Download The Document - Home | Global CCS Institute

Copy and paste this link to your website, so they can see this document directly without any plugins.



Keywords

degradation, that, with, oxidative, solution, thermal, plant, amine, products, solvent, from, Figure, rate, flue, mole, have, concentration, loading, this, capture, temperature, these, been, than, combustion, absorber, oxygen, stripper, such, post

Transcript

Australian National Low Emissions Coal
Research and Development
Project:
Environmental Impacts of Amine-based CO2 Post
Combustion Capture (PCC) Process
Activity 3: Process Modelling for Amine-based PostCombustion Capture Plant
Deliverable 3.2 Progress Report
Revision Description Issue Date
01 Draft January 2012
02 Revision April 2012
03 Revision 2 June 2012
Prepared by
CSIRO – Advanced Coal Technology Portfolio
This page is intentionally left blank
Project Team
Do Thong
Narendra Dave
Paul Feron
Merched Azzi
Enquiries
Merched Azzi
CSIRO Energy Technology
+61 2 94905307
Merched.azzi@csiro.au
Copyright and Disclaimer
© 2012 CSIRO To the extent permitted by law, all rights are reserved and no part of
this publication covered by copyright may be reproduced or copied in any form or by
any means except with the written permission of CSIRO.
Important Disclaimer
CSIRO advises that the information contained in this publication comprises general
statements based on scientific research. The reader is advised and needs to be aware
that such information may be incomplete or unable to be used in any specific situation.
No reliance or actions must therefore be made on that information without seeking prior
expert professional, scientific and technical advice. To the extent permitted by law,
CSIRO (including its employees and consultants) excludes all liability to any person for
any consequences, including but not limited to all losses, damages, costs, expenses
and any other compensation, arising directly or indirectly from using this publication (in
part or in whole) and any information or material contained in it.
ACKNOWLEDGEMENTS
The authors wish to acknowledge financial assistance provided through Australian
National Low Emissions Coal Research and Development (ANLEC R&D). ANLEC R&D
is supported by Australian Coal Association Low Emissions Technology Limited
and the Australian Government through the Clean Energy Initiative.
The authors would like to thank Barry Hopper for his constructive and detailed
comments that helped improve the report.
The authors extend their thanks to the CSIRO Library Services for their combined and
extraordinary efforts.
5
CONTENTS
ACKNOWLEDGEMENTS ........................................................................................... IV
CONTENTS .................................................................................................................. 5
EXECUTIVE SUMMARY .............................................................................................. 6
1. PROJECT OBJECTIVES AND CURRENT MILESTONE .................................. 10
2. STATE OF KNOWLEDGE ................................................................................. 10
3. MONOETHANOLAMINE (MEA) DEGRADATION - REVIEW ............................ 13
3.1 Oxidative Degradation of MEA ............................................................................... 13
3.2 Thermal Degradation of MEA ................................................................................. 19
4. DEGRADATION OF 2-AMINO-2-METHYL-1-PROPANOL (AMP) - REVIEW ... 23
4.1 Oxidative Degradation of AMP ............................................................................... 24
4.2 Thermal Degradation of AMP ................................................................................. 29
5.2 Thermal Degradation of PZ .................................................................................... 36
6. DEGRADATION OF METHYL-DIETHANOLEAMINE (MDEA) .......................... 41
6.1 Oxidative Degradation of MDEA ............................................................................. 41
6.2 Thermal Degradation of MDEA .............................................................................. 47
7 AMINE BLENDS................................................................................................ 50
7.1 Oxidative and Thermal Degradation of MDEA/MEA Blend .................................... 52
7.2 Oxidative and Thermal Degradation of AMP/PZ Blend .......................................... 53
7.3 Oxidative and Thermal Degradation of MDEA/PZ Blend ....................................... 53
8. EMISSIONS OF DEGRADATION PRODUCTS: ASPEN-PLUS SIMULATIONS56
8.1 Aspen-Plus Simulation Task ................................................................................... 57
8.2 Impact of Wash Tower Performance – MEA Base Case ....................................... 57
8.3 Atmospheric Emissions of AMP/PZ Solvent ........................................................... 64
8.4 Atmospheric Emissions of MDEA/MEA Solvent ..................................................... 74
11. ANTICIP ATED VOLATILE DEGRADATION EMISSIONS ................................ 96
12. RANKING OF SOLVENTS .............................................................................. 102
13. CONCLUSIONS .............................................................................................. 103
14. RECOMMENDATIONS & FUTURE WORK ..................................................... 106
15. REFERENCES ................................................................................................ 108
6
EXECUTIVE SUMMARY
The scientific literature concerning the formation of oxidative and thermal degradation
products of MEA, MDEA, AMP, PZ and their select blends have been reviewed in this report.
Despite high overall activity in this field, the amount of experimental work carried out for fully
characterising and quantifying the degradation of amine solvents and the applicability of
these findings to an industrial scale amine-based post combustion capture plant in terms of
predicting the atmospheric emissions of solvent degradation products has been found to be
far less than anticipated beforehand.
The wide spread of reported reaction conditions and applied analytical methods make direct
comparison of both the laboratory and pilot plant based degradation studies rather difficult.
Nevertheless the laboratory studies, pilot plant scale experiments and public domain
technical information from various commercial technology vendors on degradation of amino
solvents clearly indicates that in an industrial environment of post combustion CO2 capture,
these solvents will most certainly undergo both oxidative and thermal degradation. The
extent of degradation and the type of degradation products formed are found to depend
upon the structure of amine and the process operating conditions. Of these conditions we
highlight the concentration of amine, its CO2 loading, absorber reaction temperature, solvent
regenerator temperature, content of oxygen, SOX, NOX and particulate matter in flue gas,
composition of particulate matter (Fe, Ni, V, P, Cr, CO etc), catalytic effect of the material of
construction of plant equipment towards degradation etc. The plant operating practices, such
as the process control and how often a solvent is reclaimed, will also decide both the extent
of solvent degradation and the type of degradation products formed.
In general, potential degradation products of amine solvents (combination of oxidative and
thermal degradation) are likely to be one or more of the following:
Ammonia, primary amines / alkanolamines, secondary amines / alkanolamines, tertiary
amines / alkanolamines, aldehydes (formaldehyde, acetaldehyde), carboxylates, amides,
piperazines, piperazinones, oxazolidpnes, nitrosamines, imidazolidones, N,N-distributed
ureas and nitramines. Other compounds may be added to the list if the ongoing research
has identified any additional product.
The exact chemical structure of these degradation products depends upon the chemical
structure of parent amine and the degradation reaction pathways it has followed.
Whilst there are a number of pilot plant scale CO2 capture campaigns ongoing around the
world, there has been so far not even a single study that has attempted to close the material
balance around plant where formation of degradation products within the plant and the
atmospheric emissions of these products are fully accounted. Nevertheless, these
campaigns do confirm the following:
a) More degradation products are formed in an industrial plant environment than what
various research groups have so far determined in their laboratory studies.
b) In an industrial environment, it is the oxidative degradation that is contributing more
towards overall solvent degradation than the thermal degradation.
c) Vapour phase atmospheric emissions of heat stable salts and thermal degradation
products of amines are likely to be minimal to the point of no concern.
d) Formation of nitroso compounds in the industrial plant environment is a reality since both
Boundary Dam and ITC pilot plants in Canada confirm formation and detection of
7
1,2,3,6-tetrahydro-1-nitrosopyridine during their MEA and MEA/MDEA campaigns. It
should be noted that the Boundary Dam pilot plant was originally built by Union Carbide
and later refurbished by Fluor Ltd for SaskPower as per the Fluor Econamine technology
and it used proprietary corrosion inhibitors for both amine campaigns. Thus, Strazisar et
al., 2003, are correct in asserting that they detected nitrosamines in the lean amine
solution at Kerr-MCGee/ABB Lumus technology based post combustion CO2 capture
plant at Trona, California, and these compounds may have been formed up to 2.91 µmol
per mL of solution. It should be noted that the Trona plant also uses proprietary corrosion
inhibitors despite the lean amine (MEA) concentration being less than 20% w/w.
e) Corrosion inhibitors currently being recommended and perhaps used by commercial
technology vendors for post combustion CO2 capture may be acting as catalyst towards
the solvent degradation. Certainly, the corrosion inhibitors containing Copper, Vanadium,
Cobalt and other metals used in the reducing environment of gas processing industry are
catalyst for solvent degradation in the oxidative environment of post combustion capture.
Laboratory studies of solvent degradation by Rochelle and other investigators confirm
this.
f) The wash tower downstream of absorber plays critical role in controlling atmospheric
emissions of various amine solvents and their volatile degradation products from a CO2
capture plant. The AspenPlus process simulation results clearly indicate that the wash
tower performance is affected by ambient conditions, particularly the cooling water
temperature. Operating this tower at temperature as low as possible in practice can
substantially reduce the emissions of volatile degradation products. No doubt other
process or equipment performance improvement measures such as intercooling the
absorber or separating the condensable species downstream of the absorber (using
reflux condenser) prior to water washing the CO2 lean flue gas will certainly help in
reducing atmospheric emissions.
Published information on operating performance of the capture plant at Trona, California,
and the technology related information from MHI Ltd clearly states that the flue gas
impurities viz. the particulate matter, SOX and NOX contribute significantly towards solvent
degradation and atmospheric emissions of amine solvents. In fact, the MHI data
categorically shows that by reducing SOX content at inlet to the absorber from 3 ppmv to 1
ppmv reduces atmospheric emissions of both KS-1 and MEA solvents by more than half.
Recently published information from MHI further states that in order to eliminate formation of
aerosols in the CO2 lean exhaust stream, the SOX content of flue gas at inlet to the absorber
should be less than 0.1 ppmv. Similarly, to avoid the sludge build up in absorber and
resulting foaming and flooding of the absorber as well as to reduce the metal catalysed
degradation of solvent, the particulate level in the lean amine solvent should not be allowed
to exceed 1 ppm by weight. All of this means that for stable performance of CO2
absorption/desorption system with high efficiency, minimum solvent degradation and
minimum atmospheric emissions, the direct contact cooler and the water wash tower in post
combustion CO2 capture plant must be performing efficiently. This is particularly important in
Australian context since Australian power stations do not have de-SOX and de-NOX systems.
In terms of choosing an amine solvent for amine-based post combustion capture with
minimal environmental adverse impact, one should note the guidelines below:
i. Secondary amines have highest risk of nitrosamine formation, followed by tertiary
amines, while primary amines have the lowest risk of nitrosamine formation.
ii. All other things being equal, solvents with low vapour pressure are safer than solvents
with high vapour pressure.
8
iii. All other things being equal, a more stable solvent that will resist degradation is safer
than a less stable one since the more stable solvent will have lower emissions of
degradation products.
Using the above guidelines, the amine solvents considered in this report can be ranked in
order of their likely maximum adverse impact to minimal adverse impact as:
PZ > AMP > MEA > MDEA
The available publication from the ongoing pilot plant programs in different parts of the world
shows that the characterisation and quantifications of gas and liquid streams at inlet and
outlet to the absorber, stripper and water wash towers downstream of both absorber and
stripper has not yet been fully understood. At this stage, there has been only one pilot plant
based degradation study (RWE/LINDE/BASF Niderauβem Pilot Plant, Section 9.3) that has
attempted to close material balance with respect to amine consumption and included
formation of degradation products and their atmospheric emissions in the material balance.
Unfortunately, the material balance does not close with acceptable accuracy and in addition,
the liquid phase emissions have not been quantified.
It must be pointed out that the above study has been performed at a lignite-fired power plant
in Europe where the coal based power plants are equipped with state of the art de-SOX, deNOX and particulate filtration systems which is not the case with Australian coal-fired power
plants. Thus, there is a strong parameter of flue gas quality that needs to be considered
before using a European solvent degradation study for anticipating the solvent degradation
and resulting atmospheric emissions from a post combustion CO2 capture plant linked to an
Australian coal-fired power plant. In addition, the ambient conditions for both summer and
winter seasons in Europe are markedly different from Australian ambient conditions for both
inland and coastal locations. Thus, impact of ambient conditions on performance of the
water wash tower must also be accounted for when using a European study in Australian
context.
Since, the Australian flue gas quality and ambient conditions are considerably different from
equivalent European or US/Canadian situation, both direct contact cooler (DCC) and the
water wash towers downstream of absorber and stripper will need to be designed
accordingly. It will be preferable, if the flue gas desulphurisation as well as the removal of
oxides of nitrogen can be carried out in the DCC tower itself for an Australian post
combustion capture plant. This has potential to reduce the cost of CO2 capture from an
Australian coal-fired power plant and may even have a positive impact on the net power
plant efficiency.
In light of the above findings, a future work program should include the following:
I. Full characterisation and quantification of degradation products of amine solvents
both in the gas and liquid streams in an Australian pilot plant scale CO2 capture plant
accompanied by a closed material balance around the input and output streams of
the plant. It is expected that this pilot plant will be operating at steady state in the
gas-liquid flow regimes representative of currently operating industrial scale post
combustion CO2 capture plants.
II. Process improvements around the DCC tower and the water wash towers
downstream of both absorber and stripper in an Australian pilot plant scale to
minimise first the adverse impact of flue gas impurities on amine solvents and then
atmospheric emissions of amine solvents and their degradation products.
9
III. Development of mathematical models from first principle (for example, using the
molecular modelling principles) – bottom up approach – which account for the
solvent degradation kinetics and yields of the degradation products observed
through steps I and II above to improve predictive capability of anticipated
atmospheric emissions from a full scale amine solvent based post combustion
capture plant linked to an existing black coal-fired power plant in Australia.
10
1. PROJECT OBJECTIVES AND CURRENT MILESTONE
Aqueous amino solvent based acid gas removal processes for the natural and synthetic gas
processing industry have been universally recognised to have potential for immediate
reduction in the emissions of carbon dioxide (CO2) from fossil fuel-fired power plant flue gas
streams. As a result, commercially available technologies that use these solvents have
recently acquired prominence as a short term technological solution to curb globally rising
CO2 levels in the atmosphere. Recent laboratory and pilot plant data show that these
solvents may undergo oxidative and thermal degradation during the post combustion capture
application resulting into the formation of volatile and non-volatile organic compounds
including nitrosamine and nitramine compounds. Some of these compounds can potentially
be released to the atmosphere in both the vapour phase and the droplet phase along with
the parent solvents that may cause adverse impacts on the environment. The objective of
this project is therefore:
1. To identify the impact of process operating conditions on the degradation of amino
solvents during the post combustion capture application,
2. To identify and summarise the major degradation products of amino solvents, and
3. To estimate the likely emissions of amino solvents and their degradation products
into the atmosphere, if these solvents are utilised in a large scale post combustion
plant connected to an black coal-fired power plant.
This report attends to the above objectives by undertaking a detailed literature survey on
the oxidative and thermal degradation of aqueous amine solvents as observed during
laboratory experiments and pilot plant trials. It also reports the likely emissions of these
solvents and their degradation products to the atmosphere using available in the public
domain or if it is possible to estimate them by process modelling using commercial tools
such as the Aspen-Plus process simulation software. Since, the overall project objective is to
understand the environmental impact of amino solvents when implemented for post
combustion CO2 capture at large scale from a black coal-fired power plant flue gas stream,
the emphasis in this report has been given to those amino solvents that are being used at
present either commercially or at the technology demonstration scale for such an
application. These solvents are monoethanolamine (MEA), methyl-diethanolamine (MDEA),
2-amino-2-methyl-1-propanol (AMP), piperazine (PZ) and their blends.
2. STATE OF KNOWLEDGE
Aqueous alkanolamines have been used for decades in the gas processing industry for acid
gas removal such as carbon dioxide (CO2), hydrogen sulphide (H2S), carbonyl sulphide
(COS), carbon disulfide (CS2) and mercaptans. Monoethanolamine (MEA), Diethanolamine
(DEA), methyl-diethanolamine (MDEA), di-isopropylamine (DIPA) and diglycolamine (DGA)
have been the solvents of choice for this purpose1. For the last two decades, the sterically
hindered amine such as 2-amino-2-methyl-1-propanol (AMP) has been suggested as an
attractive solvent for acid gas removal due to its high equilibrium loading capacity and low
energy requirement for regeneration2. Flexisorb, Kerr-McGee, Fluor Econamine, MHI,
Ucarsol and BASF-activated-MDEA are some of the examples of commercial alkanolamine
based acid gas removal process technologies3. In spite of resistance of these alkanolamines
to chemical breakdown, the plant and laboratory reports indicate that, on prolonged use
particularly in the oxidative environment, these solvents can transform into products from
which they are not easily regenerated. This phenomenon, commonly referred to as ‘amine
degradation’, not only reduces the acid gas removal capacity but also leads to operational
problems such as foaming, corrosion, high solution viscosity, fouling and decreased plant
11
equipment life. Figure 1 below depicts the zones or equipment sections in a gas absorber
and a solvent regenerator (stripper) system of an aqueous amine based gas treatment
process where solvent degradation is observed to occur when processing post combustion
flue gas4.
Figure 1 – Schematic of a typical amine based gas absorption-stripping system
The oxidative degradation is a chemical reaction of amine with oxygen dissolved in the
amine solution that has entered the system either with the gas stream or due to the airingress. It occurs primarily in the absorber sump, absorber packing, piping leading to the
cross exchanger (lean/rich solvent heat exchanger) and the cross exchanger. Any oxygen
remaining after the cross exchanger will flash at the top of the stripper. In the case of post
combustion capture of CO2 from a coal-fired power plant flue gas, oxygen is present in the
gas stream at around 4 to 8% by volume. The solvent hold up time in a commercial scale
absorber sump is generally around 5 to 10 minutes allowing oxidation reactions sufficient
time to degrade the solvent and cause the solvent loss5. In addition to oxygen, the impurities
in flue gas, viz. particulate matter (metal oxides), HCL, HF, SOX and NOX present particularly
in coal combustion off gas streams, also contribute towards the amine degradation. NOX
components of flue gas do react with amine species to form nitrosamines and nitramines
that are known to be carcinogenic. Oxidation reactions, in general, involve fragmentation of
N-C bond in amine and formation of O-C bond resulting into a number of degradation
products that in turns can react with each other or with more oxygen or the amine itself to
produce more degradation products4. These reactions are invariably catalysed by the metal
ion impurities (Fe3+, Fe2+ or Cu+ etc) present in the flue gas as is usually the case with coalfired power plant flue gas6. The laboratory data show that an amine solution loaded with
dissolved CO2 is much more prone to oxidative degradation than the unloaded one. In
general, the rate of oxidative degradation of amines seems to depend upon the temperature,
the concentrations of amine, metal ions, dissolved oxygen and CO2 in the solution and the
concentration of impurities such as SOX in the flue gas7. Whilst oxidation of amines is not
12
well characterised for most amines since the primary degradation products are still being
identified, the oxidative degradation of MEA is known to produce ammonia, aldehydes,
ketones, amides and carboxylic acids6-9. Some of these degradation products have high
volatility and can be emitted to the atmosphere both in the vapour and droplet phases during
the post combustion CO2 capture and thus could potentially affect the environment
adversely. The laboratory and plant data also show that the oxidative degradation products
of amine react with the materials of construction of the plant equipment and produce heat
stable salts (HSS) which accumulate in the process system over time causing foaming,
fouling and corrosion10.
The thermal degradation of amine involves degradation of its carbamate form which occurs
due to exposure to 100 ̊C or higher temperatures during CO2 stripping. At these
temperatures, the kinetic rates of degradation reactions that are suppressed in the absorber
due to its lower temperature get elevated and cause enhanced solvent loss. Thermal
degradation primarily occurs in the stripper packing, stripper sump, reboiler, solvent
reclaimer (if present) and piping leading from the stripper to the cross exchanger. Novel
stripper designs that do not include packing, such as flash-based designs, are still subject to
thermal degradation. As with the absorber, the stripper sump hold-up is between 5 to 10
minutes, allowing adequate time for significant degradation. It is generally understood from
the laboratory studies that temperature, CO2 loading of the amine solution and concentration
of amine strongly influence the rate of thermal degradation of amine. The CO2 loading
seems to have first order effect on the degradation rate whereas the amine concentration
has more than first order effect11. Commonly known amines such as MEA, DEA and MDEA
have been used for decades in the refinery off-gas and natural gas processing industries
where thermal degradation is more predominant and hence, the thermal degradation
products of these amines and their blends are generally well characterised. However, full
characterisation of the thermal degradation products of amines in recent use, such as AMP
and PZ, are yet to be determined. The laboratory investigations of thermal degradation of
MEA show that oxazolidone, 1-(2-hydroxyethyl)-2-imidazolidone (HEIA), N-(2-hydroxyethyl)ethylenediamine (HEEDA), MEA urea, N-(2-hydroxyethyl)-diethylenetriamine (trimer of
MEA), 1-[2-[(2-hydroxyethyl)amino]ethyl]-2-imidazoline (cyclic urea of MEA trimer), N-(2hydroxyethyl) triethylenetetramine (quatramer of MEA) and higher polymeric products are
formed11. Since most of these products are high boiling and low volatility compounds, they
are less likely to be emitted to the atmosphere in the vapour phase during the post
combustion capture of CO2.
