gas

2011

PTQ supplement

gas cov copy 3.indd 1 9/3/11 13:04:31

 Creating excellence from raw materials
you need a specialist
W W W. T O P S O E . C O M

Add value to your natural gas

with Topse Technologies
Haldor Topse offers you the technology, catalyst and experience for profitable use of your natural gas feed
whether you are seeking a new plant or a revamp of an existing plant.

- hydrogen - ammonia
- synthesis gas to chemicals - methanol
- synthesis gas to fuels - DME

Through research-driven technologies, Haldor Topse offers impartial counselling on the selection of the best technology
and catalyst for your plant.

www.ptqenquiry.com
for further information

haldor.indd 1 8/3/11 13:27:06

 ptqPETROLEUM TECHNOLOGY QUARTERLY

2011
gas
www.eptq.com

5 Amines in bad odour

Chris Cunningham

7 Enhanced sulphur recovery from lean acid gases

Angela Slavens and Justin Lamar Black & Veatch
Djordje L Nikolic and Theo Brok Shell Global Solutions International

19 Cryogenic monitoring
Jerry Worsley Sensornet

25 Glycol dehydration of high-acid gas streams

John Carroll Gas Liquids Engineering
Nathan Hatcher and Ralph Weiland Optimized Gas Treating

35 Controlling corrosion in amine treatment units

Jaya Rawat, Peddy V C Rao and N V Choudary
Bharat Petroleum Corporation Ltd

41 A promoter for selective H2S removal: part I

Gerald Vorberg, Ralf Notz and Torsten Katz BASF SE
Wieland Wache and Claus Schunk Bayernoil Raffineriegesellschaft

48 Estimating gas hydrate inhibitor loss: a case study
Alireza Bahadori Curtin University

The Pearl gas-to-liquids project, Ras Laffan, Qatar. Photo: Shell

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.

contents/ed com copy 8.indt 1 9/3/11 14:40:37

 MER-0308 LO-CAT Chemist PTQ Spread.pdf 1 1/7/11 3:55 PM

MERICHEM COMPANY

CM

MY

CY

CMY

merichem.indd 2 8/3/11 13:29:37

 Sweet Solutions.

Problem: Sour Gas

Sweetest Solution: LO-CAT
H2S Removal/Recovery
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.
that are meeting their treating needs
by turning H2S into elemental sulfur
efficiently and effectively with LO-CAT.
Learn how sweet it is, visit
www.merichem.com/LO-CAT

Our extensive experience includes Merichem: A global provider of focused

treatment of gasification syngas, natural hydrocarbon-treating technology,
gas, amine acid gas, sour water stripper chemical and service solutions.
gas, fuel gas, and Claus tail gas.

LO-CATs flexible technologies and

worldwide services are backed by over
three decades of reliability. Join 200+
licensees in 29 countries on 6 continents

P: 847.285.3850 | E: mgtsales@merichem.com | www.merichem.com

www.ptqenquiry.com
for further information

merichem.indd 3 8/3/11 13:29:51

 Simplify
sulfur recovery
and cut your costs
Low-temperature tail-gas hydrogenation catalysts
that deliver superior and cleaner performance from simplified operations, and significantly
lower CO2 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.

Single source ISO 9001 technology and service provider

www.ptqenquiry.com
www.axens.net
for further information
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

axens.indd 1 10/12/09 12:59:39

 T tq
p 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
PETROLin
tonnes M TECHBiodiesel
EU2007. NOLOGY QU ARTERLY
production in the
US has also increased dramatically in the
Amines in bad
odour
past few years from 2 million gallons in 2000
to approximately 450 million gallons in 2007.
According to the National Biodiesel Board,

A
171 companies own biodiesel manufacturing
Editor mine capture of acid gases and amine plants themselves are not
plants and are actively marketing biodiesel.1.
Chris Cunningham unexpected topics for discussion by industry experts in this issue of PTQ
editor@petroleumtechnology.com
The global biodiesel market is estimated Gas. However, recent news has focused not so much on amine technology
to reach 37 billion gallons by 2016, with an for dealing with acid gas components of natural gas as on its application to the
Production Editor
average annual growth rate of 42%. Europe downstream capture of gas combustion products, meaning, of course, CO2. With
Rachel Zamorski
will continue to be the major biodiesel
production@petroleumtechnology.com carbon issues looming over the hydrocarbons industry in all of its forms, the story
market for the next decade, followed closely of the Mongstad gas power plant in Norway, which includes a flagship project to
by the USEditor
Graphics market. demonstrate large-scale capture and storage (CCS) of CO2, is causing no end of
Although
Rob Fris high energy prices, confusion for climate change policy makers in Europe.
graphics@petroleumtechnology.com
increasing global demand, drought DONG Energy and Statoil have been working at Mongstad to build a ther-
and other factors are the primary driv- mal power plant, with the aim of integrating it into Statoils neighbouring oil
Editorial
ers for higher food prices, food refinery, while DONG operates and owns the plant. It has been built to supply
tel +44 844 5888 773
competitive feedstocks
fax +44 844 5888 667 have long been Mongstad and a production platform in the North Sea with electricity and
and will continue to be a major heat, and to supply a gas treatment plant with electricity. So far so good.
concern forSales
Advertising theManager
development of biofu- One of the chief ideas behind Mongstad is to strengthen the refinerys posi-
els. To compete, the industry has
Paul Mason tion as an efficient supplier of refined products. At maximum utilisation, the
sales@petroleumtechnology.com
responded by developing methods to power plant can deliver 280 MW of electricity and about 350 MW of heat. Now
increase process efficiency, utilise or it is up and running, it is no doubt meeting all of those targets. All of the CO2
Advertising Sales
upgrade by-products and operate with it currently produces is released to the atmosphere, which is fine at present
Bob Aldridge
lower quality lipids as feedstocks.
sales@petroleumtechnology.com 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
Feedstocks
Advertising Sales Office first delayed to 2018 and most recently, on Statoils advice, the Norwegian
Biodiesel
tel +44 844refers to a diesel-equivalent
5888 771 government has delayed it further. According to Norways Petroleum and
fax +44 844 5888 662
fuel consisting of short-chain alkyl Energy Ministry, that advice centres not on the technical issues of making a
(methyl or ethyl) esters, made by the CCS project work but on whether amine technology can be used at all, on
Publisher
transesterification of triglycerides, grounds of health and safety. There is a need for greater knowledge of the
Nic Allen
commonly known as vegetable oils or
publisher@petroleumtechnology.com effects of using amine technology on health and the environment, says the
animal fats. The most common form ministry. Now the government wants to look at alternative options for imple-
uses methanol, the cheapest alcohol
Circulation menting CCS.
available,
Jacki Watts to produce methyl esters. While this has, not unexpectedly, put Statoil in bad odour with local environ-
circulation@petroleumtechnology.com
The molecules in biodiesel are primar- mentalists, who were just about prepared to put up with another major
ily fatty acid methyl esters (FAME), gas-fired power plant provided the CO2 it produces could be effectively
Crambeth Allen Publishing Ltd
usually created by transesterification contained, the ramifications travel much further. The EU is preparing to imple-
Hopesay, Craven Arms SY7 8HD, UK
between
tel +44 844 fats
5888 and776methanol. Currently, ment a large-scale programme to demonstrate CCS. Following the European
biodiesel
fax +44 844is5888 produced
667 from various tradition for these matters, individual countries will demonstrate their own
vegetable and plant oils. First-genera- flagship projects. The most successful are likely to become standards for future
tion
ISSN food-based
1362-363X feedstocks are straight implementation among the 27 states of the EU.
vegetable oils such as soybean oil and That programme is scheduled to begin in 2015, which puts Mongstad out the
animal fats such as tallow, lard, yellow
Petroleum Technology Quarterly (USPS 0014-781)
running. To make matters worse for the Norwegian government, it was being
grease,
is publishedchicken fatannual
quarterly plus andCatalysis
the by-products
edition relied on to supply a key demonstration of CCS in a major gas power plant,
by Crambeth Allen Publishing Ltd and is distributed
of the
in the USA byproduction
SPP, 75 Aberdeenof Omega-3
Rd, Emigsville, PA fatty which leaves a significant gap in the programme once it begins. A project
acids from fish
17318. Periodicals postageoil. Soybean
paid at Emigsville PA.oil and touted by the national government as a Moon landing when it was first
Postmaster: send address changes to Petroleum
rapeseeds oil are
Technology Quarterly c/o POthe common source announced looks to be firmly Earth-bound.
for biodiesel
Box 437, Emigsville,production
PA 17318-0437 in the US and Why has there been further delay to the Mongstad project? Statoil is increas-
Back numbers available from the Publisher
Europe in inc
at $30 per copy quantities
postage. that can produce ingly worried, apparently, about the production of nitrosamines as a
enough biodiesel to be used in a component of the process of amine capture of CO2 and the effects of these
commercial market with currently substances on human health, which many view with more than a pinch of
applicable technologies. scepticism.
First-generation feedstocks for
chris cunningham

GAS 2011 5

contents/ed com copy 8.indt 2 9/3/11 13:07:16

 PROCESS INSIGHT
Selecting the Best Solvent for Gas Treating
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 CO2, H2S 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 CO2 and H2S from sour gas
and is effective at low pressure. Depending on the conditions, MEA
can remove H2S to less than 4 ppmv while removing CO2 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
CO2, H2S, 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 CO2 and H2S 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 H2S 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.
Bryan Research & Engineering, Inc.
P.O. Box 4747 Bryan, Texas USA 77805
979-776-5220 www.bre.com sales@bre.com
www.ptqenquiry.com
for further information
bre.indd 1
Tertiary Amines
A tertiary amine such as MDEA is often used to selectively remove
H2S, especially for cases with a high CO2 to H2S ratio in the sour gas.
One benet of selective absorption of H2S is a Claus feed rich in H2S.
MDEA can remove H2S to 4 ppm while maintaining 2% or less CO2 in
the treated gas using relatively less energy for regeneration than that
for DEA. Higher weight percent amine and less CO2 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 CO2 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 H2S.
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.
8/3/11 13:36:38
 Enhanced sulphur recovery from
lean acid gases
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
T
he exploitation of increasingly
difficult natural gas reserves
has risen in recent years,
requiring the removal of a number
of harder to remove sulphur
species, as well as H2S, CO2 and
H2O. In addition, maximum limits
for sulphur content in pipeline gas
continue to tighten. As a result, the
treating requirements for gas field
development projects have signifi-
cantly increased in complexity,
often requiring a combination of
process steps and units. As the
development of difficult sour gas
fields is expected to further increase
in the future, strategic integration
of various gas-treating process units
is necessary to achieve an overall
optimised flow sheet with lowest
lifecycle cost. The ability to combine
technologies and process units in
the optimum configuration results
in a competitive advantage for
addressing challenges and opportu-
nities posed by undeveloped sour
gas fields.
In 2010, a study by Black &
Veatch compared several alterna-
tive Claus sulphur recovery unit
(SRU)/tail gas-treating unit (TGTU)
configurations 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 configuration test cases
had an H2S 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
www.eptq.com
b&v.indd 1
content and affirmed that, at such a
low H2S concentration, reliable
Claus unit operation can be
difficult.
The 2010 study considered the
use of the acid gas enrichment
(AGE) process to increase acid gas
H2S 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
It is important to
consider the overall
sour gas treatment
flow scheme
rather than simply
evaluating the acid gas
processing in isolation
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 configuration 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 flow 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 flow 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
flowsheet.
This article compares three differ-
ent Sulfinol AGRU options for
development of the overall flow
scheme for treatment of a sour
natural gas stream containing
significant 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 H2S and CO2 in the
solvent regeneration step of the
AGRU. All Sulfinol solvents remove
COS to a large extent in the main
absorber of the AGRU. With
Sulfinol-X, deep removal of COS
can be achieved due to the presence
of piperazine, which enhances COS
hydrolysis. With Sulfinol-M, most
of the COS would be absorbed due
to the larger number of trays
GAS 2011 7
9/3/11 13:01:07
 utilising some of the acid gas treat-
Sour natural gas feed stream used for comparison of test cases
ment schemes presented in the 2010
article. The optimum configura-
Component Sour natural gas, lbmoles/hr Sour natural gas, mole fraction tions, in terms of relative Capex
H2O 10.63 0.126
CO2 784.69 9.300 and Opex, are presented.
H2S 270.00 3.200
N2 421.88 5.000 Test cases
C1 6618.37 78.440 The sour natural gas stream used
C2 210.94 2.500
C3 101.26 1.200 for comparison of the various proc-
iC4 4.21 0.050 ess configuration test cases is shown
C4 4.21 0.050 in Table 1. Sulphur content is 96.1
iC5 1.26 0.015 t/d. About 2% of the acid gas
nC5 1.26 0.015
C6 3.37 0.040 sulphur is present as mercaptans;
CH4S 4.06 0.048 therefore, recovery of mercaptan
C2H6S 1.34 0.016 sulphur is important if high sulphur
Total 8437.47 100.000 recovery is to be achieved.
Temperature, F 95
Pressure, psia 870 The treated gas specifications
used for the comparison test cases
are shown in Table 2. The specifica-
Table 1 tions considered are typical for
pipeline-quality natural gas. Table 3
and Table 4 describe the three
Treated gas specifications
process configuration test cases
compared in this article, each of
Gas stream Specification Value which is illustrated in Figures 13.
Sales gas H2S, ppmv <5
Total sulphur, ppmv <20
The Case 1 process configuration
CO2, mole% <1.0 is shown in Figure 1. Sour natural
Fuel gas H2S, ppmv <100 gas is processed in a Sulfinol-X
Tail gas ex-TGTU absorber H2S, ppmv <250 AGRU, producing a lean acid gas
stream containing H2S and
Table 2 mercaptans. Sulfinol-X technology
employs piperazine accelerator in
Test cases for comparison addition to MDEA, sulpholane and
water. This solvent formulation
meets both low H2S and CO2 speci-
Test case Description fications, as well as low
1 Sulfinol-X main gas treating, conventional SRU/SCOT1 using MDEA
2 Sulfinol-X main gas treating, SRU/SCOT using Sulfinol-M acid gas enrichment specifications of trace sulphur
and tail gas treating, routing of enrichment absorber overhead to SCOT species (such as mercaptans, COS
3 Sulfinol-M main gas treating with integrated enrichment step and tail gas and organic sulphides). The
absorber, conventional SRU/SCOT using Sulfinol-M, routing of enrichment Sulfinol-X unit has a standard
absorber overhead to SCOT
configuration consisting of a main
absorber, hydrocarbon flash vessel
Table 3 with associated absorber and a
regeneration section. The hydrocar-
Test case process configurations
bon flash minimises co-absorbed
hydrocarbons in the acid gas. The
lean acid gas stream from the
Test case AGRU solvent AGE solvent TGTU solvent AGE overhead routing AGRU regenerator is processed in a
1 Sulfinol-X MDEA
2 Sulfinol-X Sulfinol-M Sulfinol-M To TGTU, per Case 4 from 2010 article
conventional two-bed Claus SRU.
3 Sulfinol-M Sulfinol-M Sulfinol-M To TGTU, per Case 4 from 2010 article SRU tail gas flows to an MDEA-
based Shell Claus Offgas Treatment
(SCOT)1 TGTU, where residual
Table 4 sulphur species in the SRU tail gas
are converted to H2S, which is
employed. Since Sulfinol is a hybrid treated flash gas would not be a subsequently removed in the down-
chemical/physical solvent, for the differentiator between the three stream TGTU absorber. The TGTU
reasons described above, COS cases studied in this article; hence, regenerator regenerates the rich
captured in the absorber will not be COS has been omitted from the MDEA solvent, and the overhead
present in the acid gas. Therefore, it analysis. from this tower is routed to the
is anticipated that the small varia- The resulting acid gas from each front of the SRU for the recovery of
tions of COS in the sales gas and AGRU configuration is processed sulphur in this stream.

