intrin-
sically safe passive sensing cable
with a design life of more than 30
years, resulting in high reliability
and low maintenance
The system is fully automated
and so lowers operating costs with
less risk of human error. It can
interface with existing supervisory
control and data acquisition
(SCADA) and ESD or fre and gas
control systems using standard
protocols.
The benefts to an asset manager
of such a system are clear. It can
improve the safety of both the
infrastructure and of personnel,
which can reduce insurance and
litigation costs, fulfl legal and plan-
ning requirements, and enhance a
companys safety reputation. It can
also enhance plant reliability
through reduced downtime and
inspection time, while fewer
nuisance and false alarms can result
in increased security and improved
productivity.
Principle of leak detection
using DTs
DTS is based on temperature meas-
urements using fbre optic cables
and can be used to detect both
liquid and gaseous leaks. The DTS
system illuminates the glass core of
the optical fbre with a 10-
nanosecond laser pulse, which
corresponds to a 1m pulse. As the
optical pulse propagates down the
fbre, it undergoes scattering even
in the absence of impurities and
sensornet.indd 1 9/3/11 14:43:08
structural defects. Part of this scat-
tered radiation is known as Raman
scattering. As this vibration energy
is a well-defned function of
temperature, the ratio of the signals,
in conjunction with the time of
fight of an optical pulse, is used to
determine the temperature of the
fbre at a given point.
DTS can be used on long-distance
pipelines, with repeater stations situ-
ated every 80 km if required. This
can be utilised for send-out lines or
carbon capture and storage. More
importantly for LNG applications,
DTS can be equipped with multi-
plexing devices (also known as a
mux) that can monitor several lines
from a single unit. These could
include delivery, recirculation and
burn-off gas lines, as well as multi-
ple tanks, spillways and impounding
pits. This can make DTS technology
a versatile and cost-effective option.
Fibre is the sensor
DTS systems consist of a limited
number of components, which are
the DTS hardware and the sensing
cable. The hardware is normally
situated within the control room
environment. However, a robustly
designed data acquisition device
such as Sensornets Oryx system,
which is made for deployment in
harsh environments, can be posi-
tioned in the feld, allowing the
monitoring of hard-to-reach places.
The fbre optic cable is almost as
important as the DTS itself, because
it is the fbre that provides the
information. It is important to use
the right cable for the proposed
scope of work. For leak detection
alone, it is possible to install a rela-
tively standard fbre optic cable.
However, the user should be
prepared for a one-time use only.
Tests have shown that standard
cable and fbre types detect the leak
at its initial stage, but when
subjected to temperatures of less
than -40C for even the shortest
periods of time the fbre and jacket
construction deteriorate and become
unusable in the future. It may be
acceptable in certain applications
for these cables to be sacrifcial. For
long-term use and cool-down moni-
toring, cables and fbres with
specialist coatings are required.
Cables have been designed that are
capable of consistently measuring
temperatures down to -175C.
Pipeline monitoring
Monitoring the delivery, receiving
or recirculation lines from the vessel
to the tank or process facility is crit-
ical, as this is one of the largest
sections of an infrastructure (which
can be up to several kilometres in
length, often with limited access)
that can be exposed to the greatest
amount of stress. Stories of pipe-
lines being lifted from their
supports due to performing the
cool-down process too quickly are
not uncommon. The inclusion of
fbre optic monitoring can help
complete the cool-down process in
a speedy but safe manner. Sections
of fbre are strategically placed
directly onto the surface of the
carrier pipe below the insulation
layer. These are positioned on the
crown and the base of the pipe. By
comparing the temperature of these
during cool-down, the product can
be delivered quickly and safely.
www.ptqenquiry.com
The same fbre can be routed along the length of pipe
outside the insulation layer and be used for leak detec-
tion. This fbre monitors the outside temperature of the
pipe. During leak conditions, the product returns to its
natural gaseous phase and cools the skin of the insula-
tion to near cryogenic temperatures. The system can be
installed on different pipe confgurations, including
bare pipes, foam-insulated pipes and vacuum-insulated
piping, and has been independently verifed by both
the TUV Nord and US Federal Energy Regulatory
Commission (FERC) for use on vacuum-insulated pipe
systems. The same monitoring lines can be extended to
cover the spillways and impounding pits two prime
locations for leak detection and monitoring.
Storage tank monitoring
For storage tanks, the sensing cable can be installed on
both the circumference (either within the annulus or in
the outer spillway) and the foundations of the tank.
This enables the system to check for leaks and monitor
the heating element to prevent frost heave. An addi-
tional fbre loop installed vertically up the wall of the
tank will also allow cool-down monitoring.
This solution can prove a very cost-effective alterna-
tive to resistance temperature detector sensors and
provides complete coverage and thus confdence for
the tank operator.
In older tank designs, DTS can be used to provide a
more reliable upgrade to the leak detection system by
installing the cable in the guttering in the perimeter of
tanks, under valve spill trays and in tank overfow
channels. The spill channels and impounding collection
channels are prime locations for products to collect in
the event of a leakage. Installing sensing cables at low
points in these locations provides an excellent early
warning system.
DTS in the real world
In compliance with Health and Safety Executive
requirements, the UKs National Grid Isle of Grain
LNG facilities asked Sensornet to provide a leak detec-
tion system. The monitoring system would be
non-intrusive, permitting the continued operation of
the importation and recirculation pipelines during its
installation and commissioning.
Traditional use of fow meters based on differential
fow measurement was considered inadequate to
provide the required leak detection monitoring for
emergency shutdown purposes due to their relatively
long response time. Other monitoring methods were
unable to provide distributed measurements over the
entire length of the pipeline. Therefore, a robust, non-
intrusive leak detection method was needed to provide
the required safety integrity level.
The system was designed with two cable loops and
two instrumentation units, capable of 1 out of 2 voting.
Each DTS system provides bidirectional monitoring
from either side of the loop, and the 24-way relay
modules provide hard wire contact relays to the client
SCADA system in instances such as system error,
where the rate of a temperature change exceeds a set
Creativity is the art we ap-
ply to achieve superior de-
sign and developments in
technology.
For over 70 years we have designed and supplied
cost-effective technology, process plants and equip-
ment for the oil & gas and refiningindustry around
the world. With our expertise we provide tailor-
made solutions from studies and revamps to skid-
mounted units and complete turnkey plants.
SIIRTEC NIGI Engineering Contractors
ViaAlgardi,2 20148,Milan (Italy)
Tel:+39-0239223.1 Fax:+39-0239223.010 Web: www.sini.it
Contact:marketing@sini.it
Oil & gas treatment
separation, dehydration and HC dew point control,
LPG and NGL recovery, sweetening,
mercaptan removal
Acid gas treatment & sulphur recovery
acid gas enrichment, Claus, ammonia Claus,
oxygen-enriched Claus, tail gas treatment,
sulphur degassing
Advanced process controls
Special process equipment
Italian design
A masterpiece
PTQ_86x270:Layout 1 18-03-2010 17:58 Pagina 1
www.ptqenquiry.com
temperature, where the minimum
temperature exceeded a fbre break,
or where there is overheating in the
instrument cabinet.
The system provides non-
intrusive protection and was
retroftted by installation engineers
in an operational plant without
having to shut down or stop opera-
tion. The independent fbre optic
sensing cable loops covering both
pipelines provide added resilience
to the installation. Additional multi-
plexer channels ensure future
expansion and fexibility, while self-
contained test points were
introduced at regular intervals on
the cable loops. These test points
are also used after the site accept-
ance tests (SAT) for periodic
testing.
To a tight deadline, sensing cables
were installed in the pipeline corri-
dor within culverts and on the
750m jetty section at Isle of Grain
using specialist access equipment.
Sensornet deployed its Sentinel
DTS-XR20 system, which is opti-
mised to provide a temperature
resolution of 0.1C in 10 seconds
along the entire length of the pipe-
line at 1.5m intervals. The high
specifcation of the XR20 allows the
plant to monitor the pipeline with
accuracy and integrity.
The most valuable alarm criteria
when detecting leaks are tempera-
ture rate of change and temperature
deviation of a single point. In the
event a cryogenic leak occurs within
the process plant, there will be a
signifcant temperature drop (typi-
cally greater than 100C).
When the cryogenic fuid comes
into contact with the sensing cable,
this will be detected instantly
(down to 10 seconds) and
pinpointed to within 1m. Since the
location of the leak can be detected
instantly, rapid action can be taken
to mitigate any risk and control the
situation.
Temperature rate of change at a
specifc location can indicate that a
signifcant temperature event has
occurred in a relatively short space
of time. This excludes any day/
night or seasonal variations and
concentrates on shorter periods.
The volume of the leak can be esti-
mated from the physical length of
A robustly designed
data acquisition
device can be
positioned in the
feld, allowing the
monitoring of
hard-to-reach places
sensing cable affected by the
temperature change, and the rate of
leakage can be calculated by analys-
ing the distance the leak travels
over time.
Temperature deviation of a single
point allows comparisons of a
number of temperature values
within any given zone, which can
be set from 1m to 30km. Each
temperature point within a zone
normally has similar ambient prop-
erties, so any signifcant variation
of a single temperature point
compared with the average temper-
ature of the zone can indicate that
an event has occurred.
The DTS leak detection system
can also be confgured to provide a
number of useful temperature
comparisons. Absolute temperature
can vary, so it is not always a useful
fgure when used directly for leak
detection. Obvious exclusions to the
rule include cryogenic products
such as LNG, which can reduce
surrounding the temperature to
between -120C and -165C in the
event of a leak.
Each DTS unit is an autonomous
device and an unlimited number
can be connected to the SCADA or
distributed control system (DCS)
network via standard communica-
tions protocols such as Modbus,
Ethernet or Relay Contacts. This
allows the system to be used on
pipelines varying from a few
hundred metres up to thousands of
kilometres. Once networked, each
device will only transmit alarm and
regular health check data, ensuring
the network is not overloaded with
baseline fully distributed data.
Should these data be required for
historical analysis, they are
available at any time from the hard
drive located on the DTS system.
Environmental risk reduction
The fnancial cost of a leak can be
enormous; however, the greatest
cost is to the environment.
Environmental concerns are key not
only to the local authorities and
green groups, but also to the pipe-
line operators, whose responsibility
is to ensure that risk is kept to a
minimum and, should a leak
occur, to react immediately and
effectively.
Political sensitivities, pressure
from environmental groups and
costly litigation procedures have
acted as a warning to operators to
ensure safety and environmental
responsibilities are fulflled and that
approval is gained to construct
pipelines through sensitive or
populated areas in the future.
A system that monitors mission-
critical infrastructure, and obtains
meaningful and valuable informa-
tion quickly and accurately, to allow
an asset to be as effective, safe and
risk free as possible, is the desired
outcome.
Next-generation monitoring
Many methods for leak detection
already exist. These include compu-
tational methods that are heavily
reliant upon statistical analysis;
mass balance techniques that
compare input and output values
and visual inspection by walking or
fying the line. Pigging is common
and may also be used to analyse
other aspects of the line. However,
pigging can be disruptive to the
operation and can carry its own
risks. DTS is the only leak detection
technique that can monitor an LNG
facility in real-time and provide
precise temperature and leak infor-
mation to the nearest metre.
Jerry Worsley is the Industrial Monitoring
Manager for Sensornet. He has over 15
years experience in the specifcation and
deployment of fbre optic systems in oil
and gas, construction and environmental
monitoring, and focuses on the
implementation of distributed sensing
systems in the downstream process sector.
www.ptqenquiry.com
burkhardt.indd 1 20/9/10 10:10:44
www.ptqenquiry.com
cansolv.indd 1 8/3/11 13:53:57
Glycol dehydration of high-acid
gas streams
G
lycol dehydration is a proc-
ess that presents some
unique challenges from both
technical and computational stand-
points. In the frst place, modern
designs almost invariably use tower
internals consisting of structured
packing rather than the more tradi-
tional bubble cap trays. Structured
packing offers a lower pressure
drop and considerably higher
capacity than trays, and it is well
suited to handling the very low L/
G ratios common in dehydration.
However, until now, the height of
the packing was estimated using
rules of thumb, not good engineer-
ing science. Mass transfer rate-based
modelling, on the other hand, uses
science and therefore offers greater
reliability of design. The other chal-
lenge of dehydration using any
glycol is thermodynamic: water is
the component of interest but it is
probably natures most perversely
non-ideal chemical, which makes it
challenging to devise a truly accu-
rate model for the vapour-liquid
equilibrium, especially in systems
with a high-acid gas content.
Traditionally, sweetened gases
have been the main candidates for
dehydration, mostly using triethyl-
ene glycol (TEG) prior to entering
the transmission pipelines. More
recently, there has been much inter-
est in sour gas injection as a means
of disposing of gas streams of too
low quality for sulphur recovery. In
the context of carbon capture, the
recovered CO
2
is compressed, lique-
fed and injected into underground
formations. Handling these sour
and high CO
2
streams requires
dehydration prior to compression
and/or liquefaction.
A model based on mass and heat transfer closely predicts plant performance data
for glycol dehydration of high-acid gas streams
John CArroll Gas Liquids Engineering
nAthAn hAtCher and rAlph WeilAnd Optimized Gas Treating
There are other facets of glycol
dehydration that are interesting just
from an applied science viewpoint.
One of them is the heat transfer situ-
ation that ensues in a regenerator
using both stripping gas and a
reboiler (Stahl column). When the
hot gas hits the bottom of the pack-
ing in the wash section atop the
column, it fnds itself suddenly
going from an environment in which
it is saturated with water in equilib-
rium with a predominantly TEG
stream into an environment where it
is grossly under-saturated with
respect to the pure water stream in
the wash section. This causes
extremely rapid humidifcation, and
the humidifcation process extracts
the necessary heat of vapourisation
as sensible heat from the liquid
water phase. Sudden humidifcation
can drop the wash water tempera-
ture by 3040F or even more.
This article addresses the effcacy
of a new mass and heat transfer
rate-based model in terms of how
well it refects known phase behav-
iour and how closely it predicts
known plant performance data
using both bubble cap trays and
packed columns without recourse
to height equivalent to a theoretical
plate (HETP) or height of transfer
units (HTU) estimates and other
rules of thumb.
phase equilibrium
Our phase equilibrium concerns are
twofold: accurate calculation of the
equilibrium water content of high-
and low-pressure gases containing
very high levels of CO
2
and/or H
2
S;
and calculation of the solubility of
all gases, including water, in the
dehydration solvent itself. These
two topics are inter-related because
both require the accurate assess-
ment of the interactions between
components in the vapour phase;
however, the solubility calculation
is more demanding.
Normally, to model the solubility
of a gas in a liquid, one uses the
Henrys Law approach. A thermo-
dynamically complete version of
Henrys Law
2,3
is:
a
~
i
x
i
H
ij
exp
>v
i