In the cross exchanger, the amine solvent temperature could rise to as high as 100 ̊C or
more, specially near its stripper ends, and cause the thermal degradation of amine. This is
also the primary area where the oxidative degradation could be ongoing due to the presence
of dissolved oxygen in the CO2 rich amine. The study of combined oxidative and thermal
degradation is an area of research that is decidedly lacking whereas all research to date has
focussed on either oxidative or thermal degradation of amines separately. The combined
effect of oxidative and thermal degradation needs to be fully understood to ascertain the
degradation characteristics of an amino solvent.
Oxidative and thermal degradation of amines contributes towards the solvent loss, decrease
in the solvent capacity to purify the gas stream, hydrodynamic instability of the operating
systems (foaming, increased viscosity etc.), fouling of the heat transfer surfaces and
corrosion of the plant equipment, thus contributing towards higher cost of gas purification. In
addition to the cost, the environmental impact of atmospheric emissions of these
degradation products should be addressed with the right scientific knowledge. Hence,
identifying the process operating conditions that initiate the oxidative and thermal
degradation of amines, identifying the exact nature of degradation products formed and their
concentrations in the solvent circulating within the absorber/stripper system and determining
13
the likely emissions of these products into the atmosphere is extremely important,
particularly when large scale post combustion carbon dioxide emission reduction is targeted
for coal-fired power stations to achieve the greenhouse gas (GHG) mitigation globally.
To achieve the above objective, the oxidative and thermal degradation of select amines viz.
MEA, MDEA, AMP and PZ is investigated and their anticipated atmospheric emissions
estimated when aqueous solutions of these amines are used for recovering 90% CO2 from a
black coal-fired power plant flue gas. MEA is most commonly used amine for CO2 removal
commercially. The Kerr-McGee/ABB Lummus Crest process uses 10-20% w/w aqueous
MEA12. The Fluor Daniel EconamineTM process uses 30% w/w aqueous MEA with an
inhibitor to resist carbon steel corrosion and is especially tailored for treating flue gas with
oxygen content as high as 15% v/v13. BASF and Ucarsol processes use aqueous MDEA
solution with a rate promoter such as PZ14. Kansai Electric Power Co./Mitsubishi Heavy
Industries Ltd KEPCO/MHI process uses sterically hindered amine, KS-1, which apparently
does not require corrosion inhibitors and other additives15. Whilst the exact nature of KS-1
solvent is proprietary, AMP with a rate promoter, such as PZ, could be an analogous
solvent16.
Identification of the oxidative and thermal degradation products of the above selected
amines in this report is based on the experimental data from the relevant literature studies.
Likely emissions to atmosphere of these solvents and their degradation products is reported
where available in the public domain or if it is possible to estimate by process modelling
using the commercial tools such as the Aspen-Plus process simulation software. This review
is necessary as a first step to enhance the accuracy of the process modelling results.
3. MONOETHANOLAMINE (MEA) DEGRADATION - REVIEW
Since early 1960s MEA has been in use for removal of CO2 from gas streams due to its
biodegradability, low cost and high reactivity towards CO2. Its thermal degradation in the
reducing environment of natural gas processing and oxidative degradation during CO2
removal from air supply of nuclear submarines has been studied for long with a view to
develop the corrosion inhibitors for protecting plant equipment and the anti-foaming agents
for stabilising equipment performance. However, the proposed application of this solvent to
globally remove 70 to 90% of CO2 from commercial coal-fired power station flue gas streams
has raised concerns about the environmental impact of its degradation products. As a result,
several investigators have undertaken extensive characterisation and quantifications of the
atmospheric emissions of MEA and its degradation products based on laboratory
experiments and pilot plant operations. These studies conclude that both oxidative and
thermal degradations of MEA occur during CO2 removal from a coal-fired power plant flue
gas stream and the degradation increases with temperature. The following sections describe
the current status of knowledge on the oxidative and thermal degradation of MEA and the
atmospheric emissions of its degradation products as observed at least during the pilot plant
trials. It should also be noted that MEA has been treated in this report as a base line solvent
for the post combustion CO2 capture against which the degradation performance of other
selected amines is compared.
3.1 Oxida tive Degrada tion of MEA
Despite the latest progress achieved over the last five years in the understanding of the
oxidative degradation mechanism for MEA where the electron and hydrogen abstractions
were proposed as pathways towards the degradation6 of MEA, there are still gaps in the
knowledge that require to be elucidated. Fe, Cu, Cr, Ni and V ions were found to catalyse
these paths. Figure 2 shows formation of the aminium and peroxide radicals via the electron
abstraction mechanism which results into ammonia, formaldehyde and hydroxyacetaldehyde
14
formation as primary degradation products. Figure 3 shows the hydrogen abstraction
mechanism towards formation of ammonia and formaldehyde as primary products. It is well
known that aldehydes react with oxygen to form the carboxylic acids. Since MEA solutions
have pH in the range 9 to 12, the carboxylic acids dissociate to form heat stable salts with
MEA. The presence of aldehydes and organic acids in the degraded MEA solutions has
been documented in a study by The Dow Chemical Co17. In addition, ammonia formed
during the degradation process reacts with MEA in the presence of oxygen to form amides
and alkylamines6-11. These reactions can be summarised as below:
C2H7NO + 0.5O2  HOCH2CHO + NH3--------------------------- (1)
C2H7NO + 0.5O2  2HCHO + NH3---------------------------------- (2)
HCHO + 0.5O2  HCOOH----------------------------------------- (3)
C2H7NO + 1.5O2  2HCOOH + NH3-------------------------------- (4)
C2H7NO  CH3CHO+ NH3---------------------------------- (5)
CH3CHO + 0.5O2  CH3COOH-------------------------------------- (6)
2C2H7NO + O2  2CH3COOH + 2NH3-------------------------- (7)
C2H7NO + 0.5O2 + NH3  HCONH2 + CH3NH2 + H2O-------- (8)
C2H7NO + 2O2  HOCOCOOH + NH3 + H2O-------------------- (9)
C2H7NO + 2O2 + NH3  H2NCOCONH2 + 3H2O------------------ (10)
C2H7NO + 0.5O2  H2NCH2COH + H2O-------------------------- (11)
2C2H7NO + 0.5O2  HOCH2CH2NHCHO + CH3NH2 + H2O--- (12)
2C2H7NO + 0.5O2  HOCH2CH2NHCOCH3 + NH3 + H2O----- (13)
In general, the stoichiometric equation given below represents the above reactions,
MEA + ∆O2  NH3 + Degradation Products
Figure 2 – Electron abstraction mechanism for the oxidative degradation of MEA
15
Figure 3 – Hydrogen abstraction mechanism for the oxidative degradation of MEA
Thus, the oxidative degradation of MEA involves a complex mixture of series and parallel
reactions with ammonia as a primary product. Knowing the rate of ammonia evolution and
the rate of dissolved oxygen consumption, the overall oxygen stoichometry (∆) can be
determined for the oxidative degradation of MEA as:
Oxygen Stoichiometry (∆) = (Rate of O2 consumption / Rate of NH3 evolution)
Its value varies from 0.5 to 2.5 depending on what degradation product is being formed6, 7. It
should be noted that the stoichiometry for acetaldehyde formation is 0.0, which means it can
be formed by any free radical and does not require presence of oxygen. Vevelstad et al18
have recently used continuum solvation models and performed quantum mechanical
calculations to determine the heat of reaction (∆H) and the Gibbs free energy change (∆G)
for the reactions 1 to 13 summarised above. These authors have not specified the
temperature range at which these calculations were made. Consequently, these data have
limited use towards building non-mechanistic kinetic rate models of the oxidative degradation
of MEA and determining the degradation product distribution.
Laboratory studies of the oxidative degradation of MEA in sparged and agitated reactors by
Goff6 indicate that the degradation of MEA is a function of temperature and concentrations of
metal ions (Fe, Cu, Cr, Ni, V etc), MEA and dissolved oxygen in the solution. The
degradation can be O2 mass transfer limited under laboratory conditions and the NH3
evolution increases linearly with the oxygen concentration in gas stream up to 17 % by
volume. Their results also indicate that the MEA solution with lower CO2 loading has higher
rate of oxidative degradation but lower oxygen stoichiometry in comparison with the MEA
solution which has higher CO2 loading. Under industrial conditions, the degradation of MEA
can be both kinetics and O2 mass transfer limited depending upon the liquid to gas ratio
maintained in the gas/liquid system and the hydrodynamics of the system. NH3 evolution
16
generally increases with higher MEA concentrations. At aqueous concentrations above 7
molar MEA (~ 42% w/w), the degradation rate appears to be O2 mass transfer controlled and
for solutions below 2 molar MEA (~ 12% w/w) appears to be kinetics controlled. Intermediate
solution concentrations exhibit the effects of both kinetics and O2 mass transfer control. Goff
and Rochelle19 reported that in industrial applications the oxidative degradation related MEA
loss could be 0.29 to 0.73 kg of MEA per ton of CO2 captured. Table 1 below summarises
the results of various laboratory studies for the oxidative degradation of MEA reported in the
literature.
Table 1 – Comparison of key studies for the oxidative degradation of MEA
Study T
( ̊C)
MEA
(molar)
Gas
Liquid
Volume
(Litre)
Space
Time
(min)
Degradation
Rate
(milimoles/hr)
KG
(milmoles/
hr/bar)
Rooney et al17 82 5.4 Air 1.0 181.8 0.03 to 0.07 0.2 – 0.4
Blachly et al20 55 5.4 Air 0.3 1.0 0.02 to 0.14 0.1 – 0.8
Girdler21 80 3.1 50% O2 0.1 1.0 0.36 to 1.32 0.7 – 2.6
Hofmeyer et al22 75 5.4 Pure O2 0.2 0.35 5.0 5.0
Chi et al23 55 7.0 Air 0.4 0.05 0.1 to 2.32 0.6 – 12.9
Goff et al24 55 7.0 Air 0.3 0.04 0.48 to 5.0 2.7 – 27.8
Goff et al19 55 7.0 Air 0.5 0.06 0.25 to 8.25 1.4 – 45.8
Space Time = (Liquid Volume / Gas flow rate)
KG = Apparent mass transfer co-efficient = (Degradation rate / Partial Pressure of O2 in gas)
In the case of a coal-fired power plant, the impurities such as SOX, NOX and particulate
matter (metal oxides) are invariably present in the flue gas stream in addition to un-reacted
oxygen. SOX are known to degrade MEA as per the reactions 14 to 16 below forming
organo-sulphates and thiovanic acid25. The particulate matter has potential to not only act as
a degradation catalyst but also form sludge within the gas/liquid contactor which causes
corrosion and erosion of the plant equipment.
2(C2H5ONH2) + SO3 + H2O  (C2H5ONH3)2SO4----------------------- (14)
2(C2H5ONH2) + SO2 + 0.5O2 + H2O  (C2H5ONH3)2SO4------------- (15)
C2H5ONH2 + SO2  C2H4O2S + NH3 + 0.5O2--------------------------- (16)
There is limited information available in the public domain on the mechanism of degradation
of MEA due to presence of NOX in the flue gas. However, Pedersen et al26-29 have recently
shown that degradation of MEA in the presence of NOX resulted into the formation of
nitrosamines in their AminoxTM rig and in a high-pressure autoclave. In these experiments,
the authors exposed 30 to 40 wt% MEA solution at CO2 absorber (44 ̊C) and stripper (120 ̊C)
conditions to a flue gas containing 3.5 vol% CO2, up to 14 vol% oxygen and as much as 100
ppmv NOX. These investigators observed that NO is oxidised to NO2 in the absorber inlet
and roughly 20% of NO2 is absorbed into the liquid phase due to the following reaction
taking place in the presence of oxygen in flue gas:
NO + NO2 + 0.5O2 + H2O  HNO2 + HNO3------------------ (17)
17
The above reaction is known to occur even at room temperature43. Jackson et al30 indicate
that HNO2 acts as a nitrosating agent in alkaline environment (pH >7) and the secondary
amines such as Diethanolamine (DEA) are reactive towards the nitrosamines formation in
the presence of NOX. Results of Pedersen et al indicate that ammonia is the primary
degradation product of MEA and its production is strongly correlated with the NOX
concentration in flue gas. Whilst these investigators expected MEA being a primary amine
not to form stable nitrosamine, they however detected 0.5 ppm (by weight)
nitrosodiethanolamine (NDELA) under the absorber environment after 100 hours of solution
exposure to NOX (25 to 50 ppmv). Pedersen et al believe that NDELA may have been
formed from DEA, a secondary amine, that may have been either present in the MEA
solution as an impurity or formed as a result of NOX induced degradation of MEA which has
undergone the nitrosation reactions with NO and HNO2. Figures 4 to 6 show formation of
NDELA and DEA in MEA solution as a function of its exposure time and the concentrations
of O2 and NOX in flue gas. These results clearly show that even at NOX concentration as low
as 5 ppmv, NDELA is formed. Other investigators31-33 do confirm the conclusion of Pedersen
el al that for the formation of stable nitrosamines, the presence of a secondary amine in the
reaction mixture is necessary.
Formation of nitrosamines at a concentration of 2.91 µmol/mL of lean MEA solution has also
been confirmed by Strazisar34 et al who analysed lean MEA, fresh MEA and reclaimer
bottom streams at Kerr MCcGee technology based CO2 capture plant of IMC Chemicals Inc,
Trona, California. This plant has been in operation since 1978 and uses 15 to 18% w/w MEA
solution to capture CO2 from a coal-fired boiler flue gas slip stream. In addition to
nitrosamines, organo-sulphates were also detected in the lean MEA solution. Unfortunately,
Strazisar et al did not attempt to identify individual nitrosamine chemical in their study. Whilst
this study has been first of its kind in the sense that actual power plant flue gas was used to
understand oxidative degradation of MEA, because neither the spent MEA solution nor the
reclaimer bottom stream or CO2 lean flue gas was analysed, the quantification of volatile
degradation product such as NH3 was not available. While quantitative data on the rates of
oxidative degradation of MEA can not be determined from this study, it does confirm the
formation of carboxylic acids in the degraded solution as observed by other laboratory based
studies.
Figure 4 – Formation of NDELA in MEA solvent (Legend: 7O2 / 25NOX denotes flue gas
with 7 volume % Oxygen and 25 ppmv NOX)
18
Figure 5 - Formation of DEA in MEA solvent (Legend: 7O2 / 25NOX denotes flue gas
with 7 volume % Oxygen and 25 ppmv NOX)
Figure 6 – Predicted (solid lines) and observed (dots) concentrations of NDELA in
MEA solvent AminoxTM absorber condition.
Whilst previous laboratory studies on the oxidative degradation of MEA assessed the effect
of oxygen mass transfer rate, CO2 loading of MEA, MEA concentration and catalytic effect of
the metal ions in solution and related these parameters to the evolution of NH3 as a measure
of MEA loss, the research group from University of Regina has focussed on measuring MEA
degradation rate as a function of temperature, CO2 loading of MEA and concentrations of
oxygen and sulphur dioxide (SO2) dissolved in solution to evaluate the contributions of SO2
and O2 to the degradation of MEA during CO2 capture from power plant flue gas streams3538. Uyanga and Idem37 have developed a data regression based empirical rate expressions
for oxidative degradation of MEA on the basis of such studies which are valid over the
temperature range of 328 – 413 K, MEA concentrations in the range of 3 – 7 moles per litre
of solution, SO2 concentrations in flue gas ranging from 6 to 196 ppmv and oxygen
concentrations in the range of 6 – 100 mole%. Both linear (Equation 18) and non-linear
(Equation 19) power-law rate expressions have been provided as kinetic rate models
assuming that all reactions occur in the liquid phase. These expressions are listed below. To
eliminate any mass transfer resistance, the liquid phase (reaction mixture) was vigorously
stirred at 505 rpm during the laboratory experiments.
19
-RMEA = 0.7189 exp[-(20752/8.314*T)] [MEA]1.359[SO2]2.0[CO2]-0.333[O2]0.03------- (18)
-RMEA = 0.00745 exp[-(45258/8.314*T)] [MEA]1.9[CO2]-0.333{[SO2]3.4 + [O2]2.8}------- (19)
Where RMEA is rate of degradation of MEA (moles/litre/hr), [MEA], [CO2], [SO2] and [O2] are
molar concentrations (moles/litre) of MEA, CO2, SO2 and O2 in solution and T is absorber
temperature (K). The first expression accounts for the scenario in which the concentrations
of all the species in the kinetic rate model are greater than zero. The second rate expression
is a more-general model and allows the SO2 and/or O2 concentration to be zero or greater
than zero.
Recently, Supap et al38 have improved upon the above mentioned degradation rate models
and suggested the following non-linear rate expression that is claimed to better represent the
experimental results:
-RMEA = {6.74 x109e –(29403/RT)[MEA]0.02([O2]2.91+[SO2]3.52)}/{1+1.18x[CO2]0.18}------- (20)
Where, R is universal gas constant (8.314 J/mole/K). Unlike previous models, this model
shows an improvement in that any of the parameters, i.e. [O2], [SO2] and [CO2] terms can be
removed without affecting the usability of the model.
From the process operations point of view, it is now accepted that minimising the absorber
temperature rise, reducing the O2, SO2, NOX and particulate content of flue gas, maintaining
optimal CO2 loading of MEA solution and MEA concentration, using appropriate materials of
construction and the select use of corrosion inhibitors are all necessary to minimise oxidative
degradation of MEA. Bicine, Ethylene diamine-tetra-acetic acid (EDTA), sodium sulphite
(Na2SO3), potassium sodium tartarate tetrahydrate (KNaC4H4O6.4H2O), hydroxylamine
(NH2OH), diethylene triamine penta acetic acid (DTPA), dimercaptothiadiazole (DMTD) and
a mixture of these compounds have been suggested in the literature4,6,7,11,39,40 as the
corrosion inhibiting compounds for the amine based post combustion CO2 capture
apparatus. However, not all of them are effective as observed by Rochelle and his
researcher group4,6,7,11.
Oxidative degradation related emissions of MEA and its degradation products observed in
the pilot plant scale trials or noted in commercial scale (300 to 600 tonnes per day CO2)
operations and predictions of such emissions by means of process engineering simulation
tools such as the Aspen-Plus have been discussed and summarised in the relevant sections
of this report.
3.2 Thermal Degrada tion of MEA
The thermal degradation of MEA occurs predominantly in the stripper packing and reboiler
due to exposure to high temperature. Davis and Rochelle41 indicate that thermal degradation
is minor when reboiler temperature is held below 110 ̊C but it accelerates above 130 ̊C.
These authors indicate that carbamate polymerisation due to high temperature is the main
cause of thermal degradation of MEA which is compounded when CO2 loading of the
solution is increased. Figure 7 shows the reaction pathway for thermal degradation of MEA
as envisaged by Davis and Rochelle11, 41.
20
Figure 7 – Reaction mechanism for thermal degradation of MEA
According to this reaction mechanism, MEA carbamate cyclises internally through
dehydrolysis step to form oxazolidone. Another molecule of MEA attacks oxazolidone at the
ketone group to form MEA urea. MEA molecule can also attack oxazolidone to form N-(2hydroxyethyl)-ethylenediamine (HEEDA). HEEDA then reacts with dissolved CO2 in spent
solution to form a HEEDA carbamate similar to MEA and ends up as 1-(2-hydroxyethyl)-2imidazolidone (HEIA) or reacts with oxazolidine to form the trimer of MEA. The trimer of MEA
in turn reacts with dissolved CO2 in spent solution to form cyclic urea of MEA trimer. This
polymerisation can continue indefinitely through the quatramer of MEA and its corresponding
cyclic urea. Figure 8 shows the reaction sequence for MEA thermal degradation with Table 2
showing values for various rate constants.
Figure 8 – Thermal degradation reaction sequence for MEA
21
Table 2 – Thermal degradation reaction rate constants and activation energies
Rate Constant Pre-Exponential Constant
(L/day/mole)
Activation Energy
(Kcal/mole)
K1 1.05E16 34.4
K2 2.15E16 33.3
K3 3.28E15 31.5
K4 3.58E16 33.0
K-4 4.47E15 32.6
K5 3.65E15 31.3
K-5 4.56E14 31.3
Figure 9 shows thermal degradation product distribution for 7 molal MEA at 135 ̊C and CO2
loading of 0.4 mole per mole of MEA as measured experimentally by Davis11. However,
HEEDA/HEIA product distribution trend could change with temperature as observed by
Davis11 at 100 ̊C for the same solution concentration and CO2 loading where more HEEDA
formed than HEIA.