8 GAS 2011 www.eptq.com

b&v.indd 2 9/3/11 14:41:41

 Treated
natural
gas

Regenerator
Main
absorber
To fuel
Sour
natural
gas

Flash
drum

Claus converters

Incinerator

Thermal
Air reactor
Fuel

Air
Sulphur Sulphur Sulphur

Hydrogenation
reactor
Quench TGTU Regenerator
Fuel column absorber
Inline
Air
burner/mixer

WHB

Figure 1 Case 1 process configuration: Sulfinol-X gas treating with conventional SRU/SCOT

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% H2S 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 difficult due to
a low thermal reactor temperature
and acid gas flame instability. Cases
2 and 3 utilise enrichment to
increase acid gas strength for
improved Claus operation.
The Case 2 process configuration
is shown in Figure 2. Sour natural
gas is processed in a Sulfinol-X acid
gas removal unit similar to that in
Case 1. Acid gas from the Sulfinol-
X unit is subsequently processed in
an enrichment absorber, which
absorbs H2S from the acid gas using
Sulfinol-M solvent. Sulfinol-M tech-
nology employs MDEA, sulpholane
and water, without piperazine, as is
the case with Sulfinol-X. This results
in AGE and TGTU absorbers, which
remove H2S and mercaptans while
minimising CO2 absorption.
Reduced CO2 solubility is desirable
in the AGE and TGTU absorbers to
prevent excessive CO2 flow through
the SRU and TGTU, which increases
their size. AGE absorber overhead
gas, containing most of the CO2 in
the acid gas as well as other compo-
nents not absorbed by the
Sulfinol-M solvent, flows 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

www.eptq.com GAS 2011 9

b&v.indd 3 9/3/11 16:22:19

 Treated
natural
gas

Regenerator
Main
absorber
To fuel
Sour
natural
gas

Flash
drum

Claus converters

Incinerator

Thermal
Air reactor
Fuel

Air
Sulphur Sulphur Sulphur

Hydrogenation
reactor

Fuel
Quench TGTU AGE
Inline
column absorber absorber
Air burner/mixer

Hydrogenation
reactor

Regenerator
Fuel
Inline
Air burner/mixer

WHB

Figure 2 Case 2 process configuration: Sulfinol-X gas treating with SRU using enrichment, enrichment absorber overhead routed to SCOT

mercaptans in this stream to H2S,
for absorption in the downstream
TGTU absorber, which also employs
Sulfinol-M solvent. Concentrated
acid gas flows from the Sulfinol-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
efficiency when acid gas for enrich-
ment contains sulphur species other
than H2S. With conventional acid
gas enrichment, these other sulphur
species are not absorbed using
selective treating solvents and
therefore flow unrecovered to the
incinerator. Routing the enrichment
absorber overhead to the TGTU
allows conversion of these other
sulphur species to H2S, 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

10 GAS 2011 www.eptq.com

b&v.indd 4 9/3/11 16:22:29

 Treated
natural
gas

Main Enrichment
Regenerator
absorber absorber
To fuel
Sour
natural
gas

Flash
drum
Hot flash drum

Claus converters

Incinerator

Thermal
Air reactor
Fuel

Air
Sulphur Sulphur Sulphur

Hydrogenation
reactor

Fuel
Inline
Air burner/mixer

Hydrogenation
reactor
Quench TGTU
Fuel column absorber
Inline
Air burner/mixer

WHB

Figure 3 Case 3 PFD: Sulfinol-M gas treating with integrated enrichment step and tail gas treating, enrichment absorber overhead
routed to SRU/SCOT

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
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 signifi-
cant losses in overall sulphur
recovery efficiency.
It should be noted that the
Sulfinol-M solvent circulation
system is integrated between the
AGE and TGTU absorbers similar
to the Case 4C configuration in the
2010 study; 100% of the semi-rich
solvent from the TGTU absorber is
cascaded to the AGE absorber. Case

www.eptq.com GAS 2011 11

b&v.indd 5 9/3/11 16:22:38

 )1
BTI

HBT

AGRU
4PVS
(Sulfinol-X) 5SFBUFE

OBUVSBM OBUVSBM
HBT HBT 3FDZDMF
BDJE
HBT
 3JDI -FBO   4UBDL
  HBT
436 5(56 
"DJE
HBT UBJM
HBT UBJM
HBT
AGRU   
436 TGTU Incinerator
regenerator
(MDEA)
(Sulfinol-X)

4VMQIVS 4VMQIVS
QSPEVDU QJU
WFOU
 

Figure 4 Sulphur balance for Case 1: Sulfinol-X gas treating with conventional SRU/SCOT (sulphur flows in lbmoles/hr)

)1
BTI

HBT

"(36
4PVS
4VMOPM9
5SFBUFE

OBUVSBM OBUVSBM
HBT HBT
 3JDI -FBO 
 

"(36
SFHFOFSBUPS
"(&
BCTPSCFS
PWFSIFBE
4VMOPM9


"DJE
HBT 4UBDL
 HBT
436 5(56 
UBJM
HBT UBJM
HBT
 
"(& 436 5(56 *ODJOFSBUPS
4VMOPM.
4VMOPM.

4VMQIVS
3JDI "DJE
HBT 4VMQIVS 4FNJ -FBO
QJU
WFOU
  QSPEVDU SJDI 

 

"NJOF
SFHFOFSBUPS

4VMOPM.

Figure 5 Sulphur balance for Case 2: Sulfinol-X gas treating with SRU using Sulfinol-M enrichment, AGE overhead routed to SCOT
(sulphur flows in lbmoles/hr)

12 GAS 2011 www.eptq.com

b&v.indd 6 9/3/11 13:01:55


5SFBUFE
OBUVSBM
HBT
AGRU
4PVS
(Sulfinol-M)
OBUVSBM
HBT -FBO
 
3JDI

"DJE
HBT
AGRU
regenerator
(Sulfinol-M)
Figure 6 Sulphur balance for Case 3: Sulfinol-M gas treating with integrated enrichment and tail gas treating, AGE overhead routed to
SCOT (sulphur flows in lbmoles/hr)
4C was the optimal case from the
2010 article, hence this configura-
tion was selected without further
analysis of Case 4A or 4B configu-
rations in this evaluation.
The Case 3 process configuration,
which integrates acid gas enrich-
ment within the AGRU, is shown in
Figure 3. Sour natural gas is proc-
essed in a Sulfinol-M AGRU and
the resulting acid gas is processed
in a conventional two-bed Claus
SRU. The Sulfinol-M AGRU
contains an enrichment step, which
consists of an additional hot flash
vessel and an enrichment absorber.
In this enrichment section, rich
solvent is heated and flashed at
reduced pressure. The CO2-rich
flash vapour flows to an enrichment
absorber, where lean Sulfinol-M
solvent absorbs H2S and some
mercaptans. The enrichment
absorber overhead flows to a sepa-
rate hydrogenation section of the
TGTU, for conversion of mercaptans
to H2S, similar to the Case 2 flow-
sheet. Also similar to Case 2, the
Sulfinol-M solvent circulation
system is integrated between the
AGE and TGTU absorbers, as per
the Case 4C configuration in the
2010 study. Case 3 employs a single
regenerator for regeneration of the
www.eptq.com
b&v.indd 7
)1
BTI

HBT


&OSJDINFOU
BCTPSCFS
PWFSIFBE
 4UBDL
HBT

436 5(56
UBJM
HBT UBJM
HBT
  
436 TGTU Incinerator
(Sulfinol-M)
4VMQIVS
4VMQIVS -FBO 3JDI QJU
WFOU
QSPEVDU   

Sulfinol-M solvent from all three removal specification that can be
absorbers. Concentrated acid gas achieved with Sulfinol-X, the speci-
flows from the Sulfinol-M regenera- fication of <1 mole% CO2 selected
tor to the SRU/TGTU for sulphur for this article can be easily
recovery. achieved.
Compared with Case 2, Case 3
has one solvent system instead of Comparison of sulphur recovery
two, which eliminates one regenera- efficiency
tor and reduces the equipment The overall sulphur recovery for
count in the solvent storage and each of the test cases is given in
drain systems, as well as reducing Table 5. The H2S removal and
operational complexity. The formu- sulphur recovery steps in each case
lation of Sulfinol-M allows this have been modelled on a consistent
solvent to be used both for the basis. As a result, differences
selective and non-selective removal between cases reflect their particu-
of H2S in the presence of CO2, lar process characteristics, without
depending on the process condi- influence from the inconsistencies
tions. While Sulfinol-M cannot associated with variations in param-
reach the LNG-like deep CO2 eters such as H2S removal efficiency,
Comparison of sulphur recovery efficiency for test cases
Case 1 2 3
HP flash gas to incinerator, lbmol/hr 0.083 0.080 0.031
TGTU H2S 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 SO2, 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%
Table 5
GAS 2011 13
9/3/11 13:02:04
 Comparison of solvent systems
Solvent system comparison for test cases
Solvent circulation rates and other
solvent system design data for each
Case 1 2 3 of the test cases are shown in
Main absorber
Lean solvent, gpm 1236 1236 1265 Table 6.
TGTU absorber Solvent flow rate to the main
Lean solvent, gpm 325 1015 1012 absorber is about the same in all
Inlet H2S, mole% 1.35 1.31 1.35 cases, with only a slightly higher
Inlet CO2, mole% 48.43 50.74 49.31
Overhead H2S, ppmv 248 249 248 requirement for the Sulfinol-M case
CO2 slip, % 93.6 88.9 88.2 (Case 3). In Cases 2 and 3, lean
AGE absorber Sulfinol-M solvent flows to the
Lean solvent, gpm - - - TGTU absorber with subsequent
Inlet H2S, mole% - 24.10 19.55
Inlet CO2, mole% - 69.62 74.90 cascading of the semi-rich TGTU
Overhead H2S, ppmv - 6526 6562 absorber bottoms to the AGE
CO2 slip, % - 88.1 88.7 absorber. Cases 2 and 3 have the
Main treating regenerator same lean solvent flows to the
Acid gas H2S, mole% 24.20 24.10 53.15
Total lean solvent, gpm 1284 1284 2338 TGTU absorbers and the same
Regeneration steam, lb/hr 55 888 55 888 129 387 semi-rich solvent flows to the AGE
AGE/TGTU regenerator absorbers. Thus, Case 2 has essen-
Acid gas H2S, mole% 28.46 57.22 - tially the same total solvent flow as
Total lean solvent, gpm 325 1015 -
Regeneration steam, lb/hr 19 221 53 949 - Case 3, with the only difference
Total solvent circulation, gpm 1609 2299 2338 being a minor incremental increase
Total regeneration steam, lb/hr 75 109 109 837 129 387 in flow for the Sulfinol-M main
absorber in Case 3. Sulfinol-M
requires more steam for regenera-
Table 6 tion than Sulfinol-X; therefore, Case
3 has a higher solvent regeneration
reboiler duty or circulation rate. performance. This is because the steam consumption rate than Case
Overall sulphur recovery reported 2010 study considered a lower 2. Case 1, because it does not
in Table 5 considers the sulphur TGTU absorber overhead H2S speci- perform acid gas enrichment, has
balance envelope around the entire fication than that in the Sulfinol the lowest total solvent circulation
gas treatment facility (AGRU/ schemes of this study (approxi- and steam consumption rates.
AGE/SRU/TGTU) and includes
sulphur lost in the treated natural Comparison of SRU/TGTU train size
gas stream, while sulphur recovery Most SRU and TGTU Most SRU and TGTU equipment
of the sulphur complex figures sizes are primarily determined by
reported in Table 5 considers only equipment sizes are the process gas volumetric flow
the sulphur balance around the rate. SRU, TGTU and incinerator
sulphur production facilities (AGE/ primarily determined volumetric gas rates for each test
SRU/TGTU). It is important to note case are given in Table 7.
that the 2010 study reported by the process gas Case 1 requires the largest SRU/
sulphur recovery efficiency figures TGTU train size due to the fact that
for the sulphur recovery complex
volumetric flow rate it processes SRU feed gas with the
only (ie, the second set of figures in lowest H2S concentration; flow rates
the table) and not for the overall are similar to those for Case 1 from
gas treatment facility. Furthermore, mately 180 ppmv versus 250 ppmv, the 2010 study. Cases 2 and 3
the sulphur complex recovery effi- respectively). significantly reduce the SRU size
ciency figures reported in the last All three cases are similar in because these alternatives minimise
row of Table 5 should not be used sulphur recovery efficiency (see CO2 flow through the SRU by slip-
in direct comparison with the Table 5). The sulphur balance ping CO2 from the enrichment step
recovery efficiency figures reported for each case is illustrated in with the use of Sulfinol-M solvent.
in 2010 for assessment of solvent Figures 46. Since the enrichment absorber over-
head is routed to the TGTU in
SRU/SCOT train size comparison for test cases Cases 2 and 3, the TGTU and incin-
erator sizes are not substantially
smaller than the Case 1 size.
Case 1 2 3
SRU tail gas, lbmol/hr 1870 1063 1108
TGTU feed gas, lbmol/hr 1870 1798 1727 Capital and operating
Incinerator process gas feed, lbmol/hr 1713 1648 1576 cost comparison
Table 8 shows estimated capital and
Table 7 operating costs for each case. Rough

14 GAS 2011 www.eptq.com

b&v.indd 8 9/3/11 14:42:33

 Model for our technologies.
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.