(P P
j
o
)H
=
i
P
i
v
(1)
RT
where
i
: Activity coeffcient
x
i
: Mole fraction of component i in
the solvent (its solubility)
H
ij
: Henrys constant for solute i in
solvent j, kPa/mol frac
v
-
i
Partial molar volume of i in

solvent j at infnite dilution,
m
3
/kmol
P
j
o
: Vapour pressure of the solvent,
kPa
P: Total pressure, kPa
R: Universal gas constant,
8.314 kJ/kmolK
T: Absolute temperature, K
y
i
: Mole fraction of component i in
the vapour
i
v
Fugacity coeffcient for
component i in the vapour,
unitless
Most people believe that Henrys
Law is only applicable to dilute
solutions, but the form in Equation
1 can be applied without restriction.
It would be unusual to do so, but
this equation can even be applied
to mixtures that are not typically
considered to refer to the solubility
of a gas in a liquid at all; for exam-
ple, methanol and water.
ogtrt.indd 1 8/3/11 14:50:14
The numerical value of the
Henrys constant is a function of
both the solute and the solvent.
Thus, the Henrys Law constant, H
ij
,
is different for methane in water
than it is for methane in methanol,
and H
ij
is different still for ethane
in water. Furthermore, for every
solute-solvent pair, the H
ij
is a func-
tion of the temperature. When
comparing different solutes in the
same solvent, the larger the Henrys
constant the lower the solubility of
the solute.
The activity coeffcient accounts
for the effect of concentration on
the activity of the component in the
liquid phase. The infnite dilution
defnition of the activity coeffcient
is used here; its signifcance is that
at infnite dilution, the activity coef-
fcient defned this way is unity.
The exponential term is the
Poynting correction. It accounts for
the effect of pressure on the refer-
ence fugacity, and it is important
(signifcantly different from unity)
only at high pressure. For a small
molecule such as methane, hydro-
gen sulphide or carbon dioxide in
water, the partial molar volume at
infnite dilution is approximately
0.032 m
3
/kmol. Thus, at 7000 kPa
and 20C (293K), the Poynting
correction for these systems is about
1.1. Neglecting the Poynting effect
altogether at these conditions means
the solubility is under predicted by
approximately 10%.
A common defnition used in
describing the solubility of a gas in
a liquid is partial pressure. The
partial pressure, P
i
, of a component
in a gas mixture is defned as:

P
i
=y
i
P (2)
The non-idealities in the vapour
phase are embodied in the fugacity
coeffcient. As an approximation,
the fugacity is to the partial pres-
sure as the compressibility factor is
to the molar volume. Included in
this quantity is the effect of pres-
sure, temperature and composition
on the gas phase non-ideality. For
an ideal gas mixture, the fugacity
coeffcient of all components, i, in
the mixture is unity:
i
v
=1 (3)
However, for a non-ideal mixture,
this quantity can be less than,
greater than or equal to unity. At
relatively low pressure, it is gener-
ally less than unity and it can be as
small as 0.1 (and even smaller for
liquids). At high pressure and high
temperature, the fugacity coeffcient
can be greater than unity; however,
it is rarely greater than two.
Therefore, it is very important to
include this contribution in the
solubility model.
The fugacity coeffcient in the
vapour phase can be calculated
using an equation of state such as
the Soave-Redlich-Kwong (SRK) or
Peng-Robinson (PR) equations.
These cubic equations of state have
been used for many years for
modelling the fuid properties and
phase equilibrium in petroleum
systems containing light hydrocar-
bons and a few associated
non-hydrocarbons such as carbon
dioxide, hydrogen sulphide and
nitrogen. However, they are less
successful with water and other
polar substances such as alcohols
and glycols. Typically, the equa-
tions of state must be modifed to
predict accurately the pure compo-
nent properties, and the mixing
rules must account for the complex
interactions.
At low pressure, both the
Poynting correction and the fugac-
ity coeffcients are approximately
unity. Furthermore, if the solubility
is low, Equation 1 reduces to the
more familiar form of Henrys Law:

x
i
H
ij
=y
i
P (4)
or equivalently:
x
i
H
ij
=P
i
(5)
Physically, this equation says that
the solubility of a gas increases
directly as the partial pressure of
the component in the gas increases.
However, this idealised version is
limited to about 200300 kPa and
gases with solubilities less than
about 0.1 mol%.
The ProTreat simulation tools
dehydration model uses the Peng-
Robinson equation of state (EOS)
for the vapour phase and currently
offers a four-suffx Margules equa-
tion activity coeffcient model based
on the data of Bestani & Shing
1
for
the liquid phase, as reported by
Clinton et al.
4
A similar model
based on the less conservative data
of Parrish et al
6
is planned for a
future release.
There are two important aspects
GPSA ref. Mole % (dry basis) Temp, F Pres, psia H
2
O lb/MMscf
CH
4
CO
2
H
2
S Measd ProTreat
Ex20-1 100 0 0 150 1000 220 216
Ex20-2 80 20 0 160 2000 172 188
Fig20-9 0 100 0 100 500 132 125.3
750 110 102.5
1000 125 100.7
2000 215 215.1
3000 238 247.8
5.31 94.69 0 100 850 88 96.9
1125 81 99.2
1500 128 148.6
2000 139 184.2
Fig20-16 89 11 0 100 2000 40.6 41.1
89 11 0 160 1000 286 283.9
80 20 0 100 2000 40.6 45.1
80 20 0 160 1000 282 292.5
80 20 0 160 2000 172 188.5
92 0 8 130 1500 111 103.5
72.5 0 27.5 160 1367 247 252.6
83 0 17 160 1000 292 293.4
30 60 10 100 1100 81 81.2
9 10 81 100 1900 442 264.4
5.31 94.69 0 77 1500 109.2 95
5.31 94.69 0 122 2000 164.6 234.5
Saturated water content of gases
Table 1
ogtrt.indd 2 8/3/11 14:50:24
to thermodynamic modelling of
phase equilibrium in glycol dehy-
dration systems: the water content
of the treated gas; and the hydro-
carbon, acid gas and especially the
benzene, toluene, ethylvenzene and
xylene (BTEX) content of the water-
laden glycol. Table 1 compares
ProTreat model results with Gas
Processors Suppliers Association
(GPSA) Data Book entries for satu-
rated water content. Generally,
ProTreat reproduces measured
values of water content to within
the accuracy of the data. The Peng-
Robinson EOS that performs these
saturated water content calculations
applies a large number of interac-
tion parameters (k
ij
s) for the
interactions between water and the
various gases, as well as between
the gases themselves.
Other components whose solubil-
ity in TEG is pertinent are the acid
gases and hydrocarbons, especially
the BTEX components. Vapour-
liquid equilibrium constants
(K-values) for BTEX are available in
GPA RR-131, and the data there
have been used to ft the ProTreat
solubility model for these species.
The data indicate that at typical
contactor conditions, approximately
1030% of the aromatics in the gas
stream may be absorbed in the TEG
solution. ProTreat results conform
closely to the conclusions of RR-
131, as they should, because
ProTreats solubility model has
been regressed to the actual meas-
ured BTEX solubilities.
Process simulation
The GPSA Data Book contains a nice
example of dehydration with TEG
(Example 2011). The gas is water
saturated at 600 psia, with other
particulars noted in Figure 1. Two
cases are detailed, both requiring
two theoretical stages. One uses
bubble cap trays, which, at a tray
effciency of 2530%, translates into
68 actual trays. The other case uses
10ft of an unspecifed structured
packing. ProTreat has provision for
a separate Stahl column, shown
immediately below the stripper in
Figure 1, but the stripper can also
be simulated without this column if
desired. Two condenser outlet
streams permit the withdrawal of
wet stripping gas and/or water
vapour from the system (Stream 19)
and the removal of a specifable
portion of condensed water (Stream
20), with the remainder returned
as refux.
Table 2 shows the effect of the
actual tray count on the water
content of the dehydrated gas.
ProTreat simulation indicates six
trays are adequate to reduce the
water content from 88.7 lb/MMscf
to the target level of <7 lb/MMscf
(32F dew point). Tower diameter
for 70% food is 3ft. These values
are in line with GPSA Data Book
results. In summary, the available
data indicate that the model is accu-
rately refecting literature data on
the vapour-liquid equilibrium
(VLE) and general TEG dehydra-
tion experience, as reported by
GPSA.
Dehydration column performance
Until now, only an equilibrium-stage
model has been available for calcu-
lations involving the performance of
structured packing. However, the
HETP of the particular packing is
1 1
9 8 7
2 7
1 9
2 6
1 0
2 0
2 1
Contactor
Stripper
Lean/rich
Cooler Controller Recycle
D e h yd ra te d g a s
2 2
5
2 3
O u tle t
O u tle t
2 4
6
1 2
< 7 lb H
2 0
/M M sc f
( < 0 . 0 1 4 7 4 m o le % H
2 0
)
9 9 % wt T E G
3 g a l T E G /lb H
2 0
re m o ve d ( 5 . 1 g p m )
Two id e a l sta g e s
D ia m e te r = 3 . 4 ft
2 5 -3 0 % e ffic ie n c y
6 . 8 b u b b le c a p tra ys, o r
1 0 ft o f stru c tu re d p a c k in g
3 0 M M S C F D g a s
1 0 0 F
6 0 0 p sia
Wa te r sa tu ra te d 1
C B
3
3 5 0 M B T U /h r
re b o ile r d u ty
S trip p in g
g a s
Stahl column
2
4
Figure 1 GPSA Data Book example 20-11
Number Water
of trays lb/MMscf
5 8.5
6 6.7
7 5.7
8 5.1
Water content vs tray count
Table 2
ogtrt.indd 3 8/3/11 14:50:33
degree crimp angle) packings affects
dehydration performance. For
Sulzer Mellapak structured pack-
ings, the packing designation (eg,
M250.X) is a close indicator of the
specifc area, in this case 250 m
2
/m
3
.
Simulations were all run with 40ft
of packing and the absorber was
sized for 70% food (911ft diameter
depending on crimp size). The
absorber was set up to dehydrate 49
000 lbmol/h of wet sweet methane
(trace CO
2
and H
2
S) using 250 gpm
of 99.95 wt% TEG.
The gas-phase temperature
profles for the various packing
sizes shown in Figure 2 indicate
that there can be a signifcant
temperature bulge in dehydration
columns caused by the phase
change (condensation) of the water
being absorbed. Water absorption
generates sensible heating equiva-
lent to its latent heat of
condensation. Also, when small-
crimp packings are used, the
temperature bulge is closer to the
bottom of the column because the
larger surface area allows water to
absorb so much faster. Figure 3
shows the water content of the gas
at various positions along the
height of the column. It is evident
that, after traversing the bottom
20ft of M350.X packing, the gas is
about as dry as 99.95 wt% TEG at
100F can get it. With M125.X pack-
ing, on the other hand, water is still
being removed even after the gas
has passed through 40ft of packing.
So the required bed height is very
much a function of the packing size.
It is not that rules of thumb cannot
be made to work; rather, it is that
the right rules of thumb, at least for
packing, depend on too many
parameters (not just packing type
and size but also on the gas and
solvent fuxes through the column)
and this makes them rough at best
and unreliable at worst. Until now,
the answer to this dilemma has
been to over-build the columns;
however, in a competitive environ-
ment, surely being able to avoid
over-design gives the knowledgea-
ble contractor and the astute
internals vendor a commercial
advantage. Mass and heat transfer
rate-based simulation is the preci-
sion tool that allows this to be
done.
Figure 4 shows that the blanket
rule of thumb of 68 trays for
dehydration can be a gross over-
simplifcation. The number of trays
depends at least on the dryness to
be achieved; that is, the dryness of
the solvent. If the target dew point
is not too stringent (say, 10 or 20lb
H
2
O/MMscf), 6 or 8 trays seem
adequate. But in very low dew
point applications, such as LNG
plants, two or three times that
number of trays may be required to
get to the desired dryness. With
99.97% TEG, for example, it is
possible to get to below 1 lb/MMscf
surely related to packing size. For
structured packing, size can be
expressed in terms of specifc surface
area and crimp size, characteristics
that are geometrically related. Under
otherwise identical process condi-
tions, one should expect that large
crimp packing will require a deeper
bed to give the same performance as
a relatively short bed of small crimp
packing simply if for no other reason
than because the surface area of the
small crimp material is so much
larger.
Figures 2 and 3 are simulations of
how packing size within the
Mellapak X Series (roughly 60-
2 0
0
5
1 0
1 5
2 5
3 0
3 5
1 0 0 0 1 1 0 1 2 0 1 3 0 1 4 0

D
i
s
t
a
n
c
e

f
r
o
m

t
o
p

o
f

b
e
d
,

f
t
Temperat ure, F
4 0
M 1 7 0 . X
M 3 5 0 . X
M 1 2 5 . X
M 2 . X
M 2 5 0 . X
Figure 2 Temperature profles and the effect of packing size in the Mellapak X Series
2 0
0
5
1 0
1 5
2 5
3 0
3 5
0 1 1 0 1 0 0 1 0 0 0

D
i
s
t
a
n
c
e

f
r
o
m

t
o
p

o
f

b
e
d
,

f
t
Wat er i n gas, Lb/ MMscf
4 0
M 1 7 0 . X
M 3 5 0 . X
M 1 2 5 . X
M 2 . X
M 2 5 0 . X
Figure 3 How water removal depends on packing size and packed bed depth
ogtrt.indd 4 8/3/11 14:50:43
Uni que Sol uti on for Your
Petrochemi cal Appl i cati on
Best for polyethylene, polypropylene,
ethylene glycol and ethylene oxide production
The ultimate contact-less labyrinth sealing
technology
No piston rings - no wear
Insensitive to particles and contaminated gases
Gas-tight design for zero emissions
Highly reliable components for highest
availability of your compressor
Count on our experience:
Thousands of compressors in operation,
hundreds of long-term customers
Your Beneft:
Lowest Li fe Cycl e Costs
Like this image? Get more: www.recip.com/screensaver
Ad_Laby-A4.indd 3 07.01.10 09:43
www.ptqenquiry.com
burckhardt2.indd 1 9/3/10 14:25:09
water content, but even after 16
trays, water is still being removed.
For the particular conditions
simulated in this study, it should be
mentioned that high TEG viscosity
is a consideration and will nega-
tively affect internals performance
compared with light hydrocarbons,
for example. At the solvent mois-
ture levels encountered in these
simulations, viscosity is not signif-
cantly affected by water content,
and at the lowest temperature (feed
solvent at 100F) the viscosity is
about 19 cP (for reference, corn
syrup is 50100 cP). This does not
necessarily rule out using trays,
although it does tend to make
structured packing more attractive.
Stahl columns and stripping gas
A Stahl column gives an extra stage
of regeneration by taking the
solvent from the reboiler and
contacting it with a fow of dry
inert gas. Stahl columns are essen-
tial when the dried gas must have a
very low dew point. Figure 5 shows
the effect of stripping gas rate (SCF
per gallon of TEG solvent) on the
TEG purity and on the water
content of the treated gas for
Example 20-11 from the GPSA Data
Book scaled up by a factor of fve
and using six bubble cap trays in
the absorber and a 10ft bed of
FlexiPac 1.6Y in the regenerator
(includes a 2ft refux wash section).
In terms of scale, the absorber is
nearly 7ft in diameter and is drying
gas that is water saturated at 600
psia and 100F. The regenerator is
only 15in in diameter.
The simulation results in Figure 5
show that using even a modest fow
of stripping gas can reduce the
water content in the treated gas by
more than a factor of two. Stripping
gas permits more water removal
from the solvent and increases the
dry TEG from 98.8 wt% to 99.8
wt%. This particular treated gas is
dry enough by transmission line
standards, but it is a long way from
dry enough for an LNG plant, for
example. However, this demon-
strates the principle that a Stahl
column can serve a useful purpose
it should be used when treated
gas needs to be drier than usual.
Observations from a mass transfer
rate model
The mass and heat transfer rate
model uses real trays, both in
number and mechanical detail, and
real packing in terms of actual bed
depths of a specifc packing with
defned geometry, including crimp
angle, crimp size, surface treatment,
specifc surface area and vendor.
The simulation of a packed column
begins by fnely segmenting the
packed height into a large number
of thin cross-sectional slices to
approximate the continuous nature
of contacting in packing. As a
result, one gets to observe mass and
heat transfer effects on a fairly
detailed scale.
Figure 6 shows the vapour and
liquid temperature profles through
a 10ft-deep bed of 2in metal Nutter
Rings in a regenerator operating at
1 psig. The dehydrator is treating
440 MMscfd of wet, sweet gas with
a TEG fow of 1000 gpm, and the
bone-dry nitrogen stripping gas
fow to the Stahl column is 1.9 scf/
gal. The bed starts with a 2ft-deep
wash section for TEG recovery
followed by an 8ft-deep stripping
section for water removal. The
stripper was simulated by dividing
it into 40 segments, each having a
3in depth. Finer segmentation is,
of course, possible, but it
adds nothing to the detail and
very little more to the accuracy of
the simulation. We will traverse
the regeneration column, starting
with the condensate (essentially
pure water), which enters the
column from the condenser at
180F and is heated by the gas
stream (stripping gas and water),
which is at 198F.
1 0
1 0 0
1
0 8 1 2 1 6