Figure 9 – Thermal degradation product distribution for 7 molal (30 % w/w) MEA
at 135 ̊C and 0.4 mole/mole CO2 loading
Figure 10 shows the effects of MEA concentration, CO2 loading and stripper reboiler
temperature on the amine loss due to thermal degradation as observed by Jason and
Rochelle41. These results clearly indicate that the rate of degradation quadruples at every 17
̊C rise in the stripper reboiler temperature or approximately every time the pressure of
stripper is doubled. Decreasing the CO2 loading is roughly a first order effect causing similar
decrease in the degradation rate. Increasing MEA concentration from 30 to 40% w/w shows
slightly more than first order effect. In practice, increasing MEA concentration will increase
the reboiler temperature due to elevation in the boiling point of solution. Using this
information and Aspen-Plus software, Jason and Rochelle11, 41 have developed the stripper
performance curve (Figure 11) that depicts optimum operating conditions for MEA stripper
with respect to energy and MEA consumption costs ($ per metric ton of CO2) when CO2
capture is carried out using 30% w/w MEA solution and the product CO2 pressure is 150
atm. The results show that for post combustion CO2 capture with 30% w/w MEA, optimum
stripper bottom pressure and temperature are 3.5 atm and 120 ̊C respectively when MEA
22
loss due to degradation is accounted. In this case, as per Closmann125 the cost of MEA loss
due to degradation is approximately US $1.5 per metric ton of CO2.
Figure 10 – MEA loss over time as a function of MEA concentration, CO2 loading
and stripper reboiler temperature
Figure 11 – Energy and MEA costs as a function of stripper pressure
30% w/w MEA with 0.2 CO2 lean and 0.52 CO2 rich loadings
Recently, Lepaumier et al42 have compared oxidative and thermal degradation
characteristics of MEA observed in the laboratory environment with those observed in the
pilot plant at Esbjergvaerket in Denmark. This pilot plant captures 1 ton/hr of CO2 from a slip
stream (5000 Nm3/h) of flue gas from a 400 MW coal-fired power station using 30% w/w
MEA. Whilst these investigators concur with the observations of Davis and Rochelle41, they
identified several more degradation products in the MEA solution from pilot plant than in the
laboratory solution though initial MEA concentration, CO2 loading, reaction temperature and
aging of both solutions were identical. This implies that MEA degradation pathways could be
different in an actual plant environment from those of the laboratory environment where
mostly sparged or continuously stirred batch reactors made of glass or steel are used with
either air/CO2 mixture or synthetic flue gas free of fly ash. Process factors such as the use of
demineralised water in solution preparation and the lack of metal ions in solution due to fly
ash free synthetic gas are most likely concealing the relative effect of catalysis on various
23
MEA degradation reactions in the laboratory environment. Accordingly, in addition to
Oxazolidone, and N-(2-hydroxyethyl)ethylenediamine (HEEDA) and N-(2hydroxyethyl)imidazolidinone, (HEIA), Lepaumier et al42 detected N-(2-aminoethyl)-N’-(2hydroxyethyl)imidazolidone (AEHEIA) in the laboratory solution which they could not detect
in the pilot plant solution. Similarly, HEEDA was not detected at all in the pilot plant sample.
The pilot plant sample however had a number of high molecular weight amide derivatives
resulting from the reaction of MEA with carboxylic acids, viz. formic, acetic, glycolic and
oxalic acid which were not at all present in the laboratory sample. This implies that the
carboxylic acids, usually referred to as “Heat Stable Salts”, are not stable and can react
further in an industrial plant environment to give more complex compounds. Lepaumier et
al42 identified these additional degradation products as N-(2-hydroxyethyl)acetamide (HEA),
2-hydroxy-N-(2-hydroxyethyl)acetamide (HHEA), N,N’-bis(2-hydroxyethyl)oxalamide
(BHEOX), 4-(2-hydroxyethyl)piperazin-2-one (HEPO) and N-(2-hydroxyethyl)-2-(2hydroxyethylamino) acetamide (HEHEAA) in the pilot plant solution sample.
It should be noted that in contrast to Strazisar et al34, Lepaumier et al42 analysed the lean
MEA, the CO2 rich MEA and the reclaimer bottom solutions to identify both oxidative and
thermal degradation products of MEA in an actual operating process environment of
Esbjergvaerket pilot plant. Lepaumier et al results for MEA degradation are essentially
similar to those of Strazisar et al34 where these investigators determined the degradation
products by simply analysing the lean MEA solution circulating between the stripper and
absorber of Trona, California, based Kerr McGee plant. This implies that the degradation
products once formed may not get reversed to MEA either in the absorber or the stripper and
continue to build with time.
In summary, the oxidative and thermal degradation of MEA is a complex phenomenon that
leads to a number of different organo-chemicals that may not yet have been fully
characterised and quantified. In an actual plant environment, the degradation products
represent a combined effect of oxidative and thermal degradation which is perhaps different
from the sum total of effects measured separately in the laboratory environment. For
increasing the confidence in assessing and predicting the degradation products of a selected
solvent, it is important to have knowledge about the laboratory studies as well as the
correspondent results from pilot plant measurements
4. DEGRADATION OF 2-AMINO-2-METHYL-1-PROPANOL (AMP) - REVIEW
AMP (2-Amino-2-Methyl-1-Propanol) is a sterically hindered primary amine that forms bicarbonate ions on reaction with the dissolved CO2 in solution as against the carbamate ions
produced by MEA in the same situation.
R-NH2 + CO2 + H2O  R-NH3+ + HCO3- ----------- (21)
2R-NH2 + CO2  R-NH3+ + R-NH-COO------------- (22)
As a result, its CO2 carrying capacity is twice that of MEA on mole per mole basis. However,
AMP is a slow reacting amine and PZ (Piperazine) is normally added to the solution to
accelerate CO2 absorption reaction44-48. AMP is also less corrosive compared to MEA its
corrosion potential can be controlled by adding simple inorganic salts such as Na2SO3 and
NaVO3 as shown experimentally by Veawab49. Aqueous solution of AMP (3 molar) and PZ (2
molar) mixture is supposedly CESAR1 solvent50 that has been trialled at 1 ton/hr post
combustion CO2 capture pilot plant at Esbjergvaerket in Denmark. This solvent gives better
performance in terms of reduced rebolier energy requirement in comparison with MEA.
Mitsubishi Heavy Industries (MHI) technology solvent, KS-1, is also a mixture of primary
24
hindered amine and a rate promoter such as PZ, though the exact details of this solvent are
proprietary knowledge.
4.1 Oxida tive Degrada tion of AMP
The public domain literature information on the oxidative degradation of hindered amines,
particularly AMP, is very limited. Only very recently, Wang and Jens51 have made available
experimental data on the oxidative degradation of AMP. They performed oxidative
degradation experiments on 3 to 5 molar aqueous solution of AMP in a 200 mL glass
autoclave under pure oxygen (250 to 350 kPa) at 100 to 140 ̊C and air/CO2 mixture at 50 to
70 ̊C for 16 days. Amine concentration was determined periodically by the cation-exchange
chromatography and the degradation products identified by the anion chromatography, gas
chromatography-mass spectrometry and Fourier transform infra red spectroscopy. Figures
12 to 19 show their results. These results show that the oxidative degradation rate of AMP is
close to that of Methyldiethanolamine (MDEA) but less than that of MEA under identical
conditions. AMP is not a stable amine despite steric hinderance and the Oxazolidone
formation can not be prevented. However, the steric hinderance of amine function stops ring
opening of oxazolidone to further high molecular weight degradation products as seen with
MEA. Degradation of AMP is oxygen mass transfer limited and strongly depends upon
oxygen pressure. Transition metal ions (Fe2+, Fe3+ and Cu2+) have no obvious effect on
oxidative degradation rates and that explains low corrosion potential of AMP observed by
Veawab et al49, 52. However the oxidative degradation rates seem to increase with CO2
loading as shown in Figure 18. Various carboxylates, ammonia, acetone, 2,4-Lutidine
(Dimethyl pyridine) and 4,4-deimethyl-2-oxazolidone are the main oxidative degradation
products of AMP53.
Figure 12 – Comparison of oxidative degradation of various amines at 120 ̊C,
5 molar aqueous solutions, 350 kPa pure Oxygen and 16 days time
25
Figure 13 – Effect of AMP concentration on oxidative degradation,
250 kPa Oxygen and 120 C̊
Figure 14 – Effect temperature on oxidative degradation of AMP,
250 kPa pure oxygen
Figure 15 – Effect of Oxygen pressure at 120 ̊C
26
Figure 16 – Effect of solution agitation rate on oxidative degradation with
corresponding anion determination, 120 ̊C and 250 kPa oxygen pressure
Figure 17 – Effect of metal ions on oxidative degradation of AMP,
120 ̊C and 250 kPa Oxygen pressure
27
Figure 18 – Effect of CO2 loading on oxidative degradation of AMP,
120 ̊C and 250 kPa Oxygen pressure
Figure 19 – Effect of CO2 loading on oxidative degradation product distribution,
5 mole/kg of AMP degraded at 120 ̊C and 250 kPa Oxygen pressure
28
Voice and Rochelle54 and Lepaumier et al55, 56 have recently investigated the oxidative
degradation of several new amines for their potential to form volatile products including
ammonia. They confirm that the steric hinderance nature of amine function in AMP causes
less formation of volatile products including ammonia at the CO2 absorber temperature of
most likely industrial practice (~ 55 to 80 ̊C) when compared with the performance of MEA.
This implies that in industrial practice, the atmospheric emissions of volatile degradation
products due to use of AMP as a solvent will be less than that for MEA as a solvent.
At present, there is no laboratory or pilot plant experimental data available for either
identifying or quantifying the oxidative degradation products of AMP due to SOX and NOX.
However, the laboratory data from Oh et al57, Choi et al58 and Seo et al59 for absorption of
CO2 from a gas mixture containing SO2 indicate that the aqueous solution of 30% w/w AMP
reacts with SO2 as per the following reaction:
SO2 +H2O +2AMP  2AMPH+ + SO32------------- (23)
This implies that AMP may be consumed by SOX impurities in flue gas during absorption of
CO2. Mimura et al60 have proposed the use of AMP at 40 C̊ with and without rate promoter
(Piperazine) to simultaneously absorb CO2 and NO2 from a gas stream. The published data
shows that in the absence of oxygen in the gas phase, 30% w/w AMP solution without any
CO2 loading or a rate promoter has the liquid side mass transfer co-efficient value of 3.8 x
10-8 mol/s.Pa.m2 for NO2 absorption which increases to 6.5 x 10-8 mol/s.Pa.m2 when
Piperazine concentration in the solution is increased from zero to 3% w/w. This implies that
addition of a rate promoter such as Piperazine accelerates AMP-NO2 reaction. However, to
accurately assess the rate of AMP-NO2 reaction in a CO2 absorber, data with CO2 loaded
amines are required.
In light of the above literature information, it should be noted that Mimura et al60 are the
originators of MHI process technology that uses KS-1 hindered amine solvent with a rate
promoter and in this context, it is interesting to note that the published information61-64 on
MHI process technology states clearly that the KS-1 solvent loss due to the formation of heat
stable salts is minimised, if SOX concentration at inlet to the absorber is kept as low as 0.1
ppmv. These publications also point out the relationship of atmospheric emissions (both
vapour and droplet phase) of KS-1 solvent to the SO3 concentration of flue gas at inlet to the
absorber as shown in Table 3. In addition, the relevant MHI publications clearly state that
none of NOX in flue gas gets absorbed in the direct contact cooler upstream of the absorber
but 1 to 3% of the inlet NOX is absorbed by KS-1 solvent producing however low levels of
heat stable salts.
Table 3 – Atmospheric emissions of KS-1 solvent v/s SO3 content of flue gas
SO3 Concentration of flue gas @ absorber inlet KS-1 Emissions
Zero 0.4 ppmv
1 ppmv 9.1 ppmv
3 ppmv 23.2 ppmv
29
4.2 Thermal Degrada tion of AMP
Similar to the oxidative degradation of AMP, data on the thermal degradation of
AMP has become only recently available though it is very limited. Figure 20 below shows the
thermal degradation behaviour observed by Wang and Jens53 in the nitrogen environment. It
shows that AMP does not degrade in a non-oxidative environment at least up to 140 ̊C.
Figure 20 – Thermal degradation of AMP
However, comparing Figure 18 with Figure 20 shows that the presence of dissolved CO2
does influence oxidative degradation of AMP. Therefore CO2 loading should similarly
influence thermal degradation of AMP somewhat. This type of degradation is most likely to
be much less than that observed for MEA because AMP does not produce carbamate ions
and hence, carbamate polymerisation related degradation products are unlikely to be
formed. This is confirmed recently by Freeman et al65 who have compared apparent first
order degradation constant for thermal degradation of 7 molar AMP at CO2 loading of 0.4
mole per mole of amine and 135 ̊C with the MEA solution of identical concentration under
identical degradation conditions of CO2 loading and temperature. They found that this
constant for AMP was 21 x 10-9 sec-1 whereas for MEA it was 134 x 10-9 sec-1, thus
confirming at least 6 times lower thermal degradation tendency for AMP in comparison with
an equivalent MEA solution, when CO2 is dissolved in the solution.
In summary, the oxidative and thermal degradation of AMP is less likely compared with
MEA, when it is used for removing CO2 from oxygen containing flue gas. Various
carboxylates, ammonia, acetone, 2,4-Lutidine (Dimethyl pyridine) and 4,4-deimethyl-2oxazolidone are the main oxidative degradation products of AMP. Increasing the CO2
loading of AMP causes oxidative and thermal degradation of AMP to increase. Increase in
the temperature increases its rate of oxidative degradation though the thermal degradation
of AMP is less likely to be an issue until the solution temperature is more than 140 ̊C and the
solution is loaded with CO2. Transition metal ions (Fe2+, Fe3+ and Cu2+) have no obvious
effect on the oxidative degradation rates and that explains low corrosion potential of AMP. In
addition, it should be noted that AMP has no α-hydrogen and hence it is unable to form an
imine which is assumed to be a first step in the electron abstraction based oxidative
degradation mechanism for the primary amines (AMP is a sterically hindered primary amine)
as shown in Figure 2.
30
5. Degradation of Piperazine (PZ)
PZ is a cyclic diamine (molecular formula of C4H10N2) that shows enhanced kinetic
absorption rates for CO2, almost 1.5 to 2 times that of MEA. Hence, it is being used for quiet
sometime as a rate promoter for the slow reacting amine systems such as MDEA and AMP4.
The concentration of PZ in blends, when used as a rate promoter, is usually low (between
0.5 - 2.5 molal) because it is not highly water soluble at room temperature. When exposed to
water, either in aqueous solution or absorbed from air, PZ easily hydrates to hexahydrate
(PZ.6H2O). In aqueous solution with CO2, the two amino functions on PZ can react and
create numerous PZ-based species. The PZ species present in solution, as shown in Figure
21, include PZ carbamate (PZCOO-), PZ dicarbamate [PZ(COO-)2], protonated PZ (H+PZ),
diprotonated PZ (H2+2PZ) and protonated PZ carbamate (H+PZCOO-)4. The speciation of PZ
in solution is a function of its CO2 loading as shown in Figure 22. Since PZ has two amine
groups – in simple terms – double alkalinity compared to MEA, the CO2 loading of its
solution is usually described in terms of moles of CO2 per mole of alkalinity which is one half
of traditional loading, i.e. mole of CO2 per mole of amine4. In the absence of CO2 in solution,
PZ exists primarily as free PZ with a small portion of H+PZ. As the CO2 concentration
increases, carbamate of PZ, PZCOO-, begins to form with the H+PZ concentration increasing
and the free PZ concentration decreasing. Once a CO2 loading of about 0.28 mole CO2 per
mole alkalinity is reached, PZCOO- reaches its maximum concentration and H+PZ begins to
level out4. Leading up to this loading, the concentration of H+PZCOO- is steadily increasing,
lagging the production of both H+PZ and PZCOO-. After a loading of 0.28 mole CO2 per mole
alkalinity, the concentrations of H+PZ, PZCOO- and free PZ are decreasing while H+PZCOOis increasing. A loading of 0.5 mole CO2 per mole alkalinity represents the amount of CO2
needed in order to have half of the PZ basic groups react with CO2, while the other half are
protonated. This is essentially the maximum realistic loading. At this point, the solution is
primarily made of zwitterions H+PZCOO- with a small concentration of H+PZ present.
Through the entire range of loading, the concentration of PZ(COO-)2 is very small,
demonstrating that this species is not a preferred form of PZ. The concentrations of H+PZH+
and PZ(COO-)2 are not represented in Figure 22 because their concentrations are nearly
negligible. Whilst in the laboratory environment maximum CO2 loading of 0.5 mole per mole
of alkalinity is achievable with PZ, for industrial post combustion capture situation as
applicable to a coal-fired power plant, the maximum loading is unlikely to be more than 0.41
CO2 mole per mole alkalinity due to limitations of equipment size and maximum
concentration of CO2 possible in flue gas4, 66.
31
Figure 21 – Speciation of PZ in aqueous solutions
Figure 22 – Speciation of 8 m (molal) PZ at 40 ̊C
Whilst the solid-liquid transition temperature – the temperature at which a liquid solution will
first precipitate when cooled slowly - for plain aqueous solution of PZ increases with
concentration of PZ (Figure 23), Figure 24 shows that increasing the CO2 loading of solution
reduces the solid-liquid transition temperature4, 67-71. Thus, aqueous solutions of PZ with
concentrations as high as 8 to 10 molal (40 to 50% w/w) can be employed for CO2
absorption, if the solution is correctly loaded initially. This has the potential to reduce the size
of gas/liquid absorption system. Unfortunately, though PZ has boiling point lower than that
for MEA (146.5 ̊C v/s 170 ̊C), its volatility is similar to MEA due to non-ideal behaviour of
32
solution which has potential to result into the vapour phase losses of PZ comparable to
MEA, when used for CO2 emissions reduction4.
Figure 23 – Solid-liquid transition temperature for PZ with experimental data from
various investigations
Figure 24 – Solid-liquid transition temperature for CO2 loaded PZ solutions
5.1 Oxidative Degradation of PZ
Among various international research groups, the oxidative degradation of PZ has been
investigated exhaustively by Professor Garry Rochelle and his research students due to their
interest in investigating 8 molal (~40% w/w) PZ solution as one of the cost and energy
efficient alternatives to 30% w/w MEA. Two bench scale reactors, the Teflon oxidation
reactor (TOR) and the Integrated solvent degradation apparatus (ISDA), were used to study
oxidative degradation4. The TOR is a batch reactor. It was used to investigate the oxidative
degradation of PZ at low gas flow rate (100 mL/min) and rapidly oxidize PZ solutions with
higher levels of oxygen than expected in an industrial CO2 absorber. It was equipped with an
agitator (1400 rpm) to increase the mass transfer of oxygen in the solution. The ISDA
system cyclically created oxidizing and thermally degrading conditions in a single system. It
circulated 200 mL/min of solvent and simulated degradation conditions observed in an
absorber/stripper configuration designed for CO2 capture. Figure 25 shows the schematics
of the ISDA system. Technical description of both systems has been given elsewhere4, 72.
33
The oxidative degradation characteristics of 8 molal PZ were studied in these bench scale
reactors from 55 to 125 ̊C with inlet gas containing 40 to 98% oxygen and 2 to 6% CO2. The
CO2 loading of 8 molal PZ was kept 0.6 mole per mole of amine (or 0.3 mole CO2 per mole
alkalinity). Metal salts as aqueous sulphates were added to the PZ solution to study the
catalytic effect of iron (1 mM Fe2+), stainless steel metals or SSM (0.4 mM Fe2+, 0.1 mM Cr3+
and 0.05 mM Ni2+) and copper (4 mM Cu2+) on the rate of oxidative degradation4. Here as
well as in the entire report, the units, mM, stand for milimoles of the particular species in the
solution.
Figure 25 – Integrated solvent degradation apparatus for oxidative degradation study
The results of these investigations showed that the activation energy of PZ loss during
oxidative degradation is 40 ± 4.5 kJ per mole of PZ. PZ loss has a first order relationship
with partial pressure of oxygen in the inlet gas. The major identified oxidation products of PZ
are Ethylenediamine (EDA), N-Formyl Piperazine (FPZ), formate, ammonia, oxalate and
oxalyl amides4. Minor oxidative degradation products include acetate, acetyl amides, other
formyl amides, nitrite and nitrate when PZ is heavily oxidised. The rate of generation of total
formate (RTF) i.e., the sum of formate and formyl amides, was seen as a good indicator of
the overall level of PZ oxidation and this relationship was determined by data regression as:
RTF = 0.1098 x RPZ---------------- (24)
Table 4 below shows the results of oxidative degradation of 8 molal PZ with 0.3 mole CO2
loading per mole of PZ in the presence of iron, stainless steel metals and copper4.
Table 4 – Effect of metals on oxidative degradation of PZ @ 55 ̊C
Metals
Average PZ Loss or Product formation Rate
(mM/kg Solution/hr/kPa O2)
RPZ (measured) RTF (measured) RPZ (estimated)
None 6.2 ± 2.5 0.07 ± 0.04 0.6 ± 0.3
Iron 7.7 ± 7.0 0.08 ± 0.03 0.7 ± 0.3
Stainless steel 3.1 ± 4.3 0.09 ± 0.03 1.2 ± 0.9
Copper 84.4 ± 31 9.7 ± 4.0 107 ± 65
34
These results imply that iron and stainless steel metals (Fe2+, Cr3+ and Ni2+ ions in solution)
do not catalyse degradation of PZ, however, copper (Cu2+) is a strong catalyst for PZ
oxidation. Comparing the oxidative degradation behaviour of PZ in the presence of iron and
stainless steel with that of 30% w/w MEA under similar conditions shows that PZ is 5.5 to 6
times more resistant to degradation than MEA. A similar comparison between PZ and MEA
in the presence of copper shows that PZ is 2.6 times more resistant to copper catalysed
oxidative degradation than MEA is. This implies that carbon steel and stainless steel as
materials of construction for post combustion CO2 absorber are suited for PZ application but
copper based corrosion inhibitors such as copper borates or carbonates used for preventing
corrosion in sour gas treatment industry cannot be used for PZ based post combustion
capture plant.