Build on our technologies.

Call us, we inform you: +49 (0) 69 58 08-0
www.lurgi.com
www.ptqenquiry.com
A member of the Air Liquide Group for further information
1122_e

15_lurgi.indd 1 8/3/11 13:42:28

 instead of MDEA. In the 2010 study,
Cost comparison for test cases
the AGE/SRU/TGTU configuration
of Case 4C resulted in slightly
Case 1 2 3
Fuel lower capital cost than the conven-
Consumption, MMBtu/hr 32.3 26.8 30.6 tional SRU/TGTU configuration of
Yearly cost 1.7 1.4 1.6 Case 1, which did not include
Power enrichment. Here, Case 2 is slightly
Consumption, kW 1314 1348 1490
Yearly cost 0.7 0.7 0.8 higher in capital cost than Case 1.
600 psig saturated steam The difference between the results
Export, lb/hr 19 940 17 481 17 301 of the 2010 study and the results
Yearly credit 1.4 1.2 1.2 reported here is explained primarily
50 psig saturated steam
Import, lb/hr 58 964 99 383 118 769 by the fact that the hydrogenation
Yearly cost 3.5 5.9 7.0 step was split into two reactors
Maintenance and non-utility operating costs instead of combined into a single
Yearly cost (10% TIC) 8.7 9.2 9.4 reactor, as in the 2010 Case 4
Cost summary
Total yearly operation and maintenance cost 13.2 16.0 17.6 configurations.
Total installed cost 86.6 92.2 93.9 The important comparison is
Net present cost 171.9 195.6 207.6 between Cases 2 and 3, both of
Notes: Costs are reported in millions of US dollars.
which produce acid gas with a H2S
1. Costs were prepared using the following information: concentration high enough to ensure
* $6.00/MMBtu LHV fuel stable Claus SRU performance. In
* $0.064/kWh power
* $6.96/1000 lb 50 psig sat. steam both of these cases, the SRU, TGTU
* $8.31/1000 lb 600 psig sat. steam and incinerator portions are similar
* 355 operating days per year
* 15% annual rate of return in size and cost; therefore, the
* 25-year plant life difference in capital cost is in
the AGRU/AGE/TGTU treating
sections. Although Case 3, with its
Table 8 single Sulfinol-M solvent system,
requires fewer equipment items in
order of magnitude capital costs are makes the acid gas enrichment these sections, some of them are
provided on a US Gulf Coast basis achieved in the other cases desira- larger and therefore slightly more
and include engineering, procure- ble, in spite of their larger amine expensive than the equivalent equip-
ment and installation of all process units with higher energy ment items in Case 2. Sulfinol-M
and utility equipment; storage, consumption. requires more regeneration energy
metering and loading facilities; than Sulfinol-X, making the regener-
plus, construction indirects, spares, ation system larger. Additionally,
engineering and commissioning. Alternative acid gas Sulfinol-M requires a larger absorber
Items such as owners costs, land in the AGRU.
costs, escalation and contingency enrichment schemes Despite the fact that this analysis
are excluded. Operating costs are shows a slight preference for Case 2
discounted for each year of plant must be considered to over Case 3, the capital cost differ-
life and are added to capital cost to ential is small and quite possibly
form each cases net present cost.
achieve high sulphur within the margin of error of the
The conventional SRU/TGTU
arrangement paired with Sulfinol-X
recovery efficiency estimates. When operating costs are
considered, there is a greater depar-
treating for Case 1 serves as a when organic sulphur ture between the net present cost of
benchmark for comparison between the two options, with Case 2 still
the other cases, similar to the way compounds are emerging as the preferred configu-
in which the Case 1 conventional ration. However, the small
SRU/TGTU arrangement of 2010 present difference in net present cost
served as a benchmark for the suggests that a detailed analysis
enrichment cases studied at that The Case 2 Sulfinol-X design is between these two options should
time. The Sulfinol-X design of Case essentially the same as Case 1; be carried out when selecting the
1 represents an optimum approach however, in Case 2, the lean acid preferred scheme for a specific sour
if considering only the AGRU and gas from the AGRU is enriched to natural gas-treating application.
not the impact that the resulting improve Claus SRU operating
lean acid gas has on the down- performance, while also giving the Summary and conclusions
stream SRU. However, as benefit of reduced SRU size. The A 2010 analysis by Black & Veatch
ascertained in the 2010 study, Claus AGE/SRU/TGTU for Case 2 is acknowledged that acid gas enrich-
SRU operational difficulty caused similar to Case 4C from the 2010 ment should be considered when
by low acid gas H2S concentration study, but with Sulfinol-M solvent upstream AGRU configurations

16 GAS 2011 www.eptq.com

b&v.indd 9 9/3/11 13:02:27

 produce lean acid gas with H2S
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 efficiency 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
high recovery efficiency, and
employs maximum solvent integra-
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
for
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...

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
Sulfinol gas-treating schemes to
examine the lifecycle cost of the
entire sour natural gas treatment
facility.
In the current analysis, Case 2,
Sulfinol-X main gas treating with
Sulfinol-M acid gas enrichment and
tail gas treating, results in a slightly
lower lifecycle cost than Case 3,
which is a fully integrated Sulfinol-
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 specific
sour natural gas treating applica-
tion. Additionally, it is important
not to overlook the somewhat
intangible operational and mainte-
nance benefits associated with
operating a single solvent system in
Case 3, versus two separate systems
in Case 2, when selecting the best
configuration to meet a sour gas
processors needs.
Remove Solids and Hydrocarbons from
your solvent with MPRs High Capacity,
Easily Regenerated Media.
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
filters, 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 removing solids to
submicron size particles, while also removing hydrocarbons
(including diesel, gas oils, paraffins, BTEX and various petroleum
intermediates) from amine solutions.
Eliminating cartridge filters and activated carbon means eliminating filter replacement,
disposal and labor
SSX and HCX processes are available exclusively from MPR Services, Inc..

Sulfinol and SCOT are trademarks of Shell. For more information, call 281-377-7424
or visit our website at www.mprservices.com

MPR Services, Inc., A Tessenderlo Kerley, Inc. Company.

Further reading
1 McNamara H J, Schilk J A, US Patent No.
4 356 161, Washington, DC, US Patent and
Trademark Office, 1982.
1201 FM 646, Dickinson, TX 77539-3014 U.S.A.
5FMFQIPOF

t
'"9

E-mail: JOGPNQSTFSWJDFT!ULJOFUDPN
t
NQSFVSPQF!ULJOFUDPN
t
NQSNJEEMFFBTU!ULJOFUDPN
2011 Tessenderlo Kerley, Inc.

www.ptqenquiry.com
www.eptq.com for further information GAS 2011 17

b&v.indd 10 9/3/11 13:02:38

 Sweet
Sour

Gas processing expertise in a variety of avors.

8IFO
JU
DPNFT
UP
JOOPWBUJWF
QSPDFTT
EFTJHO
BOE
CSPBE

FYQFSJFODF
#MBDL

7FBUDI
JT
UIF
QSPWJEFS
PG
DIPJDF
:PV
DBO
DPVOU
PO
VT
UP
VOEFSTUBOE
ZPVS
TQFDJ
D
CVTJOFTT

ESJWFST
BOE
EFMJWFS
VOCJBTFE
UBJMPSFE
HBT
USFBUJOH
BOE

QSPDFTTJOH
TPMVUJPOT
GPS
ZPVS
VOJRVF
OFFET
8IFUIFS
JUT

POF
PG
PVS
JOOPWBUJWF
QSPQSJFUBSZ
UFDIOPMPHJFT
PS
B
IPTU
PG

PUIFS
UFDIOPMPHJFT
MPPL
UP
#MBDL

7FBUDI
GPS
CFTUWBMVF

BOE
JOUFHSBUFE
TPMVUJPOT
GPS
BOZ
DPOEJUJPOT

8FSF
CVJMEJOH
B
XPSME
PG
EJGGFSFODF
5PHFUIFS

$POTVMUJOH







&OHJOFFSJOH







$POTUSVDUJPO







0QFSBUJPO




I X X XCWDPN

www.ptqenquiry.com
for further information

11_GOC_PTQ's Gas Supplement_AD.indd 1 2/21/11 2:41 PM

b&v.indd 1 8/3/11 14:17:05
 Cryogenic monitoring
Fibre optic distributed temperature sensing has several advantages over
conventional real-time data sensors in cryogenic leak detection at LNG terminals
Jerry Worsley
Sensornet
E
nvironmental and safety issues
are becoming ever more
important in the liquefied
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 significant
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 benefits 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
www.eptq.com
sensornet.indd 1
any cool-down or leak event. By
contrast, using fibre optic DTS, the
location of such an incident can be
identified almost instantly and
remedial action taken. In fibre optic
DTS, the fibre is the sensor and the
measurement is totally non-
intrusive. No welds are required to
the pipe or vessel and the fibre can
be easily strapped to the outside of
the pipe with no interruption to the
operational process. The fibre
provides its own communications
path and no additional tie cables
are required. Most importantly,
DTS using optical fibres 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 profiling applications,
including LNG, liquefied petroleum
gas (LPG), ethylene and CO2 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 benefits 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 fire and gas
control systems using standard
protocols.
The benefits 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, fulfil 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 fibre optic cables
and can be used to detect both
liquid and gaseous leaks. The DTS
system illuminates the glass core of
the optical fibre with a 10-
nanosecond laser pulse, which
corresponds to a 1m pulse. As the
optical pulse propagates down the
fibre, it undergoes scattering even
in the absence of impurities and
GAS 2011 19
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-defined function of
temperature, the ratio of the signals,
in conjunction with the time of
flight of an optical pulse, is used to
determine the temperature of the
fibre 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
20 GAS 2011
sensornet.indd 2
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 field, allowing the
monitoring of hard-to-reach places.
The fibre optic cable is almost as
important as the DTS itself, because
it is the fibre 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 fibre optic cable.
However, the user should be
prepared for a one-time use only.
Tests have shown that standard
cable and fibre 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 fibre and jacket
construction deteriorate and become
unusable in the future. It may be
acceptable in certain applications
for these cables to be sacrificial. For
long-term use and cool-down moni-
toring, cables and fibres with
www.ptqenquiry.com
for further information
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
fibre optic monitoring can help
complete the cool-down process in
a speedy but safe manner. Sections
of fibre 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.eptq.com
9/3/11 11:36:46
 PTQ_86x270:Layout 1 18-03-2010 17:58 Pagina 1