D
r
y

g
a
s

w
a
t
e
r

c
o
n
t
e
n
t
,

l
b
/
M
M
s
c
f
Number of t rays
0 . 1
9 9 . 9 5 %
9 9 . 5 0 %
9 9 . 9 7 %
9 9 . 9 0 %
Figure 4 Water removal to low dew points requires deeper beds and more trays
99.4
100.0
99.8
99.6
99.2
99.0
98.8
0 0.5 1.0 1.5 2.0 2.5 3.0
%
t
w

,
y
t
i
r
u
p

l
o
c
y
l
G

Stripping gas rate, SCF/gallon
98.6
4
7
6
5
3
32
1
f
c
s
M
M
/
b
l

,
s
a
g

y
r
D

0
|b/MMscF
TEG wt
Figure 5 Effect of stripping gas fow on solvent dryness and water content of gas
ogtrt.indd 5 8/3/11 14:50:53
As the condensate trickles down
through the bed, it continues to be
heated by the wet stripping gas, but
within the frst 18in its temperature
reaches a peak (194.5F) and then
suddenly plummets to about 160F.
The gas becomes hotter as we
descend through the column, and
eventually it experiences a rapid
climb to 294F. The question is: why
do the phase temperatures trend
this way? The answer for the gas
phase is relatively simple: feed,
preheated to 300F, enters the
column at the 2ft level and fashes
into its vapour and liquid parts.
The fact that the vapour tempera-
ture changes radically at the feed
point is simply the result of the hot
feed meeting a cold refux water
stream. But why does the refux
water cool so much?
When the vapour leaves the top of
the stripping section, its water
content is very nearly in equilibrium
with the liquid in the stripping
section. The liquid there is better
than 95 wt% TEG with only 5 wt%
(1718 mol%) water. So, although
saturated with water, it is saturated
with only an 18 mol% water phase.
When it hits the frst refux segment,
this vapour is suddenly exposed to
a pure, hot water stream (fowing
from the wash section), so it is seri-
ously under saturated against pure
water. The consequence then is for
water to evaporate as fast as possi-
ble so as to resaturate the hot gas,
now against pure hot water. This is
a typical humidifcation operation
and has some interesting character-
istics. The evaporation process is not
mass transfer rate limited. Rather, it
is limited by the rate at which the
latent heat of vapourisation
demanded by the humidifcation
process can be drawn from the bulk
liquid to the gas-liquid interface.
This process is one of heat transfer
and, indeed, the rate of humidifca-
tion right above the feed point is
controlled by essentially conductive
heat transfer across the liquid flm
running over the packing. The proc-
ess is heat transfer rate limited. The
mass transfer driving force for
humidifcation is so high that
enough heat is drawn from the
refux water to cool it, in this case by
approximately 33F.
As we traverse lower into the
column and approach the bottom of
the stripping section, liquid starts
to meet much hotter vapour coming
from the reboiler. (The bubble point
of TEG-water mixtures is quite
sensitive to water content when the
water content is low.) This cools the
vapour and heats the liquid, and
again some of the heat transfer is
the result of water transferring from
the vapour back into the liquid in
the bottom couple of segments of
the stripping section.
Figures 7a and b are intended to
show how the use of stripping gas
affects these temperature profles.
The plots are for the same 10ft bed
of Flexipac 1.6Y structured packing
referenced in Figure 5. Even when
stripping gas is not used, the
condensed water leaving the refux
4
0
1
1
3
5
6
7
8
9
1 5 0 1 7 5 2 0 0 2 2 5 2 5 0 2 7 5 3 0 0 3 2 5 3 5 0

D
i
s
t
a
n
c
e

f
r
o
m

t
o
p

o
f

b
e
d
,

f
t
Temperat ure, F
1 0
Va p o u r
L iq u id
Figure 6 Temperature profles in regenerator with stripping gas via Stahl column
4
0
1
2
3
5
6
7
8
125 150 175 200 225 250 275 300 325 350 375

p
o
t

m
o
r
f

e
c
n
a
t
s
i
D

t
f

,
d
e
b

f
o
Temperature, F
Vapour
Liquid
Figure 7 Stripping gas is not necessary for anomalous temperature profles
(a) No stripping gas (b) Stripping gas at 3 scf/gal
4
0
1
2
3
5
6
7
8
140 165 190 215 240 265 290 315 340 365

f
o

p
o
t

m
o
r
f

e
c
n
a
t
s
i
D

t
f

,
d
e
b
Temperature, F
Vapour
Liquid
ogtrt.indd 6 9/3/11 14:45:02
section meets a much hotter gas,
and a good part of the temperature
equilibration takes place by water
evaporation from the refux stream.
This is, again, a humidifcation
process and the refux water cools
as a result of the demand for latent
heat for vapourisation. In Figure 7b,
the presence of stripping vapour
dilutes the gas leaving the stripping
section, which results in slightly
more driving force for humidifca-
tion together with the small
additional capacity of the gas for
water. (How much the stripping
gas adds to the driving force and
capacity depends on the relative
fows of stripping gas versus
reboiler vapour.) Again, the refux
water temperature peaks and then
drops from 197F to about 141F,
for a total cooling of 56F, the same
as for the case without stripping
gas. However, with stripping gas
(dry nitrogen), the temperature
bulge in the refux section is a
consequence of the dilution by
nitrogen. We note in passing that,
to maintain water balance, the no
stripping gas case required a 212F
condenser temperature versus 180F
for the case with stripping gas.
Factors that affect dehydration of
sweet gases
To learn more about the practical
limits of glycol dehydration, a
sensitivity analysis was conducted,
starting with the 3 SCF/gallon
stripping gas rate at 1.72 MMBtu/
hr reboiler duty referenced in the
preceding section. First, the reboiler
duty was increased to 2.5 MMBtu/
hr, which amounted to an increase
from 1124 to 1634 Btu/gallon. This
change allowed the glycol purity to
increase from 99.79 to 99.96 wt%
TEG. The water content of the
dehydrated gas was predicted to
drop from 2.7 to 2.0 lb/MMscf,
assuming six bubble cap trays for
contacting. This is not much of an
improvement for nearly a 50%
increase in reboiler energy. Are
there enough absorber trays to take
advantage of the greatly improved
TEG dryness? The number of trays
was increased from 6 to 8 in the
glycol contactor, while maintaining
the lean glycol purity of 99.96 wt%.
This dropped the water content by
an order of magnitude, from 2.0 to
0.2 lb/MMscf. Twelve trays gave
no improvement. So much for rules
of thumb!
Finally, at 12 trays of dehydrator
contacting, some of the reboiler
duty was traded for stripping gas.
Lowering the reboiler duty from 2.5
to 2.0 MMBtu/hr and doubling the
stripping gas rate (from 3 to 6 SCF/
gallon) resulted in a predicted
water content of 0.13 lb/MMscf (a
dew point well below the hydrate
formation temperature). Simulated
glycol purity was 99.975 wt%. So it
appears that stripping gas is a more
effective way to improve dehydra-
tion system performance.
There is much that can be learned
by playing with a mass transfer rate
model, even for such a seemingly
mundane process as glycol dehy-
dration. The mass and heat transfer
rate model allows one to probe the
limits of what is in practice
possible; for example, trading
expensive reboiler duty for possibly
cheaper stripping gas, or optimising
dehydration unit performance in a
demanding LNG setting. Since this
kind of tool is so refective of the
real engineering physics, the real
chemistry and the real processing
happening in an amine, dimethyl
ether of polyethyleneglycol
(DMPEG) or glycol plant, using the
tool as a virtual laboratory, pilot
plant and full-scale virtual plant is
very attractive.
Dehydration of nearly pure acid
gases: H
2
S
Down-well disposal of acid gases
(so-called acid gas injection)
requires the gas to be compressed
to a very high pressure. If the gas is
wet, compressing it will cause
liquid water to drop out and this
liquid will be saturated with acid
gas at high pressure. Unless one is
8 0
1 0 0
1 2 0
6 0
5 1 0 1 5 2 0

W
a
t
e
r

i
n

t
r
e
a
t
e
d

g
a
s
,

l
b
/
M
M
s
c
f
Number of t rays
4 0
C O
2
S w e e t
H
2
S
Figure 8 Effect of tray count on water content of methane, CO
2
and H
2
S dehydrated
with 99.5 wt% TEG at 15 psig and 100F using bubble trays
8 0
1 2 0
1 3 0
1 0 0
6 0
7 0
1 1 0
9 0
5 0
1 2 1 6 2 0 2 4 2 8