The effect of temperature on the rate of PZ loss (RPZ) due to oxidative degradation over the
temperature range 55 to 125 ̊C in the presence of stainless steel material was determined
as:
RPZ = 9.29e-5389/T ---------------------- (25)
Where RPZ has units of mili-mole PZ/kg of solution/hr/kPa O2 and T has units of K̊.
With the oxidative degradation characteristics of PZ established in the bench scale reactors,
Freeman et al4, 65, 66 have estimated that for a 500 MWe (installed) power plant, a post
combustion CO2 capture plant based on 8 molal PZ solvent and operating at 90% CO2
capture with 120 ̊C as the stripper operating temperature, will have the oxidative degradation
loss of PZ as 2.1 moles (or 181 g) per metric ton of CO2 captured.
Whilst the ISDA system allowed the designated research group to study the oxidative
degradation of PZ at temperatures as high as 125 ̊C – bordering on to the region of thermal
degradation – and thereby understand the extent of likely oxidative degradation of CO2
loaded lean PZ solution in the stripper ends of cross exchanger (see Figure 1), The obtained
results have the following limitations:
• The impact of CO2 loading on the extent of oxidative degradation and the resulting
product distribution were not studied over the CO2 carrying capacity [0.79 mole CO2
per kg (H2O + PZ)] of 8 molal PZ solvent.
• The ISDA system is not a true absorber/stripper system in the sense that CO2 was
not cycled in and out of the solvent.
• Thermal reactor did not flash oxygen as is the case with industrial stripper and the
laboratory investigations had the situation of pure thermal degradation products also
undergoing high temperature oxidation.
• Accelerated degradation (98% O2 / 2% CO2) did not represent the industrial practice
where oxygen concentration seldom increases beyond 15% by volume in flue gas
and where CO2 concentration could be as high as 14% by volume, hence the
degradation product distribution could be skewed or extreme and unrepresentative of
what could occur in an industrial post combustion capture plant. This has potentially
a significant impact on the likely atmospheric emissions of PZ and its degradation
products when either PZ is used as a singular solvent or blended with other amines.
In light of the above limitations, the oxidative degradation results for PZ reported here should
be viewed with caution. In addition, there is currently no data available on the impact of SOX
on piperazine solvent, though NOX impact towards formation of relevant nitrosamines has
been investigated by Ashouripashaki and Rochelle73 and Jackson and Attala74.
35
Ashouripashaki and Rochelle73 have studied the kinetics of nitrosamine formation over 21 to
75 ̊C range in 8 molal PZ that had CO2 loading of 0.3 moles per mole of alkalinity. They used
Teflon Oxidation Reactor (TOR) referred earlier both in the low gas flow (100 mL/min) and
the high gas flow (8 Litre/min) modes. The gas stream contacting PZ solution contained 98%
by volume O2 and 2% by volume CO2. It was further spiked with 55 ppmv NO2. The PZ
solution had metal ions representing the stainless steel as material of construction added in
the form of water soluble sulphates. Mostly mono-nitrosopiperazine (MNPZ) formation was
detected in the liquid phase though Di-nitrosopiperazine (DNPZ) could have also formed.
The nitrosation reaction had activation energy of 47 kJ/mole of PZ and the order of reaction
was one with respect to both PZ and NO2 concentrations. Addition of formaldehyde in PZ
solution as a `corrosion inhibitor catalysed the nitrosation reaction slightly. Figures 26 shows
that the formation of nitrite in solution increases with time due to absorption of NO2 and there
is corresponding rise in the concentration of MNPZ due to reaction of PZ with nitrite at 55 ̊C
in the low gas flow reactor. Figure 27 shows the similar result for the high gas flow reactor.
These results clearly show that the yield of formation of nitrosation products is very low in
the laboratory environment.
Figure 26 – Formation of Nitrite from NO2 absorption and MNPZ production with time
100 mL/min, 55 ̊C, 50 ppmv NO2, 0.4 mM Fe, 0.1 mM Cr & 0.05 mM Ni
Figure 27 – Formation of Nitrite from NO2 absorption and MNPZ production with time
7.65 L/min, 55 ̊C, 50 ppmv NO2, 0.4 mM Fe, 0.1 mM Cr & 0.05 mM Ni
Low level of absorption of NO2 and corresponding low yield of MNPZ in the high gas flow
batch reactor of the identical reactor volume can be attributed to smaller residence time for
gas in the reactor. This implies that for the industrial scale absorption operation, the packing
36
height, liquid hold up in the absorber, liquid to gas ratio in the absorber and the liquid
residence time up to the stripper end of cross exchanger will decide the yield of MNPZ in
addition to NO2 concentration in the flue gas stream and the absorber temperature.
Jackson and Attala74 used 15% w/w PZ solution at 60 ̊C and synthetic flue gas with 8000
ppmv NOX to study the formation of nitroso products. The gas stream composition was 13%
by volume CO2, 5% by volume oxygen and the rest nitrogen. In addition to MNPZ and
DNPZ, N-Oxopiperazine and Piperazine nitramine were detected by Attala in the degraded
liquid sample. Figure 28 shows that Nitrosapiperazines are thermally stable at the operating
temperature of stripper and therefore their formation is unlikely to be reversed in the reboiler.
It should however be noted that 8000 ppmv NOX concentration in the synthetic flue gas
stream is way above what is usually noted with the industrial scale coal-fired power plant flue
gas stream (roughly 200 to 300 ppmv in the absence of FGD). Allowing for complete
conversion of NOX (NO + NO2) to Nitrosapiperazines, it is unlikely that CO2 rich piperazine
solution will carry nitroso compounds more than the stoichiometric equivalent weight of 300
ppmv NOX.
Figure 28 – Thermal resistance of MNPZ at 160 ̊C
At present, the quantification of individual products of NOX related degradation of PZ is not
available in the public domain literature. Additionally, there is no information available in the
public domain on the inhibitors for nitrosamines formation during post combustion capture of
CO2.
5.2 Thermal Degrada tion of PZ
Thermal degradation of PZ has been extensively studied by Professor Garry Rochelle and
his research4, 75, 76, 77 group at Texas University due to their interest in 8 molal PZ as a
solvent to replace 30% w/w MEA. They studied the thermal degradation characteristics of PZ
over the temperature range 135 to 175 ̊C, PZ concentration range 4 to 20 molal and CO2
loading from zero to 0.47 mole per mole alkalinity. They also studied the degradation effect
of metal ions in solution at 175 ̊C. Their results indicate that thermal degradation of PZ
follows first order and therefore the concentration of PZ in solution, CPZ, at any given time, t,
can be expressed as:
CPZ = CPZO × e-kt
Where, CPZO is the initial concentration of PZ. The reaction rate constant, k (sec-1), could be
defined as:
k = a × e-E/RT
37
Where, E is the activation energy for the thermal degradation step, R is the universal gas
constant and T is ̊K. Table 5 below shows experimentally determined degradation rate
constants for PZ solutions of different concentration with different CO2 loadings at different
temperatures. The activation energy E remains around 184 to 191 kJ per mole for PZ.
Table 5 – Measured k values for 4 to 20 molal PZ at different temperatures
PZ
Concentration
(molal)
CO2 Loading
(mole/mole alkalinity)
k x 10-9 (sec-1)
135 oC 150 oC 165 oC 175 oC
4 0.3 36 114
8 0 - 0.1 0.8 7.0
8 0.1 - - 18 66
8 0.2 - - - 79
8 0.3 1.0 6.1 31 132
8 0.4 - 7.9 41 171
8 0.47 - - - 24
12 0.3 - - 50 156
20 0.3 - 24 - 269
The above results indicate that unloaded PZ solution as well as those with low CO2 loading
(<0.2 mole per mole alkalinity) are fairly resistant to thermal degradation at normal stripper
temperatures in the range 120 to 135 ̊C. There is a slight increase in the rate constant, k, for
increased PZ concentration in the range 4 to 12 molal. In comparison with changes in CO2
loading and temperature, which produce over two orders of magnitude change in, k, the
effect of PZ concentration is definitely minor. The change in ‘k’ for 20 molal PZ is significant
in comparison with 4 to 12 molal data, and probably reflects the drastic changes in solution
composition and its degradation product distribution since the solution has high
concentration (62% w/w PZ) and high viscosity (>300 cP). The effect of CO2 loading is more
complex as the rate constant, k, seems to rise with loading up to 0.4 mole per mole alkalinity
but then drops at 0.47 value for 175 ̊C. Based on modelling work, Freeman and Rochelle75
explain that the concentration of free PZ and PZ carbamate (PZCOO-) are low above 0.4
mole CO2 per mole alkalinity while concentration of bicarbonate (HCO3-) reaches a
significant level (see also Figure 22). The presence of CO2 in the form of HCO3- appears to
deactivate the catalytic effect of CO2 when present in the form of other species, as seen at
lower loadings.
Table 6 below compares the first order thermal degradation rate constants for PZ, AMP,
MDEA and MEA at various concentrations and CO2 loading to indicate their relative
degradation tendencies4, 65, 75, 76, 77. This data shows that PZ is most thermal degradation
resistant amine followed by AMP, a hindered amine, then MDEA, a tertiary amine and finally
MEA, a primary amine. The thermal degradation activation energy for 7 molal AMP is 112 kJ
per mole and for 7 molal MEA it is approximately 134 to 156 kJ per mole.
38
Table 6 – First order k values for thermal degradation of various amines
Concentration
(molal)
CO2 Loading
(mole/mole alkalinity)
k x 10-9 (sec-1)
Amine
120 oC 135 oC 150 oC
PZ 8 (40% w/w) 0.3 - 1.0 6.1
AMP 7 (38% w/w) 0.4 8 21 86
MDEA 7 (45% w/w) 0.1 - - 283
MDEA 7 0.2 42 438
MEA 3.5 (18% w/w) 0.2 13 41
MEA 3.5 0.4 22 109
MEA 7 (40% w/w) 0.2 28 90 397
MEA 7 0.4 29 134 828
To study the effect of metal ions on thermal degradation of PZ, Freeman et al65 subjected 8
molal PZ with 0.3 CO2 loading to 5 mM Cu2+/0.1 mM Fe2+ and 0.25 mM Fe2+/0.6 mM
Cr3+/0.25 mM Ni2+ ions in solution at 175 ̊C and found no catalytic effect of metals on the
thermal degradation rates. Thus, iron, stainless steel and copper materials of construction
are suitable in the stripper for the post combustion capture application of PZ. However as
mentioned in the previous section, the oxidative degradation of PZ is accelerated by
presence of copper ions in the solution, therefore copper based construction materials and
corrosion inhibitors must be avoided for CO2 absorption using PZ.
Taking into account, little catalytic effect of iron and steel materials on both oxidative and
thermal degradation of PZ, particularly in the concentration range of 4 to 8 molal with CO2
loading up to 0.4 mole per mole alkalinity, a significant message from the data shown in
Tables 5 and 6 is that PZ has potential for a least corrosive and stable reboiler operation at
150 ̊C. Assuming 10 minutes of liquid residence time in the stripping section and using the
known CO2 carrying capacity of 8 molal PZ as 0.79 mole CO2 per kg of (H2O + PZ), Chen
and Rochelle78 have estimated the thermal degradation loss of PZ at 150 ̊C as 0.043 mM PZ
per mole of CO2 captured, i.e. 84 gm per tonne of CO2.
The characterisation and quantification of thermal degradation products at 135 to 175 ̊C
carried out by Freeman4 in the laboratory environment for concentrated aqueous PZ loaded
with CO2 shows that N-formylpiperazine (FPZ), N-(2-aminoethyl)piperazine (AEP) and NH4+
(ammonium) are major products. Other products include 2-imidazolidone (IMID), N-(2hydroxyethyl) Piperazine (HEP), N-Ethyl Piperazine (EPZ), Ethylenediamine (EDA), NMethyl Piperazine (MPZ), formate, acetate, and other heat stable salts such as oxalate and
glycolate. In the absence of CO2, PZ degradation does occur but half as fast as the loaded
solution. Without CO2 loading, IMID and formates including FPZ are not formed, instead
ethylenediamine (EDA) is formed. Table 7 shows the thermal degradation products formed
from 8 molal CO2 loaded PZ after 20 weeks at 165 ̊C and their distribution. It shows that only
74% and 63% of N and C lost from initial PZ can be accounted from mass balance with
known degradation products. In other words, the remaining lost N and lost C is present in the
degradation product mixture as unidentified products. Of the total degradation products,
FPZ, NH4+ (ammonium) and AEP account for 57% and 45% of N and C lost. The detailed
results for other temperature, PZ concentration and CO2 loading conditions are given in the
tabulated form in the PhD dissertation by Freeman4. The significance of choosing only 165 ̊C
thermal degradation condition for presenting Freeman’s results in this report is explained
further below.
39
Table 7 – Thermal degradation mass balance for PZ
(8 molal PZ, 0.3 mole CO2 per mole alkalinity loading, 165 ̊C)
Compound Concentration (mM/kg of solution)
Lost N
(%)
Lost C
(%)
PZ lost in 20 weeks 1729
CO2 lost in 20 weeks 888
N-Formyl Piperazine (FPZ) 558 33 36
NH4+ 496 14 -
N-(2-Aminoethyl) Piperazine (AEP) 118 10 9
Total 57 45
2-Imidazolidone (IMID) 109 6 4
N-(2-hydroxyethyl) Piperazine (HEP) 61 4 5
N-Ethyl Piperazine (EPZ) 54 3 4
Ethylenediamine (EDA) 53 3 1
N-Methyl Piperazine (MPZ) 21 1 1
Formate 108 - 1
Total Acetate 36 - 1
Total 74 63
Freeman4 has postulated that the thermal degradation of PZ proceeds through a variety of
SN2 type substitution reactions that result in the ring opening and polymerisation products. In
the first step of mechanism, 1-[2-[(2-aminoethyl) amino] ethyl] PZ (AEAEPZ) is believed to
be formed from a ring opening SN2 reaction of PZ with H+PZ. The limiting rate was found to
require protonated amine, but not CO2. AEAPZ then reacts with dissolved CO2 to form urea
of AEAEPZ which is suspected to create other products. Protonated AEAEPZ in the solution
produces N-(2-Aminoethyl) Piperazine (AEP) and ethylenediamine (EDA) as well as NH4+
ions. EDA reacts for creating 2-Imidazolidone (IMID) while Hoffman elimination and
subsequent hydration of AEP produces N-(2-hydroxyethyl) Piperazine (HEP). More details of
the thermal degradation reaction mechanisms are given in the PhD dissertation by
Freeman4.
Table 6 not only indicates relative thermal degradation tendencies of various amines in
terms of their apparent first order degradation rate constants at a particular temperature of
interest, this data can also be used to determine the maximum equivalent stripper
temperature (MEST) for any amine for which the thermal degradation rate is required to
remain the same as that of 30% w/w MEA, the base case amine solvent currently used
commercially for post combustion capture. For 7 molal or 30% w/w MEA, the optimum
stripper temperature with respect to minimum cost of energy for steam stripping and
minimum cost of solvent replacement has been estimated by Davis11 as approximately 120
̊C (see Figure 11). This result is supported by the historical operation of MEA systems
successfully at temperature between 115 and 120 ̊C. Assuming that the rich loading for MEA
in the reboiler is 0.4 mole CO2 per mole of MEA – same as mole CO2 per mole alkalinity – in
this case, its thermal degradation rate is expressed by its apparent first order rate constant,
2.91 x 10-8 sec-1. Therefore, maximum thermal degradation temperature for an amine of
interest that matches this apparent first order rate constant value is its maximum equivalent
stripper temperature (MEST), if the degradation rate needs to not exceed that of MEA.
Figure 29 below show`s this concept of MEST limit applied to 8 molal PZ and the stripper
temperature determined for this solvent as approximately 163 ̊C77.
40
Figure 29 – Determination of MEST for 8 molal PZ with 0.3 mole CO2/mole alkalinity
Figure 29 also demonstrates that the thermal stability of 8 molal PZ is two orders of
magnitude more than 7 molal MEA. This methodology of determining MEST for any amine
involves the assumption that all amines degrade thermally in a manner that follows first order
rate or the degradation data can be approximated as first order. It does not take into account
the cost of amine involved, for example in the case of an expensive amine such as PZ (cost
per kg at least 2.5 times more than that of MEA) it may be prudent to operate at lower
temperature than the MEST point in order to reduce the degradation rate even lower than
that encountered with 7 molal MEA. However, this methodology of determining MEST
provides a reasonable starting for the stripper design and suggests maximum propensity of
an amine to thermally degrade. Table 8 presents the MEST values for a select number of
amines and their blends4.
Table 8 – Estimated MEST for select amines and their blends for CO2 capture
(@ k =2.91x10-8 Sec-1, Source: Freeman’s PhD Dissertation4)
Amine Concentration
(Molal)
CO2 Loading
(Mole/mole
alkalinity)
Temperature
Range
(oC)
Activation
Energy
(kJ/mole)
MEST
(oC)
PZ 8 0.3 135 -175 184 163
PZ 8 0.4 135-175 191 162
AMP 4.81 0.5 135 140
AMP 7 0.4 120-150 112 137
MDEA 3.6 0.5 135 120
MDEA 7 0.1 150 128
MDEA 7 0.2 100-135 128
MDEA 8.4 0.4 135 119
MEA 3.5 0.4 100-135 129 122
MEA 3.5 0.5 100-135 86 111
MEA 7 0.2 120-150 123 121
MEA 7 0.4 100-150 157 121
MEA 7 0.5 100-150 162 117
MEA 11 0.2 120-135 125
MEA 11 0.4 120-135 116
MEA 11 0.5 120-135 112
PZ/AMP 6/4 0.4 135-150 134
MDEA/PZ 7/2 0.11 135-150 138
MEA/PZ 7/2 0.4 100-150 84 104
MEA/AMP 7/2 0.4 100-150 146 123
41
The MEST results for the blended CO2 capture results shed interesting light on the
interaction of PZ and other amines in a blend at high temperature. Davis11 as well as
Closmann72 report that PZ is preferentially degraded when blended with both MEA and
MDEA. Their analysis of ‘k’ values indicates that PZ is degraded 12 to 15 times faster than
MDEA and 4 to 8 times faster than the total amine concentration when in 7 molal MDEA/2
molal PZ blend. In 7 molal MEA/2 molal PZ blend, PZ degrades only 1.2 to 1.7 times faster
than MEA and 1.1 to 1.9 times faster than total amine. Based on overall amine loss, the
blend of MEA and PZ is seen as significantly less stable with MEST value 34 ̊C lower than
that for MDEA/PZ blend because MEA and PZ are both highly unstable in this blend. In the
MDEA/PZ blend, however, MDEA degrades at similar rates as it does alone while PZ loss is
accelerated.
In summary, PZ is a solvent with high resistance to both oxidative and thermal degradation
and can be used at high concentration to compact the CO2 absorption/desorption system
and thereby reduce its cost, provided the process operating conditions are maintained within
its solubility limits. Its degradation products such as ammonia and ethylenediamine can
potentially emit to atmosphere. It does degrade in the presence of NOX to produce mononitrosopiperazine. Currently, there is little information available on its degradation due to
SOX, particularly in the coal-fired power plant flue gas. Though, iron and steel do not seem to
catalyse its degradation particularly below 150 ̊C and these materials could be used in the
construction of absorber and stripper, it is very susceptible to degradation by copper.
Therefore, copper based materials of construction and corrosion inhibitors are detrimental to
efficient performance of PZ as a solvent. Unfortunately, PZ gets preferentially degraded
when mixed with other amines such as MDEA and MEA.
6. DEGRADATION OF METHYL-DIETHANOLEAMINE (MDEA)
MDEA is a tertiary amine and unlike MEA, a primary amine, forms bicarbonate ions in
solution as shown below instead of carbamate ions when CO2 is absorbed in its aqueous
solution. This gives MDEA the CO2 carrying capacity per mole twice that of MEA.
MDEA + CO2 + H2O  MDEAH+ + HCO3-
MDEA has been used for decades in the gas processing industry for acid gas removal (a
mixture of CO2, H2S, COS, CS2, mercaptans etc) where feed gas composition is typically
75% C1 to C5 alkanes and no oxygen, SOX or NOX. Under these conditions, solvent
degradation is dominated by thermal and CO2 catalysed processes. MDEA, being a tertiary
amine, is slow reacting and needs a catalytic rate promoter such as piperazine, PZ, or a fast
reacting amine such as, monoethanolamine, MEA, in solution to accelerate CO2 absorption.