The same fibre can be routed along the length of pipe

outside the insulation layer and be used for leak detec-
tion. This fibre 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-
Italian design
tion to near cryogenic temperatures. The system can be
installed on different pipe configurations, including
A masterpiece
bare pipes, foam-insulated pipes and vacuum-insulated
piping, and has been independently verified 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 fibre 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 confidence 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 Creativity is the art we ap-
tanks, under valve spill trays and in tank overflow ply to achieve superior de-
sign and developments in
channels. The spill channels and impounding collection
technology.
channels are prime locations for products to collect in
For over 70 years we have designed and supplied
the event of a leakage. Installing sensing cables at low cost-effective technology, process plants and equip-
points in these locations provides an excellent early ment for the oil & gas and refining industry around
warning system. the world. With our expertise we provide tailor-
made solutions from studies and revamps to skid-
DTS in the real world mounted units and complete turnkey plants.
In compliance with Health and Safety Executive
requirements, the UKs National Grid Isle of Grain
LNG facilities asked Sensornet to provide a leak detec-
Oil & gas treatment
tion system. The monitoring system would be separation, dehydration and HC dew point control,
non-intrusive, permitting the continued operation of LPG and NGL recovery, sweetening,
the importation and recirculation pipelines during its mercaptan removal
installation and commissioning. Acid gas treatment & sulphur recovery
Traditional use of flow meters based on differential acid gas enrichment, Claus, ammonia Claus,
oxygen-enriched Claus, tail gas treatment,
flow measurement was considered inadequate to sulphur degassing
provide the required leak detection monitoring for
emergency shutdown purposes due to their relatively Advanced process controls
long response time. Other monitoring methods were Special process equipment
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
SIIRTEC NIGI Engineering Contractors
modules provide hard wire contact relays to the client Via Algardi, 2 20148, Milan (Italy)
SCADA system in instances such as system error, Tel: +39-0239223.1 Fax: +39-0239223.010 Web: www.sini.it
Contact: marketing@sini.it
where the rate of a temperature change exceeds a set

www.ptqenquiry.com
for further information
www.eptq.com GAS 2011 21

sensornet.indd 3 9/3/11 11:36:57

 temperature, where the minimum
temperature exceeded a fibre break,
or where there is overheating in the
instrument cabinet.
The system provides non-
intrusive protection and was
retrofitted by installation engineers
in an operational plant without
having to shut down or stop opera-
tion. The independent fibre optic
sensing cable loops covering both
pipelines provide added resilience
to the installation. Additional multi-
plexer channels ensure future
expansion and flexibility, 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
specification 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
significant temperature drop (typi-
cally greater than 100C).
When the cryogenic fluid 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
specific location can indicate that a
significant 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
22 GAS 2011
sensornet.indd 4
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 significant 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 configured to provide a
number of useful temperature
comparisons. Absolute temperature
can vary, so it is not always a useful
figure when used directly for leak
A robustly designed
data acquisition
device can be
positioned in the
field, allowing the
monitoring of
hard-to-reach places
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 financial 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 fulfilled 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
flying 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 specification and
deployment of fibre 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.eptq.com
9/3/11 11:37:10
 www.ptqenquiry.com
for further information

burkhardt.indd 1 20/9/10 10:10:44

 www.ptqenquiry.com
for further information

cansolv.indd 1 8/3/11 13:53:57

 Glycol dehydration of high-acid
gas streams
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
G
lycol dehydration is a proc-
ess that presents some
unique challenges from both
technical and computational stand-
points. In the first 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 CO2 is compressed, lique-
fied and injected into underground
formations. Handling these sour
and high CO2 streams requires
dehydration prior to compression
and/or liquefaction.
www.eptq.com
ogtrt.indd 1
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 finds 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 humidification, and
the humidification process extracts
the necessary heat of vapourisation
as sensible heat from the liquid
water phase. Sudden humidification
can drop the wash water tempera-
ture by 3040F or even more.
This article addresses the efficacy
of a new mass and heat transfer
rate-based model in terms of how
well it reflects 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 CO2 and/or H2S;
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 Law2,3 is:
>vi (P Pjo)H (1)
a~ixiHij exp = P v
RT i i
where
i: Activity coefficient
xi: Mole fraction of component i in
the solvent (its solubility)
Hij: Henrys constant for solute i in
solvent j, kPa/mol frac
-
vi Partial molar volume of i in
solvent j at infinite dilution,
m3/kmol
Pjo: Vapour pressure of the solvent,
kPa
P: Total pressure, kPa
R: Universal gas constant,
8.314 kJ/kmolK
T: Absolute temperature, K
yi: Mole fraction of component i in
the vapour
iv Fugacity coefficient 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.
GAS 2011 25
8/3/11 14:50:14
 can be greater than unity; however,
Saturated water content of gases
it is rarely greater than two.
Therefore, it is very important to
GPSA ref. Mole % (dry basis) Temp, F Pres, psia H2O lb/MMscf include this contribution in the
CH4 CO2 H2S Measd ProTreat
Ex 20-1 100 0 0 150 1000 220 216 solubility model.
Ex 20-2 80 20 0 160 2000 172 188 The fugacity coefficient in the
Fig 20-9 0 100 0 100 500 132 125.3 vapour phase can be calculated
750 110 102.5 using an equation of state such as
1000 125 100.7
2000 215 215.1 the Soave-Redlich-Kwong (SRK) or
3000 238 247.8 Peng-Robinson (PR) equations.
5.31 94.69 0 100 850 88 96.9 These cubic equations of state have
1125 81 99.2 been used for many years for
1500 128 148.6
2000 139 184.2 modelling the fluid properties and
Fig 20-16 89 11 0 100 2000 40.6 41.1 phase equilibrium in petroleum
89 11 0 160 1000 286 283.9 systems containing light hydrocar-
80 20 0 100 2000 40.6 45.1 bons and a few associated
80 20 0 160 1000 282 292.5
80 20 0 160 2000 172 188.5 non-hydrocarbons such as carbon
92 0 8 130 1500 111 103.5 dioxide, hydrogen sulphide and
72.5 0 27.5 160 1367 247 252.6 nitrogen. However, they are less
83 0 17 160 1000 292 293.4 successful with water and other
30 60 10 100 1100 81 81.2
9 10 81 100 1900 442 264.4 polar substances such as alcohols
5.31 94.69 0 77 1500 109.2 95 and glycols. Typically, the equa-
5.31 94.69 0 122 2000 164.6 234.5 tions of state must be modified to
predict accurately the pure compo-
Table 1 nent properties, and the mixing
rules must account for the complex
The numerical value of the 1.1. Neglecting the Poynting effect interactions.
Henrys constant is a function of altogether at these conditions means At low pressure, both the
both the solute and the solvent. the solubility is under predicted by Poynting correction and the fugac-
Thus, the Henrys Law constant, Hij, approximately 10%. ity coefficients are approximately
is different for methane in water A common definition used in unity. Furthermore, if the solubility
than it is for methane in methanol, describing the solubility of a gas in is low, Equation 1 reduces to the
and Hij is different still for ethane a liquid is partial pressure. The more familiar form of Henrys Law:
in water. Furthermore, for every partial pressure, Pi, of a component
solute-solvent pair, the Hij is a func- in a gas mixture is defined as: xiHij = yiP (4)
tion of the temperature. When
comparing different solutes in the Pi = yiP (2) or equivalently:
same solvent, the larger the Henrys
constant the lower the solubility of The non-idealities in the vapour xiHij = Pi
(5)
the solute. phase are embodied in the fugacity
The activity coefficient accounts coefficient. As an approximation, Physically, this equation says that
for the effect of concentration on the fugacity is to the partial pres- the solubility of a gas increases
the activity of the component in the sure as the compressibility factor is directly as the partial pressure of
liquid phase. The infinite dilution to the molar volume. Included in the component in the gas increases.
definition of the activity coefficient this quantity is the effect of pres- However, this idealised version is
is used here; its significance is that sure, temperature and composition limited to about 200300 kPa and
at infinite dilution, the activity coef- on the gas phase non-ideality. For gases with solubilities less than
ficient defined this way is unity. an ideal gas mixture, the fugacity about 0.1 mol%.
The exponential term is the coefficient of all components, i, in The ProTreat simulation tools
Poynting correction. It accounts for the mixture is unity: dehydration model uses the Peng-
the effect of pressure on the refer- Robinson equation of state (EOS)
ence fugacity, and it is important iv = 1 (3) for the vapour phase and currently
(significantly different from unity) offers a four-suffix Margules equa-
only at high pressure. For a small However, for a non-ideal mixture, tion activity coefficient model based
molecule such as methane, hydro- this quantity can be less than, on the data of Bestani & Shing1 for
gen sulphide or carbon dioxide in greater than or equal to unity. At the liquid phase, as reported by
water, the partial molar volume at relatively low pressure, it is gener- Clinton et al.4 A similar model
infinite dilution is approximately ally less than unity and it can be as based on the less conservative data
0.032 m3/kmol. Thus, at 7000 kPa small as 0.1 (and even smaller for of Parrish et al6 is planned for a
and 20C (293K), the Poynting liquids). At high pressure and high future release.
correction for these systems is about temperature, the fugacity coefficient There are two important aspects

26 GAS 2011 www.eptq.com

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 (kijs) 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 fit the ProTreat
solubility model for these species.
The data indicate that at typical
contactor conditions, approximately
1030% of the aromatics in the gas
<7 lb H20/MMscf
(<0.01474 mole% H20)
12
99%wt TEG
3 gal TEG/lb H20 removed (5.1 gpm)
Contactor
9 8
Recycle
Two ideal stages
Diameter = 3.4 ft
11
25-30% efficiency
30 MMSCFD gas 6.8 bubble cap trays, or
100F 10 ft of structured packing
600 psia
Water saturated
1 reboiler duty
Figure 1 GPSA Data Book example 20-11
www.eptq.com
ogtrt.indd 3
Water content vs tray count
Number Water
of trays lb/MMscf
5 8.5
6 6.7
7 5.7
8 5.1
Table 2
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
efficiency of 2530%, translates into
68 actual trays. The other case uses
10ft of an unspecified structured
packing. ProTreat has provision for
a separate Stahl column, shown
immediately below the stripper in
Dehydrated gas
CB 7 23
Controller Cooler
19
27
Lean/rich
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 specifiable
portion of condensed water (Stream
20), with the remainder returned
as reflux.
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% flood is 3ft. These values
are in line with GPSA Data Book
results. In summary, the available
data indicate that the model is accu-
rately reflecting 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
Outlet
22 24
5
6
Outlet
Stripper
21
350 MBTU/hr
20
3 2
Stahl column
Stripping
26 gas
10 4
GAS 2011 27
8/3/11 14:50:33

0 temperature bulge in dehydration
5 M170.X
Distance from top of bed, ft
10
15
20
25
30
35
40
1000
Figure 2 Temperature profiles and the effect of packing size in the Mellapak X Series
surely related to packing size. For
structured packing, size can be
expressed in terms of specific 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-
0 advantage. Mass and heat transfer
5 M170.X
Distance from top of bed, ft
that there can be a significant
M125.X columns caused by the phase
change (condensation) of the water
M2.X
being absorbed. Water absorption
M250.X
generates sensible heating equiva-
M350.X
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
110 120 130 140 20ft of M350.X packing, the gas is
Temperature, F 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
degree crimp angle) packings affects has passed through 40ft of packing.
dehydration performance. For So the required bed height is very
Sulzer Mellapak structured pack- much a function of the packing size.
ings, the packing designation (eg, It is not that rules of thumb cannot
M250.X) is a close indicator of the be made to work; rather, it is that
specific area, in this case 250 m2/m3. the right rules of thumb, at least for
Simulations were all run with 40ft packing, depend on too many
of packing and the absorber was parameters (not just packing type
sized for 70% flood (911ft diameter and size but also on the gas and
depending on crimp size). The solvent fluxes through the column)
absorber was set up to dehydrate 49 and this makes them rough at best
000 lbmol/h of wet sweet methane and unreliable at worst. Until now,
(trace CO2 and H2S) using 250 gpm the answer to this dilemma has
of 99.95 wt% TEG. been to over-build the columns;
The gas-phase temperature however, in a competitive environ-
profiles for the various packing ment, surely being able to avoid
sizes shown in Figure 2 indicate over-design gives the knowledgea-
ble contractor and the astute
internals vendor a commercial
M125.X rate-based simulation is the preci-
sion tool that allows this to be

M2.X done.
10 M250.X Figure 4 shows that the blanket
M350.X rule of thumb of 68 trays for
15 dehydration can be a gross over-
simplification. The number of trays
20
depends at least on the dryness to
be achieved; that is, the dryness of
25
the solvent. If the target dew point
is not too stringent (say, 10 or 20lb
30
H2O/MMscf), 6 or 8 trays seem
35 adequate. But in very low dew
point applications, such as LNG
40 plants, two or three times that
0 1 10 100 1000 number of trays may be required to
Water in gas, Lb/MMscf get to the desired dryness. With
99.97% TEG, for example, it is
Figure 3 How water removal depends on packing size and packed bed depth possible to get to below 1 lb/MMscf

28 GAS 2011 www.eptq.com

ogtrt.indd 4 8/3/11 14:50:43

 Unique Solution for Your
Petrochemical Application
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 Benefit:
Lowest Life Cycle Costs

www.ptqenquiry.com
for further information

Like this image? Get more: www.recip.com/screensaver

$GB/DE\$LQGG 
burckhardt2.indd 1 9/3/10 14:25:09
 water content in the treated gas by
100
more than a factor of two. Stripping
gas permits more water removal
from the solvent and increases the
Dry gas water content,

dry TEG from 98.8 wt% to 99.8

10 wt%. This particular treated gas is
lb/MMscf

dry enough by transmission line

standards, but it is a long way from
dry enough for an LNG plant, for
1 99.97% example. However, this demon-
99.95% strates the principle that a Stahl
99.90% column can serve a useful purpose
99.50% it should be used when treated
0.1 gas needs to be drier than usual.
0 8 12 16
Number of trays
Observations from a mass transfer
rate model
Figure 4 Water removal to low dew points requires deeper beds and more trays The mass and heat transfer rate
model uses real trays, both in
water content, but even after 16 solvent from the reboiler and number and mechanical detail, and
trays, water is still being removed. contacting it with a flow of dry real packing in terms of actual bed
For the particular conditions inert gas. Stahl columns are essen- depths of a specific packing with
simulated in this study, it should be tial when the dried gas must have a defined geometry, including crimp
mentioned that high TEG viscosity very low dew point. Figure 5 shows angle, crimp size, surface treatment,
is a consideration and will nega- the effect of stripping gas rate (SCF specific surface area and vendor.
tively affect internals performance per gallon of TEG solvent) on the The simulation of a packed column
compared with light hydrocarbons, TEG purity and on the water begins by finely segmenting the
for example. At the solvent mois- content of the treated gas for packed height into a large number
ture levels encountered in these Example 20-11 from the GPSA Data of thin cross-sectional slices to
simulations, viscosity is not signifi- Book scaled up by a factor of five approximate the continuous nature
cantly affected by water content, and using six bubble cap trays in of contacting in packing. As a
and at the lowest temperature (feed the absorber and a 10ft bed of result, one gets to observe mass and
solvent at 100F) the viscosity is FlexiPac 1.6Y in the regenerator heat transfer effects on a fairly
about 19 cP (for reference, corn (includes a 2ft reflux wash section). detailed scale.
syrup is 50100 cP). This does not In terms of scale, the absorber is Figure 6 shows the vapour and
necessarily rule out using trays, nearly 7ft in diameter and is drying liquid temperature profiles through
although it does tend to make gas that is water saturated at 600 a 10ft-deep bed of 2in metal Nutter
structured packing more attractive. psia and 100F. The regenerator is Rings in a regenerator operating at
only 15in in diameter. 1 psig. The dehydrator is treating
Stahl columns and stripping gas The simulation results in Figure 5 440 MMscfd of wet, sweet gas with
A Stahl column gives an extra stage show that using even a modest flow a TEG flow of 1000 gpm, and the
of regeneration by taking the of stripping gas can reduce the bone-dry nitrogen stripping gas
flow 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

D r y
ga s,
lb/ M M sc G

Glyc ol
pur iU y w t %

 
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
 5&(
XU  the regeneration column, starting
MC..TD' with the condensate (essentially
 
      
pure water), which enters the
Stripping gas rate, SCF/gallon column from the condenser at
180F and is heated by the gas
stream (stripping gas and water),
Figure 5 Effect of stripping gas flow on solvent dryness and water content of gas which is at 198F.