W
a
t
e
r

i
n

t
r
e
a
t
e
d

g
a
s
,

l
b
/
M
M
s
c
f
Packi ng dept h, f t
4 0
C O
2
S w e e t
H
2
S
Figure 9 Effect of tray count on water content of methane, CO
2
and H
2
S dehydrated with
99.5 wt% TEG at 15 psig and 100F using Mellapak M170.X structured packing
ogtrt.indd 7 8/3/11 14:51:14
prepared to build compressors and
other equipment from unobtainium,
it is paramount that the water be
removed from the gas before
compression. Therefore, it is of
interest to compare dehydration of
otherwise pure but water-saturated
H
2
S with the equivalent sweet gas
volume. As Figure 8 shows, it is
diffcult to lower the water content
to the same level as high-pressure
pipeline gas, for example, simply
because of the treating pressure.
The lines in this plot are simulation
results for bubble cap trays in a
10ft-diameter column treating 30
MMscfd of either pure methane,
CO
2
or H
2
S saturated with water at
120F. Lean solvent is 200 gpm of
99.5 wt% TEG at 100F. Figure 9
shows similar results for Mellapak
M 170.X structured packing.
Under otherwise identical condi-
tions, the water content of H
2
S in
equilibrium with the lean TEG is
somewhat higher than it is for
either CO
2
or even methane. This
has to do with the very polar nature
of H
2
S, which binds water more
tightly to itself than CO
2
does. But
this is information that can be
gleaned more easily directly from
VLE calculations. The mass transfer
rate-based dehydration model,
however, shows that in this instance
H
2
S is pretty much completely
dehydrated after about 12 trays
(about 20ft of packing), whereas
CO
2
and methane require 15 trays
(2426ft of packing) to reach their
ultimate dryness.
In an overall sense, the acid gases
appear to be no harder than hydro-
carbons to dehydrate, although H
2
S
is very polar, which leaves it a little
wetter than most other gases.
However, there is nothing particu-
larly challenging about dehydrating
the acid gases, except for their high
solubility in the TEG solvent. For
the conditions corresponding to
Figures 8 and 9 (15 psig and 120F),
the methane, CO
2
and H
2
S content
of the rich solvent is simulated to
be 8.4 ppm, 0.66 mol% and 3.32
mol%, respectively. Thus, the water-
laden solvent contains quite a bit of
H
2
S and this must be removed
before the solvent is regenerated to
a suitable state of dryness, because
it is a certainty that this much H
2
S
cannot be vented to atmosphere
with the water vapour stripped
from the solvent.
Its high solubility in TEG is really
the main diffculty with dehydrat-
ing very high concentration H
2
S;
otherwise, in terms of just water
absorption, the acid gases are no
more challenging than any other
gas from a purely process stand-
point. The diffculty is accurately
representing the gas-phase non-
idealities that arise from the highly
polar nature of water and H
2
S.
Conclusions
Rules of thumb are fraught with
danger because often the rules
simply do not apply. A blanket tray
effciency of 25% is close to the
truth most of the time. But in deep
water removal, 25% is optimistic
and, unless one adds several addi-
tional safety trays, failure will
threaten. The situation with pack-
ing, be it random or structured, is
much worse. Quoting or recom-
mending a single HETP or HTU is
foolhardy. The right value depends
on the packing as well as on the
operating conditions and the gas
dryness sought. Rules of thumb had
their place when the best one could
do was an equilibrium-stage calcu-
lation, and reliance had to be placed
on experience as expressed (and
mis-expressed) in such rules of
thumb. Today, we have available
powerful heat and mass transfer
rate-based simulation tools, and
reliance on rules of thumb and
other approximations and guesti-
mates is no longer warranted.
The beauty of the heat and mass
transfer rate-based approach to
simulation is that one never has to
worry about tray effciencies,
artifcial residence times, HETPs,
HTUs and other rules of thumb.
The mass transfer rate-based tower
model does not use rules of thumb
it uses science and good engi-
neering to predict performance.
These results were all obtained
without any correction factors
whatsoever. They are true predic-
tions in every sense of the word.
ProTreat is a mark of Optimized Gas Treating
Inc.
References
1 Bestani B, Shing K S, Infnite dilution activity
coeffcients of water in TEG, PEG, glycerol and
their mixtures in the temperature range 50 to
140C, Fluid Phase Equilibria, 50, 1989.
2 Carroll J J, What is Henrys Law?, Chem. Eng.
Prog., 87, 9, 1991, 4852.
3 Carroll J J, Use Henrys Law for
multicomponent mixtures, Chem. Eng. Prog.,
88, 8, 1992, 5358.
4 Clinton P, Hubbard R A, Shah H, A review
of TEG-water equilibrium data and its effect
on the design of glycol dehydration units,
Laurence Reid Gas Conditioning Conference,
Norman, OK, 2008.
5 Engineering Data Book, Gas Processors
Suppliers Association, 12th Ed, vol II, 20, 2004.
6 Parrish W R, Won K W, Baltatu M E, Phase
behavior of the triethylene glycol-water system
and dehydration/regeneration design for
extremely low dew point requirements, 65th
Annual Convention of the GPA, San Antonio,
TX, 1986.
John J Carroll is the Director, Geostorage
Process Engineering for Gas Liquids
Engineering, Ltd in Calgary, Canada. He
was recently appointed Honored Visiting
Professor at Southwest Petroleum University
in Chengdu, China. He holds bachelor and
doctoral degrees in chemical engineering
from the University of Alberta, Edmonton.
Nathan Hatcher is Vice-President, Technology,
Optimized Gas Treating, Buda, Texas. He has
spent most of his career in amine treating and
sulphur recovery, and has extensive experience
in the practical application of process
simulation to operations, troubleshooting and
training. He holds a BS in chemical engineering
from the University of Kansas.
Email: nate.hatcher@ogtrt.com
Ralph Weiland founded Optimized Gas
Treating in 1992 and has been active in Canada,
Australia and the US in research in gas treating
since 1965. He developed the frst mass
transfer rate-based model for amine columns
for Dow Chemical and the Windows-based
ProTreat process simulation package. He holds
BA.Sc, MA.Sc, and PhD degrees in chemical
engineering from the University of Toronto.
Email: ralph.weiland@ogtrt.com
The mass transfer
rate-based tower
model uses science
and good engineering
to predict
performance
ogtrt.indd 8 11/3/11 11:21:17
Serving you along
the entire LNG Value Chain.
Rely on our expertise in process engineering as well as in the planning, project development and execution
of turnkey process plants. Proft from our competence along the entire LNG Value Chain:
Process plants Small and mid-scale LNG receiving terminals
BUhifU`[Ug`]eiYZUWh]cb LNG loading facilities for ship, truck and rail
BUhifU`[UgdfY!hfYUhaYbhB;@YlhfUWh]cb LCNG fuelling stations
B]hfc[YbfY^YWh]cb LNG fuelling systems
<Y`]iafYWcjYfmUbX`]eiYZUWh]cb Local storage and re-gasifcation
Coil-wound and plate-fn heat exchangers LNG tanker trailers for road and rail
Coldboxes Submerged combustion vaporizers
@B;ghcfU[YhUb_g @B;WUff]YfVc]`!cZZ[UgfY!`]eiYZUWh]cbgmghYag
Designing Processes - Constructing Plants.
Linde AG
Engineering Division
Dr.-Carl-von-Linde-Strae 6-14, 82049 Pullach, Germany
Tel. +49.89.7445-3784, Fax +49.89.7445-4928
E-Mail: natural-gas-plants@ linde-le.com, www.linde.com
z4151.indd 1 07.09.2010 09:54:10
www.ptqenquiry.com
linde.indd 1 8/3/11 13:57:52
Controlling corrosion in amine
treatment units
T
he amine unit plays a vital
role in the petroleum refning,
gas processing, coal gasifca-
tion and ammonia manufacturing
industries. With advances in hydro-
processing technologies to reduce
sulphur levels in gasoline and
diesel fuels, the requirement for an
effcient, well-established and relia-
ble separation system for the
removal of gases such as H
2
S and
CO
2
has become an important
requirement for refners. Amine
treatment has proved to be the
principal commercially established
method for gas/liquid purifcations
by removal of H
2
S and CO
2
.
In many cases, the removal of
only H
2
S is required; CO
2
remains
in the system, where it can be
managed by optimising the reaction
rates of amines and treating gases.
The treatment involves the removal
of H
2
S and CO
2
gases from fue
gases and LPG, with the help of
amines such as monoethanolamine
(MEA), diethanolamine, (DEA)
methyldiethanolamine (MDEA) and
diisopropanolamine (DIPA), which
have a tendency to absorb both the
gases. In an absorber column, sour
fue gas or LPG comes into contact
with lean amine. The treated gas/
LPG goes for end use and the H
2
S-
rich amine then goes to the
regenerator column.
During this process, major prob-
lems of corrosion and instability of
operation raise a signifcant threat
to an amine gas-treating plant,
resulting in unscheduled break-
downs and outages. Major corrosion
failures in these units have been
attributed to free acid gas and high
temperatures. Within this process,
contaminant byproducts called heat
A range of measures can be taken to minimise corrosion in amine units
JAyA RAwAT, Peddy V C RAo and N V ChoudARy
Bharat Petroleum Corporation Ltd
stable salts (HSS) are formed and
they gradually build up beyond
tolerable limits in the amine circula-
tion loop. Amine plant operational
problems, such as excessive foam-
ing, corrosion and capacity
reduction, are often attributed to
the accumulation of amine HSS.
Signifcant amine loss has been
observed in the operation of these
units because of its high foaming
characteristics.
Amine treatment process
The treatment involves the removal
of H
2
S and CO
2
from fuel gas and
LPG with the help of amine solu-
tions, which tend to absorb both
gases. A process fow diagram of
the unit is shown in Figure 1. In the
absorber column, the sour fuel gas
or LPG comes into contact with
lean amine, which absorbs H
2
S and
CO
2
. The treated gas/LPG goes for
end use and the H
2
S- and CO
2
-rich
amine then goes to the regenerator
column. In the regenerator column,
the rich amine solution is stripped
of its absorbed sour gases, with
steam as the heating medium, so
that the regenerated amine can be
reused in the absorber.
The steam strips out the absorbed
H
2
S and CO
2
present in the amine
solutions according to the following
reactions:
R
2
NH
3
S R
2
NH + H
2
S (1)
(R
2
NH
3
) CO
3
R
2
NH + CO
2
+ H
2
O (2)
where R is a CH
2
CH
2
OH group.
The liberated acid gases and
steam from the top are cooled in
the overhead condensers. The
uncondensed gas goes to the
sulphur recovery unit via a pres-
sure control valve and condensed
liquid is pumped back to the
column as a total refux. From the
F e e d
g a s
Tre a te d g a s
A c id g a s
S o u r c o n d e n sa te
Absorber
Flash
drum
Reflux
drum
Lean
cooler
Condenser
Stripper
Lean/rich
exchanger
R ic h
a m in e
L e a n a m in e
N a O H
Wa ste
Wa te r
P ro d u c t
Figure 1 Process fow diagram for amine treatment unit
bpcl.indd 1 8/3/11 14:53:49
regenerator bottom, lean amine
exchanges heat with feed-rich
amine before entering the
absorbers.
Absorption process
The absorption process is used to
remove a component (solute) from
the gas stream by contacting the
gas with a liquid solution (see
Figure 2 and Equation 3):
The amine solvent is the water
solution of an alkanolamine used to
remove the H
2
S, mercaptans and
CO
2
. The most common amines
used include MEA, DEA, DIPA,
MDEA and diglycolamine (DGA).
These amines are very water solu-
ble because they contain hydroxyl
groups. The nitrogen or ammonia
group (NH
3
) reacts with the acid
gas to form a water-soluble salt.
Typical properties of the amines
and their structures are given in
Table 1.
Some of the important criteria for
the selection of amines are:
High solubility
Low volatility and viscosity
Easy recovery
Least corrosive and least cost.
Corrosion problems in amine units
Corrosion in the amine system is
mainly caused by acid gases, strong
acids, contamination such as amine
HSS and amine degradation prod-
ucts. H
2
S formed during some
reactions also causes corrosion in
iron-based metallurgies by forming
iron sulphide (FeS). CO
2
corrosion
is caused by the reduction of undis-
sociated carbonic acid in turbulent
areas, where a passive ferrous
carbonate scale is unable to form.
Carbonic acid can undergo the
following reactions:
H
2
CO
3
+ e- HCO
3
- +H (4)
H
2
CO
3
CO
2
+ H
2
O (CO
2
gas evolution) (5)
The organic acids and many
amine degraded products, high
velocities and turbulence break
down an FeS protective layer and
remove iron from the metal. The
corrosion product is water stable
until it contacts the high H
2
S
concentrations in the absorber. FeS
is formed and precipitates to create
deposits, while acid is returned to
resume the corrosion cycle.
Corrosion in the overhead
systems is caused by an accumula-
tion of high concentrations of H
2
S,
nitrogen compounds and CO
2
along
with a variety of acids. Rich amine
causes pitting of the exchangers
and piping, as well as the reboiler
section, where free acid gases and
higher temperatures are the main
driving forces of corrosion
Formation of heat stable amine salts
Heat stable amine salts (HSAS) are
formed by the reaction of inorganic
contamination of the amine solution
with strong and weak acids. Acids
increase the corrosion, fouling and
emulsifcation potential and reduce
the gas-treating capacity. HSAS are
heat stable because these acids are
not removed in the regenerator.
Acids with a pKa of less than 5 will
form HSAS, mainly formic, cyanic
and sulphuric acids. Formic acid is
very corrosive to carbon steel and
stainless steel in the absence of
oxygen, while MDEA formate is
more corrosive than formic acid
alone. Although low concentrations
of HSAS are not corrosive, corro-
sion, fouling and foaming increase
as the acids concentration increases.
The reaction of amines with weak
acid gases gives thermally reversi-
ble salts as follows:
MDEA + H
2
S MDEAH + HS- (6)
The reaction of amines with acids
stronger than acid gases produces
HSAS:
MDEA + H MDEAH
+
+ X- (7)
where HX is a strong acid and X-
denotes an acid anion such as
formate, acetate, propionate, glyco-
late, oxalate, sulphate, thiosulphate,
thiocyanate, phosphate, nitrate and
chloride. Alkalies such as NaOH
and KOH also react with strong
acids and contribute to the forma-
tion of HSAS.
Salts formed by either of these
reactions lower the pH of the solu-
tion and increase conductivity,
thereby making the system more
acidic and prone to corrosion.
This article deals with the differ-
ent types of corrosion mechanisms
in an amine treatment unit and the
control strategies to deal with them,
N a tu ra l
g a s
L e a n so lve n t
S we e t g a s
R ic h so lve n t
L iq u id
G a s
Figure 2 Schematic presentation of
absorption process
S.No. Amine Chemical structure Molecular Base strength,
solvent weight pKa
Monoethanolamine (MEA) 61.8 9.52

Diethanolamine (DEA) 105.14 8.88

Diglycolamine (DGA) or 105.14 9.50
aminoethoxyethanol

Methyldiethanolamine (MDEA) 119.16 8.56

Diisopropanolamine (DIPA) 133.19 8.82
Amine solvents
Table 1
H
2
N
O H
O H H O
H
N
H O
O
H
2
N
O H
H O
N
O H H O
H
N
H
2
N
O H
O H H O
H
N
H O
O
H
2
N
O H
H O
N
O H H O
H
N
Cooling
Lean amine + H
2
S + CO
2
amine rich in H
2
S + CO
2

Heating (3)
bpcl.indd 2 8/3/11 14:53:56
including selection criteria for
amine solvents and metallurgical
guidelines for the construction of
amine plants.
Results and discussion
Corrosion studies were carried out
with rich and lean solutions of DEA
under varying operating conditions
with various metallurgies, such as
carbon steel, stainless steel and
Monel. The solvents used for the
experiments were DEA, MEA and
MDEA. Weight loss studies were
carried out on the process samples
using carbon steel coupons accord-
ing to ASTM G1-90 and ASTM G
3172. The experimental setup
consisted of a reactor, condensers
and a constant temperature bath.
Various corrosion coupons of carbon
steel were placed with the help of
specially designed coupon holders
in the reactor and condenser section
in such a way as to observe corro-
sion in the liquid as well as vapour
phase. The liquid temperature was
maintained at 120 +/- 1C. A flming
amine-based corrosion inhibitor was
used and its dosages with various
samples were optimised. The
weights of the coupons were meas-
ured before and after the test, and
corrosion rates were then calculated.
Effect of corrosion rate with
various metallurgies
The results of the corrosion studies
were obtained from various amine
solutions, including DEA, MEA and
MDEA, along with various metall-
urgies (see Table 2). Amine
solutions at 15% along with 5000
ppm of dissolved H
2
S were used
for the studies.
It was observed that corrosion
rates are lower with Monel and
stainless steel metallurgies,
compared to carbon steel. Further
corrosion rates were observed to be
lower with MDEA and MEA,
compared to DEA.
Effect of fuid velocities on
corrosion rates
Corrosion rates vary with fuid
velocities and amine types. A fuid
velocity higher than 1.5 m/sec is, in
the case of a rich amine solution,
more corrosive than a lean amine
solution. When amine solutions are
loaded with acid gases, their
tendency to cause erosion and
corrosion increases, so that, in the
event of higher fuid velocities
along with turbulence, the passive
layer of FeS becomes damaged,
which results in higher corrosion
rates. The recommendation is to
operate at lower velocities with rich
amine solutions rather than lean
amine solutions
Effect on corrosion rate of varying
temperatures
The corrosion rates of carbon steel
were studied at temperatures vary-
ing between 60C and 120C, with
various loadings of H
2
S. It was
observed that the corrosion rate
rises with an increasing H
2
S concen-
tration. Furthermore, corrosion
rates are greater at higher tempera-
tures. At 120C, a corrosion rate of
40 mil/y was obtained with a H
2
S
loading of 5%. Hence, at higher
temperatures, with greater percent-
ages of H
2
S, the use of a corrosion
inhibitor is recommended.
Effect of corrosion inhibitor
Corrosion inhibitors of the flming
amine type were found to be the
most suitable for reducing corrosion.
Concentrations of corrosion inhibitor
Metals Corrosion rates, MPY
DEA solvent MEA solvent MDEA solvent
Carbon steel 15.7 10.6 8.4
304 SS 10.5 5.8 3.4
316 SS 9.5 3.8 1.2
Monel 5.6 4.5 0.9
Corrosion rates of various alloys with amine solvents
Table 2
30
35
25
20
15
10
5
Lean amine Rich amine
Y
P
M

,
e
t
a
r

n
o
i
s
o
r
r
o
C

Amine types
0
Ve|ocity 1.0 m/sec
Ve|ocity 1.25 m/sec
Ve|ocity 1.5 m/sec
Ve|ocity 2.0 m/sec
Figure 3 Corrosion rates of rich and lean amine at various fuid velocities
30
45
40
35
25
20
15
10
5
60 90 120
Y
P
M

,
e
t
a
r

n
o
i
s
o
r
r
o
C

Temperature, C
0
H
2
S (1}
H
2
S (2}
H
2
S (5}
Figure 4 Corrosion rates at various H
2
S conc and temperature
bpcl.indd 3 8/3/11 14:54:08
amine loading and increasing the
circulation rate
Circulation rate increases the corro-
sive action of HSAS, fashing acid gas
amine degradation products, high
temperature, acid gas loading and
high amine concentration.
Use of corrosion inhibitors
Injection of flming amine-type
inhibitor should be carried out in
the regenerator overhead, directly
upstream of the problem area,
usually via the reboiler feed line
Inhibitor injection rates and
dosages should be based on contin-
uous corrosion monitoring.