A blend of 7 molal MDEA with 2 molal PZ is often commercially used in the gas processing
industry72. Whilst thermal degradation of MDEA has been studied by Chakma and Meisen79
under non-oxidative environment, the detailed information on its oxidative degradation has
become available only recently due to interest of Garry Rochelle and his research group72,
80, 81 at the University of Texas at Austin, USA, in assessing blends of MDEA and PZ as
suitable alternative for MEA as a solvent for the post combustion capture of CO2.
6.1 Oxida tive Degrada tion of MDEA
Early work on the oxidative degradation of MDEA, though of a limited nature, has been by
Rooney et al17 who found that 30% w/w aqueous solution of MDEA, when exposed at 80 ̊C
to 100 mL/min of a gas stream containing 50% by volume CO2 and 50% by volume O2, had
only 1.6% amine loss compared to 33% amine loss for 15% w/w MEA solution under
identical conditions. The degraded MDEA solution contained heat stable salts, primarily the
42
acetate, formate and glycolate. They also found that 50% w/w MDEA solution had less heat
stable salts in comparison with 30% w/w MDEA solution, when 5.5 mL/min of air was
bubbled through solutions for 28 days. Tables 9 and 10 below present their results.
Table 9 – Results of oxidation studies of 30% w/w MDEA at 80 ̊C
Days 0 7 14 21 28
Acetate (ppm) <10 82 to 120 202 to 205 332 to 353 437 to 449
Formate (ppm) <10 80 to 107 121 to 126 167 to 187 220 to 249
Glycolate (ppm) <10 278 to 186 411 to 430 561 to 593 667 to 704
Table 10 – Results of oxidation studies of 50% w/w MDEA at 80 ̊C
Days 0 7 14 21 28
Acetate (ppm) <10 21 54 83 111
Formate (ppm) <10 93 155 215 236
Glycolate (ppm) <10 224 338 431 512
To study the effect of CO2 loading on oxidative degradation of MDEA, both 30 and 50% w/w
MDEA solutions were loaded with 0.25 mole CO2 per mole amine and 5.5 mL/min of air was
bubbled through the solution at 80 ̊C for 28 days. Tables 11 and 12 show the results of these
investigations.
Table 11 – Results of oxidation studies of CO2 loaded 30% w/w MDEA at 80 ̊C
Days 0 7 14 21 28
Acetate (ppm) <10 24 81 158 265
Formate (ppm) <10 60 101 152 209
Glycolate (ppm) <10 156 261 379 504
Table 12 – Results of oxidation studies of CO2 loaded 50% w/w MDEA at 80 ̊C
Days 0 7 14 21 28
Acetate (ppm) <10 12 35 65 93
Formate (ppm) <10 70 143 225 312
Glycolate (ppm) <10 120 225 341 431
These results show that CO2 loaded 50% MDEA has better resistance to degradation than
CO2 loaded 30% w/w MDEA. In addition to the heat stable salts, Rooney et al17 detected
approximately 1600 ppm DEA in the MDEA solution samples with or without CO2 loading
after 4 weeks. When the solutions were heated in nitrogen blanket to 80 ̊C, no DEA was
detected which meant that oxygen was contributing towards DEA formation.
The research group at University of Regina, Canada, has studied the oxidative degradation
of MDEA invariably as a blended amine with MEA. Several studies82-86 on the kinetics and
product distribution from the oxidative degradation of MDEA/MEA blends have been
43
reported. The temperature range for their investigations was typically 55-120 ̊C, the overall
amine concentration was 5-9 moles/Litre, the oxygen pressure was 250 kPa and CO2
loading was 0 to 0.4 mole per mole of total amine. Supap et al, Bello and Idem and Lawal
and Idem reported vast amount (80 to 100) of degradation products. The results do not
seem to reflect the systematic variations in reaction conditions, neither are they supported by
plausible reaction schemes. The expected demethylation of MDEA to diethanolamine (DEA)
and formation of methyl aminoethanol (MAE) are identified for some but not all experiments.
Some general trends however do appear, for example MDEA is more prone to oxidative
degradation than MEA and it is preferentially oxidised in the blend. The stability of
MDEA/MEA blends to oxidative degradation decreases in the order MDEA/MEA/O2 >
MDEA/MEA/O2/CO2 > MDEA/MEA/CO2.
Closmann72 studied oxidative degradation of MDEA as a single amine and as a blend with 2
molal PZ. He used both the low gas flow and the integrated solvent degradation apparatus
(ISDA) mentioned at Section 6.1. The temperature condition for the low gas flow reactor
varied from 55 to 70 ̊C whereas for the ISDA system (Figure 25), the oxidative degradation
temperature remained at 55 ̊C and the thermal degradation section temperature varied up to
130 ̊C. To study the oxidative degradation at accelerated rate a gas mixture of 98% by
volume O2 and 2% by volume CO2 was used in the headspace above the solution. Metal ion
catalysis effect towards the oxidative degradation was assessed using aqueous solutions of
FeSO4.7H2O, Cr(SO4)3.23H2O and NiSO4.6H2O at concentrations of 0.4 mM Fe2+, 0.1 mM
Cr3+ and 0.05 mM Ni2+ respectively. 7 molal MDEA at a nominal initial loading of 0.1 mole
CO2 per mole alkalinity (or 0.1 mole CO2 per mole of MDEA) was the solvent utilised in all
experiments. In the ISDA system, the solvent circulation rate was maintained at 200 mL/min.
Closmann72 identified major oxidative degradation products of MDEA as primary amine
monoethanol amine (MEA), secondary amines Methyl-aminoethanol (MAE) and
Diethanolamine (DEA), the amino acids bicine, glycine, and hydroxyethyl sarcosine (HES),
formyl amides of MAE and DEA, ammonia and heat stable salts formate, glycolate, acetate
and oxalate. He used the production rate of total formate (formate plus formyl amides of
MAE and DEA) as an indicator to differentiate the impact of various degradation parameters
on the oxidative degradation. Closmann’s results for the oxidative degradation in low gas
flow reactor show that in the temperature range 55 to 70 ̊C, MDEA degradation rate was 0.4
mM/hr and alkalinity (nitrogen) loss rate was less than 0.1 mM/hr. The total formate
production rate was also less than 0.03 mM/hr in this range despite the Fe2+ concentration in
solution at 1 mM. This implies that in an industrial scale MDEA based CO2 capture plant,
oxidative degradation of MDEA is most likely to shift to the cross exchanger where
temperatures in excess of 70 ̊C are possible and carbon steel as a material of construction
for the absorber has no significant adverse impact. When the ISDA system was used to
study the effect of temperature cycling on the solvent degradation with the oxidative reactor
temperature in the system at 55 ̊C but the thermal reactor temperature varying from 55 to
120 ̊C, the rates of MDEA and alkalinity loss and the rate of total formate production was as
given in Table 13. In this case, 7 molal MDEA carried stainless steel metal salts at the
concentration of 0.4 mM Fe2+, 0.1 mM Cr3+ and 0.05 mM Ni2+. These results help in
understanding the effect of temperature on oxidation rates. The results clearly indicate that
MDEA degradation increased as the solvent was cycled to a higher temperature. MDEA and
alkalinity loss rates increased with thermal reactor temperature as did the rates of formate
and total formate production. This trend supports the conclusion that degradation of MDEA is
a high-temperature oxidative process. However, the data also demonstrate that the increase
in formate production between 100 and 120 ̊C is lower than the increase observed between
90 and 100 ̊C, indicating that the oxidation processes become dissolved oxygen limited
between 100 and 120 ̊C.
44
Table 13 – Oxidative degradation of 7 molal MDEA in ISDA system
0.1 mole CO2/mole alkalinity loading, 98% O2/2% CO2 gas
Temperature
(oC)
MDEA
Loss
(mM/hr)
Alkalinity
Loss
(mM/hr)
Formate
Production
(mM/hr)
Total Formate
Production
(mM/hr)
55 0.9 0.9 0.005 0.01
80 0.9 1.0 0.034 0.039
90 2.9 2.2 0.12 0.15
100 4.1 3.1 0.18 0.31
120 4.6 4.9 0.28 0.34
To study the effect of dissolved oxygen concentration, Closmann ran the ISDA system’s
oxidative section at 55 ̊C and the thermal section at 90 ̊C but with 98% air/2% CO2 gas
mixture in one case and 98% N2/2% CO2 in other case. The results are shown in Table 14.
Both formate and total formate production rates relate to oxygen concentration.
Table 14 – Effect of dissolved O2 on degradation of 7 molal MDEA in ISDA system
0.1 mole CO2/mole alkalinity loading, 1mM Fe2+
Gas Mixture
MDEA
Loss
(mM/hr)
Alkalinity
Loss
(mM/hr)
Formate
Production
(mM/hr)
Total Formate
Production
(mM/hr)
98% N2/2% CO2 1.9 1.6 0.013 0
98% Air/2% CO2 0.2 0.2 0.044 0.058
98% O2/2% CO2 2.9 2.,2 0.12 0.15
Using experimental observations of the ISDA system and the assumptions given below,
Closmann72 has proposed the oxidative degradation model (Equation 26) for 7 molal MDEA
over the temperature range 55 to 130 ̊C:
• Degradation of MDEA follows the first order reaction rate kinetics in terms of
the dissolved oxygen concentration,
• The ISDA system behaves as an ideal plug flow reactor (PFR),
• The PFR operates almost isothermally,
• Very little degradation occurs at low temperature (< 70 ̊C) and hence, the
oxidative reactor in the ISDA system merely saturates the MDEA solution with
oxygen,
• Most of the oxidative degradation occurs in the thermal reactor of the ISDA
system, and
• The PFR volume is that of the thermal reactor.
∆ [Prod] = {(PO2*S*Q*t) / (KH*VTOT)}*{1-exp (-k1*VTR/Q)} ----------- (26)
Where various terms in Equation 26 have the following meaning:
∆ [Prod] = The quantity of degradation product formed, mM
PO2 = The partial pressure of Oxygen in the gas stream, kPa
45
S = Stoichiometric coefficient for Oxygen for a particular product formation
Q = Solvent flow rate, m3/hr
t = Time for the oxidative degradation, hr
KH = Henry’s coefficient for oxygen solubility at the operating temperature of
oxidative reactor in the ISDA system
VTOT = Total volume of the ISDA system, m3
K1 = First order reaction rate constant for product formation at temperature
T1, hr
-1
VTR = Volume of the thermal reactor in the ISDA system, m3
The first order reaction rate constant K1 for formation of a particular degradation product at
temperature T (oK) is related to the rate constant K0 for the base case temperature T0 by the
Arrhenius Equation 27.
K1 = K0*exp[-EA/R{1/T1 – 1/T0}] ---------- (27)
Where, EA is the activation energy for formation of a particular degradation product and R is
the universal gas constant.
The above model has essentially three parameters viz., EA, K0 and S. Table 15 lists the
values of these parameters for the formation of formate, total formate (includes amides) and
Bicine as determined by Closmann72 which gave him the best fit between the experimentally
observed and predicted initial rates of formation of these degradation products from 7 molal
MDEA over the temperature range 55 to 130 ̊C.
Table 15 – 7 molal MDEA oxidative degradation model results
Parameter K0 (hr-1)
EA
(kJ/mole)
S
(mole Product per mole O2)
Formate 2.6 151 0.09
Total Formate 3.4 152 0.14
Bicine 0.32 244 0.35
Closmann utilised the above model to estimate the limit for MDEA solution exposure to high
temperature in the cross heat-exchanger (Figure 1) and in piping to the steam stripper. An
assumption was made that most of the dissolved oxygen will strip out of the solution at the
top of the stripper and the residence time for solution in the cross exchanger plus piping is
30 seconds. Accordingly, for 10% oxygen consumption towards degradation in the cross
exchanger/piping region of 7 molal MDEA based CO2 capture plant, the average
temperature tolerance was estimated to be 104 ̊C. For 20% and 50% oxygen consumption,
the temperature tolerance increases to 108 and 116 ̊C. Finally using the oxidative
degradation model results, Closmann estimated that for a 500 MW coal-fired power plant, 7
molal MDEA based post combustion CO2 capture plant that has 90% CO2 capture duty will
have 0.29 moles of MDEA loss per ton of CO2 captured, if the CO2 carrying capacity of the
solution is 0.5 mole CO2 per kg of MDEA solvent.
The oxidative degradation model of Closmann described above has limitations as below:
46
• The ISDA system is not a true absorber/stripper system in the sense that CO2 was
not cycled in and out of the solvent.
• Thermal reactor did not flash un-reacted oxygen as is the case with industrial stripper
and the laboratory investigations had the situation of pure thermal degradation
products also undergoing high temperature oxidation.
• Accelerated degradation (98% O2 / 2% CO2) did not represent the industrial practice
where oxygen concentration seldom increases beyond 15% by volume in flue gas
and where CO2 concentration could be as high as 14% by volume, hence the
degradation product distribution seen by Closmann could be skewed or extreme and
unrepresentative of what could occur in an industrial post combustion capture plant.
This has potentially a significant impact on the likely atmospheric emissions of MDEA
and its degradation products.
• The oxidative degradation model assumes isothermal plug flow reactor behaviour
which is rarely the case for an industrial absorber/stripper system.
• The oxidative degradation model looks at the rate of degradation of MDEA in terms
of only oxygen consumption and involves estimation of floating parameters EA, K0
and S. The initial values (guess) for these parameters were obtained by plotting the
rate of degradation of MDEA or the rates of degradation products formation as a
function of temperature and/or oxygen partial pressure. These values were improved
upon in steps using the error minimisation mathematical techniques till the predicted
rates of degradation product formation matched with the experimentally measured
rates. In that sense, the parametric estimation is at best a curve fitting exercise and
the parameter values derived from the laboratory investigations may not hold for the
industrial situation.
• The oxidative degradation model is valid for estimating only the initial rate of
degradation products formation and not the accumulation of degradation products
over time. This model is also applicable over only a limited temperature range and
assumes the maximum concentration of dissolved oxygen in the spent solvent as
that decided by Henry’s law at 55 ̊C which assumes absorber operating isothermally
at that temperature which is not usually the case with industrial absorbers.
In summary, the oxidative degradation of MDEA produces primary amine monoethanolamine
(MEA), secondary amines Methyl-aminoethanol (MAE) and Diethanolamine (DEA), the
amino acids bicine, glycine, and hydroxyethyl sarcosine (HES), formyl amides of MAE and
DEA, ammonia and heat stable salts formate, glycolate, acetate and oxalate. At present, the
study of oxidative degradation of MDEA is still progressing and there is no data available to
explain the impact of SOX and NOX on the oxidative degradation, though Mimura et al60
indicate that 30% w/w MDEA absorbs NO2 and CO2 simultaneously with the liquid side mass
transfer co-efficient value of 1.5 x 10-7 mol/s/Pa/m2. A patented CANSOLV process87 uses
an aqueous amine solution for high selective absorption of SO2 from large variety of gas
streams. The experimental evidence by Closmann72 of dealkylation of MDEA to form
secondary amines, DEA and MAE, points towards the potential for the formation of nitrosocompounds (N-nitroso diethanolamine and 2-methylnitrosoamino-ethanol) in an industrial
situation. Moreover, there is a possibility that MDEA could degrade to diethylamine which
can form diethylnitrosamine and diethylnitramine.
47
6.2 Thermal Degrada tion of MDEA
Thermal degradation of MDEA was studied first by Chakma and Meisen79 purely in terms of
understanding the applicability of 20 to 50% w/w MDEA as a solvent for CO2 capture from
the commercial gas treatment operations where oxidative environment is non-existent. They
used 600 mL stirred autoclave to expose 20 to 50% w/w aqueous MDEA solution to CO2
(partial pressure 1.38 to 4.24 MPa) mixed with N2 at 100 to 200 ̊C. Thermal degradation
products were identified by gas chromatography and mass spectrometry. These researchers
found that MDEA did not degrade even after 240 hours at 200 ̊C when it was exposed to
only nitrogen gas but the presence of CO2 immediately catalysed thermal degradation.
Degradation however remained slow at temperatures below 120 oC even after 300 hours.
The degradation products formed at temperatures well in excess of 120 oC were same as
those at 120 C̊. Higher temperatures merely accelerated the rate of degradation. Overall
MDEA degradation followed first order rate kinetics up to 170 ̊C but beyond that the kinetics
moved to higher order. The overall degradation rate constant increased with initial solution
concentration up to about 3.5 moles MDEA per litre but then started to decrease with higher
concentration because with increasing MDEA concentration, availability of water to form
protonated MDEA started to decrease, thus slowing down the degradation reaction.
Increasing CO2 partial pressure increased its solubility and therefore accelerated
degradation rate.
Overall, Chakma and Meisen79 determined that CO2 catalysed thermal degradation of MDEA
produced 2-dimethylaminoethanol (DMAE) as a primary product which then further degraded
into various other products as per the reaction equations given below. Table 16 lists the
degradation products that were identified with GCMS techniques by these researchers.
MDEA + CO2 + H2O ↔
𝐾𝑒𝑞
MDEAH+ + HCO3- --------------- (28)
MDEA + MDEAH+ →
𝐾2
DMAE + EO + DEA -------------- (29)
2DMAE + CO2 →
𝐾3
TMA + EO + MAE ----------- (30)
2MAE + CO2 →
𝐾4
DMP ------------------------- (31)
EO + H2O →
𝐾5
EG -------------- (32)
DEA + CO2 + MAE →
𝐾6
HMP ----------------------- (33)
DEA + EO →
𝐾7
TEA ------------- (34)
HMP + EO →
𝐾8
BHEP ------------ (35)
TEA + DEA + CO2 →
𝐾9
TEHEED ------------------ (36)
DMP + EO →
𝐾10
HMP ------------- (37)
DEA + CO2 →
𝐾11
HEOD ------------ (38)
2DEA + CO2 →
𝐾12
THEED ----------- (39)
THEED →
𝐾13
BHEP ------ (40)
Where Keq is the equilibrium rate constant for reaction of MDEA with dissolved CO2 and rest
of the K are individual reaction rate constants. Table 17 lists the individual Arrhenius
frequency factor and activation energy values for these rate constants.
48
Table 16 – CO2 catalysed thermal degradation products of MDEA
Compound Name Abbreviation
Methanol Me-OH
Ethylene Oxide EO
Trimethylamine TMA
N,N-Dimethylethanamine DMEA
Ethylene Glycol or 1,2 Ethanediol EG
2 - Dimethylamino Ethanol DMAE
4 - Methyl Morpholine MM
1,4 Dimethyl Piperazine DMP
Methyl Diethanolamine MDEA
Diethanolamine DEA
1 - (2 - hydroxyethyl) – 4 – Methyl Piperazine HMP
Unidentified
Triethanolamine TEA
N,N-bis-(2 – hydroxyethyl) – Piperazine BHEP
3 – (hydroxyethyl) – 2 – Oxazolidone HEOD
N,N,N-tris-(hydroxyethyl) Ethylenediamine THEED
Tetra – (hydroxyethyl) Ethylenediamine TEHEED
Table 17 – The Arrhenius parameters for MDEA degradation rate constants
Rate
Constant
Units KFreq
(Sec-1)
Activation Energy
(kJ/mole)
K2 Litre-ions/mole2/hr 2.34 x 104 57.4
K3 Litre2/ mole2/hr 4.77 x 105 50.9
K4 Litre2/ mole2/hr 2.42 x 105 60.6
K5 Hr-1 2.60 x 104 56.2
K6 Litre2/ mole2/hr 5.82 x 104 52.4
K7 Litre/ mole/hr 2.70 x 106 65.4
K8 Litre/ mole/hr 2.28 x 104 54.4
K9 Litre2/ mole2/hr 6.13 x 104 56.3
K10 Litre/ mole/hr 3.47 x 104 79.4
K11 Litre/ mole/hr 5.39 x 106 75.2
K12 Litre/ mole/hr 2.00 x 107 79.4
K13 Hr-1 3.80 x 1011 12.8
Using the above kinetic model, Chakma and Meisen79 showed that the predicted values for
the degradation product concentrations matched very closely with the experimentally
observed values. However, it should be noted that these researchers used the degradation
temperatures and the CO2 partial pressures way above the values encountered with post
combustion capture mainly to accelerate the degradation chemistry and obtain the results
within hours. Since thermal degradation involves a number of competing reactions in series
as well as in parallel, the acceleration of degradation by utilising very high temperatures and
CO2 partial pressures may have impacted on the product distribution, even if the chemical
nature of the products remains same with what was observed by these researchers at 120
̊C. Interestingly, these researchers confirm that there was no appreciable thermal
49
degradation even in the presence of CO2 at 120 C̊ and therefore, operating CO2 stripper for
MDEA based post-combustion capture process around 120 ̊C will leave no major cause for
thermal degradation products formation and their likely atmospheric emissions. This is also
being pointed out by Freeman4 in her concept of MEST for stripper in Table 8 for 3.6 to 8.4
molal MDEA solutions loaded with CO2.