30 GAS 2011 www.eptq.com

ogtrt.indd 5 8/3/11 14:50:53

 As the condensate trickles down
through the bed, it continues to be 0
heated by the wet stripping gas, but 1

Distance from top of bed, ft

within the first 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 flashes
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 reflux water
stream. But why does the reflux
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 first reflux segment,
this vapour is suddenly exposed to
a pure, hot water stream (flowing
1
3
4
5
6
7
8
9 Vapour
Liquid
10
150 175 200 225 250 275 300 325 350
Temperature, F
Figure 6 Temperature profiles in regenerator with stripping gas via Stahl column
As we traverse lower into the the vapour back into the liquid in
column and approach the bottom of the bottom couple of segments of
the stripping section, liquid starts the stripping section.
to meet much hotter vapour coming Figures 7a and b are intended to
from the reboiler. (The bubble point show how the use of stripping gas
of TEG-water mixtures is quite affects these temperature profiles.
sensitive to water content when the The plots are for the same 10ft bed
water content is low.) This cools the of Flexipac 1.6Y structured packing
vapour and heats the liquid, and referenced in Figure 5. Even when
again some of the heat transfer is stripping gas is not used, the
the result of water transferring from condensed water leaving the reflux


D ist a nc e
f rom
t op
of


from the wash section), so it is seri-

ously under saturated against pure

water. The consequence then is for
be d,
f t

water to evaporate as fast as possi- 

ble so as to resaturate the hot gas, 
now against pure hot water. This is 
a typical humidification operation 7BQPVS

and has some interesting character- -JRVJE

istics. The evaporation process is not          
mass transfer rate limited. Rather, it Temperature, F
is limited by the rate at which the
latent heat of vapourisation

demanded by the humidification
process can be drawn from the bulk 

D ist a nc e
f rom
t op

liquid to the gas-liquid interface. 

This process is one of heat transfer
of
be d,
f t


and, indeed, the rate of humidifica- 
tion right above the feed point is

controlled by essentially conductive

heat transfer across the liquid film 7BQPVS
running over the packing. The proc- 
-JRVJE
ess is heat transfer rate limited. The 
mass transfer driving force for           
humidification is so high that Temperature, F
enough heat is drawn from the
reflux water to cool it, in this case by Figure 7 Stripping gas is not necessary for anomalous temperature profiles
approximately 33F. (a) No stripping gas (b) Stripping gas at 3 scf/gal

www.eptq.com GAS 2011 31

ogtrt.indd 6 9/3/11 14:45:02


100
Water in treated gas,
120
lb/MMscf
80
60
40
5 10
Figure 8 Effect of tray count on water content of methane, CO2 and H2S dehydrated
with 99.5 wt% TEG at 15 psig and 100F using bubble trays
section meets a much hotter gas,
and a good part of the temperature
equilibration takes place by water
evaporation from the reflux stream.
This is, again, a humidification
process and the reflux 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 humidifica-
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
flows of stripping gas versus
reboiler vapour.) Again, the reflux
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
130
120
Water in treated gas,
110
100
lb/MMscf
90
80
70
60
50
40
12
Figure 9 Effect of tray count on water content of methane, CO2 and H2S dehydrated with
99.5 wt% TEG at 15 psig and 100F using Mellapak M170.X structured packing
32 GAS 2011
ogtrt.indd 7
Sweet
CO2
H2S
15 20
Number of trays
bulge in the reflux 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,
Sweet
CO2
H2S
16 20 24 28
Packing depth, ft
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 reflective 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: H2S
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
www.eptq.com
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
H2S with the equivalent sweet gas
volume. As Figure 8 shows, it is
difficult 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,
CO2 or H2S 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 H2S in
equilibrium with the lean TEG is
somewhat higher than it is for
either CO2 or even methane. This
has to do with the very polar nature
of H2S, which binds water more
tightly to itself than CO2 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
H2S is pretty much completely
dehydrated after about 12 trays
(about 20ft of packing), whereas
CO2 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 H2S
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, CO2 and H2S 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
H2S and this must be removed
before the solvent is regenerated to
a suitable state of dryness, because
it is a certainty that this much H2S
www.eptq.com
ogtrt.indd 8
cannot be vented to atmosphere
with the water vapour stripped
from the solvent.
Its high solubility in TEG is really
the main difficulty with dehydrat-
ing very high concentration H2S;
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 difficulty is accurately
representing the gas-phase non-
idealities that arise from the highly
polar nature of water and H2S.
Conclusions
Rules of thumb are fraught with
danger because often the rules
simply do not apply. A blanket tray
efficiency of 25% is close to the
truth most of the time. But in deep
water removal, 25% is optimistic
The mass transfer
rate-based tower
model uses science
and good engineering
to predict
performance
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 efficiencies,
artificial 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, Infinite dilution activity
coefficients 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 first 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
GAS 2011 33
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. Profit from our competence along the entire LNG Value Chain:

Process plants Small and mid-scale LNG receiving terminals

  BUhifU`[Ug`]eiYZUWh]cb
  BUhifU`[UgdfY!hfYUhaYbhB;@YlhfUWh]cb
  B]hfc[YbfY^YWh]cb
  <Y`]iafYWcjYfmUbX`]eiYZUWh]cb
Coil-wound and plate-fin heat exchangers
Coldboxes
 @B;ghcfU[YhUb_g
LNG loading facilities for ship, truck and rail
LCNG fuelling stations
LNG fuelling systems
Local storage and re-gasification
LNG tanker trailers for road and rail
Submerged combustion vaporizers
 @B;WUff]YfVc]`!cZZ[UgfY!`]eiYZUWh]cbgmghYag

Designing Processes - Constructing Plants.

www.ptqenquiry.com
Linde AG for further information
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

linde.indd 1 8/3/11 13:57:52
 Controlling corrosion in amine
treatment units
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
T
he amine unit plays a vital
role in the petroleum refining,
gas processing, coal gasifica-
tion and ammonia manufacturing
industries. With advances in hydro-
processing technologies to reduce
sulphur levels in gasoline and
diesel fuels, the requirement for an
efficient, well-established and relia-
ble separation system for the
removal of gases such as H2S and
CO2 has become an important
requirement for refiners. Amine
treatment has proved to be the
principal commercially established
method for gas/liquid purifications
by removal of H2S and CO2.
In many cases, the removal of
only H2S is required; CO2 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 H2S and CO2 gases from flue
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
flue gas or LPG comes into contact
with lean amine. The treated gas/
LPG goes for end use and the H2S-
rich amine then goes to the
regenerator column.
During this process, major prob-
lems of corrosion and instability of
operation raise a significant 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
www.eptq.com
bpcl.indd 1
Treated gas
Absorber
Feed
gas
Flash
drum
Rich
amine
Lean/rich
exchanger
Figure 1 Process flow diagram for amine treatment unit
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.
Significant 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 H2S and CO2 from fuel gas and
LPG with the help of amine solu-
tions, which tend to absorb both
gases. A process flow 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 H2S and
CO2. The treated gas/LPG goes for
end use and the H2S- and CO2-rich
Acid gas
Condenser
Reflux
drum
Stripper
Water
Sour condensate
NaOH Product
Waste
Lean
Lean amine
cooler
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
H2S and CO2 present in the amine
solutions according to the following
reactions:
R2NH3S R2NH + H2S (1)
(R2NH3) CO3 R2NH + CO2 + H2O (2)
where R is a CH2CH2OH 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 reflux. From the
GAS 2011 35
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):
Cooling
Lean amine + H2S + CO2 amine rich in H2S + CO2
Heating (3)
The amine solvent is the water
solution of an alkanolamine used to
remove the H2S, mercaptans and
CO2. 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 (NH3) 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-
Sweet gas
Lean solvent
Liquid
Natural
gas
Gas
Rich solvent
Figure 2 Schematic presentation of
absorption process
ucts. H2S formed during some
reactions also causes corrosion in
iron-based metallurgies by forming
iron sulphide (FeS). CO2 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:
H2CO3 + e- HCO3- +H (4)
H2CO3 CO2 + H2O (CO2 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 H2S
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 H2S,
nitrogen compounds and CO2 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
emulsification 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 + H2S 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-

Amine solvents denotes an acid anion such as
formate, acetate, propionate, glyco-
H2N
S.No. Amine Chemical structure
OH Molecular Base strength, late, oxalate, sulphate, thiosulphate,
solvent H2N H
N
weight pKa thiocyanate, phosphate, nitrate and
OH
Monoethanolamine (MEA)
HO OH 61.8 9.52 chloride. Alkalies such as NaOH
H
N
Diethanolamine (DEA) HO O OH
105.14 8.88 and KOH also react with strong
HO acids and contribute to the forma-
O
Diglycolamine (DGA) or 105.14 9.50 tion of HSAS.
aminoethoxyethanol HO H2N
Salts formed by either of these

HO H2N
N
OH reactions lower the pH of the solu-
Methyldiethanolamine (MDEA)
HO OH 119.16 8.56 tion and increase conductivity,
N
thereby making the system more
acidic and prone to corrosion.
Diisopropanolamine (DIPA) H 133.19 8.82
N This article deals with the differ-
HO OH
H ent types of corrosion mechanisms
N
HO OH in an amine treatment unit and the
Table 1 control strategies to deal with them,

36 GAS 2011 www.eptq.com

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 filming
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 H2S 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 fluid velocities on
corrosion rates
Corrosion rates vary with fluid
velocities and amine types. A fluid
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
www.eptq.com
bpcl.indd 3
Corrosion rates of various alloys with amine solvents
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
Table 2

7FMPDJUZ

NTFD
 7FMPDJUZ

NTFD

C or r os i on
r a t e ,
M P Y
7FMPDJUZ

NTFD
 7FMPDJUZ

NTFD





-FBO
BNJOF 3JDI
BNJOF
Amine types
Figure 3 Corrosion rates of rich and lean amine at various fluid velocities
loaded with acid gases, their various loadings of H2S. It was
tendency to cause erosion and observed that the corrosion rate
corrosion increases, so that, in the rises with an increasing H2S concen-
event of higher fluid velocities tration. Furthermore, corrosion
along with turbulence, the passive rates are greater at higher tempera-
layer of FeS becomes damaged, tures. At 120C, a corrosion rate of
which results in higher corrosion 40 mil/y was obtained with a H2S
rates. The recommendation is to loading of 5%. Hence, at higher
operate at lower velocities with rich temperatures, with greater percent-
amine solutions rather than lean ages of H2S, the use of a corrosion
amine solutions inhibitor is recommended.
Effect on corrosion rate of varying Effect of corrosion inhibitor
temperatures Corrosion inhibitors of the filming
The corrosion rates of carbon steel amine type were found to be the
were studied at temperatures vary- most suitable for reducing corrosion.
ing between 60C and 120C, with Concentrations of corrosion inhibitor

)4


)4


Cor rosi on
r at e ,
MPY

)4








  
Temperature, C
Figure 4 Corrosion rates at various H2S conc and temperature
GAS 2011 37
8/3/11 14:54:08