Important properties of corrosion
inhibitors
Low to no foaming tendency
Good solids dispersion
Good solubility in water
Should provide good adherence
to FeS protective layer on the metal
surface
Effective against organic acid,
CO
2
, H
2
S and amine degradation
product corrosion
Compatibility with all the alloys
in the system.
Inhibitor injection points and
corrosion monitoring
Commonly added to the regenera-
tor overhead, the inhibitor returns to
the tower with the refux and then is
distributed throughout the unit
Water wash the overhead
condenser with a refux slipstream,
routinely purge the ammonia in the
refux accumulator and inject the
inhibitor in the reboiler feed line
Pump and water wash to control
overhead corrosion usually caused
by ammonia salts deposition
Corrosion monitoring by ER and
LP probes. LP probes should use a
data logger to average multiple
readings
Routine amine tests should be
performed for amine concentration,
acid gas loading and foaming
tendency
Rigorous analysis for acids and
metals should be conducted every
46 weeks.
Summary
Corrosion in amine treatment units is
critical as it involves a check on vari-
ous operating and mechanical
parameters, such as the selection of
solvent, operating conditions and
metallurgies. It can be minimised by:
A suitable selection of solvent
Monitoring HSAS and metals and
minimisation of HSAS formation
The selection of proper corrosion
inhibitor with optimised dosage
and its injection at proper locations
in the unit
Water washing the overhead
with refux
Continuous corrosion monitoring.
Acknowledgement
The authors express their sincere thanks
to BPCL management for their constant
encouragement and support.
References
1 Dupart M S, Bacon T R, Edwards D J,
Hydrocarbon Processing, 1993, 311.
2 Verma N, Verma A, Fuel Processing
Technology, 90, 2009, 483489.
3 Fontana M G, Corrosion Engineering, McGraw
Hill Publication, 1987.
4 Rooney P R, Dupart M S, Bacon T R,
Hydrocarbon Processing, 1997, 6571.
Jaya Rawat is a Deputy Manager, R&D, in the
Corporate R&D Centre of Bharat Petroleum
Corporation Ltd, Surajpur, India. She holds a
doctorate in chemistry and has published 50
research papers and fve patents.
Email: jayarawat@bharatpetroleum.in
Peddy V C Rao is Senior Manager, R&D, with
Bharat Petroleum Corporation Ltd. He holds a
doctorate in chemistry from Indian Institute of
Technology, Bombay. India.
Email: raopvc@bharatpetroleum.in
N V Choudary is a Chief Manager, R&D, with
Bharat Petroleum Corporation Ltd. He holds a
doctorate in chemistry, has published over 70
research papers and fled over 50 patents.
Email: choudarynv@bharatpetroleum.in
varying from 5 to 20 ppm were used
for the studies, along with DEA and
a 5% H
2
S loading. A reduction in the
corrosion rate was observed at all
temperatures in the range 60120C.
The corrosion rate was very high
without a corrosion inhibitor (22
mil/y at 120C), whereas a signif-
cant reduction in the corrosion rate
was observed after dosing the corro-
sion inhibitor at 10 ppm.

Corrosion control strategies in
amine treatment units
Minimising HSAS formation
HSAS build-up tolerance is
specifc to the amine solution used
With increased HSAS concentra-
tion (above 4%), fouling, foaming
and corrosion problems increase.
With MEA and DEA (10% of amine
concentration), 23% HSAS does
not create severity, but, in the case
of MDEA, 2% HSAS creates severe
corrosion.
Optimising fuid velocities
High velocities and turbulence
damage the protective FeS layer.
The process of forming scale results
in higher corrosion rates, as erosion
forces removal of the scale
To reduce erosion, amine veloci-
ties should be no greater than
1 m/sec (3 ft/s) in exchangers and
2.1 m/s (7 ft/s) in piping.
Acid loadings
Acid gas loading, circulation rate
and amine concentration should also
be adjusted to minimise corrosion
Amine concentration should be
increased accordingly before rich
1 5
2 0
2 5
1 0
5
6 0 9 0 1 2 0

C
o
r
r
o
s
i
o
n

r
a
t
e
,

M
P
Y
Temperat ure, C
0
C o rro sio n in h ib ito r ( 0 p p m )
C o rro sio n in h ib ito r ( 5 p p m )
C o rro sio n in h ib ito r ( 1 0 p p m )
Figure 5 Corrosion rates with different dosages of corrosion inhibitor
bpcl.indd 4 8/3/11 14:54:15
A promoter for selective H
2
S
removal: part I
T
he selective removal of H
2
S
has become an important
topic over past decades. This
is driven by several factors, one
being the production of an H
2
S-
enriched, and thus high-quality,
Claus gas in MDEA-based acid gas
enrichment (AGE) units.
Other felds where selective gas
treatment would also be benefcial
are natural gas and refnery appli-
cations. For sour gas felds, this
application is becoming increas-
ingly attractive due to limited sweet
gas resources. For refneries, debot-
tlenecking issues and an increased
fexibility for processing different
crudes are the most important
drivers.
From an operational perspective,
savings in energy and circulation
rate, as well as a reduction in equip-
ment sizing, are the obvious
benefts of enhanced selective
treatment. In addition, tight envi-
ronmental regulations and sulphur
specifcations are attributed to this
subject.
The principles for the selective
removal of H
2
S with amine-based
solvents follow three major routes:
Hindered amines, controlling the
selectivity primarily in the absorber
Various design options and
absorber internals, affecting the
difference in CO
2
and H
2
S mass
transfer kinetics
Promoted tertiary amines, focus-
ing more on enhanced regeneration
and thus leading to lower H
2
S
loadings.
With respect to the fnal point,
the advanced promoter system
presented in this article can be a
considerable leap forward for more
fexible selective designs. Very low,
A new MDEA promoter achieves very low H
2
S lean loadings and the option for
fexible design of acid gas enrichment units by varying absorber heights
GErAlD VorbErG, rAlf NoTz and TorSTEN KATz BASF SE
WiElAND WAcHE and clAuS ScHuNK Bayernoil Raffneriegesellschaft
achievable lean loadings are an
option to adjust selectivity by vary-
ing the absorber height without
losing control over a tight H
2
S spec-
ifcation with a suffcient safety
margin.
This article gives an overview of
the principles, while part II (see
PTQ, Q2 2011) shows the ability of
this promoter system to drop H
2
S
lean loadings in a refnery amine
system.
Acid gas removal with amine-
based solvents is a mature and
widespread application in the oil
and gas industry. Besides specifc
design variations, acid gas removal
units (AGRU) always follow the
principle of an absorber-regenerator
confguration. First, acid gases are
removed from the fuid stream in
the absorber by the liquid solvent
typically at 2050C and elevated
pressures up to 80 bar, depending
on the feed gas conditions. In a
second step, the dissolved acid
gases are desorbed in a regenerator
at inverse conditions. Desorbed
acid gases can be further processed
in Claus sulphur recovery units
(SRU), reinjected for enhanced oil
recovery (EOR) or simply fared.
When focusing on the two main
acid gas components, CO
2
and H
2
S,
process designers differentiate
between total acid gas removal and
selective sulphur removal or simply
selective removal. As the name
implies, selective sulphur removal
selectively removes H
2
S, while
other acid gases, for instance CO
2
,
are slipped into the treated gas.
Consequently, selective removal
has a different focus compared with
total acid gas removal processes,
such as BASFs aMDEA process,
where simultaneous removal of H
2
S
and CO
2
is intended. This means
that suitable solvents have different
characteristics.
H
2
S selectivity
Defnitions
In industry, various expressions for
H
2
S selectivity are used to judge the
selectivity of an absorption process
with regard to H
2
S compared to
CO
2
. In the following, the most
important examples are given (cH
2
S
and cCO
2
stand for molar concen-
tration in gas streams):
H
2
S selectivity (rigorous defnition)
H
2
S selectivity is calculated as:
CO
2
co-absorption (or CO
2

pick-up)
CO
2
co-absorption in the absorber
is specifed as:
CO
2
_co-absorption = 1-
C
CO2
treated gas
C
CO2
feed gas
CO
2
/H
2
S ratio comparison (often
used for acid gas enrichment)
The decrease in CO
2
/H
2
S ratio indi-
cates the effciency of an acid gas
enrichment unit:
A proper comparison of these
expressions for different solvents or
applications must be evaluated with
great care, since the absolute level
of H
2
S in the treated gas is less
considered. In other words, H
2
S
selectivity is well defned, but not
H
2
S_selectivity =
(C
H2S
feed gas - C
H2S
treated gas)/C
H2S
feed gas
(C
CO2
feed gas - C
CO2
treated gas)/C
CO2
feed gas
Ratio
feed gas
=
C
CO2
feed gas
vs Ratio
acid offgas
C
CO2
acid offgas
C
H2S
feed gas C
H2S
acid offgas
basf.indd 1 9/3/11 12:35:20
www.eptq.com
particularly suitable for judging H
2
S specifcations in
treated gas.
The following example, calculated according to the
rigorous defnition, might show the difference:
Feed gas: H
2
S 1 v% CO
2
5 v%
Treated gas case 1: CO
2
1.5 v% H
2
S 1 vppm H
2
S selectivity = 1.428
Treated gas case 2: CO
2
1.5 v% H
2
S 50 vppm H
2
S selectivity = 1.421
Dependency on operating conditions
Selective treatment with amine-based solvents gener-
ally takes advantage of the rapid reaction of H
2
S
compared to the kinetically hindered reaction of CO
2
;
CO
2
frst has to react with water to form carbonic acid.
In particular, tertiary amines are often used for selec-
tive applications, as they are unable to form
carbamates, the only fast reaction involving CO
2
.
These are the reactions of tertiary amines in aqueous
solutions:
Reaction of water and amine (fast)
R1R2R3N + H
2
O R1R2R3NH
+
+ OH
_
2 H
2
O H
3
O
+
+ OH
_
H
2
S reaction (fast)
H
2
S + H
2
O HS
_
+ H
3
O
+
CO
2
reactions (overall reaction slow):
CO
2
+ 2H
2
O HCO
3
_
+ H
3
O
+
(slow)
HCO
3
_
+ OH
_
H
2
O + CO
3
2-
(fast)
In this reaction system, CO
2
co-absorption, and thus
H
2
S selectivity, is heavily infuenced by reaction condi-
tions such as the temperature and concentration of CO
2

and H
2
S in the liquid phase. This means higher pressure
and temperature, as well as a higher CO
2
/H
2
S ratio in
the feed gas, favour CO
2
co-absorption and lower H
2
S
selectivity.
Applications
Selective H
2
S removal is predominantly required and
applied in the following segments:
Natural gas (NG)
Refnery off-gas (ROG)
Claus tail gas (TG)
AGE.
AGE has probably been the most important driver for
the development of selective solvents in recent years. It
requires utmost H
2
S selectivity in order to get a H
2
S-rich
feed gas for the downstream Claus SRU or for down-
hole reinjection. For instance, a minimum H
2
S
concentration of 25 vol% is a guiding value for SRUs.
Higher H
2
S concentrations are naturally more desirable.
Upgrading and debottlenecking measures triggered
the move to more selective treatment in refneries by
replacing unselective with selective generic solvents. In
fact, the majority of refnery expansions lead to a higher
total sulphur nameplate capacity by processing more
sour crudes. Without signifcant mechanical changes,
capacity increases of more than 25% have been achieved,
as mainly H
2
S, instead of CO
2
and H
2
S, has been proc-
essed along the entire acid gas removal chain. Moreover,
B
O
R
S
I
G
BORSIG GROUP
Egellsstrasse 21
D-13507 Berlin/Germany
Phone: +49 (30) 4301-01 Fax: +49 (30) 4301-2236
E-mail: info@borsig.de www.borsig.de
Leading Technology for
Innovative Solutions
Process Gas Waste Heat
Recovery Systems
for ammonia, methanol, hydrogen, gas-to-liquid plants, coal
gasification plants, direct reduction plants in iron metallurgy,
waste heat boilers in nitric acid, caprolactam and formalde-
hyde plants, waste heat boilers downstream of the partial
oxidation of oil and natural gas
Quench Coolers
for ethylene cracking furnaces:
- Transfer Line Exchangers (TLE)
- BORSIG Linear Quenchers (BLQ)
Scraped Surface Exchangers
Other BORSIGProducts:
Compressors, Compressor Valves and Blowers, Membrane
Technology, Boiler and Power Plant Technology, Industrial
and Power Plant Services
www.ptqenquiry.com
basf.indd 2 9/3/11 12:35:56
tighter sulphur regulations in many
countries have been fulflled in
combination with improved sulphur
recovery rates by implementing
Claus tail gas treatment (TGT),
where H
2
S is selectively removed
and returned to the Claus inlet.
From todays market perspective,
selective acid gas removal in natu-
ral gas is an increasingly interesting
application. On the one hand, the
utilisation of more sour gas felds is
increasing rapidly, while the access
to sweet natural gas resources
becomes more and more limited
(sour in colloquial terms means H
2
S
>5 vol%). On the other hand, for
sales gas specifcation, operators
request not the highest level but a
dedicated or even an adjustable
level of selectivity. This means that
CO
2
slip, and thus H
2
S selectivity,
needs to be controlled to match, for
instance, a certain calorifc value for
the sales gas. CO
2
treated gas speci-
fcations between 1.5 and 2.5 vol%
are most common. But what does
this mean for selective designs?
In fact, the quality and quantity
of production wells change over the
years, but the treated gas specifca-
tion should remain within a narrow
window at all times. This task is a
challenge for both process design
and solvent activity. Proper selec-
tive designs should be able to
handle:
Various turndown rates
Changing environmental condi-
tions, in particular temperature
Various feed gas scenarios.
Table 1 shows an overview of the
application felds and their major
differences in conditions and
requirements. The listed criteria are
crucial for the solvent as well as for
design selection.
Solvents for H
2
S selective treatment
Three types of solvents are most
suitable for selective treatment:
generic methyldiethanolamine,
MDEA, (tertiary amine) and blends,
promoted MDEA and severely ster-
ically hindered amines.
Generic MDEA (tertiary amine)
and blends
Among selective amines, MDEA is
the most common due to its many
benefcial properties, in particular
the inability to enter into the
carbamate reaction with CO
2
. Its
ability to change its characteristics
either for total acid gas removal or
selective removal by adding an acti-
vator or a promoter has increased
its range of application because of:
Availability and pricing
Fewer corrosion issues than with
primary and secondary amines
Good absorption performance
Suitability for generic designs
Selectivity adjustment by blend-
ing with other amines
Low regeneration demand for
absorbed CO
2
.