In addition to Chakma and Meisen79, Closmann72 has also studied thermal degradation of
MDEA in the temperature range 120 to 150 ̊C for 7 molal solution concentration and CO2
loading up to 0.2 mole per mole alkalinity. Table 18 below shows MDEA loss rate. Based on
the earlier work by Chakma and Meisen79, Closmann estimated that the degradation kinetics
followed first order rate and accordingly estimated the overall degradation rate constant for 7
molal MDEA as a function of temperature. For 150 ̊C, Closmann determined the first order
rate constant as 8.06 x 10-7 sec-1 which closely matches the value of Chakma and Meisen. In
addition to this, Closmann’s estimate of 60 kJ per mole MDEA for the activation energy for
thermal degradation is in line with Chakma and Meisen’s estimate of 57.4 kJ per mole
MDEA. However, there is some disagreement between these two bodies of work in relation
to the products of thermal degradation. Table 19 lists the products observed by Closmann in
his thermal degradation work where ethylene oxide, ethylene glycol and other compounds
were not observed as products of degradation.
Table 18 – MDEA loss rate in thermally degraded 7 molal MDEA
Loading
(Mole CO2/ mole Alkalinity)
Temperature
(oC)
MDEA Loss
(Milimoles/hr)
0 150 0.0 ± 0.37
0.1 120 0.0 ± 0.5
0.1 135 0.5 ± 0.3
0.1 150 1.2 ± 1.5
0.2 120 1.6
0.2 135 2.2
0.2 150 3.0 ± 0.7
Table 19 – Thermal degradation products of 7 molal MDEA
Compound Name Abbreviation
N,N-Dimethylethanamine DMEA
2-Dimethylamino Ethanol DMAE
N-(2-hydroxyethyl)-N-methyl Formamide MAE-Amide
Diethanolamine DEA
1,4 Dimethyl Piperazine DMP
N,N-bis-(2-hydroxyethyl) Formamide DEA-Amide or AEMP
1-(2-hydroxyethyl)–4–Methyl Piperazine HMP
Triethanolamine TEA
N,N-bis-(2–hydroxyethyl)-Glycine Bicine
N,N-bis-(2-hydroxyethyl)-Piperazine BHEP
N,N,N-tris-(hydroxyethyl) Ethylenediamine THEED
Methyl-N,N,N-tris-(hydroxyethyl) Ethylenediamine MTHEED
In summary, thermal degradation of MDEA follows first order rate kinetics at temperatures
below 150 to 170 ̊C, with secondary amines such as DEA and MAE as the degradation
50
products. Above 170 ̊C, thermal degradation of MDEA follows higher order. MDEA is more
resistant to thermal degradation in comparison with MEA.
7 AMINE BLENDS
The previous sections have summarised both thermal and oxidative degradation
characteristics of common amine solvents that are commercially used for CO2 capture.
Table 20 lists their CO2 carrying capacity, heat of reaction (absorption) with CO2 and the
reaction rate constant for CO2 capture88-92.
Table 20 – Comparison of CO2 capture characteristics of common amines
Amine Maximum Theoretical CO2
Carrying Capacity
(mole CO2/mole amine)
Heat of
Absorption
(MJ/kg CO2)
Reaction Rate
Constant @ 25oC
(m3/mol/sec)
MEA 0.5 1.9 6.0
MDEA 1 1.1 5 x 10 -3
AMP 1 1.6 0.58
PZ 2 1.6 70.0
It can be seen from this data that while MEA is a fast reacting amine, its high heat of
absorption for CO2 and low CO2 carrying capacity are distinct disadvantages. Similarly,
whilst MDEA has lower heat of absorption in comparison with MEA and twice the CO2
carrying capacity, it is handicapped by slow reaction kinetics as typified by 3 orders of
magnitude lower reaction rate constant. Piperazine, on the other hand, has all favourable
characteristics, provided it is within its solid/liquid transition temperature limit shown in Figure
24. AMP, a hindered amine has better characteristics for CO2 capture compared to MDEA.
Therefore, blending MDEA with MEA or Piperazine (PZ) or blending AMP with PZ can
produce a solvent that has CO2 absorption rate as well as CO2 carrying capacity higher than
either MEA or MDEA or AMP. Accordingly, Dubois and Thomas89, 92 have experimentally
measured improvements (Figure 30 and Figure 31) in the CO2 absorption efficiency at 25 ̊C
over that of 30% w/w MDEA solvent by using different single and blended amines for CO2
removal in a given contactor where the inlet gas stream contained 10% v/v CO2. Figure 32
shows the absorption enhancement factor or the “activation ratio” as a result of a particular
level of blending in 30% w/w MDEA. Similarly, Choi et al45, 93 have experimentally measured
CO2 absorption rates at different temperatures and the absorption activation ratio at these
temperatures (Tables 21 and 22) for 30% w/w AMP and its blends with MDEA and PZ.
Figure 30 – CO2 absorption efficiency (Aexp%) with single amine solutions
in a fixed size contactor, [PZEA = (Piperazinyl -1)-2-ethylamine]
51
Figure 31 - CO2 absorption efficiency (Aexp%) with mixed amine solutions for
30% w/w MDEA in a fixed size contactor, [PZEA = (Piperazinyl -1)-2-ethylamine]
Figure 32 – CO2 absorption enhancement or “Activation Ratio” for various additives in
30% w/w MDEA in a fixed size contactor, [PZEA = (Piperazinyl -1)-2-ethylamine]
Table 21 – Specific absorption rates of CO2 in blends of AMP at various temperatures
CO2 Absorption Rate kmole of CO2 per m2 per sec
Temperature ( ̊C) 30 40 50 70
Solvent (% w/w)
AMP (30) 3.03 3.75 4.25 4.92
AMP (30) / MDEA (1) 3.08 3.86 4.45 5.13
AMP (30) / MDEA (3) 3.15 3.95 4.60 5.45
AMP (30) / MDEA (5) 3.27 4.07 4.68 5.74
AMP (30) / PZ (1) 3.09 3.90 4.44 5.17
AMP (30) / PZ (3) 3.21 4.03 4.63 5.57
AMP (30) / PZ (5) 3.34 4.18 4.89 6.02
52
Table 22 – Absorption “Activation Ratio” for AMP/MDEA & AMP/PZ blends
Temperature (oC) 30 40 50 70
Solvent (% w/w)
AMP (30) 1.00 1.00 1.00 1.00
AMP (30) / MDEA (1) 1.02 1.03 1.05 1.04
AMP (30) / MDEA (3) 1.04 1.05 1.08 1.11
AMP (30) / MDEA (5) 1.08 1.09 1.10 1.17
AMP (30) / PZ (1) 1.02 1.04 1.05 1.06
AMP (30) / PZ (3) 1.06 1.08 1.09 1.14
AMP (30) / PZ (5) 1.10 1.12 1.15 1.23
These results clearly show that blending a primary amine or a fast reacting cyclic diamine
with a tertiary or hindered amine has the potential to reduce the liquid circulation rate, the
size of absorber and the associated equipment as well as the overall energy demand for
CO2 capture in an industrial situation. In essence, it can improve the techno-economic
viability of an amine based post combustion capture plant for carbon capture and storage.
This could be the reason why blended amine based CO2 capture technologies for gas
processing have been lately investigated for the post combustion capture application. The
MHI Process94 that uses a blend of AMP and PZ as a preferred solvent, the BASF process95
that uses the PZ activated MDEA as a solvent and the PSR process96-98 from University of
Regina that uses a blend of MDEA and MEA are the examples of blended amine based
capture processes. With these upcoming new post combustion capture processes in mind,
the following sections delve into the oxidative and thermal degradation of three blended
amine solvents; viz. a blend of MDEA and MEA, a blend of MDEA and PZ, and a blend of
AMP and PZ.
7.1 Oxida tive and Thermal Degrada tion of MDEA/MEA Blend
Combined oxidative/thermal degradation of MDEA/MEA blend has been studied by the
research group82, 83 from University of Regina, Canada. Table 23 below shows their
laboratory based experiment conditions. The degradation product characterisation and
identification was done using the gas chromatography and mass spectrometry techniques.
Table 23 – Oxidative and thermal degradation conditions for MDEA/MEA blends
Temperature (oC) 100 120 120
MEA Concentration (moles/Litre) 5 5 7
MDEA Concentration (moles/Litre) 2 2 2
CO2 Loading (mole/mole total amine) 0.18 0.502 0.43
O2 Pressure (kPa) 250 250 250
The degradation products seemed to vary with varying the ratio of MEA and MDEA
concentration in the blends. More degradation products were formed with increasing the
solution temperature and increasing the CO2 loading. MDEA seemed to get preferentially
degraded when blended with MEA, though as a standalone amine, it is more resistant to
degradation in comparison with MEA. 2-(methylamino)ethanol, 1-amino-2-propanol, 1,3propanediamine and 1,2-propanediamine were identified as major degradation products with
a host of minor degradation products including 3-methylpyridine, N,N-dimethylurea, Nhydrohycarbaminic acid etc. No kinetics of combined degradation was determined and no
53
product distribution was quantified. One of the strongest limitations has been no degradation
experiments in the joint presence of CO2 and O2 at conditions that typify a CO2 absorber
operation (typically 40 to 70 ̊C) and no identification of the expected formation of DEA and
MAE. Similarly, N,N-dimethyl-2-ethanolamine (DMAE) and the carbamate polymerisation
products were also not identified in the degradation product mixture.
Whilst, the University of Regina researchers82, 83, 85 state to use MDEA/MEA blend in the
ratio 1:4, there is no information provided by them on the impact of SOX and NOX on the
degradation characteristics of the blend. Degradation of MEA by SOX species has been
narrated in the earlier sections and formation of NDELA in the presence of NOX has been
confirmed by Pedersen et al26, 29 as described in the earlier sections. Given that MDEA is a
tertiary weak amine and SOX species are strongly acidic, degradation of MDEA by reaction
with SO2 and SO3 is inevitable. There is indeed literature evidence for the flue gas
desulphurisation using tertiary or hindered amines, for example the CANSOLV process87.
Similarly, Mimura et al60 indicate that 30% w/w MDEA absorbs NO2 and CO2 simultaneously
with the liquid side mass transfer co-efficient value of 1.5 x 10-7 mol/s/Pa/m2. However, the
experimental evidence by Closmann72 for the formation of secondary amines, DEA and
MAE, points towards the potential for the formation of nitroso-compounds (N-nitroso
diethanolamine and 2-methylnitrosoamino-ethanol) in an industrial situation that could have
certainly adverse environmental impact. Moreover, there is a possibility that MDEA could
degrade to diethylamine which can form diethylnitrosamine and diethylnitramine14. Thus, the
blend of MEA and MDEA may have potentially adverse impact in terms of the atmospheric
emissions perhaps to the extent same as the sum total of individual impact.
7.2 Oxida tive and Thermal Degrada tion of AMP/PZ Blend
Currently, there is no data available on the oxidative and thermal degradation of AMP/PZ
blend though the kinetics of CO2 absorption in this blend has been investigated. Since both
AMP and PZ have individually lower tendency towards degradation in comparison with MEA,
it could be speculated that the combined oxidative and thermal degradation products of this
blend could be a product mixture that represents the sum total of individual oxidative and
thermal degradation products at least in terms of characterisation if not in quantification. It
should be though noted that in an industrial post combustion capture situation, PZ in the
presence of NOX will produce MNPZ as indicated by Freeman4 (Figures 26 and 27) and
other additional nitroso compounds of PZ such as DNPZ, N-Oxopiperazine and Piperazine
nitramine may also get formed as indicated by Jackson and Attala30, 74 in their laboratory
situation.
7.3 Oxida tive and Thermal Degrada tion of MDEA/PZ Blend
The oxidative and thermal degradation of MDEA/PZ blend has been studied extensively by
Closmann72 using the low gas flow reactor (TOR) and the integrated solvent degradation
assembly (ISDA) described in the previous sections. The TOR was used to study the
oxidative degradation separately whereas the ISDA was used to study the combined effect
of oxidative and thermal degradation. In addition, they used cylindrical batch reactors to
exclusively study the effect of thermal degradation. For all of their experiments a blend of 7
molal MDEA and 2 molal PZ was used. Table 24 below shows their experimental conditions.
As shown in Table 24, both unloaded and loaded blends of MDEA and PZ were tested for
assessing the degradation. Closmann argues that the low CO2 loadings of 0.1 and 0.14 mole
per mole alkalinity with 55 to 70 ̊C TOR temperature relate to the absorber condition in an
industrial situation (a coal-fired power plant post combustion capture scenario with 90% CO2
recovery) whereas the thermal reactor temperature of 100 to 120 ̊C and the CO2 loading
54
conditions of 0.3 and 0.43 mole per mole alkalinity relate to the rich amine stream
undergoing regeneration in the stripper reboiler.
The oxidative degradation at 55 ̊C in the TOR produced mostly formates (0.013 mM/hr). The
MDEA and PZ loss rate was not measured at this temperature. At 70 ̊C, both formyl amides
and bicine were formed in addition to the formates. MDEA and PZ initial loss rates at this
temperature were 0.17 ± 0.35 and 0.07 ± 0.05 mM/hr respectively. In terms of overall molar
quantity, the secondary amines (DEA+ MAE) represented the largest degradation product,
formed at a rate of 0.2 mM/hr. The total formate production rate (formates plus formyl
amides) was 0.08 mM/hr whereas the formates only production rate and bicine formation
rate were 0.031 and 0.024 mM/hr respectively. This signifies that ammonia produced during
oxidative degradation does not react with the heat stable salts at lower temperature and the
emissions of amides could be lowered in industrial practice by controlling the temperature
bulge in the absorber.
Table 24 – 7 molal MDEA / 2 molal PZ blend degradation conditions
Reactor
Type
Temperature
(oC)
Gas
Composition
(%O2/%CO2)
Initial CO2
Loading
(Mole CO2/mole
alkalinity)
Metal Ion
Additives
(mM)
TOR 55 92.5/7.5 0.3 1 Fe2+
TOR 55 92.5/7.5 0.24 0.1 Fe2+, 0.6 Cr3+, 0.1 Ni2+
TOR 55 92.5/7.5 0.23 0.1 Fe2+, 5 Cu2+
TOR 70 92.5/7.5 0.14 0.4 Fe2+, 0.1 Cr3+, 0.1 Ni2+
Thermal
Reactor
100, 120 0.25, 0.23, 0.43
Thermal
Reactor
100, 120 0.1, 0.2
Thermal
Reactor
100, 120 0.18 1 Fe2+
Thermal
Reactor
135 0.12, 0.23
Thermal
Reactor
135, 150 0.11, 0.26
Thermal
Reactor
135, 150 0.0, 0.02, 0.3
Thermal
Reactor
120, 135, 150 0.0, 0.1, 0.25
ISDA 55/120 98.0/2.0 0.14 0.4 Fe2+, 0.1 Cr3+, 0.1 Ni2+
ISDA 55/90 98.0/2.0 0.14 0.4 Fe2+, 0.1 Cr3+, 0.1 Ni2+
ISDA 55/100 98.0/2.0 0.14 0.4 Fe2+, 0.1 Cr3+, 0.1 Ni2+
ISDA 55/125 98.0/2.0 0.14 0.4 Fe2+, 0.1 Cr3+, 0.1 Ni2+
Closmann72 ranks amines in terms of their tendency towards high total formate production
as 7 molal MEA > 7 molal MDEA/2 molal PZ > 7 molal MDEA. He also detected that the
formyl amides specific to PZ degradation were formed at higher level suggesting PZ
degrades faster when blended with MDEA though one would expect less PZ degradation
due to its higher resistance to the oxidative degradation in comparison with MDEA. There is
no proper explanation in the open literature why MDEA and PZ, though individually robust
55
solvents, get preferentially degraded when blended with MEA and MDEA respectively. This
could be considered as an open subject for further investigations.
The thermal degradation experiments conducted separately in the thermal reactor showed
the degradation products listed in Table 25.
Table 25 – Thermal degradation (only) products of 7 molal MDEA/2 molal PZ
Compound Name Abbreviation
2 – Dimethylamino Ethanol DMAE
Methyl Paperazine 1-MPZ
N-(2-hydroxyethyl)-N-methyl Formamide MAE-Amide
Diethanolamine DEA
1,4 Dimethyl Piperazine DMP
3 – (hydroxyethyl) – 2 – Oxazolidone HEOD
4 – Methyl – 1 – piperazine-ethanamine MPZEA
N,N-bis-(2-hydroxyethyl) Formamide DEA-Amide or AEMP
1 - (2 - hydroxyethyl) – 4 – Methyl Piperazine HMP
N,N-bis-(2–hydroxyethyl)-Glycine Bicine
2-[[2-(1-piperazinyl)ethyl]amino]ethanol PEAE
N,N-bis-(2 – hydroxyethyl) – Piperazine BHEP`
2-[[2-(4-methyl-1-piperazinyl)ethyl]amino]ethanol MPZEA-OH
Methyl-N,N,N-tris-(hydroxyethyl) Ethylenediamine MTHEED
These results showed that MDEA/PZ blend is thermally stable up to 150 ̊C in the absence of
CO2 loading similar to MDEA is up to 120 ̊C. The formation of amides of DMAE and DEA
imply that ammonia is definitely a product of thermal degradation that reacts with carboxylic
acids which may have been formed in the MDEA/PZ blend despite lack of oxygen due to the
formation of oxygen in-situ as a result of CO2 induced degradation of MDEA as observed by
Lawal and Idem82 in their studies. Closmann has used the thermal degradation results to
develop a “Universal activation energy” based theoretical model that estimates the
degradation product formation. It has an underlying assumption that the rate limiting step in
the degradation of the MDEA/PZ blend is the first order protonation of MDEA, i.e. the
formation of MDEAH+ and that the overall degradation rate of the blend is independent of the
concentration of PZ. With these assumptions, Closmann proposes the following equations to
predict the rates of losses of MDEA and PZ:
𝑑[𝑀𝐷𝐸𝐴]
𝑑𝑡
= -KT*[MDEA]T*α ---------- (41)
KT = K408*exp {104300
8.314
* ( 1
408
- 1
𝑇
)} --------- (42)
Where, 𝑑[𝑀𝐷𝐸𝐴]
𝑑𝑡
is the MDEA loss rate in moles per hour, K408 is the first order degradation
rate constant at 408 K in hr-1, T is the reaction temperature in oK, 104300 J/mole is the
“universal activation energy” for the degradation of MDEA/PZ blend and 8.314 J per mole
per oK is the universal gas constant, [MDEA]T is molal concentration of MDEA in the solution
at temperature T (oK) and time t (hr), and α is initial moles of CO2 loading per mole alkalinity
of the blend solution.
56
The PZ loss rate, 𝑑[𝑃𝑍]
𝑑𝑡
, is calculated as:
𝑑[𝑃𝑍]
𝑑𝑡
= 1.13 * 𝑑[𝑀𝐷𝐸𝐴]
𝑑𝑡
------------ (43)
Closmann calculated the value for K408 as 0.001751 hr-1.
The combined oxidative and thermal degradation experiments in the ISDA system
conducted by Closmann showed that the degradation products of 7 molal MDEA/2 molal PZ
blend over the temperature range 55 to 125 ̊C are formate, acetate, oxalate, DMAE, a
mixture of MAE and DEA, 1-MPZ, DMP, AEP, N-formyl PZ, MAE-amide, DEA-amide,
Bicine, HES and Glycine. In this case, the loss of MDEA, PZ and alkalinity as a function of
time, t, were regressed as:
MDEA loss (milimole) = 3452.23*exp (-0.001*t) ------- (44)
PZ loss (milimole) = 969.68*exp (-0.002*t) ---------- (45)
Alkalinity loss (milimole) = 5050.13*exp (-0.00056*t) ------- (47)
Using equations 26 and 27, Closmann72 calculated the first order rate constant and the
activation energy for the total formates production during combined oxidative and thermal
degradation as 10 hr-1 and 80 kJ per mole of total amine at 363 oK. Accordingly, the upper
temperature limits for 10%, 20% and 50% oxygen consumption by the CO2 rich MDEA/PZ
blend in the cross exchanger and the piping leading to the stripper (Figure 1) was calculated
by Closmann as 92 ̊C, 97 ̊C and 106 ̊C respectively, if the solution residence time in this
section of the CO2 absorber / regeneration system is approximately 30 sec. Closmann has
further calculated that for a 500 MW coal-fired power plant using 7 molal MDEA/2 molal PZ
blend with initial CO2 loading of 0.14 mole per mole alkalinity, if the flue gas contains oxygen
at 5 kPa and 90% CO2 capture is required, then the overall amine loss due to the combined
oxidative and thermal degradation will be 0.27 moles per ton of CO2 captured.
In summary, the behaviour of 7 molal MDEA/2 molal PZ is in between the behaviour of 7
molal MDEA and 8 molal PZ solvents. Currently, there is no data available on the impact of
SOX/NOX on the degradation characteristics of the MDEA/PZ blend but the comments made
on MDEA and PZ as separate amines for the impact of SOX/NOX on their degradation and
subsequent likely atmospheric emissions upon usage of these amines for the post
combustion capture should be kept in mind.