25
20
Corrosion rate, MPY
15
10
5 tendency
0 metals should be conducted every
Figure 5 Corrosion rates with different dosages of corrosion inhibitor
varying from 5 to 20 ppm were used
for the studies, along with DEA and
a 5% H2S 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 signifi-
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
specific 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 fluid 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
38 GAS 2011
bpcl.indd 4
Corrosion inhibitor (0 ppm)
Corrosion inhibitor (5 ppm)
Corrosion inhibitor (10 ppm)
60 90 120
Temperature, C
amine loading and increasing the
circulation rate
Circulation rate increases the corro-
sive action of HSAS, flashing acid gas
amine degradation products, high
temperature, acid gas loading and
high amine concentration.
Use of corrosion inhibitors
Injection of filming 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,
CO2, H2S 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 reflux and then is
distributed throughout the unit
Water wash the overhead
condenser with a reflux slipstream,
routinely purge the ammonia in the
reflux 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
Rigorous analysis for acids and
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 reflux
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 five 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 filed over 50 patents.
Email: choudarynv@bharatpetroleum.in
www.eptq.com
8/3/11 14:54:15
 A promoter for selective H2S
removal: part I
A new MDEA promoter achieves very low H2S lean loadings and the option for
flexible 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 Raffineriegesellschaft
T
he selective removal of H2S
has become an important
topic over past decades. This
is driven by several factors, one
being the production of an H2S-
enriched, and thus high-quality,
Claus gas in MDEA-based acid gas
enrichment (AGE) units.
Other fields where selective gas
treatment would also be beneficial
are natural gas and refinery appli-
cations. For sour gas fields, this
application is becoming increas-
ingly attractive due to limited sweet
gas resources. For refineries, debot-
tlenecking issues and an increased
flexibility 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
benefits of enhanced selective
treatment. In addition, tight envi-
ronmental regulations and sulphur
specifications are attributed to this
subject.
The principles for the selective
removal of H2S 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 CO2 and H2S mass
transfer kinetics
Promoted tertiary amines, focus-
ing more on enhanced regeneration
and thus leading to lower H2S
loadings.
With respect to the final point,
the advanced promoter system
presented in this article can be a
considerable leap forward for more
flexible selective designs. Very low,
www.eptq.com
basf.indd 1
achievable lean loadings are an where simultaneous removal of H2S
option to adjust selectivity by vary- and CO2 is intended. This means
ing the absorber height without that suitable solvents have different
losing control over a tight H2S spec- characteristics.
ification with a sufficient safety
margin. H2S selectivity
This article gives an overview of Definitions
the principles, while part II (see In industry, various expressions for
PTQ, Q2 2011) shows the ability of H2S selectivity are used to judge the
this promoter system to drop H2S selectivity of an absorption process
lean loadings in a refinery amine with regard to H2S compared to
system. CO2. In the following, the most
Acid gas removal with amine- important examples are given (cH2S
based solvents is a mature and and cCO2 stand for molar concen-
widespread application in the oil tration in gas streams):
and gas industry. Besides specific
design variations, acid gas removal H2S selectivity (rigorous definition)
units (AGRU) always follow the H2S selectivity is calculated as:
principle of an absorber-regenerator
configuration. First, acid gases are H S_selectivity = (CH2S feed gas - CH2Streated gas)/CH2Sfeed gas
removed from the fluid stream in
2
(CCO2 feed gas - CCO2treated gas)/CCO2feed gas
the absorber by the liquid solvent
typically at 2050C and elevated CO2 co-absorption (or CO2
pressures up to 80 bar, depending pick-up)
on the feed gas conditions. In a CO2 co-absorption in the absorber
second step, the dissolved acid is specified as:
gases are desorbed in a regenerator
at inverse conditions. Desorbed C treated gas
CO2_co-absorption = 1- CO2
acid gases can be further processed CCO2 feed gas
in Claus sulphur recovery units
(SRU), reinjected for enhanced oil CO2/H2S ratio comparison (often
recovery (EOR) or simply flared. used for acid gas enrichment)
When focusing on the two main The decrease in CO2/H2S ratio indi-
acid gas components, CO2 and H2S, cates the efficiency of an acid gas
process designers differentiate enrichment unit:
between total acid gas removal and
selective sulphur removal or simply Ratio C feed gas
= CO2
C acid offgas
vs Ratioacid offgas CO2
feed gas
selective removal. As the name CH2S
feed gas CH2S acid offgas
implies, selective sulphur removal
selectively removes H2S, while A proper comparison of these
other acid gases, for instance CO2, expressions for different solvents or
are slipped into the treated gas. applications must be evaluated with
Consequently, selective removal great care, since the absolute level
has a different focus compared with of H2S in the treated gas is less
total acid gas removal processes, considered. In other words, H2S
such as BASFs aMDEA process, selectivity is well defined, but not
GAS 2011 39
9/3/11 12:35:20

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
BORSIG GROUP
Egellsstrasse 21
D-13507 Berlin/Germany
Phone: +49 (30) 4301-01
E-mail: info@borsig.de
basf.indd 2
particularly suitable for judging H2S specifications in
treated gas.
BORSIG
The following example, calculated according to the
rigorous definition, might show the difference:
Feed gas: H2S 1 v% CO2 5 v%
Treated gas case 1: CO2 1.5 v% H2S 1 vppm H2S selectivity = 1.428
Treated gas case 2: CO2 1.5 v% H2S 50 vppm H2S selectivity = 1.421
Dependency on operating conditions
Selective treatment with amine-based solvents gener-
ally takes advantage of the rapid reaction of H2S
compared to the kinetically hindered reaction of CO2;
CO2 first 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 CO2.
These are the reactions of tertiary amines in aqueous
solutions:
Reaction of water and amine (fast)
_
R1R2R3N + H2O R1R2R3NH+ + OH
_
2 H2O H3O + OH
+
H2S reaction (fast)
_
H2S + H2O HS + H3O+
CO2 reactions (overall reaction slow):
_
CO2 + 2H2O HCO3 + H3O+ (slow)
_ _
HCO3 + OH H2O + CO32- (fast)
In this reaction system, CO2 co-absorption, and thus
H2S selectivity, is heavily influenced by reaction condi-
tions such as the temperature and concentration of CO2
and H2S in the liquid phase. This means higher pressure
and temperature, as well as a higher CO2/H2S ratio in
the feed gas, favour CO2 co-absorption and lower H2S
selectivity.
Applications
Selective H2S removal is predominantly required and
applied in the following segments:
Natural gas (NG)
Refinery 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 H2S selectivity in order to get a H2S-rich
feed gas for the downstream Claus SRU or for down-
hole reinjection. For instance, a minimum H2S
concentration of 25 vol% is a guiding value for SRUs.
Higher H2S concentrations are naturally more desirable.
Upgrading and debottlenecking measures triggered
the move to more selective treatment in refineries by
replacing unselective with selective generic solvents. In
fact, the majority of refinery expansions lead to a higher
total sulphur nameplate capacity by processing more
sour crudes. Without significant mechanical changes,
capacity increases of more than 25% have been achieved,
as mainly H2S, instead of CO2 and H2S, has been proc-
Fax: +49 (30) 4301-2236
www.borsig.de essed along the entire acid gas removal chain. Moreover,
www.ptqenquiry.com
for further information www.eptq.com
9/3/11 12:35:56
 tighter sulphur regulations in many
countries have been fulfilled in
combination with improved sulphur
recovery rates by implementing
Claus tail gas treatment (TGT),
where H2S 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
Characterisation of H2S selective applications
Natural gas Acid gas enrichment Refinery offgas Tail gas
Typical feed gas specification
pH2S, mbar/psi Up to 10 000/145 Below 500/7.25 Up to 2000/29 Below 100/1.45
pCO2, mbar/psi Up to 10 000/145 Below 2000/29 Up to 1000/14.5 Below 500/7.25
Typical treated gas specification
cH2S, vppm 510 5100 5100 550
CO2 slip Flexible Very high Medium High
System hold-up Large Medium Medium Small
p = partial pressure/c = concentration

utilisation of more sour gas fields is

increasing rapidly, while the access
to sweet natural gas resources
becomes more and more limited
(sour in colloquial terms means H2S
>5 vol%). On the other hand, for
sales gas specification, operators
request not the highest level but a
dedicated or even an adjustable
level of selectivity. This means that
CO2 slip, and thus H2S selectivity,
needs to be controlled to match, for
instance, a certain calorific value for
the sales gas. CO2 treated gas speci-
fications 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 specifica-
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 fields and their major
differences in conditions and
requirements. The listed criteria are
crucial for the solvent as well as for
design selection.
Solvents for H2S 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
beneficial properties, in particular
Table 1
the inability to enter into the
carbamate reaction with CO2. 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 CO2.
Promoted MDEA
H2S is a stronger acid than CO2 and
thus has a stronger binding energy
at the stripper temperature.
Consequently, amine regeneration
of H2S down to trace levels becomes
progressively more difficult. In
particular, designs with very low
H2S specifications in the low ppm
range may use regeneration
promoters in order to release H2S
more easily from the solution. In
this regard, MDEA acidified with
phosphoric, sulphuric or other acids
becomes regenerated more easily.
Even the unintentional acidification
by formation of heat stable salts
(HSS) and their acids acetates,
formates, sulphates and so on
leads to a very efficient H2S
regeneration boost. However, the
level of acidification 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 CO2 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
H2S partial pressures, such as acid
gas enrichment, hindered amines
are better positioned and highly
efficient. Moreover, the relatively
high price level of hindered amines
makes it more economical in small
plants (AGE) than in large (natural
gas) plants.
Design specifics for selective
treatment
For the design, various aspects have
to be taken into account:
H2S specification in the treated
gas
CO2 specification/slip in the
treated gas
Feed gas conditions (composition,
pressure and temperature), which
may vary over the plants lifetime
Required CO2/H2S 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:
CO2 reaction kinetics and number
of theoretical stages, mass transfer

www.eptq.com GAS 2011 41

basf.indd 3 9/3/11 12:36:11

 and internals selection, and temper-
ature and selectivity.
CO2 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 CO2 absorption is
reduced to a certain extent.
However if the absorber height is
too short, H2S may not be absorbed
sufficiently. In todays selective
designs, very tight H2S specifica-
AdvAmine & AdvaSulf
your best tools
in sour gas treatment
Prosernat offers
a global range
of advanced technical solutions
for sour gas treatment
and sulphur recovery
Tour FRANKLIN - 100/101 TERRASSE BOIELDIEU
F-92042 Paris La Dfense Cedex - France
Phone: + 33 1 47 67 20 00 - Fax: + 33 1 47 67 20 07
e-mail: sales@prosernat.com
http://www.prosernat.com
42 GAS 2011
basf.indd 4
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 flexible 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
AdvAmine
HiLoadDEA
MDEAmax
energizedMDEA
AdvaSulf
Claus
Tail Gas Treatment, Clauspol II,
Sulfreen, Sultimate
Sulphur degassing, Aquisulf
www.ptqenquiry.com
for further information
force for the mass transfer at the
absorber overhead.
Mass transfer and internals selection
The pathways of mass transfer for
CO2 and H2S are quite different.
While CO2 absorption is affected by
mass transfer limitations in the
liquid phase, the instantaneous H2S
reaction in the solvent moves the
major H2S 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 flow 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 flexibility 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 CO2 co-absorp-
tion and resulting selectivity, but the
influence 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, CO2 at a high partial pres-
sure will be absorbed more easily
in the amine solution. In addition,
the heat of CO2 and H2S absorption
expedites the formation of an
www.eptq.com
9/3/11 12:36:25
 adverse temperature profile in the
absorber (see Figure 1). The contin- 50
uing acceleration of this
45

Absorber height, segments

phenomenon has a negative influ-
ence on overall gas capacity and
H2S selectivity. Besides the specific
heat of absorption and the amount
of absorbed CO2 and H2S, further
parameters, such as the specific
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 CO2 and H2S
start to increase the complexity of
the system. It is therefore important
to understand the temperature
profile 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 flexibility for selective natu-
ral gas plants is driven by:
Variation in absorber height
through multiple amine feed points
Low lean amine loading to
achieve H2S specification with suffi-
cient safety margin
Flexibility in circulation rate to
manage the temperature profile in
the absorber and control CO2 mass
transfer
Sufficient reboiler energy to keep
the lean loading low at various
circulation rates.
Promoted MDEA solvent for natural
gas plants
Most natural gas plants have CO2
partial pressures >1.5 bara (>21.8
40
35
30
25
20
15
10
5
0
40
Figure 1 Absorber, liquid-phase temperature profiles red: feasible T-profile at
appropriate L/G ratio; blue: T-bulge profile at too low L/G ratio
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 profile 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 flow
rate. Temperature bulge effects in
45 50 55 60 65 70 75 80 85
Temperature, C
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
refineries
The AGRU setup of a refinery is
very complex and the different
requirements of the absorbers prefer
an all-fits-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

Absorber Desorber
6 Top 10
9
5 Acidified MDEA
8 MDEA
MDEA
7
4 Acidified MDEA
6
3 5
4
2
3
2
1 High acid gas Low acid gas
concentration 1 concentration
0 0
7.5 8.0 8.5 9.0 9.5 10.0 10.5 11.0 11.5 Bottom 7.5 8.0 8.5 9.0 9.5
pH pH

Figure 2 Example of pH profiles in an absorber (left) and a desorber (right) with generic MDEA and acidified MDEA

www.eptq.com GAS 2011 43

basf.indd 5 9/3/11 12:36:36

 120C
H2S partial pressure

MDEA
MDEA + promoter
Promoted MDEA

50C
Stripper
sump
Absorber
head

H2S loading

Figure 3 Influence of the acidic promoter formulation on the isothermal equilibrium curves at 50C and 120C (H2S partial pressure
regime of 1 mbar)

_
today in more than 40% of all refin- loadings. In this respect, the HS anion, described in the follow-
ery amine systems. In addition, the promoter formulation provides an ing reaction:
relatively high refill demand comes option to design selective units with
H2S + MDEA HS- + MDEAH+
along with pricing and availability. a high degree of flexibility by vary-
With respect to regeneration energy, ing absorber heights, while low CO2 desorption can be described
the unintentional and sometimes acid gas lean loadings keep H2S by the following net equilibrium
quick formation of HSS to concen- outlet concentrations at a very low reaction:
trations of up to 3 wt% is supportive level.
-
to some extent. However, corrosion The following describes the mech- CO2 + MDEA + H2O HCO3 + MDEAH+
issues force refiners to neutralise anism of achieving low acid gas
those acids with caustic solution. lean loadings. The purpose of desorption is to
strip H2S and CO2 from the rich
New-generation promoter system Principles solution to obtain a regenerated
A new promoter has been devel- Desorption mechanism in solvent. As a consequence of CO2
oped and incorporated: BASFs the stripper and H2S mass transfer from the
selective sMDEA+ technology. This H2S desorption follows an equilib- liquid phase into the gas phase, the
promoter formulation is highly effi- rium reaction between amine, above two equilibrium reactions
cient in achieving very low H2S lean protonated amine and the dissolved move towards the left side and the
pH value of the solution increases
accordingly from the top to the
bottom of the stripper.
MDEA 0.5 0.0060
H2S lean loading, mol/mol