Promoted MDEA
H
2
S is a stronger acid than CO
2
and
thus has a stronger binding energy
at the stripper temperature.
Consequently, amine regeneration
of H
2
S down to trace levels becomes
progressively more diffcult. In
particular, designs with very low
H
2
S specifcations in the low ppm
range may use regeneration
promoters in order to release H
2
S
more easily from the solution. In
this regard, MDEA acidifed with
phosphoric, sulphuric or other acids
becomes regenerated more easily.
Even the unintentional acidifcation
by formation of heat stable salts
(HSS) and their acids acetates,
formates, sulphates and so on
leads to a very effcient H
2
S
regeneration boost. However, the
level of acidifcation has a certain
limit, as there is an impact on the
pH value of the amine solution and
the equilibrium curve of the acid
gas solubility in the solvent.
Severely sterically hindered amines
Sterically hindered amines, such as
tertiarybutylaminoethoxyethanol,
incorporate two useful properties
at low acid gas partial pressures;
as a secondary amine, the reaction
is fast and acid gas capture capac-
ity is higher compared to tertiary
amines. In contrast to generic
secondary amines, steric hindrance
due to the shape of the molecule
reduces the CO
2
equilibrium load
and thus increases the selectivity.
However, both advantages are
progressively reduced at increas-
ing acid gas partial pressures, such
as those in most natural gas appli-
cations (see Table 1). Thus, for
highly selective applications at low
H
2
S partial pressures, such as acid
gas enrichment, hindered amines
are better positioned and highly
effcient. Moreover, the relatively
high price level of hindered amines
makes it more economical in small
plants (AGE) than in large (natural
gas) plants.
Design specifcs for selective
treatment
For the design, various aspects have
to be taken into account:
H
2
S specifcation in the treated
gas
CO
2
specifcation/slip in the
treated gas
Feed gas conditions (composition,
pressure and temperature), which
may vary over the plants lifetime
Required CO
2
/H
2
S ratio in the
acid off-gas
Ambient conditions
Flexibility/turndown rate
Footprint and size of the plant.
For MDEA-based or other
solvents, for which selectivity is
controlled by kinetic effects, several
factors are important for design:
CO
2
reaction kinetics and number
of theoretical stages, mass transfer
Natural gas Acid gas enrichment Refnery offgas Tail gas
Typical feed gas specifcation
pH
2
S, mbar/psi Up to 10 000/145 Below 500/7.25 Up to 2000/29 Below 100/1.45
pCO
2
, mbar/psi Up to 10 000/145 Below 2000/29 Up to 1000/14.5 Below 500/7.25
Typical treated gas specifcation
cH
2
S, vppm 510 5100 5100 550
CO
2
slip Flexible Very high Medium High
System hold-up Large Medium Medium Small
p = partial pressure/c = concentration
Characterisation of H
2
S selective applications
Table 1
basf.indd 3 9/3/11 12:36:11
tions, in the very low ppm range,
are required. This means that a
design with a large safety margin is
highly advisable. As a consequence,
process designers apply multiple
amine feed points to the absorber
to adjust the mass transfer area. In
some cases, even fexible weir
heights are designed as a sophisti-
cated solution. An even higher
degree of selectivity through reduc-
tion of the tower height could be
achieved by reducing the lean load-
ings and by increasing the driving
force for the mass transfer at the
absorber overhead.
Mass transfer and internals selection
The pathways of mass transfer for
CO
2
and H
2
S are quite different.
While CO
2
absorption is affected by
mass transfer limitations in the
liquid phase, the instantaneous H
2
S
reaction in the solvent moves the
major H
2
S mass transfer resistance
into the gas phase. As a conse-
quence, an adequate design tool has
to consider these mass transfer
resistances for different compo-
nents. Rate-based simulation tools,
which also need detailed input for
the column internals, are capable of
predicting operational conditions
with the required accuracy. They
can even predict operating condi-
tions depending on the fow regime.
The results often depend on the
choice of internals: trays versus
packing.
Trays are very robust and, with
respect to multiple amine feed
points, the design is easy to imple-
ment. In addition, fouling or
plugging is a minor issue. On the
other hand, a packing design usually
enables smaller column diameters to
be selected and provides more safety
and fexibility for changing L/G
ratios. However, the installation of
multiple amine feed points is more
complex. Lower liquid hold-up and
lower back-mixing tendencies
increase the selectivity substantially
compared to trays operated in their
standard envelope.
Temperature and selectivity
Feed gas and amine temperatures
are crucial factors for CO
2
co-absorp-
tion and resulting selectivity, but the
infuence from a designers perspec-
tive is limited to the available
cooling capabilities. In regions where
seasonal ambient temperatures may
differ by more than 40C (72F) and
peaks may reach 50C (122F) or
more this is a particular challenge.
By taking the most severe condi-
tions for instance, high amine
and feed gas temperatures as an
example, CO
2
at a high partial pres-
sure will be absorbed more easily
in the amine solution. In addition,
the heat of CO
2
and H
2
S absorption
expedites the formation of an
and internals selection, and temper-
ature and selectivity.
CO
2
reaction kinetics and number of
theoretical stages
With respect to kinetics, shorter
absorber columns reduce the over-
all mass transfer area and overall
retention time, so CO
2
absorption is
reduced to a certain extent.
However if the absorber height is
too short, H
2
S may not be absorbed
suffciently. In todays selective
designs, very tight H
2
S specifca-
A d vA m i n e & A d va S u lf
yo u r b e st to o ls
i n so u r g a s tre a tm e n t
Ad vAm i n e
HiLoadDEA
MDEAmax
energizedMDEA
Ad va S u lf
Claus
Tail Gas Treatment, ClauspolII,
Sulfreen, Sultimate
Sulphur degassing, Aquisulf
P ro se rn a t o ffe rs
a g lo b a l ra n g e
o f a d va n ce d te ch n i ca l so lu ti o n s
fo r so u r g a s tre a tm e n t
a n d su lp h u r re co ve ry
To u r F R A N K L I N - 1 0 0 /1 0 1 T E R R A S S E B O I E L D I E U
F -9 2 0 4 2 P a ris L a D fe n se C e d e x - F ra n c e
P h o n e : +3 3 1 4 7 6 7 2 0 0 0 - F a x: +3 3 1 4 7 6 7 2 0 0 7
e -m a il: sa le s@ p ro se rn a t. c o m
h ttp : //www. p ro se rn a t. c o m
www.ptqenquiry.com
basf.indd 4 9/3/11 12:36:25
adverse temperature profle in the
absorber (see Figure 1). The contin-
uing acceleration of this
phenomenon has a negative infu-
ence on overall gas capacity and
H
2
S selectivity. Besides the specifc
heat of absorption and the amount
of absorbed CO
2
and H
2
S, further
parameters, such as the specifc
heat capacity of gas and liquid
phases, heat transfer as well as the
L/G ratio, play an important role.
It is obvious that at this point cross-
over effects between CO
2
and H
2
S
start to increase the complexity of
the system. It is therefore important
to understand the temperature
profle and location of the tempera-
ture bulge, in particular its distance
from the very sensitive absorber
top, to predict the absorption
performance of an operating plant.
Taking these issues into account,
design fexibility for selective natu-
ral gas plants is driven by:
Variation in absorber height
through multiple amine feed points
Low lean amine loading to
achieve H
2
S specifcation with suff-
cient safety margin
Flexibility in circulation rate to
manage the temperature profle in
the absorber and control CO
2
mass
transfer
Suffcient reboiler energy to keep
the lean loading low at various
circulation rates.
Promoted MDEA solvent for natural
gas plants
Most natural gas plants have CO
2

partial pressures >1.5 bara (>21.8
psi). At these conditions, MDEA-
based solvents are closer to
hindered amines from a selectivity
perspective. In addition, the rela-
tively high total acid gas partial
pressures of natural gas plants
reduce the acid gas capacity advan-
tage of hindered amines. Moreover,
temperature profle and bulge
control at high ambient tempera-
tures often determine the circulation
rates of many AGE units and result
in relatively low acid gas loadings
in the amine solution in the
absorber bottom. The required
over-circulation may become even
worse for hindered amine designs,
since the advantage of higher acid
gas absorption capacity is partly
compensated by the lower heat
capacity of a smaller solvent fow
rate. Temperature bulge effects in
hindered amines are comparatively
more sensitive regarding the
absorption performance. In some
cases, additional side coolers may
help to increase the performance of
these processes. Last but not least,
the size of todays natural gas
trains, with an amine hold-up of
several hundred tonnes, makes the
use of MDEA-based solvents much
more economical.
Promoted MDEA solvent for
refneries
The AGRU setup of a refnery is
very complex and the different
requirements of the absorbers prefer
an all-fts-one solution. Although
acid gas partial pressures are differ-
ent from those of natural gas plants,
MDEA-based solvents are well
established and are represented
6
5
4
3
2
1
7 . 5 8 . 0 8 . 5 9 . 0 9 . 5 1 0 . 0 1 0 . 5 1 1 . 0 1 1 . 5
B o tto m
pH
Absorber
0
A c id ifie d M D E A
M D E A
H ig h a c id g a s
c o n c e n tra tio n
1 0
8
9
4
5
6
7
3
2
1
7 . 5 8 . 0 8 . 5 9 . 0 9 . 5
pH
Desorber
0
M D E A
L o w a c id g a s
c o n c e n tra tio n
To p
Figure 2 Example of pH profles in an absorber (left) and a desorber (right) with generic MDEA and acidifed MDEA
3 0
5 0
4 5
4 0
3 5
2 5
2 0
1 5
1 0
5
4 0 4 5 5 0 5 5 6 0 6 5 7 0 7 5 8 0 8 5
A
b
s
o
r
b
e
r

h
e
i
g
h
t
,

s
e
g
m
e
n
t
s
Temperat ure, C
0
Figure 1 Absorber, liquid-phase temperature profles red: feasible T-profle at
appropriate L/G ratio; blue: T-bulge profle at too low L/G ratio
basf.indd 5 9/3/11 12:36:36
today in more than 40% of all refn-
ery amine systems. In addition, the
relatively high refll demand comes
along with pricing and availability.
With respect to regeneration energy,
the unintentional and sometimes
quick formation of HSS to concen-
trations of up to 3 wt% is supportive
to some extent. However, corrosion
issues force refners to neutralise
those acids with caustic solution.
New-generation promoter system
A new promoter has been devel-
oped and incorporated: BASFs
selective sMDEA+ technology. This
promoter formulation is highly eff-
cient in achieving very low H
2
S lean
loadings. In this respect, the
promoter formulation provides an
option to design selective units with
a high degree of fexibility by vary-
ing absorber heights, while low
acid gas lean loadings keep H
2
S
outlet concentrations at a very low
level.
The following describes the mech-
anism of achieving low acid gas
lean loadings.
Principles
Desorption mechanism in
the stripper
H
2
S desorption follows an equilib-
rium reaction between amine,
protonated amine and the dissolved
HS
_
anion, described in the follow-
ing reaction:
H
2
S + MDEA HS
-
+ MDEAH
+
CO
2
desorption can be described
by the following net equilibrium
reaction:
CO
2
+ MDEA + H
2
O HCO
3
-
+ MDEAH
+
The purpose of desorption is to
strip H
2
S and CO
2
from the rich
solution to obtain a regenerated
solvent. As a consequence of CO
2
and H
2
S mass transfer from the
liquid phase into the gas phase, the
above two equilibrium reactions
move towards the left side and the
pH value of the solution increases
accordingly from the top to the
bottom of the stripper.
Since CO
2
is a weaker acid
at typical stripper temperatures
compared with H
2
S, CO
2
is prefera-
bly released in the upper part of the
stripper, whereas a signifcant
amount of H
2
S as the stronger
acid is released in the lower part
of the stripper. Releasing CO
2
and
obtaining low residual CO
2
concen-
trations becomes easier than
achieving low H
2
S residual loading,
which requires much more strip-
ping stream and reboiler energy.
Low acid gas concentrations in
the lean solvent are directly linked
to the maximum achievable gas
purities in the absorber overhead:
unfortunately, most gases have very
tight H
2
S specifcations for
1 5
2 5
2 0
1 0
5

H
2
S

l
e
a
n

l
o
a
d
i
n
g
,

N
m
3
/
t
o
Regenerat i on energy
0
0 . 1 8
0 . 3 0
0 . 2 4
0 . 1 2
0 . 0 6

H
2
S

l
e
a
n

l
o
a
d
i
n
g
,

N
m
3
/
t
o
0 . 0 0
sM D E A +
M D E A
0 . 3
0 . 5
0 . 4
0 . 2
0 . 1

H
2
S

l
e
a
n

l
o
a
d
i
n
g
,

N
m
3
/
t
o
Regenerat i on energy
0 . 0
0 . 0 0 3 6
0 . 0 0 6 0
0 . 0 0 4 8
0 . 0 0 2 4
0 . 0 0 1 2

H
2
S

l
e
a
n

l
o
a
d
i
n
g
,

m
o
l
/
m
o
l
0 . 0 0 0 0
Figure 4 H
2
S regeneration lab tests (reduced desorber height)

H
2
S

p
a
r
t
i
a
l

p
r
e
s
s
u
r
e
H
2
S l oadi ng
M D E A +p ro m o te r
P ro m o te d M D E A
M D E A
S trip p e r
su m p
A b so rb e r
h e a d
5 0 C
1 2 0 C
Figure 3 Infuence of the acidic promoter formulation on the isothermal equilibrium curves at 50C and 120C (H
2
S partial pressure
regime of 1 mbar)
basf.indd 6 9/3/11 14:46:11
Fr om Engi neer i ng Thr ough Fabr i c at i on
For top performing mass transfer and tower internal components let AMISTCO
Separation Products use your existing drawings or modify them to improve your
process. Our experienced in-house engineering and full scale fabrication
capabilities can streamline your project from inception to installation.
Valve (numerous options)
Sieve or perforated
Bubble cap trays
Cartridge trays
DuaI ow
Bafe
www.amistco.com 24hr EMERGENCY 8ERVCE 1-800-839-6374
A Member of Fractionation Research, Inc.
Mesh numerous alloys & plastics
Vane (Chevron) Double Pocket Vane
Candle or Fiberbed Insertion Misthx
MIST ELIMINATORS
DISTRIBUTORS & SUPPORTS
Manufactured to customer
specihcations or engineered to
meet performance requirements.
COALESCERS
Oil water separations Haze removal
from fuels Removal of tower wet
rehux Caustic treater applications
STRUCTURED PACKING
Woven, sheet metal, and knitted
structured packing. Built to spec
or performance requirement.
RANDOM PACKING
Catalyst bed supports Basket strainers
Nozzles Outlet/Inlet baskets
Distributors Hub and header laterals
PROFILE WIRE SCREEN
Wide variety of random packing
types, sizes and materials in
stock! Ask us how our patented
SuperBlend 2-Pac can increase
your capacity and efhciency.
www.ptqenquiry.com
amistco.indd 1 8/3/11 14:02:35
6
10
9
8
7
5
4
3
2
1
1E-
5
1E-
4
1E-
3
1E-
2
1E-
1
1E
0
1E
1
m