8. EMISSIONS OF DEGRADATION PRODUCTS: ASPEN-PLUS SIMULATIONS
The preceding sections have reviewed the literature data on the oxidative and thermal
degradation of MEA, MDEA, AMP, PZ and their blends. These amines are primary, tertiary,
hindered, and secondary cyclical types and therefore, span most of the emission risk
presented by amine solvents. The degradation data as presented by various investigators
though not entirely obtained under the process conditions that are relevant to an industrial
post combustion capture environment, however, do provide a list of likely degradation
compounds. With identification of these compounds and their physical/chemical properties
as well as their kinetics of formation known, it is possible to estimate their likely atmospheric
emissions in an industrial scale post combustion capture environment using the process
modelling tools such as CO2SIM, Aspen-Plus, PRO-II or ProTreat. The previous milestone
report99 for the Task 3 of the research project, “Environmental Impact of Amine based Postcombustion Processes” submitted to the ANLEC R&D has discussed a base case scenario
57
of capturing 90% CO2 from a typical Australian black coal-fired power plant flue gas stream
using the un-inhibited 30% w/w MEA solvent based post combustion CO2 capture
technology. This report estimates for this base case atmospheric emissions, both vapour
and droplet phase, of MEA and its degradation products using the Aspen-Plus process
simulation software. The underlying assumptions and methodology employed for the
estimation of emissions have been discussed in detail in that report99. These assumptions
and methodology could be used to estimate the atmospheric emissions of blended amine
solvents such as a mixture of AMP and PZ or a mixture of MDEA and MEA.
8.1 As pen-Plus S imula tion Tas k
Using the methodology referred above and the Aspen-Plus process simulation software, the
following tasks have been carried out for the current phase of the ANLEC R&D funded post
combustion capture project:
a) Estimate the impact of wash water temperature on the atmospheric emissions of
MEA and its degradation products for the base case scenario.
b) Estimate the vapour and the droplet phase atmospheric emissions of amine
solvents when blended amine solvents B and C described below replace the uninhibited 30% w/w MEA solvent in the base case scenario.
c) Estimate the impact of wash water temperature on the atmospheric emissions
solvents B and C.
Solvent B Composition: 25% w/w AMP, 15% w/w PZ, 60% w/w Water
Solvent C Composition: 25% w/w MDEA, 5% w/w MEA, 70% w/w Water
For the tasks (b) and (c), the estimation of atmospheric emissions of the degradation
products of AMP, PZ, MDEA and MEA has not been incorporated into the current process
simulations since the characterisation and the quantification of the oxidative and thermal
degradation products of the blends of AMP/PZ and MDEA/MEA is still uncertain particularly
with respect to the impact of SOX and NOX on the degradation processes as described in the
earlier sections. As these uncertainties diminish with more laboratory and pilot plant based
solvent degradation data become available from various research organisations including
CSIRO, these Aspen-Plus process models will be updated. It is envisaged that the next
milestone report will have the results from the Aspen-Plus process models that have
incorporated full degradation information on these blends.
8.2 Impac t of Was h Tower Performance – MEA Bas e Cas e
The previous milestone report99 had estimated the atmospheric emissions of MEA and its
combined oxidative and thermal degradation products for one set of the Wash Tower
operating condition. The report had also highlighted that in a conventional MEA based post
combustion capture plant, the Wash Tower is the last line of defence against uncontrolled
emissions of volatiles to the atmosphere from the CO2 absorber.
Based on the pilot plant emissions data from Moser et al119,120 and Aas111, mentioned in the
report99 the estimated concentration of several degradation products on a dry gas basis in
the CO2 lean gas after absorber are estimated below:
DEA ~ 0.3 mg/Nm3, Formaldehyde ~ 0.35 mg/Nm3, Acetaldehyde ~ 0.35 mg/Nm3,
Acetone ~ 0.5 mg/Nm3, NH3 ~ 27 mg/Nm3, Methylamine ~ 0.3 mg/Nm3,
Acetamide ~ 0.5 mg/Nm3
58
The above chemicals could be present in vapour phase at the above stated concentration
levels either in the process stream G1 or the process stream G3 of the flow sheet shown in
Figure 33. Assuming the capture plant operating at steady state and solvent reclamation
occurring at the end of 3rd week after start-up, likely atmospheric emissions of MEA and its
oxidative as well as thermal degradation products were calculated in ASPEN simulation
using the above data. The ASPEN calculations involved introducing the above chemicals at
their above stated concentration levels in Stream G1 or Stream G3 as the case may be and
normalising the stream flow rate. The atmospheric emissions of these chemicals in the
droplet phase carryover of wash water from the Wash Tower was calculated using 0.13 m3
droplets carry-over per million m3 of CO2 Lean gas stream.
Since water balance has to be maintained around the capture plant during the steady state
operation, only one degree of freedom is available for an existing capture plant where the
Wash Tower has already been designed and installed with a set value for maximum liquid to
gas flow ratio i.e., only the wash water temperature could be varied to control emissions.
Figure 33 below shows the Aspen-Plus generated material and energy balance flow sheet
for the base case of entire CO2 capture plant where the Wash Tower is assumed to be
operating at 45 ̊C. By activating the cooler HX7 on the wash water stream WW3, it is
possible to operate the Wash Tower at lower temperature. In Australia, the inland summer
temperature for cooling water could be as much as 30 to 35 ̊C whereas in winter, it could
drop to as low as 5 ̊C. Thus, there is a possibility that the Wash Tower temperature could be
lowered to 15 ̊C safely depending upon the cooling water temperature once the heat
exchanger HX7 is activated. Such a process operating situation can potentially reduce the
level of atmospheric emissions of MEA and its degradation products significantly.
Aspen simulation results for the base case considered reactivity of DEA (which could be
present in the MEA as impurity or it is a by-product of oxidative degradation of MEA) with
CO2 in flue gas leaving the absorber as well as the possibility of maximum concentration of
the volatiles either at the outlet of the absorber (Stream G1) or at the inlet to the Wash
Tower (stream G3) depending upon how the basic design of the capture plant is carried out
to effectively counter the temperature bulge occurring in the absorber due to the heat of
absorption of CO2 in aqueous amine. For the sake of completing the exercise, the impact of
wash water temperature on the atmospheric emissions due to the volatiles being present at
the maximum level either in the stream G1 or in the stream G3 was calculated. Tables 26 to
29 show the atmospheric emission results for both the cases (G1 and G3) where the
reaction of DEA with CO2 is allowed in the Wash Tower. The Wash Tower temperature as a
result of cooling water temperature change is allowed to vary as 45 ̊C, 35 ̊C, 25 ̊C and 15 ̊C.
Figure 34 shows the impact of drop in the Wash Tower temperature on the emissions of
MEA and its degradation products.
59
Figure 33 – Base case operation of the Wash Tower (Temperature 45 oC)
40.0
1.07
4704.1
0.00
LEAN
45.0
1.09
1801.1
1.00
F4
RE2
46.4
1.10
5016.5
0.00R1
70.6
1.06
1679.7
1.00G1
46.6
1.80
5025.7
0.00
R2
HEAT1
101.9
1.77
5025.7
0.01
RICH
119.9
1.78
4736.4
0.00
L1
56.6
1.77
4736.4
0.00
L2
30.0
1.10
9.2
0.00
MAKEUP
40.0
1.75
322.5
1.00CO2
38.9
0.90
1801.1
1.00
F2
62.0
1.10
1801.1
1.00
F3
44.8
1.09
0.9
0.00
WWATER
40.0
1.07
4704.1
0.00L4
45.3
1.01
2923.4
0.00D0
35.0
1.01
2952.7
0.00D2
45.3
0.95
2973.9
0.00D3
20.0
1.01
29.3
0.00
DWATER
45.3
1.01
2973.9
0.00
D4
44.8
1.04
2000.5
0.00
WW1
44.8
1.10
2000.5
0.00
WW2
78.4
0.96
1822.3
1.00F1
44.8
1.09
1999.6
0.00WW3
44.8
1.08
1999.6
0.00
WW4
45.0
1.05
1679.7
0.84G2
45.0
1.05
1488.7
1.00
G3
45.0
1.05
191.0
0.00RE1
20.0
1.10
1.0
0.00
WATER
44.8
1.04
1488.8
1.00G4
123.8
1.02
1488.8
1.00
GAS
143.8
0.96
1822.3
1.00FLUE
45.3
1.01
50.6
0.00
DWASTE
MJIKG
ABSORBER
QC=0.000
QR=0.000
HX4
HX5
Q=-296.761
P2
W=0.103
STRIPPER
QC=-149.288
QR=371.742BLOWERW=12.027
HX6
Q=-75.824
HX3
Q=-8.811
HX2
Q=-33.879
P1
W=0.006
P3
W=0.004
DCC
QC=0.000
QR=0.000
HX7
Q=0.000
WV
CONDEN
Q=0.000
HX8
Q=-140.329
B9
W=0.000
WASH Q=0.000
HX9 Q=35.204
HX1
Q=-35.200
SE P
SEP
Q=0.000
Temperature (C)
Pressure (bar)
Mass Flow Rate (tonne/hr)
Vapor Fraction
Q Duty (MW)
W Power(MW)
4.23
SPLIT
RECLAIM
P4
W=0.000
56.6
1.77
33.2
0.00
RC1
56.6
1.77
0.0RC2
RC3
56.6
1.80
33.2
0.00
RC4
56.6
1.77
4703.2
0.00
L3
MIX
45.1
1.01
2952.7
0.00
D1
60
Table 26 – Atmospheric emissions (Stream G4) of MEA and its degradation products predicted at 45 oC
Chemical
Emissions
For Maximum Volatiles input In Stream G3 For Maximum Volatiles input In Stream G1
Vapour
Phase
Droplet
Phase
Total Vapour
Phase
Droplet
Phase
Total Vapour
Phase
Droplet
Phase
Total Vapour
Phase
Droplet
Phase
Total
mg/Nm3
dry CO2
Lean gas
mg/Nm3
dry CO2
Lean gas
mg/Nm3
dry CO2
Lean gas
mg/tonne
CO2
mg/tonne
CO2
mg/tonne
CO2
mg/Nm3
dry CO2
Lean gas
mg/Nm3
CO2 dry
Lean gas
mg/Nm3
dry CO2
Lean gas
mg/tonne
CO2
mg/tonne
CO2
mg/tonne
CO2
MEA 2.45E-02 1.10E-01 1.35E-01 7.92E+01 3.57E+02 4.37E+02 2.63E-02 1.10E-01 1.37E-01 8.53E+01 3.58E+02 4.43E+02
NITROSOMORPHOLINE 2.36E-06 4.01E-07 2.76E-06 7.63E-03 1.30E-03 8.93E-03 2.36E-06 4.02E-07 2.76E-06 7.64E-03 1.30E-03 8.95E-03
NDELA 4.75E-15 8.62E-13 8.67E-13 1.54E-11 2.79E-09 2.81E-09 4.76E-15 8.63E-13 8.68E-13 1.54E-11 2.80E-09 2.81E-09
NH3 1.13E-01 2.32E-03 1.16E-01 3.67E+02 7.52E+00 3.74E+02 8.56E-04 1.89E-05 8.75E-04 2.77E+00 6.11E-02 2.83E+00
HEEDA 0 0 0 0 0 0 0 0 0 0 0 0
OXAZOLIDONE 3.88E-12 2.08E-12 5.96E-12 1.26E-08 6.75E-09 1.93E-08 3.88E-12 2.09E-12 5.97E-12 1.26E-08 6.76E-09 1.93E-08
HEIA 0 0 0 0 0 0 0 0 0 0 0 0
TRIMER 0 0 0 0 0 0 0 0 0 0 0 0
CYCLIC UREA 0 0 0 0 0 0 0 0 0 0 0 0
POLYMER 0 0 0 0 0 0 0 0 0 0 0 0
DEA 1.08E-07 2.59E-05 2.60E-05 3.49E-04 8.39E-02 8.43E-02 1.05E-12 2.37E-10 2.38E-10 3.39E-09 7.67E-07 7.71E-07
FORMALDEHYDE 2.73E-01 6.16E-06 2.73E-01 8.85E+02 2.00E-02 8.85E+02 2.62E-01 5.91E-06 2.62E-01 8.48E+02 1.91E-02 8.48E+02
ACETALDEHYDE 2.98E-01 3.97E-06 2.98E-01 9.66E+02 1.29E-02 9.66E+02 2.88E-01 3.84E-06 2.88E-01 9.34E+02 1.24E-02 9.34E+02
ACETONE 3.32E-01 1.39E-05 3.33E-01 1.08E+03 4.50E-02 1.08E+03 3.12E-01 1.31E-05 3.12E-01 1.01E+03 4.23E-02 1.01E+03
METHYLAMINE 2.17E-01 6.80E-06 2.17E-01 7.03E+02 2.20E-02 7.03E+02 2.06E-01 6.46E-06 2.06E-01 6.68E+02 2.09E-02 6.68E+02
ACETAMIDE 6.99E-05 4.32E-05 1.13E-04 2.26E-01 1.40E-01 3.66E-01 1.05E-07 6.46E-08 1.69E-07 3.39E-04 2.09E-04 5.48E-04
61
Table 27 – Atmospheric emissions (Stream G4) of MEA and its degradation products predicted at 35 oC
Chemical
Emissions
For Maximum Volatiles input In Stream G3 For Maximum Volatiles input In Stream G1
Vapour
Phase
Droplet
Phase
Total Vapour
Phase
Droplet
Phase
Total Vapour
Phase
Droplet
Phase
Total Vapour
Phase
Droplet
Phase
Total
mg/Nm3
dry CO2
Lean gas
mg/Nm3
dry CO2
Lean gas
mg/Nm3
dry CO2
Lean gas
mg/tonne
CO2
mg/tonne
CO2
mg/tonne
CO2
mg/Nm3
dry CO2
Lean gas
mg/Nm3
CO2 dry
Lean gas
mg/Nm3
dry CO2
Lean gas
mg/tonne
CO2
mg/tonne
CO2
mg/tonne
CO2
MEA 1.26E-02 1.06E-01 1.19E-01 4.08E+01 3.44E+02 3.85E+02 1.36E-02 1.06E-01 1.20E-01 4.40E+01 3.44E+02 3.88E+02
NITROSOMORPHOLINE 1.81E-06 3.86E-07 2.20E-06 5.87E-03 1.25E-03 7.12E-03 1.82E-06 3.87E-07 2.20E-06 5.88E-03 1.25E-03 7.13E-03
NDELA 2.94E-15 8.30E-13 8.33E-13 9.52E-12 2.69E-09 2.70E-09 2.94E-15 8.31E-13 8.34E-13 9.54E-12 2.69E-09 2.70E-09
NH3 7.52E-02 2.24E-03 7.74E-02 2.43E+02 7.25E+00 2.51E+02 5.68E-04 1.82E-05 5.86E-04 1.84E+00 5.89E-02 1.90E+00
HEEDA 0 0 0 0 0 0 0 0 0 0 0 0
OXAZOLIDONE 3.02E-12 2.01E-12 5.03E-12 9.80E-09 6.50E-09 1.63E-08 3.03E-12 2.01E-12 5.04E-12 9.82E-09 6.51E-09 1.63E-08
HEIA 0 0 0 0 0 0 0 0 0 0 0 0
TRIMER 0 0 0 0 0 0 0 0 0 0 0 0
CYCLIC UREA 0 0 0 0 0 0 0 0 0 0 0 0
POLYMER 0 0 0 0 0 0 0 0 0 0 0 0
DEA 5.33E-08 2.49E-05 2.50E-05 1.73E-04 8.08E-02 8.09E-02 5.20E-13 2.28E-10 2.28E-10 1.68E-09 7.38E-07 7.40E-07
FORMALDEHYDE 2.66E-01 6.55E-06 2.66E-01 8.62E+02 2.12E-02 8.62E+02 2.55E-01 6.28E-06 2.55E-01 8.26E+02 2.03E-02 8.26E+02
ACETALDEHYDE 2.92E-01 4.38E-06 2.92E-01 9.45E+02 1.42E-02 9.45E+02 2.82E-01 4.23E-06 2.82E-01 9.14E+02 1.37E-02 9.14E+02
ACETONE 3.08E-01 1.55E-05 3.08E-01 9.96E+02 5.02E-02 9.96E+02 2.89E-01 1.45E-05 2.89E-01 9.35E+02 4.71E-02 9.35E+02
METHYLAMINE 2.09E-01 7.23E-06 2.09E-01 6.77E+02 2.34E-02 6.77E+02 1.98E-01 6.87E-06 1.98E-01 6.43E+02 2.22E-02 6.43E+02
ACETAMIDE 5.03E-05 4.16E-05 9.19E-05 1.63E-01 1.35E-01 2.98E-01 7.54E-08 6.22E-08 1.38E-07 2.44E-04 2.02E-04 4.46E-04
62
Table 28 – Atmospheric emissions (Stream G4) of MEA and its degradation products predicted at 25 oC
Chemical
Emissions
For Maximum Volatiles input In Stream G3 For Maximum Volatiles input In Stream G1
Vapour
Phase
Droplet
Phase
Total Vapour
Phase
Droplet
Phase
Total Vapour
Phase
Droplet
Phase
Total Vapour
Phase
Droplet
Phase
Total
mg/Nm3
dry CO2
Lean gas
mg/Nm3
dry CO2
Lean gas
mg/Nm3
dry CO2
Lean gas
mg/tonne
CO2
mg/tonne
CO2
mg/tonne
CO2
mg/Nm3
dry CO2
Lean gas
mg/Nm3
CO2 dry
Lean gas
mg/Nm3
dry CO2
Lean gas
mg/tonne
CO2
mg/tonne
CO2
mg/tonne
CO2
MEA 5.81E-03 1.02E-01 1.08E-01 1.88E+01 3.31E+02 3.50E+02 6.28E-03 1.02E-01 1.09E-01 2.03E+01 3.31E+02 3.52E+02
NITROSOMORPHOLINE 1.34E-06 3.72E-07 1.71E-06 4.35E-03 1.21E-03 5.56E-03 1.35E-06 3.73E-07 1.72E-06 4.36E-03 1.21E-03 5.57E-03
NDELA 1.68E-15 7.99E-13 8.01E-13 5.45E-12 2.59E-09 2.60E-09 1.69E-15 8.00E-13 8.02E-13 5.47E-12 2.59E-09 2.60E-09
NH3 4.66E-02 2.16E-03 4.87E-02 1.51E+02 6.99E+00 1.58E+02 3.51E-04 1.75E-05 3.69E-04 1.14E+00 5.68E-02 1.20E+00
HEEDA 0 0 0 0 0 0 0 0 0 0 0 0
OXAZOLIDONE 2.30E-12 1.93E-12 4.23E-12 7.43E-09 6.26E-09 1.37E-08 2.30E-12 1.93E-12 4.23E-12 7.45E-09 6.27E-09 1.37E-08
HEIA 0 0 0 0 0 0 0 0 0 0 0 0
TRIMER 0 0 0 0 0 0 0 0 0 0 0 0
CYCLIC UREA 0 0 0 0 0 0 0 0 0 0 0 0
POLYMER 0 0 0 0 0 0 0 0 0 0 0 0
DEA 2.32E-08 2.40E-05 2.40E-05 7.51E-05 7.78E-02 7.79E-02 2.27E-13 2.20E-10 2.20E-10 7.36E-10 7.11E-07 7.12E-07
FORMALDEHYDE 2.57E-01 7.03E-06 2.57E-01 8.33E+02 2.28E-02 8.33E+02 2.46E-01 6.74E-06 2.46E-01 7.98E+02 2.18E-02 7.98E+02
ACETALDEHYDE 2.83E-01 4.93E-06 2.83E-01 9.17E+02 1.60E-02 9.17E+02 2.74E-01 4.76E-06 2.74E-01 8.87E+02 1.54E-02 8.87E+02
ACETONE 2.77E-01 1.74E-05 2.77E-01 8.98E+02 5.64E-02 8.98E+02 2.60E-01 1.63E-05 2.60E-01 8.43E+02 5.29E-02 8.43E+02
METHYLAMINE 1.99E-01 7.76E-06 1.99E-01 6.45E+02 2.51E-02 6.45E+02 1.89E-01 7.36E-06 1.89E-01 6.13E+02 2.39E-02 6.13E+02
ACETAMIDE 3.45E-05 4.00E-05 7.45E-05 1.12E-01 1.30E-01 2.41E-01 5.17E-08 5.99E-08 1.12E-07 1.68E-04 1.94E-04 3.62E-04
63
Table 29 – Atmospheric emissions (Stream G4) of MEA and its degradation products predicted at 15 oC
Chemical
Emissions
For Maximum Volatiles input In Stream G3 For Maximum Volatiles input In Stream G1
Vapour
Phase
Droplet
Phase
Total Vapour
Phase
Droplet
Phase
Total Vapour
Phase
Droplet
Phase
Total Vapour
Phase
Droplet
Phase
Total
mg/Nm3
dry CO2
Lean gas
mg/Nm3
dry CO2
Lean gas
mg/Nm3
dry CO2
Lean gas
mg/tonne
CO2
mg/tonne
CO2
mg/tonne
CO2
mg/Nm3
dry CO2
Lean gas
mg/Nm3
CO2 dry
Lean gas
mg/Nm3
dry CO2
Lean gas
mg/tonne
CO2
mg/tonne
CO2
mg/tonne
CO2
MEA 2.40E-03 9.87E-02 1.01E-01 7.78E+00 3.20E+02 3.27E+02 2.60E-03 9.87E-02 1.01E-01 8.42E+00 3.20E+02 3.28E+02
NITROSOMORPHOLINE 9.58E-07 3.59E-07 1.32E-06 3.11E-03 1.16E-03 4.27E-03 9.60E-07 3.59E-07 1.32E-06 3.11E-03 1.16E-03 4.28E-03
NDELA 8.92E-16 7.71E-13 7.72E-13 2.89E-12 2.50E-09 2.50E-09 8.94E-16 7.72E-13 7.73E-13 2.90E-12 2.50E-09 2.50E-09
NH3 2.68E-02 2.08E-03 2.89E-02 8.68E+01 6.75E+00 9.36E+01 2.02E-04 1.69E-05 2.19E-04 6.55E-01 5.48E-02 7.10E-01
HEEDA 0 0 0 0 0 0 0 0 0 0 0 0
OXAZOLIDONE 1.70E-12 1.86E-12 3.56E-12 5.50E-09 6.03E-09 1.15E-08 1.70E-12 1.87E-12 3.57E-12 5.51E-09 6.04E-09 1.16E-08
HEIA 0 0 0 0 0 0 0 0 0 0 0 0
TRIMER 0 0 0 0 0 0 0 0 0 0 0 0
CYCLIC UREA 0 0 0 0 0 0 0 0 0 0 0 0
POLYMER 0 0 0 0 0 0 0 0 0 0 0 0
DEA 8.85E-09 2.32E-05 2.32E-05 2.87E-05 7.50E-02 7.51E-02 8.69E-14 2.12E-10 2.12E-10 2.82E-10 6.86E-07 6.86E-07
FORMALDEHYDE 2.47E-01 7.61E-06 2.47E-01 7.99E+02 2.46E-02 7.99E+02 2.36E-01 7.29E-06 2.36E-01 7.66E+02 2.36E-02 7.66E+02
ACETALDEHYDE 2.71E-01 5.68E-06 2.71E-01 8.78E+02 1.84E-02 8.78E+02 2.62E-01 5.49E-06 2.62E-01 8.50E+02 1.78E-02 8.50E+02
ACETONE 2.42E-01 1.96E-05 2.42E-01 7.83E+02 6.34E-02 7.83E+02 2.27E-01 1.84E-05 2.27E-01 7.35E+02 5.95E-02 7.35E+02
METHYLAMINE 1.88E-01 8.38E-06 1.88E-01 6.08E+02 2.72E-02 6.08E+02 1.78E-01 7.96E-06 1.78E-01 5.78E+02 2.58E-02 5.78E+02
ACETAMIDE 2.26E-05 3.86E-05 6.11E-05 7.31E-02 1.25E-01 1.98E-01 3.38E-08 5.78E-08 9.16E-08 1.09E-04 1.87E-04 2.97E-04
64
Figure 34 – Impact of decreasing Wash Tower temperature on MEA content
Figure 34 shows that as the Wash Tower temperature drops due to cooling the wash water
stream WW3, the MEA content of stream WW1 leaving the wash tower increases. In other
words, the atmospheric emissions of MEA decrease which is what Tables 26 to 29 shows.