Since CO2 is a weaker acid

H2S lean loading, Nm3/to

Acidified MDEA
sMDEA+ 0.4 0.0048 at typical stripper temperatures
compared with H2S, CO2 is prefera-
25 0.3 0.0036 0.30 bly released in the upper part of the
stripper, whereas a significant
H2S lean loading, Nm3/to

H2S lean loading, Nm3/to

0.2 0.0024
20 0.24 amount of H2S as the stronger
0.1 0.0012 acid is released in the lower part
of the stripper. Releasing CO2 and
15 0.18
0.0 0.0000 obtaining low residual CO2 concen-
Regeneration energy trations becomes easier than
10 0.12 achieving low H2S residual loading,
which requires much more strip-
5 0.06 ping stream and reboiler energy.
Low acid gas concentrations in
the lean solvent are directly linked
0 0.00
Regeneration energy to the maximum achievable gas
purities in the absorber overhead:
unfortunately, most gases have very
Figure 4 H2S regeneration lab tests (reduced desorber height) tight H2S specifications for

44 GAS 2011 www.eptq.com

basf.indd 6 9/3/11 14:46:11

 0HVKQXPHURXVDOOR\V SODVWLFV :LGHYDULHW\RIUDQGRPSDFNLQJ &DWDO\VWEHGVXSSRUWV%DVNHWVWUDLQHUV
W\SHVVL]HVDQGPDWHULDOVLQ
9DQH&KHYURQ'RXEOH3RFNHW9DQH 1R]]OHV2XWOHW,QOHWEDVNHWV
VWRFN$VNXVKRZRXUSDWHQWHG
&DQGOHRU)LEHUEHG,QVHUWLRQ0LVW[ 6XSHU%OHQG3DFFDQLQFUHDVH 'LVWULEXWRUV+XEDQGKHDGHUODWHUDOV
\RXUFDSDFLW\DQGHIFLHQF\

MIST ELIMINATORS RANDOM PACKING PROFILE WIRE SCREEN

From Engineering Through Fabrication

Valve (numerous options) For top performing mass transfer and tower internal components let AMISTCO
Sieve or perforated
Separation Products use your existing drawings or modify them to improve your
Bubble cap trays
Cartridge trays process. Our experienced in-house engineering and full scale fabrication
'XDORZ capabilities can streamline your project from inception to installation.
%DIH

A Member of Fractionation Research, Inc.

ZZZDPLVWFRFRPKU(0(5*(1&<6(59,&(

www.ptqenquiry.com
for further information

DISTRIBUTORS & SUPPORTS COALESCERS STRUCTURED PACKING

0DQXIDFWXUHGWRFXVWRPHU 2LOZDWHUVHSDUDWLRQV+D]HUHPRYDO :RYHQVKHHWPHWDODQGNQLWWHG
VSHFLFDWLRQVRUHQJLQHHUHGWR IURPIXHOV5HPRYDORIWRZHUZHW VWUXFWXUHGSDFNLQJ%XLOWWRVSHF
PHHWSHUIRUPDQFHUHTXLUHPHQWV UHX[&DXVWLFWUHDWHUDSSOLFDWLRQV RUSHUIRUPDQFHUHTXLUHPHQW

amistco.indd 1 8/3/11 14:02:35




Absor be r
h e i gh t ,
m
Figure 5 illustrates the corre-
sponding H2S concentration profile
 in the gas phase for a 10m packed-
 bed absorber with a feed gas partial

pressure for CO2 of ~2 bar (29 psi)
and H2S ~1 bar (14.5 psi) at 40C

(104F). As a matter of equilibrium,
 the respective H2S specification for
 MDEA is in the low hundreds
 ppmv range, while for acidified
 .%&" MDEA the H2S specification would

"DJEJFE
.%&" be around a tenth of this. In the

&
Figure 5 Absorber H2S profile [0]
instance, 510 vppm for natural gas;
see Table 1 but much looser CO2
specifications. However, this
requirement is in opposition to the
chemistry of H2S desorption,
described above.
Another effect that hampers H2S
desorption is related to the pH
value increasing steadily towards
the stripper sump. As a conse-
quence, the ratio between
protonated HS- and H2S 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 - pKaH2S
[H2S]
Considering these issues, it
becomes obvious how energy inten-
sive the purification of gases down
to very low H2S 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 H2S loading can be obtained by
shifting the equilibrium of the
reaction
_
H2S + MDEA HS + MDEAH+
towards the H2S side, by adding a
strong acid (AH) to the system:
_ -
HS + AH H2S + A
46 GAS 2011
basf.indd 7
T.%&" sMDEA+ case, the ultra-low lean
& & & & & &
H2 S concentration, mol%
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,
acidification is limited to some
extent, so acidified MDEA solvents
may require an increase in liquid
flow 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 specifications
Figure 4 shows the graph of a
regeneration lab test run with
MDEA, acidified MDEA (H3PO4)
and MDEA with the new promoter
system (sMDEA+ technology), with
the residual H2S lean loading
(Nm3/t, mol/mol) against regenera-
tion energy, Q. In the important
regime of very low loadings (<0.5
Nm3/t, <6E-3 mol/mol), a magni-
fied graph illustrates the difference.
loading in the 1E-4 mol/mol range
is able to achieve a treated gas
specification below 5 ppmv, with
sufficient safety margin and a negli-
gible impact on the H2S absorption
capacity at the absorber bottom.
To reach a similar low lean load-
ing with a state-of-the-art acidified
MDEA, a doubling of the H3PO4
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, H2S selective acid gas
removal has become an increasingly
important field for world-scale
natural gas plant designs. Contrary
to acid gas enrichment techniques,
selectivity and thus CO2 slip are
limited and require some adjusta-
bility and flexibility.
From a solvent technology
perspective, advanced H2S 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 flexibility in turndown
rates, changing feed gas specifica-
tions and conditions. In addition,
very tightly treated gas sulphur
specifications 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
www.eptq.com
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 H2S
binding energy and to lower resid-
ual H2S 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 H2S utilizing generic MDEA, 37th Annual
Laurance Reid Gas Conditioning Conference,
Norman, Oklahoma, 1987.
3 Carey T R, Hermes J E, Rochelle G T, A
THE RVP PROFIT ANALYZER
MINIVAP ON-LINE
VP of gasoline, crude oil and LPG
ASTM D5191, 5188, 6377, 6378, 6897, EN 13016-1+2,
IP394, 406, 481
Highest accuracy for best blending to official limits
Up to 2 sample streams
Automatic calibration
Fast and easy maintenance
Automatic lubrication NEW!
VP of
Grabner Instruments Messtechnik GmbH / A-1220 Vienna/Austria / Dr. Otto-Neurath-Gasse 1
Phone +43/1/282 16 27-0 / Fax +43/1/280 73 34 / info.grabner-instruments@ametek.at
ON-LINE half page.indd 1
www.eptq.com
basf.indd 8
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
www.ptqenquiry.com
for further information
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 CO2 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
03.03.2011 17:15:05
GAS 2011 47
9/3/11 12:37:10
 Estimating gas hydrate inhibitor loss:
a case study
A predictive tool is used to understand and predict the loss of methanol during
natural gas hydrate inhibition

Alireza Bahadori
Curtin University

T
he amount of methanol to be
injected as a hydrate inhibitor
must not only be sufficient 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
106 m3/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 flow, which leads
to safety hazards and substantial
economic risks.1,2
Tuned coefficients for the prediction of methanol vapourisation loss
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 significant
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 significant 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 coefficients, 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 significant impor-
tance to natural gas engineers.

Coefficient Values for pressure Values for pressure between Methanol vapourisation loss during
less than 6000 kPa 6000 and 20000 kPa
gas hydrate inhibition
A1 4.5112124 x 104 5.32442401 x 103
B1 -6.5625532 x 108 3.64817009 x 107 Equation 1 shows a definition of
C1 2.7486759 x 1012 -1.43519141 x 1012 methanol vapourisation loss:
D1 -3.26326963 x 1015 5.41358084 x 1015 methanol vapour composition to
A2 -3.6813711 x107 -4.71938915 x 106
methanol liquid composition.
B2 5.3777095 x 1011 -2.60914887 x 1010
C2 -2.25654440 x 1015 1.17661977 x 1015 Equation 2 is an Arrhenius-type
D2 2.68069426 x 1018 -4.48418384 x 1018 function to correlate methanol
A3 1.00121117 x 1010 1.39340046 x 109 vapourisation loss as a function of
B3 -1.46856806 x 1014 6.012666228 x 1012
temperature (K), wherein the
C3 6.17387798 x 1017 -3.21209807 x 1017
D3 -7.33914654 x 1020 1.237524600 x 1021 relevant coefficients (see Table 1)
A4 -9.07444577 x 1011 -1.369737341 x 1011 are correlated as a function of
B4 1.3364870 x 1016 -4.38900897 x 1014 pressure in kPa(abs) (Equations
C4 -5.6294519 x 1019 2.919844093 x 1019
36).
D4 6.696496249 x 1022 -1.137898262 x 1023
kg methanol (1)
LM = Million standard m3 gas
Table 1 Mass% methanol in water phase

48 GAS 2011 www.eptq.com

curtin.indd 1 9/3/11 12:49:34


Tuned coefficients for the prediction of
methanol loss in the condensate phase
Coefficient Value
A1 1.6812339429 x 103
B1 -1.0803155907 x 104
C1 2.2311693912 x 104
D1 -1.4729301398 x 104
A2 -1.564777661 x 106
B2 9.9874920671 x 106
C2 -2.0506540953 x 107
D2 1.3479263861 x 107
A3 4.8553968051 x 108
B3 -3.0835714347 x 109
C3 6.2988799656 x 109
D3 -4.1248186297 x 109
A4 -5.0424809928 x 1010
B4 3.1810877336 x 1011
C4 -6.4667402344 x 1011
D4 4.220547048 x 1011
Table 2
b c d
ln(LM) = a + + 2+ 3
T T T (2)
Where:
B C D
a = A1 + 1 + 21 + 31
P P P (3)
B2 C2 D2
b = A2 + + +
P P2 P3 (4)
B3 C3 D3
c = A3 + + +
P P2 P3 (5)
B4 C4 D4
d = A4 + + +
P P2 P3 (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 Our new clean-fuels
phase during gas hydrate inhibition plant is straining the
Equation 7 represents the proposed
governing equation in which four
auxiliary units!
coefficients 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 coefficients are shown in How can we boost their capacity
Table 2: without major construction?
TO MEET CLEAN FUELS require-
b c d
ln() = a1 + + 2 + 3 ments for gasoline, a medium-sized
T T T (7) 0LGZHVWHUQUHQHU\DGGHGDORZ
Where: VXOIXUIXHOVWHFKQRORJ\SODQW*DVROLQH
a = A1 + B1 + C12 + D13 (8) WKURXJKSXWZDVXQFKDQJHG+RZHYHU
WKHVKDUSLQFUHDVHLQVXOIXUUHPRYDO
b = A2 + B2 + C22 + D23 (9)
c = A3 + B3 + C32 + D33 (10) UHTXLUHGPRUHK\GURJHQIURPWKH
d = A4 + B4 + C42 + D43 (11) K\GURJHQXQLWDQGVHQWPRUHVXOIXU
gases to the amine treaters and
GRZQVWUHDPVXOIXUXQLWV7KHVH
These optimum tuned coefficients DX[LOLDU\XQLWVEHFDPHERWWOHQHFNV
RYHUGULYHQDWWKHFRVWRISURGXFW
help to cover temperatures in the
SXULW\DQGDPLQHFRQVXPSWLRQ
range 240320K and methanol
concentrations up to 0.70 mass
fraction in the aqueous phase.
The optimum tuned coefficients
shown in Table 2 can be further
retuned quickly according to
the proposed approach1,2 if

Comparison with reported data3 of predicted results of methanol vapourisation loss

Temp, Pressure, Reported methanol Calculated methanol Percentage

K kPa vapourisation loss3 vapourisation loss absolute 2QVWXG\LQJWKHK\GURJHQDPLQH
kg methanol kg methanol deviation DQGVXOIXUXQLWV$PLVWFRIRXQGPDQ\
Million standard m3 gas LM = Million standard m3 gas RSSRUWXQLWLHVIRULPSURYLQJVHSDUDWLRQ
Mass% methanol in water phase Mass% methanol in water phase HIFLHQF\DQGFDSDFLW\7KHSUREOHPV
261.15 2000 11 10.916 0.76 ZHUHVROYHGZLWKRXWPDMRUFRQVWUXFWLRQ
277.15 2000 32.7 32.217 1.47 E\DSSO\LQJPRGHUQWHFKQRORJ\WRPLVW
269.15 3000 15.9 15.229 4.22 HOLPLQDWRUVOLTXLGOLTXLGVHSDUDWRUV
281.15 3000 31.2 31.605 1.29 DQGWRZHUWUD\VDQGSDFNLQJ5HVXOWV
271.15 4000 14.6 14.639 0.26 LQFOXGHGKD]HIUHHSURGXFWDQG
283.15 4000 29.85 30.232 1.27 UHGXFHGDPLQHFRQVXPSWLRQ1RZD
273.15 5000 14.6 14.402 1.35 GLHVHOFOHDQIXHOVSODQWLVEHLQJDGGHG
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
Read more on this topic at
277.15 7000 27.8 27.605 0.70 www.amistco.com
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
3KRQH)D[
Average absolute deviation percent (AADP) 1.07% amistco@amistco.com
KU(PHUJHQF\6HUYLFH
Table 3

www.ptqenquiry.com
www.eptq.com for further information GAS 2011 49

curtin.indd 2 9/3/11 12:49:43


0.025
hydrocarbon phase, mole fraction
Dissolved methanol in liquid
more data become available in the
future.
Methanol mass fraction = 0.60 It is expected that efforts during

Methanol mass fraction = 0.02 this investigation will pave the way
0.020 to an accurate prediction of metha-
nol loss in the liquid hydrocarbon
0.015 phase () in mole fraction, as well
as methanol vapourisation loss in
various conditions, which can be
0.010
used by engineers and scientists for
monitoring the key parameters
0.005 periodically.