,
t
h
g
i
e
h

r
e
b
r
o
s
b
A

H
2
S concentration, mol%
0
Acidied MDEA
sMDEA+
MDEA
Figure 5 Absorber H
2
S profle [0]
instance, 510 vppm for natural gas;
see Table 1 but much looser CO
2
specifcations. However, this
requirement is in opposition to the
chemistry of H
2
S desorption,
described above.
Another effect that hampers H
2
S
desorption is related to the pH
value increasing steadily towards
the stripper sump. As a conse-
quence, the ratio between
protonated HS
-
and H
2
S increases
according to the well-known corre-
lation between the pH value and
pKa values of an acid/base system:
log

f
[HS-]
p= pH - pK
a
H2S
[H
2
S]
Considering these issues, it
becomes obvious how energy inten-
sive the purifcation of gases down
to very low H
2
S ppm levels can be
when amine-based solvents are
used. In this respect, the addition of
an acid to amine solvents has
become a common solution to over-
come the effect of high pH value
and to detain acid/base reactions in
the stripper sump.
In other words, a further decrease
in H
2
S loading can be obtained by
shifting the equilibrium of the
reaction
H
2
S + MDEA HS
_
+ MDEAH
+
towards the H
2
S side, by adding a
strong acid (AH) to the system:
HS
_
+ AH H
2
S + A
-
However, a decrease in the pH
value in the stripper and an increase
in the protonated amine concentra-
tion also affect the absorber pH
value (see Figure 2) and thus the
equilibrium loading. Consequently,
acidifcation is limited to some
extent, so acidifed MDEA solvents
may require an increase in liquid
fow rate to cope with this negative
side effect. This is a crucial point,
especially at elevated amine and
feed gas temperatures above 40C
(104F).
The respective isothermal equilib-
rium curves demonstrate that an
ideal promoter system needs to
have an acidic character in the
stripper bottom, but a neutral
impact in the absorber overhead
(see Figure 3). In addition, no nega-
tive side effects, such as foaming,
corrosion, thermal instability or
even evaporation, should occur.
Such a tailor-made solvent which
at least comes close to the ideal
solvent has been developed and
is described next.
Regeneration test runs and
achievable absorber specifcations
Figure 4 shows the graph of a
regeneration lab test run with
MDEA, acidifed MDEA (H
3
PO
4
)
and MDEA with the new promoter
system (sMDEA+ technology), with
the residual H
2
S lean loading
(Nm
3
/t, mol/mol) against regenera-
tion energy, Q. In the important
regime of very low loadings (<0.5
Nm
3
/t, <6E-3 mol/mol), a magni-
fed graph illustrates the difference.
Figure 5 illustrates the corre-
sponding H
2
S concentration profle
in the gas phase for a 10m packed-
bed absorber with a feed gas partial
pressure for CO
2
of ~2 bar (29 psi)
and H
2
S ~1 bar (14.5 psi) at 40C
(104F). As a matter of equilibrium,
the respective H
2
S specifcation for
MDEA is in the low hundreds
ppmv range, while for acidifed
MDEA the H
2
S specifcation would
be around a tenth of this. In the
sMDEA+ case, the ultra-low lean
loading in the 1E-4 mol/mol range
is able to achieve a treated gas
specifcation below 5 ppmv, with
suffcient safety margin and a negli-
gible impact on the H
2
S absorption
capacity at the absorber bottom.
To reach a similar low lean load-
ing with a state-of-the-art acidifed
MDEA, a doubling of the H
3
PO
4

concentration would be required.
But, as a consequence of this meas-
ure, the absorber equilibrium curve
and thus the entire acid gas capac-
ity would suffer, requiring a
compensation of approximately
512% higher amine liquid load.
Conclusions
Besides highly selective AGE appli-
cations, H
2
S selective acid gas
removal has become an increasingly
important feld for world-scale
natural gas plant designs. Contrary
to acid gas enrichment techniques,
selectivity and thus CO
2
slip are
limited and require some adjusta-
bility and fexibility.
From a solvent technology
perspective, advanced H
2
S selective
acid gas removal can be carried out
through sterically hindered amines
or tertiary amines, both in combina-
tion with other amines or a
promoter system. Additionally,
sophisticated design measures are
applied to ensure a reliable
operation.
Todays plants require a very high
degree of fexibility in turndown
rates, changing feed gas specifca-
tions and conditions. In addition,
very tightly treated gas sulphur
specifcations in the low ppm range
(<5 ppmv) are mandatory.
The high acid gas partial pressures
in most natural gas applications, and
the requirement for adjustable
selectivity, make MDEA-based
basf.indd 7 11/3/11 11:37:45
technologies attractive due to
competitive solvent prices and
generic designs. For opex, capex and
specication purposes, acidic
promoted systems are in use to
overcome the relatively high H
2
S
binding energy and to lower resid-
ual H
2
S loading in the amine.
However, current acidied state-
of-the-art systems also have their
limits, as the improved regenera-
tion will trigger an inverse effect on
the entire capacity. Alternatively,
the new, proprietary promoter
system described in this article
shows different behaviour: a
substantial increase in regeneration
ability while hardly affecting
absorption capacity.
Further reading
1 Bullin J A, Polasek J, Selective absorption
using amines, 61st GPA Conference, Tulsa,
Oklahoma, 1982.
2 Harbison J L, Handwerk G E, Selective removal
of H
2
S utilizing generic MDEA, 37th Annual
Laurance Reid Gas Conditioning Conference,
Norman, Oklahoma, 1987.
3 Carey T R, Hermes J E, Rochelle G T, A
model of acid gas absorption/stripping using
methyldiethanolamine with added acid, Gas
Separation & Purication, Jun 1991, vol 5.
4 Kohl A L., Nielsen, Gas Purication, 5th ed,
Gulf Publishing Corp, 1997.
5 Weiland R H, Dingman J C, Effect of solvent
blend formulation on selectivity in gas treating,
45th Annual Laurance Reid Gas Conditioning
Conference, Norman, Oklahoma, 1995.
6 Huffmaster M A, Stripping requirements for
selective treating with Sulphinol and amine
systems, 47th Annual Laurance Reid Gas
Conditioning Conference , Norman, Oklahoma,
1997.
7 Weiland R H, Sivasubramanian M S, Dingman
J C, Effective amine technology: controlling
selectivity, increasing slip, and reducing sulphur,
53th Annual Laurance Reid Gas Conditioning
Conference, Norman, Oklahoma, 2003.
8 A Renery for Bavaria, ofcial Bayernoil
brochure, July 2009.
Gerald Vorberg is a Senior Technology Manager
in BASFs Gas Treatment team in Ludwigshafen
and Project Leader for Selective Acid Gas
Removal. He joined BASFs Catalyst Group in
1997 as a Global Product Technology Manager
and holds a diploma in chemical engineering
from the University of Applied Sciences (FHT),
Mannheim, Germany.
Email: gerald.vorberg@basf.com
Ralf Notz is a Research Engineer at BASF SE in
Ludwigshafen. He holds a diploma in process
engineering from the University of Stuttgart
and a PhD in CO
2
capture from power plant ue
gas by reactive absorption from the Institute
of Thermodynamics and Thermal Process
Engineering at the University of Stuttgart.
Email: ralf.notz@basf.com
Torsten Katz is head of the Global Technology
Team at BASF SE and coordinates BASFs
new business development activities in gas
treatment. He studied mechanical engineering
at the Technical University of Aachen, Germany
(RWTH Aachen) and holds a PhD in evaporation
technology.
Email: torsten.katz@basf.com
Wieland Wache is a Process Engineer at
Bayernoil Renery in Vohburg, Germany.
He holds a diploma in chemistry from the
Technical University (RWTH) Aachen and a PhD
in chemical engineering on Fischer-Tropsch
synthesis and dehydrogenation of middle
distillates from the University of Bayreuth.
Email: wieland.wache@bayernoil.de
Claus Schunk is Plant Manager at Bayernoil
Renery in Vohburg and was formerly Lead
Process Engineer for the development and
implementation of the OATS units. He holds
a diploma in process engineering from the
Technical University Karlsruhe.
Email: claus.schunk@bayernoil.de
MI NI VAP ON- LI NE
G ra b n e r I n stru m e n ts M e sste ch n i k G m b H / A-1 2 2 0 Vi e n n a /Au stri a / D r. O tto -N e u ra th -G a sse 1
P h o n e +4 3 /1 /2 8 2 1 6 2 7 -0 / F a x +4 3 /1 /2 8 0 7 3 3 4 / i n fo . g ra b n e r-i n s tru m e n t s @ a m e te k . a t
THE RVP PROFIT ANALYZER
V P o f g a so lin e , cru d e o il an d L P G
AS T M D 51 91 , 51 8 8 , 6 3 7 7, 6 3 7 8 , 6 8 9 7, E N 1 3 01 6 -1 +2 ,
I P 3 94 , 4 0 6 , 4 81
H ig h e st accu racy fo r b e st b le n d in g to o fficial lim its
U p to 2 sam p le stre am s
Au to m atic calib ratio n
F a st an d e a sy m ain te n an ce
Au to m atic lu b ricatio n VP o f
N E W!
ON-LINE half page.indd 1 03.03.2011 17:15:05
www.ptqenquiry.com
basf.indd 8 9/3/11 12:37:10
Estimating gas hydrate inhibitor loss:
a case study
T
he amount of methanol to be
injected as a hydrate inhibitor
must not only be suffcient to
prevent freezing of the inhibitor
water phase, but also to provide for
the equilibrium vapour-phase
content of the inhibitor and the
amount that is soluble in the
condensate liquid phase. In this
article, a case study is presented
showing how the information
gained from a predictive tool can
be used to understand and predict
the loss of methanol during natural
gas hydrate inhibition. According
to this study, more than $3700/day
is the cost of lost methanol for 3 x
10
6
m
3
/d of natural gas.
Gas hydrate formation in natural
gas and natural gas liquids (NGL)
systems can block pipelines, equip-
ment and instruments, restricting or
interrupting the fow, which leads
to safety hazards and substantial
economic risks.
1,2

A predictive tool is used to understand and predict the loss of methanol during
natural gas hydrate inhibition
AlirezA BAhAdori
Curtin University
Methanol is the most commonly
used hydrate inhibitor in subsea
petroleum industries, gas treatment
and processing, pipelines and wells,
with worldwide usage worth
several million dollars per year.
1

Due to its high volatility, methanol
is lost in the vapour phase. Often,
when applying methanol as an
inhibitor, there is a signifcant
expense associated with the cost of
lost methanol.
The amount of methanol injected
to treat the water phase, including
the amount of inhibitor lost to the
vapour phase and the amount that
is soluble in the hydrocarbon liquid
phase, equals the total amount of
required methanol.
In addition, one of the primary
factors in the selection process is
related to the possibilities for recov-
ery, regeneration and reinjection of
the spent material. Usually, metha-
nol is not regenerated because of its
intermittent application (mainly
during start-up or shutdown).
However, when it is injected contin-
uously, as is often observed in gas
systems, it is sometimes regener-
ated.
1,2
The losses to the vapour
phase can be prohibitive, in which
case operators select monoethylene
glycol.
Considering all of these issues,
there is a signifcant need for the
development of an accurate and
simple-to-use predictive tool to
represent methanol loss during gas
hydrate inhibition. Predictive tools
to minimise the complex and time-
consuming calculation steps are also
an essential requirement. It
is apparent that mathematically
compact, simple and reasonably
accurate equations, which contain
fewer tuned coeffcients, would be
preferable for computationally inten-
sive simulations. The present study
discusses the formulation
of a novel and simple predictive tool
that could be of signifcant impor-
tance to natural gas engineers.
Methanol vapourisation loss during
gas hydrate inhibition
Equation 1 shows a defnition of
methanol vapourisation loss:
methanol vapour composition to
methanol liquid composition.
Equation 2 is an Arrhenius-type
function to correlate methanol
vapourisation loss as a function of
temperature (K), wherein the
relevant coeffcients (see Table 1)
are correlated as a function of
pressure in kPa(abs) (Equations
36).
kg methanol (1)
L
M
=
Million standard m
3
gas
Mass% methanol in water phase
Coeffcient Values for pressure Values for pressure between
less than 6000 kPa 6000 and 20000 kPa
A
1
4.5112124 x 10
4
5.32442401 x 10
3
B
1
-6.5625532 x 10
8
3.64817009 x 10
7
C
1
2.7486759 x 10
12
-1.43519141 x 10
12
D
1
-3.26326963 x 10
15
5.41358084 x 10
15
A
2
-3.6813711 x10
7
-4.71938915 x 10
6
B
2
5.3777095 x 10
11
-2.60914887 x 10
10
C
2
-2.25654440 x 10
15
1.17661977 x 10
15
D
2
2.68069426 x 10
18
-4.48418384 x 10
18
A
3
1.00121117 x 10
10
1.39340046 x 10
9
B
3
-1.46856806 x 10
14
6.012666228 x 10
12
C
3
6.17387798 x 10
17
-3.21209807 x 10
17
D
3
-7.33914654 x 10
20
1.237524600 x 10
21
A
4
-9.07444577 x 10
11
-1.369737341 x 10
11
B
4
1.3364870 x 10
16
-4.38900897 x 10
14
C
4
-5.6294519 x 10
19
2.919844093 x 10
19
D
4
6.696496249 x 10
22
-1.137898262 x 10
23
Tuned coeffcients for the prediction of methanol vapourisation loss
Table 1
curtin.indd 1 9/3/11 12:49:34
ln(L
M
) = a +
b
+
c
+
d
T T
2
T
3
(2)
Where:
a = A
1
+
B
1

+
C
1
+
D
1
P P
2
P
3
(3)

b = A
2
+
B
2

+
C
2
+
D
2
P P
2
P
3
(4)

c = A
3
+
B
3

+
C
3
+
D
3
P P
2
P
3
(5)

d = A
4
+
B
4

+
C
4
+
D
4
P P
2
P
3
(6)
The novel tools proposed in the
present work are simple formula-
tions. Furthermore, the selected
exponential function to develop the
tool leads to well-behaved (that is,
smooth and non-oscillatory) equa-
tions, enabling fast and more
accurate predictions.
Methanol loss in condensate liquid
phase during gas hydrate inhibition
Equation 7 represents the proposed
governing equation in which four
coeffcients are used (Equations
811) to correlate methanol solubil-
ities in the liquid hydrocarbon
phase () in mole fraction as a
function of temperature (T) for a
given methanol mass fraction in
the aqueous phase, where the rele-
vant coeffcients are shown in
Table 2:

ln() = a
1
+
b
+
c
+
d
T T
2
T
3
(7)
Where:
a = A
1
+ B
1
+ C
1
2
+ D
1
3
(8)
b = A
2
+ B
2
+ C
2
2
+ D
2
3
(9)
c = A
3
+ B
3
+ C
3
2
+ D
3
3
(10)
d = A
4
+ B
4
+ C
4
2
+ D
4
3
(11)
These optimum tuned coeffcients
help to cover temperatures in the
range 240320K and methanol
concentrations up to 0.70 mass
fraction in the aqueous phase.
The optimum tuned coeffcients
shown in Table 2 can be further
retuned quickly according to
the proposed approach
1,2
if
Table 2
Table 3
Coeffcient Value
A
1
1.6812339429 x 10
3
B
1
-1.0803155907 x 10
4
C
1
2.2311693912 x 10
4
D
1
-1.4729301398 x 10
4
A
2
-1.564777661 x 10
6
B
2
9.9874920671 x 10
6
C
2
-2.0506540953 x 10
7
D
2
1.3479263861 x 10
7
A
3
4.8553968051 x 10
8
B
3
-3.0835714347 x 10
9
C
3
6.2988799656 x 10
9
D
3
-4.1248186297 x 10
9
A
4
-5.0424809928 x 10
10
B
4
3.1810877336 x 10
11
C
4
-6.4667402344 x 10
11
D
4
4.220547048 x 10
11
Tuned coeffcients for the prediction of
methanol loss in the condensate phase
Temp, Pressure, Reported methanol Calculated methanol Percentage
K kPa vapourisation loss
3
vapourisation loss absolute
kg methanol kg methanol deviation
Million standard m
3
gas L
M
= Million standard m
3
gas
Mass% methanol in water phase Mass% methanol in water phase
261.15 2000 11 10.916 0.76
277.15 2000 32.7 32.217 1.47
269.15 3000 15.9 15.229 4.22
281.15 3000 31.2 31.605 1.29
271.15 4000 14.6 14.639 0.26
283.15 4000 29.85 30.232 1.27
273.15 5000 14.6 14.402 1.35
285.15 5000 29.3 28.592 2.41
275.15 6000 14.8 14.609 1.29
287.15 6000 29.2 29.218 0.06
277.15 7000 27.8 27.605 0.70
279.15 8000 16.8 16.973 1.03
283.15 9000 20.9 20.755 0.69
285.15 10 000 22.2 22.236 0.16
279.15 20 000 13.7 13.699 0.007
289.15 20 000 23 22.95 0.21
Average absolute deviation percent (AADP) 1.07%
Comparison with reported data
3
of predicted results of methanol vapourisation loss
www.ptqenquiry.com
Our new clean-fuels
plant is straining the
auxiliary units!
How can we boost their capacity
without major construction?
Read more on this topic at
www.amistco.com
TO MEET CLEAN FUELS require-
ments for gasoline, a medium-sized
Midwestern refnery added a low-
sulfur fuels technology plant. Gasoline
throughput was unchanged. However,
the sharp increase in sulfur removal
required more hydrogen from the
hydrogen unit and sent more sulfur
gases to the amine treaters and
downstream sulfur units. These
auxiliary units became bottlenecks,
overdriven at the cost of product
purity and amine consumption.
On studying the hydrogen, amine,
and sulfur units, Amistco found many
opportunities for improving separation
effciency and capacity. The problems
were solved without major construction
by applying modern technology to mist
eliminators, liquid-liquid separators,
and tower trays and packing. Results
included haze-free product and
reduced amine consumption. Now a
diesel clean-fuels plant is being added.
Prore: 281-331-595 Fax: 281-585-1Z80
amistco@amistco.com
21-rr Erergercy 3erv|ce: 800-839-3Z1
curtin.indd 2 9/3/11 12:49:43
more data become available in the
future.
It is expected that efforts during
this investigation will pave the way
to an accurate prediction of metha-
nol loss in the liquid hydrocarbon
phase () in mole fraction, as well
as methanol vapourisation loss in
various conditions, which can be
used by engineers and scientists for
monitoring the key parameters
periodically.
Results
Table 3 presents the summary of
accuracies with the proposed
predictive tools in terms of average
absolute deviation percentage with
reliable data.
3
It shows that the
proposed correlation has an
average absolute deviation percent-
age of less than 1.1%, which can be
considered to be a very small devia-
tion from reliable data. Figures 1
and 2 show the proposed methods
results and performance in the
prediction of methanol solubility in
the hydrocarbon liquid phase for a
wide range of conditions from two
different viewpoints. They show
that high temperatures and more
injected methanol in the aqueous
phase lead to higher solubility of
methanol in the liquid phase. Table
4 illustrates the accuracy of the
proposed correlation in predicting
the solubility of methanol in
paraffnic hydrocarbons in compari-
son with some reported data.
3
The
accuracy of correlation in terms of
average absolute deviations
is 2.12%.
The following is a case study
showing how the information
evolving out of a predictive tool
can be used to understand and
predict the loss of methanol during
natural gas hydrate inhibition.
According to these calculations,
$3701/day is the cost of methanol
loss, which is calculated by
this simple predictive tool for
3 x 10
6
m
3
/d of natural gas leaving
an offshore platform.
Case study
As indicated, 3 x 10
6
m
3
/d of natu-
ral gas leaves an offshore platform
at 38C and 8300 kPa (abs), and the
water content is 850 mg/Sm
3
. The
gas comes onshore at 4C and 6200
0 . 0 1 5
0 . 0 2 5
0 . 0 2 0
0 . 0 1 0
0 . 0 0 5
2 2 0 2 4 0 2 5 0 2 6 0 2 7 0 2 8 0 2 9 0 3 0 0 3 1 0 3 2 0

D
i
s
s
o
l
v
e
d

m
e
t
h
a
n
o
l

i
n

l
i
q
u
i
d

h
y
d
r
o
c
a
r
b
o
n

p
h
a
s
e
,

m
o
l
e

f
r
a
c
t
i
o
n
Temperat ure, K
0 . 0 0 0
M e th a n o l m a ss fra c tio n = 0 . 6 0
M e th a n o l m a ss fra c tio n = 0 . 0 2
1 0
6
1 0
2
1 0
4
1 0
8
1 0
1 0
1 0
2
1 0
1

S
o
l
u
b
i
l
i
t
y

o
f

m
e
t
h
a
n
o
l

i
n

c
o
n
d
e
n
s
a
t
e

h
y
d
r
o
c
a
r
b
o
n

l
i
q
u
i
d
,

m
o
l
e

f
r
a
c
t
i
o
n
Met hanol mass f ract i on i n wat er phase
1 0
1 2
Te m p e ra tu re = +5 0 C
Te m p e ra tu re = 1 5 C
Figure 1 Performance of proposed tool for prediction of solubility of methanol in the
hydrocarbon condensate phase
Figure 2 Performance of proposed tool for prediction of solubility of methanol in
hydrocarbon condensate phase from another view point
Temp, Methanol mass Calculated solubility Reported solubility Absolute
C fraction of methanol in of methanol in deviation,
hydrocarbons, hydrocarbons, %
mole fraction mole fraction
3

12 0.20 0.00099 0.001 1
30 0.20 0.00294 0.003 2
50 0.20 0.006 0.006 0
21 0.35 0.0044 0.004 1
47 0.35 0.01035 0.01 3.5
-10 0.60 0.00295 0.003 1.6
2 0.60 0.00514 0.005 2.8
43 0.60 0.02034 0.02 1.7
-40 0.70 0.0021 0.002 5
0 0.70 0.0073 0.0071 2.8
50 0.70 0.0265 0.027 1.9
Average absolute deviation, % 2.1
Comparison with reported data
3
of calculated solubility of methanol in hydrocarbons
Table 4
curtin.indd 3 9/3/11 12:49:54
Losses = 14.58 x 10
-6
(kg/m
3
)/(wt%
methanol)
Daily vapourisation losses = 14.58 x
10
-6
(3 x 10
6
)(27.5) = 1203 kg/day.
1203 kg/day ($3/kg) = $3609/day
is the cost of vapourised methanol
calculated by this predictive tool.
Solution for methanol loss in
condensate phase
Win = 850 mg/Sm
3
Wout = 152 mg/Sm
3

W = 698 mg/Sm
3
Water condensed = (3 x 10
6
) (698) =
2094 x 10
6
mg/day = 2094 kg/day
Calculate mass rate of inhibitor in
the water phase:

Estimate losses to hydrocarbon
liquid phase from proposed method
at 4C and 27.5 wt% MeOH:
a = 9.136448034367 x 10
1

b = -8.869793699719 x 10
4

c = 2.8126995933983 x 10
7

d = -3.0722013379 x 10
9

The solubility of methanol in the
hydrocarbon phase is estimated to
be around 0.0011 or 0.11 mol%:
f
3 x 10
6
Sm
3
pf
56m
3
pf
778kg
p f
1kgmole
p
= 933
day 10
6
Sm
3
m
3
140 kg kg
mol/day.
Kg mol methanol = 933(0.0011) =
1.026 moles/day.
Kg methanol = (1.026) (32) = 32.84
kg/day.
The methanol in the condensate
phase can be recovered by down-
stream water washing.
Cost per month = ($3/kg) (30.72
kg/day)(30 day/month) = $2956
per month.
The total cost of methanol loss per
day will become 92 + 3609 =
$3701/day.
References
1 Bahadori A, Vuthaluru H B, Energy & Fuels,
24, 2010, 29993002.
2 Bahadori A, Vuthaluru H B, Journal of Loss
Prevention in the Process Industries, 23, 3, 2010,
379384.
3 GPSA Engineering Data Book, 12th ed, 2004,
Gas Processors Suppliers Assoc, Tulsa, USA.
Alireza Bahadori works in the Chemical
Engineering Department of Curtin University
of Technology, Perth, Western Australia.
Email:Alireza.bahadori@postgrad.curtin.edu.au
kPa (abs), and water content is 152
mg/Sm
3.
The hydrate temperature
of the gas is 18C. Associated
condensate production is 56m
3
/
(million standard m
3
). The conden-
sate has a density of 778 kg/m and
a molecular mass of 140. The
required methanol inhibitor concen-
tration in the water phase to avoid
hydrate formation is 27.5%.
Calculate the mass rate of inhibitor
loss in the water phase and the
amount of methanol loss in the
hydrocarbon liquid phase, as well
as the amount of methanol vapouri-
sation loss. Consider the price of
methanol as $3 per kg.
Solution for methanol
vapourisation loss
X = 0.275 mass fraction of methanol
T = 277.15K
For vapourisation loss of methanol,
we use the tuned coeffcients
reported in Table 1:
a = -1.6454503534842 x 10
3
b =1.3743444476865 x 10
6
c = -3.8051922967936 x 10
8
d = 3.4980770983675 x 10
10
L
M
= 14.58
m
i
=
0.275(2094)
= 794 kg/day
1 0.275
slip (minimum removal) to meet
North American treating specica-
tions. Due to the small amount of
acid gas removal, the absorbers in
these plants were operated at small
liquid rates and very low weir
loads. Simulations fell far short of
agreeing with the eld data.
Eventually, it was realised that the
absorbers were well into the spray
regime and ProTreats mass transfer
coefcient correlations were inap-
plicable because they were
developed from data on trays oper-
ating in the froth regime. To
reproduce the performance data,
the vapour-side mass transfer coef-
cients had to be increased by
upwards of a factor of 10 or more,
and the liquid-side coefcients had
to be decreased correspondingly.
Since then, six more commercial
tower performance data sets with
absorbers at low weir load have
come to light. The only non-propri-
etary set of measurements is from a
plant operated by Signalta
Resources near Forestburg, Alberta,
Canada.
4
These eight data sets are
ppp'^imj'\hf IMJJ-+)*)2.
from plants in Canada, the US and
India and span the full spectrum of
weir loads from about 2.2 m
3
/mhr
(3 gpm/ft) to 45 m
3
/mhr (60 gpm/
ft). Each plant was simulated rst
using ProTreats database correla-
tions for tray mass transfer
coefcients (froth regime), using the
owsheet shown in Figure 1.
Simulations were repeated for vari-
ous values of a constant factor,
multiplying the gas-side mass trans-
fer coefcient (the same factor
divided the liquid coefcient) until
agreement with the performance
data was obtained.
Figure 2 shows how the results
compare with each other and how
the correction alters with changing
weir load. The plot shows a truly
remarkable degree of correlation,
especially in view of the fact that
S
p
r
a
y

R
e
g
i
m
e

C
o
r
r
e
c
t
i
o
n

F
a
c
t
o
r
We ir L o a d
N e w M e x ic o
I o wa
A lb e rta
I n d ia
N o rth Te x a s
B a se
N o rth Te x a s
N o rth Te x a s
N o rth Te x a s
?b`nk^+Ahpma^\hkk^\mbhg_Z\mhk_hklikZrk^`bf^hi^kZmbhgoZkb^lpbmap^bkebjnb]ehZ]
ppp'imj^gjnbkr'\hf
_hk_nkma^kbg_hkfZmbhg
3225 Gallows Road
Fairfax, Virginia 22037-001, USA
www.exxonmobil.com/refningtechnologies
+1-703-846-2568 fax +1-703-846-3872
technology.licensing@exxonmobil.com
So Advanced,
Yet So Si mple.
Commerci al l y Proven Rout e f or The
Product i on of Cl ean Gasol i ne f rom Coal
t hrough Coal Gasi f cat i on, Met hanol
Synt hesi s, and Met hanol Conversi on
ExxonMobil Research and Engineering Companys
Methanol- to- Gasoline (MTG) Technology, the key step
for coal to gasoline conversion, was commercially
operated in New Zealand for over ten years. Now we
offer the second generation MTG technology
which provides:
t 6MUSBDMFBOIJHIPDUBOFHBTPMJOF
t "QSPEVDUGVMMZDPNQBUJCMFXJUISFOFSZHBTPMJOF
t 4JNQMFSFMJBCMFYFECFESFBDUPSEFTJHO
t "NPSFFDPOPNJDBMMZDPNQFUJUJWFPGGFSJOH
ogt.indd 3 14/9/10 09:40:04
www.ptqenquiry.com
curtin.indd 4 9/3/11 12:50:05
Amistco Separation Products 45, 49
www.amistco.com
Axens 6
www.axens.net
Black & Veatch 18
www.bv.com
Borsig 40
www.borsig.de
Bryan Research & Engineering 15
www.bre.com
Burckhardt Compression 23, 29
www.burckhardtcompression.com
Cansolv Technologies 24
www.cansolv.com

ExxonMobil Research and Engineering Company 51
www.exxonmobil.com/refningtechnologies
Grabner Instruments 47
www.grabner-instruments.com
Haldor Topse IFC
www.topsoe.com

Linde 34
www.linde.com
Lurgi 4
www.lurgi.com
Merichem Company 23
www.merichem.com
MPR Services 17, OBC
www.mprserve.com
NPRA IBC
www.npra.org
Prosernat 42
www.prosernat.com
Siirtec Nigi 21
www.sini.it
WorleyParsons 20
www.worleyparsons.com
Alphabetical list of advertisers
For more information on these advertisers, go to www.ptqenquiry.com
ad index copy 7.indd 1 9/3/11 15:00:06
NPRA 2011 Conference Schedule
Join colleagues and customers at
NPRAs industry-leading conferences.
.!4) /.!,0%42/#(%-) #!,2%&) .%23!33/#) !4) /.
M eet the Right People
fo r B u sin e ss Su cce ss
Register today at npra.org/meetings or call us at (202) 457-0480 for more information.
Security Conference and
Exhibition
March 12
InterContinental Hotel
Houston, TX
National Safety
Conference and Exhibition
May 1011
Omni Fort Worth
Fort Worth, TX
Labor Relations/Human
Resources Conference
May 2526
Convention Center
Denver, CO
Reliability & Maintenance
Conference and Exhibition
May 2427
Convention Center
Denver, CO
Q&A and Technology
Forum
October 912
JW Marriott
JW Marriott San Antonio Hill
Country Resort & Spa
San Antonio, TX
Environmental
Conference
October 2425
New Orleans Marriott
New Orleans, LA
International Lubricants
and Waxes Meeting
November 1011
Hilton Post Oak
Houston, TX
Annual Meeting
March 2022 Marriot Rivercenter San Antonio, TX
International Petrochemical Conference
March 2729 Grand Hyatt San Antonio, TX
Public Policy Conference
May 56
Washington Court Hotel
Washington, DC
NPRA_Con_Cal_2011_PTQ.indd 1 2/2/11 12:40:22 PM
npra.indd 1 4/2/11 12:21:49
www.mprservices.com
MPR Services, Inc. Introduces
New and Improved, Compact Technology
with Limiting Use of Resources and Energy
www.ptqenquiry.com
mpr.indd 1 8/3/11 14:07:58

PTQ Gas 2011

Diunggah oleh

Informasi Dokumen

Judul Asli

Hak Cipta

Format Tersedia

Bagikan dokumen Ini

Bagikan atau Tanam Dokumen

Opsi Berbagi

Apakah menurut Anda dokumen ini bermanfaat?

Apakah konten ini tidak pantas?

Hak Cipta:

Format Tersedia

PTQ Gas 2011

Diunggah oleh

Hak Cipta:

Format Tersedia

gas

(DGA), diisopropanolamine (DIPA),

is another primary amine that removes

has been proven

, methanol, and NMP. For more information about

Partial molar volume of i in

Anda mungkin juga menyukai