0.000 Results
220 240 250 260 270 280 290 300 310 320 Table 3 presents the summary of
Temperature, K accuracies with the proposed
predictive tools in terms of average
Figure 1 Performance of proposed tool for prediction of solubility of methanol in the absolute deviation percentage with
hydrocarbon condensate phase 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-
Solubility of methanol in condensate

102
tion from reliable data. Figures 1
hydrocarbon liquid, mole fraction

and 2 show the proposed methods

104 results and performance in the
prediction of methanol solubility in
106 the hydrocarbon liquid phase for a
wide range of conditions from two
108
different viewpoints. They show
that high temperatures and more
injected methanol in the aqueous
1010 Temperature = +50C
Temperature = 15C
phase lead to higher solubility of
methanol in the liquid phase. Table
1012 4 illustrates the accuracy of the
102 101 proposed correlation in predicting
Methanol mass fraction in water phase the solubility of methanol in
paraffinic hydrocarbons in compari-
Figure 2 Performance of proposed tool for prediction of solubility of methanol in son with some reported data.3 The
hydrocarbon condensate phase from another view point accuracy of correlation in terms of
average absolute deviations
is 2.12%.
Comparison with reported data3 of calculated solubility of methanol in hydrocarbons The following is a case study
showing how the information
Temp, Methanol mass Calculated solubility Reported solubility Absolute evolving out of a predictive tool
C fraction of methanol in of methanol in deviation, can be used to understand and
hydrocarbons, hydrocarbons, % predict the loss of methanol during
mole fraction mole fraction3 natural gas hydrate inhibition.
12 0.20 0.00099 0.001 1
30 0.20 0.00294 0.003 2 According to these calculations,
50 0.20 0.006 0.006 0 $3701/day is the cost of methanol
21 0.35 0.0044 0.004 1 loss, which is calculated by
47 0.35 0.01035 0.01 3.5 this simple predictive tool for
-10 0.60 0.00295 0.003 1.6
2 0.60 0.00514 0.005 2.8 3 x 106 m3/d of natural gas leaving
43 0.60 0.02034 0.02 1.7 an offshore platform.
-40 0.70 0.0021 0.002 5
0 0.70 0.0073 0.0071 2.8 Case study
50 0.70 0.0265 0.027 1.9
As indicated, 3 x 106 m3/d of natu-
Average absolute deviation, % 2.1 ral gas leaves an offshore platform
at 38C and 8300 kPa (abs), and the
water content is 850 mg/Sm3. The
Table 4 gas comes onshore at 4C and 6200

50 GAS 2011 www.eptq.com

curtin.indd 3 9/3/11 12:49:54

 kPa (abs), and water content is 152 Losses = 14.58 x 10-6 (kg/m3)/(wt%
f3 x 10 Sm p f56m p f778kgp f1kgmolep = 933
6 3 3

mg/Sm
slip
3.
(minimum The hydrate
removal) temperature
to meet methanol) day 10 Sm3 m3
6
140 kg kg
of theAmerican
North gas is treating
18C. Associated
specica- Daily vapourisationNew losses
Mexico = 14.58 x mol/day.

Spray Regime Correction Factor

condensate
tions. Due toproduction
the small amountis 56mof 3
/ 10-6 (3 x 106)(27.5) = 1203 kg/day.
(million
acid gas standard
removal, m the). absorbers
3
The conden- in 1203 kg/day ($3/kg) = $3609/day Kg mol methanol = 933(0.0011) =
sate has
these plantsa density of 778 kg/m
were operated and
at small is the cost of vapourised methanol 1.026 moles/day.
a molecular
liquid rates and massvery of low140. weir
The calculated by this predictive
Iowa tool. Kg methanol = (1.026) (32) = 32.84
required
loads. methanol inhibitor
Simulations fell far shortconcen-
of kg/day.
tration in the
agreeing withwater thephaseeldto data.
avoid Solution for methanol loss in The methanol in the condensate
hydrate
Eventually, formation
it was realised is that27.5%.
the condensate phase phase can be recovered by down-
Calculate were
absorbers the mass wellrate
intoofthe inhibitor
spray Win = 850 mg/Sm3 stream water washing.
regime
loss inand theProTreats
water phase mass transfer
and the Wout = 152 mg/Sm3 Alberta Cost per month = ($3/kg) (30.72
North Texas
coef
amountcientofcorrelations
methanol were loss in inap-
the W = 698 mg/Sm3 North Texas India
kg/day)(30 day/month)
Base
= $2956
plicable North Texas
hydrocarbon liquid phase, aswere
because they well Water condensed = (3 x 10 ) (698) = per
6
month.
North Texas
developed
as the amount fromofdata on trays
methanol oper-
vapouri- 2094 x 106 mg/day = 2094 kg/day The total cost of methanol loss per
ating
sation loss. Consider the priceTo
in the froth regime. of Calculate mass rate of inhibitor in Weir day
Load will become 92 + 3609 =
reproduce
methanol asthe performance
$3 per kg. data, the water phase: $3701/day.
the vapour-side mass transfer coef- ?b`nk^+Ahpma^\hkk^\mbhg_Z\mhk_hklikZrk^`bf^hi^kZmbhgoZkb^lpbmap^bkebjnb]ehZ]
Solution
cients had to be increased by
for methanol mi =
0.275(2094)
= 794 kg/day References
upwards
vapourisation factor
of a loss of 10 or more, from 1plants
0.275in Canada, the US and multiplying
1 Bahadori A,the gas-side
Vuthaluru H B,mass
Energytrans-
& Fuels,
and the liquid-side
X = 0.275 mass fraction coefofcients had
methanol India and span the full
Estimate losses to hydrocarbonspectrum of fer
24, coef
2010, cient
29993002. (the same factor
to be decreased
T = 277.15K correspondingly. weir loads
liquid phase from
fromabout 2.2 m3method
proposed /mhr divided theA,liquid
2 Bahadori coef
Vuthaluru H B,cient)
Journaluntil
of Loss
Since then, six more commercial (3atgpm/ft) to 45 m 3
/mhr (60 gpm/ Prevention in the
agreement Process
with theIndustries, 23, 3, 2010,
performance
For vapourisation loss of methanol, 4C and 27.5 wt% MeOH:
tower performance data sets with ft). 379384.
we use the tuned coefficients a =Each plant was xsimulated
9.136448034367 101 rst data was obtained.
3 GPSA Engineering Data Book, 12th ed, 2004,
absorbers
reported inatTable low 1:weir load have using ProTreats database
b = -8.869793699719 x 10 4 correla- Figure 2 shows how the results
Gas Processors Suppliers Assoc, Tulsa, USA.
come to light.
a = -1.6454503534842 The only non-propri-
x 10 3 tions for tray mass
c = 2.8126995933983 x 107 transfer compare with each other and how
etary set of measurements
b =1.3743444476865 x 10 6 is from a coef
d = -3.0722013379 x 109 using the
cients (froth regime), the correction alters with changing
Alireza Bahadori works in the Chemical
plant operated
c = -3.8051922967936 by
x 108 Signalta owsheet shown of in
The solubility Figurein the
methanol 1. weir load. Department
Engineering The plot shows
of CurtinaUniversity
truly
Resources
d = 3.4980770983675 x 1010 Alberta,
near Forestburg, Simulations were repeated for
hydrocarbon phase is estimated tovari- remarkable degree of correlation,
of Technology, Perth, Western Australia.
Canada.
LM = 14.58These eight data sets are
4
ous values 0.0011
be around of a orconstant factor,
0.11 mol%: especially in view of the fact that
Email:Alireza.bahadori@postgrad.curtin.edu.au

So Advanced, Commercially Proven Route for The

Yet So Simple. Production of Clean Gasoline from Coal
through Coal Gasification, Methanol
Synthesis, and Methanol Conversion
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
6MUSB
DMFBO
IJHI
PDUBOF
HBTPMJOF

t
"
QSPEVDU
GVMMZ
DPNQBUJCMF
XJUI
SFOFSZ
HBTPMJOF

t
4JNQMF
SFMJBCMF
YFE
CFE
SFBDUPS
EFTJHO

t
"
NPSF
FDPOPNJDBMMZ
DPNQFUJUJWF
PGGFSJOH

3225 Gallows Road

Fairfax, Virginia 22037-001, USA
www.exxonmobil.com/refiningtechnologies
+1-703-846-2568 fax +1-703-846-3872
technology.licensing@exxonmobil.com

www.ptqenquiry.com
ppp'imj^gjnbkr'\hf
www.eptq.com for further information
_hk_nkma^kbg_hkfZmbhg GAS 2011 51
ppp'^imj'\hf  IMJJ-+)*)2.

curtin.indd3 4
ogt.indd 9/3/11
14/9/10 12:50:05
09:40:04
 Alphabetical list of advertisers

Amistco Separation Products 45, 49 Haldor Topse IFC

www.amistco.com www.topsoe.com

Axens 6 Linde 34
www.axens.net www.linde.com

Black & Veatch 18 Lurgi 4

www.bv.com www.lurgi.com

Borsig 40 Merichem Company 23

www.borsig.de www.merichem.com

Bryan Research & Engineering 15 MPR Services 17, OBC

www.bre.com www.mprserve.com

Burckhardt Compression 23, 29 NPRA IBC

www.burckhardtcompression.com www.npra.org

Cansolv Technologies 24 Prosernat 42

www.cansolv.com www.prosernat.com

ExxonMobil Research and Engineering Company 51 Siirtec Nigi 21

www.exxonmobil.com/refiningtechnologies www.sini.it

Grabner Instruments 47 WorleyParsons 20

www.grabner-instruments.com www.worleyparsons.com

For more information on these advertisers, go to www.ptqenquiry.com

52 GAS 2011 www.eptq.com

ad index copy 7.indd 1 9/3/11 15:00:06

 Meet the Right People
for Business Success

Join colleagues and customers at

NPRAs industry-leading conferences.
NPRA 2011 Conference Schedule
Annual Meeting International Petrochemical Conference
March 2022 Marriot Rivercenter San Antonio, TX March 2729 Grand Hyatt San Antonio, TX
Security Conference and Public Policy Conference National Safety Labor Relations/Human
Exhibition Conference and Exhibition Resources Conference
March 12 May 56 May 1011 May 2526
InterContinental Hotel Washington Court Hotel Omni Fort Worth Convention Center
Houston, TX Washington, DC Fort Worth, TX Denver, CO

Reliability & Maintenance Q&A and Technology Environmental International Lubricants

Conference and Exhibition Forum Conference and Waxes Meeting
May 2427 October 912 October 2425 November 1011
Convention Center JW Marriott New Orleans Marriott Hilton Post Oak
Denver, CO JW Marriott San Antonio Hill New Orleans, LA Houston, TX
Country Resort & Spa
San Antonio, TX

Register today at npra.org/meetings or call us at (202) 457-0480 for more information.

. !4)/.!, 0 %42/#(%-)#!, 2 %&).%23 ! 33/#)!4)/.

NPRA_Con_Cal_2011_PTQ.indd 1 2/2/11 12:40:22 PM

npra.indd 1 4/2/11 12:21:49
 Welcome to the RVMB
(Rotating Valve Multi Bed)

MPR Services, Inc. Introduces

New and Improved, Compact Technology
with Limiting Use of Resources and Energy
MPRs engineering, research and development teams continue to create
improvements in our cutting edge reclamation technology.

t
$PNQBDU
4J[F

t
$MBTT

%JWJTPO

FYQMPTJPO
QSPPG


t
3FEVDFT
8BUFS
$IFNJDBM
BOE
&OFSHZ
3FRVJSFNFOUT

t
%VF
UP
MPXFS
XBUFS
SFRVJSFNFOUT
UIBO
.13T
TUBOEBSE
NPCJMF





VOJUT
XF
HFOFSBUF
MFTT
XBTUF
CSJOF


t
$PTU
4BWJOHT


The RVMBs effectual and conservative operation gives you cost savings
BOE
CFUUFS
SFDMBNBUJPO
BCJMJUZ
XIJMF
SFNPWJOH
IFBU
TUBCMF
TBMUT
DIMPSJEFT

formate, amino acids, bicine) and other corrosive contaminates.

MPRs RVMB has proven performance and reliability.

The RVMB conserves use of resources saving you
wastewater, energy and chemical use which saves you money!

Call us for an amine analysis from our

R&D laboratory and RVMB reclamation service.

MPR Services, Inc, a member of the Tessenderlo Group,

1201 FM 646, Dickinson, TX 77539-3014 USA
Telephone 281-337-7424 Fax 281-337-6534
FNBJM

JOGPNQSTFSWJDFT!ULJOFUDPN
t
Npreurope@tkinet.com
t
NQSNJEEMFFBTU!ULJOFUDPN
www.mprservices.com
HSSX is a registered trademark of MPR Services, Inc., U.S.A. SSX, HCX, and Oxex are trademarks of MPR Services, Inc., U.S.A.
www.ptqenquiry.com Tessenderlo Kerley, Inc. 2011
for further information

mpr.indd 1 8/3/11 14:07:58