Similarly, these tables show that the atmospheric emissions of the degradation products of
MEA including ammonia and NDELA decrease. For example, assuming maximum presence
of volatiles at inlet to the Wash Tower, i.e. in Stream G3, total ammonia emissions (vapour
plus droplet phase) decrease from 0.12 mg/Nm3 of dry CO2 lean gas or 374 mg/tonne of
product CO2 at 45 ̊C to 0.03 mg/Nm3 of dry CO2 lean gas or 93.6 mg/tonne of product CO2 at
15 oC. Similarly, Nitrosomorpholine emissions decrease from 8.93 x 10-3 mg/tonne of product
CO2 at 45 ̊C to roughly half at 15 ̊C. Reduction in emissions of other compounds such as
formaldehyde, acetaldehyde, acetone, methylamine and acetamide with reduction in the
wash water temperature can be tracked through Tables 26 to 29. It should be noted that the
droplet phase emissions are based on a general rule of thumb for the droplet entrainment
from gas/liquid contactors as proposed by the Handbook of Gas Producers’ Association and
therefore, these values are purely indicative. In reality, the droplet phase emissions would
depend upon the hydrodynamics within the CO2 absorber and the design of its internals.
Nevertheless, Tables 26 to 29 provide some sort of base line for the environmental
regulatory bodies to set an emission standard for the amine based post combustion capture
technologies.
8.3 Atmospheric Emissions of AMP/PZ Solvent
AMP/PZ blend is supposedly the preferred solvent of MHI technology16, 94 and several
investigators57-60 have indicated that AMP could be used to co-absorb NO2 and CO2.
Currently, there is not much information available on the oxidative and thermal degradation
characteristics of the blend, though for the individual components of the blend some data is
available. As a result for the present case, only the atmospheric emissions of AMP and PZ
have been determined using the Aspen-Plus software. The AMP/PZ aqueous blend with
25% w/w AMP and 15% w/w PZ is almost same as 3 molar AMP and 1.5 molar PZ blend.
Therefore, as a check for the validation of Aspen-Plus based process simulations, AspenPlus (Version 7.3 updated) generated vapour-liquid equilibrium (V-L-E) data for the AMP-PZH2O system as a function of CO2 loading was compared with the experimental data available
from Bruder et al100 and Yang et al101. Figures 35 to 39 compare predictions of Aspen-Plus
1283.73
1283.74
1283.75
1283.76
1283.77
1283.78
1283.79
1283.8
0 10 20 30 40 50
M
EA
F
lo w R
at e (k g/ h) Wash Tower Temperature (°C)
MEA Component in Stream WW1
Prior WASH TOWER - Stream G3
After ABSORBER - Stream G1
65
results with the experimental results over the operating temperature range (40 to 120 ̊C) of
the AMP/PZ based post combustion capture plant.
Figure 35 – V-L-E data for 3M AMP/2M PZ/H2O system with CO2 loading at 40 oC
Figure 36 – V-L-E data for 3M AMP/2M PZ/H2O system with CO2 loading at 60 oC
Figure 37 – V-L-E data for 3M AMP/2M PZ/H2O system with CO2 loading at 80 oC
0.0001
0.001
0.01
0.1
1
10
100
1000
10000
0 0.2 0.4 0.6 0.8 1
C
O
2
Pr es su re (k Pa )
CO2 Loading (mol CO2 / mol Amine)
40°C
2010, Yang
2011, Bruder
ASPEN updated
0.001
0.01
0.1
1
10
100
1000
10000
0 0.2 0.4 0.6 0.8 1
C
O
2
Pr es su re (k Pa )
CO2 Loading (mol CO2 / mol Amine)
60°C
2010, Yang
2011, Bruder
ASPEN updated
0.001
0.1
10
1000
0 0.2 0.4 0.6 0.8 1
C
O
2
Pr es su re (k Pa )
CO2 Loading (mol CO2 / mol Amine)
80°C
2010, Yang
2011, Bruder
ASPEN updated
66
Figure 38 – V-L-E data for 3M AMP/2M PZ/H2O system with CO2 loading at 100 oC
Figure 39 – V-L-E data for 3M AMP/2M PZ/H2O system with CO2 loading at 120 oC
These results show that the Aspen-Plus (Version 7.3 Updated) software is a reasonably
reliable tool for predicting V-L-E data for AMP/PZ/H2O/CO2 system – a pre-requisite for
accurate material and energy balance. Similar checks were also made for the other blends
of AMP/PZ where AMP concentration varied from 1 molar to 3 molar and the PZ
concentration varied from 0.5 molar to 2 molar. With this validation, Aspen-Plus software
was used to develop the material and energy balance flow sheets using the same flue gas
input data and other operating process information (for example, the Direct contact cooler)
as for the base case which is described in detail in the previous milestone report. Figures 40
and 41 show material and energy flows for a capture plant with and without the condenser
for recovering the volatiles from the CO2 lean gas stream VENT1 leaving the absorber.
These figures show that the reboiler pressure and temperature are 211 kPa and 125 ̊C
respectively. The reboiler energy demand for regenerating the spent amine blend is roughly
3.8 MJ/kg of CO2 captured. It is roughly 11% less than the energy demand for the base case
(4.23 MJ/kg of CO2 – Please see Figure 33). The CO2 lean and CO2 rich amine loadings for
this case are 0.075 moles CO2 per mole of (AMP/PZ) and 0.312 moles CO2 per mole of
(AMP/PZ) respectively. Figures 42 and 43 show that content of AMP and PZ in the Wash
Tower outlet stream W1 increases as temperature of the wash water at inlet to the Wash
Tower drops. In other words, the atmospheric emissions of AMP and PZ decrease with
lowering the temperature of wash water circulating in the Wash Tower. Tables 30 to 33 list
the atmospheric emissions of various chemical species leaving the capture plant with the
CO2 lean gas stream VENT2, where there is a condenser for volatiles downstream of the
absorber and the wash water temperature is changing over a range. Tables 34 to 37 list the
atmospheric emission results when there is no condenser for the volatiles downstream of the
absorber. These results clearly signify the role of Wash Tower temperature control in
reducing the atmospheric emissions. Figures 44 and 45 also show that the vapour phase
0.01
0.1
1
10
100
1000
10000
0 0.2 0.4 0.6 0.8 1
C
O
2
Pr es su re (k Pa )
CO2 Loading (mol CO2 / mol Amine)
100°C
2011, Bruder
ASPEN updated
0.01
1
100
10000
0 0.2 0.4 0.6 0.8 1
C
O
2
Pr es su re (k Pa )
CO2 Loading (mol CO2 / mol Amine)
120°C
2011, Bruder
ASPEN updated
67
emissions of AMP, PZ, and total amines measured as ppmv decrease as the wash water
temperature at inlet to the Wash Tower decreases. These results clearly show that in an
industrial situation either a condenser downstream of absorber or inter-cooling of the
absorber for reducing the temperature rise due to heat of absorption reaction is necessary to
reduce the atmospheric emissions, if a blend of AMP/PZ is preferred solvent for CO2
capture.
The Aspen-Plus results show that when the lean solvent loading is 0.075 moles CO2 per
mole of AMP plus PZ, the lean solvent temperature is 40 ̊C and the flue gas temperature at
inlet to the absorber is 45 oC, then temperature of the CO2 lean gas at outlet of the absorber
is 55.4 ̊C and at this temperature, the gas stream carries approximately 1220 ppmv of AMP
plus PZ. Of course, the flue gas stream in this case is a typical Australian black coal-fired
power plant off gas. Using the CO2SIM process simulation software, SINTEF102 shows that
under identical absorber operating conditions except for the flue gas representing a natural
gas combined cycle power plant off gas stream and the lean solvent inlet loading being 0.1
mole CO2 per mole of total amine, the vapour phase content of AMP plus PZ in the CO2 lean
gas stream at the outlet of the absorber will be roughly 555 ppmv. When comparing the
Aspen-Plus and the CO2SIM simulation results, it should be noted that in addition to the
difference in lean solvent loading and the flue gas composition, there has been a difference
in the rich loading as well as the Liquid to Gas (L/G) ratio for the absorber. The Aspen-Plus
simulations show 0.312 moles CO2 per mole of total amine as the rich loading and 3.76 as
the mass basis L/G ratio in the absorber, whereas the CO2SIM uses the rich loading as 0.64
moles CO2 per mole of total amine and only 0.52 as the mass basis L/G ratio in the
absorber.
68
Figure 40 – AMP/PZ based CO2 capture plant material and energy flow sheet with condenser at the absorber outlet
Temperature (C)
Pressure (bar)
Mass Flow Rate (tonne/hr)
Vapor Fraction
Duty (MW)
Q Duty (MW)
W Power(MW)
40.0
1.05
6784.76
0.00
LEAN
45.0
1.09
1801.08
1.00
MEAFLUE
RE2
56.3
1.08
7089.64
0.00RICH1
55.3
1.05
1542.58
1.00VENT1
60.1
2.08
7090.00
0.00
RICH2
HEAT1
115.7
2.07
7090.00
0.01RICH3
125.0
2.11
6767.57
0.00L1
70.1
1.34
6767.57
0.00 L3
30.0
1.22
0.37
0.00
MAKEUP
43.8
1.05
4298.39
0.00
W2
43.8
1.05
4281.20
0.00
W3
43.8
1.05
17.19
0.00WWATER
0.00BLOWOUT
43.8
1.01
4298.39
0.00W1
42.0
1.02
4281.20
0.00
W4
70.0
1.05
6784.76
0.00
L7
43.8
1.01
1484.99
1.00
VENT2
110.0
1.00
1484.99
1.00
VENTGAS
HEAT2
Q
20.0
1.02
6.00
0.00WATER
46.8
1.03
1496.47
1.00G3
40.4
2.06
322.49
1.00GASCO247.0
1.04
1542.58
0.96
G2
46.8
1.03
46.12
0.00RE1 ABSORBER
QC=0.000
QR=0.000
HX4
Q=375.900
HX5
Q=-375.900
PUMP4
W=0.230
B3
PUMP3
W=0.006
HX3
Q=-8.708
B5
HX7
Q=29.342
B2
Q=0.000
STRIPPER
QC=-97.500
QR=335.965
B1
Q=-195.028
CONDENSE
Q=0.000
B6
W=0.000
B7
Q=-31.534
3.81
MJIKG
69
Figure 41 - AMP/PZ based CO2 capture plant material and energy flow sheet without condenser at the absorber outlet
Temperature (C)
Pressure (bar)
Mass Flow Rate (tonne/hr)
Vapor Fraction
Duty (MW)
Q Duty (MW)
W Power(MW)
40.0
1.05
6787.42
0.00
LEAN
45.0
1.09
1801.08
1.00
MEAFLUE
56.2
1.08
7045.02
0.00RICH1
55.4
1.05
1543.62
1.00VENT1
60.0
2.08
7046.10
0.00
RICH2
HEAT1
115.7
2.07
7046.10
0.01RICH3
125.0
2.11
6723.67
0.00L1
70.0
1.34
6723.67
0.00 L3
30.0
1.22
1.08
0.00
MAKEUP
48.7
1.05
2963.65
0.00
W2
48.7
1.05
2899.90
0.00
W3
48.7
1.05
63.75
0.00WWATER
0.00BLOWOUT
48.7
1.01
2963.65
0.00W1
42.0
1.02
2899.90
0.00
W4
69.8
1.05
6787.42
0.00
L7
48.7
1.01
1510.87
1.00
VENT2
110.0
1.00
1510.87
1.00
VENTGAS
HEAT2
Q
20.0
1.02
31.00
0.00WATER
40.6
2.06
322.45
1.00GASCO2
ABSORBER
QC=0.000
QR=0.000
HX4
Q=373.581
HX5
Q=-373.581
PUMP4
W=0.228
B3
PUMP3
W=0.004
HX3
Q=-20.753
B5
HX7
Q=27.986
B2
Q=0.000
STRIPPER
QC=-97.000
QR=335.020
B1
Q=-194.049
3.80
MJIKG
70
Figure 42 – Impact of the wash water temperature on AMP/PZ content of stream W1
(CO2 Capture plant design with condenser for volatiles downstream of absorber)
Figure 43 – Impact of the wash water temperature on AMP/PZ content of stream W1 (CO2
Capture plant design without condenser for volatiles downstream of absorber)
42488
42490
42492
42494
42496
42498
42500
0 10 20 30 40 50
AM
P/
PZ
F
lo w (k g/ h) Wash Water Temperature °C
Total APM+PZ in Stream W1
272840
272860
272880
272900
272920
272940
272960
272980
273000
273020
273040
0 5 10 15 20 25 30 35 40 45
AM
P/
PZ
fl ow R
at e (k g/ h) Wash Water Temperature °C
Total APM+PZ in Stream W1
71
Table 30 - Atmospheric emissions of AMP, PZ and other compounds at 43.8 oC
(Absorber with condenser, Wash water @ 42 oC)
Chemical
Emissions
Atmospheric Emissions (Stream VENT2)
Vapour
Phase
Droplet
Phase
Total Vapour
Phase
Droplet
Phase
Total
mg/Nm3
dry CO2
Lean gas
mg/Nm3
dry CO2
Lean gas
mg/Nm3
dry CO2
Lean gas
mg/tonne
CO2
mg/tonne
CO2
mg/tonne
CO2
AMP 8.99E+00 1.15E+00 1.01E+01 3.14E+04 4.02E+03 3.54E+04
PZ 1.52E+00 4.35E-01 1.96E+00 5.31E+03 1.52E+03 6.83E+03
SO2 2.90E+00 7.97E-05 2.90E+00 1.01E+04 2.78E-01 1.01E+04
NO2 1.13E-02 7.78E-07 1.13E-02 3.95E+01 2.72E-03 3.95E+01
NO 4.77E+02 6.67E-05 4.77E+02 1.67E+06 2.33E-01 1.67E+06
Table 31 - Atmospheric emissions of AMP, PZ and other compounds at 37.8 oC
(Absorber with condenser, Wash water @ 32 oC)
Chemical
Emissions
Atmospheric Emissions (Stream VENT2)
Vapour
Phase
Droplet
Phase
Total Vapour
Phase
Droplet
Phase
Total
mg/Nm3
dry CO2
Lean gas
mg/Nm3
dry CO2
Lean gas
mg/Nm3
dry CO2
Lean gas
mg/tonne
CO2
mg/tonne
CO2
mg/tonne
CO2
AMP 4.27E+00 1.11E+00 5.39E+00 1.49E+04 3.88E+03 1.88E+04
PZ 7.10E-01 4.22E-01 1.13E+00 2.48E+03 1.47E+03 3.95E+03
SO2 2.67E+00 8.52E-05 2.67E+00 9.33E+03 2.98E-01 9.33E+03
NO2 9.47E-03 8.15E-07 9.47E-03 3.31E+01 2.84E-03 3.31E+01
NO 4.77E+02 8.54E-05 4.77E+02 1.66E+06 2.98E-01 1.66E+06
Table 32 - Atmospheric emissions of AMP, PZ and other compounds at 31.3 oC
(Absorber with condenser, Wash water @ 22 oC)
Chemical
Emissions
Atmospheric Emissions (Stream VENT2)
Vapour
Phase
Droplet
Phase
Total Vapour
Phase
Droplet
Phase
Total
mg/Nm3
dry CO2
Lean gas
mg/Nm3
dry CO2
Lean gas
mg/Nm3
dry CO2
Lean gas
mg/tonne
CO2
mg/tonne
CO2
mg/tonne
CO2
AMP 1.79E+00 1.06E+00 2.85E+00 6.25E+03 3.71E+03 9.96E+03
PZ 2.91E-01 4.04E-01 6.94E-01 1.01E+03 1.41E+03 2.42E+03
SO2 2.41E+00 9.16E-05 2.41E+00 8.41E+03 3.19E-01 8.41E+03
NO2 7.63E-03 8.49E-07 7.63E-03 2.66E+01 2.96E-03 2.66E+01
NO 4.76E+02 1.12E-04 4.76E+02 1.66E+06 3.91E-01 1.66E+06
72
Table 33 - Atmospheric emissions of AMP, PZ and other compounds at 26.3 oC
(Absorber with condenser, Wash water at 15 oC)
Chemical
Emissions
Atmospheric Emissions (Stream VENT2)
Vapour
Phase
Droplet
Phase
Total Vapour
Phase
Droplet
Phase
Total
mg/Nm3
dry CO2
Lean gas
mg/Nm3
dry CO2
Lean gas
mg/Nm3
dry CO2
Lean gas
mg/tonne
CO2
mg/tonne
CO2
mg/tonne
CO2
AMP 9.01E-01 1.03E+00 1.93E+00 3.14E+03 3.60E+03 6.74E+03
PZ 1.45E-01 3.92E-01 5.36E-01 5.04E+02 1.37E+03 1.87E+03
SO2 2.21E+00 9.64E-05 2.21E+00 7.70E+03 3.36E-01 7.70E+03
NO2 6.39E-03 8.71E-07 6.39E-03 2.23E+01 3.04E-03 2.23E+01
NO 4.76E+02 1.37E-04 4.76E+02 1.66E+06 4.79E-01 1.66E+06
Table 34 - Atmospheric emissions of AMP, PZ and other compounds at 48.7 oC
(Absorber without condenser, Wash water @ 42 oC)
Chemical
Emissions
Atmospheric Emissions (Stream VENT2)
Vapour
Phase
Droplet
Phase
Total Vapour
Phase
Droplet
Phase
Total
mg/Nm3
dry CO2
Lean gas
mg/Nm3
dry CO2
Lean gas
mg/Nm3
dry CO2
Lean gas
mg/tonne
CO2
mg/tonne
CO2
mg/tonne
CO2
AMP 1.31E+02 7.84E+00 1.39E+02 4.59E+05 2.74E+04 4.86E+05
PZ 5.16E+01 7.97E+00 5.95E+01 1.80E+05 2.79E+04 2.08E+05
SO2 2.90E+00 6.02E-05 2.90E+00 1.01E+04 2.10E-01 1.01E+04
NO2 1.13E-02 5.52E-07 1.13E-02 3.95E+01 1.93E-03 3.95E+01
NO 4.77E+02 4.61E-05 4.77E+02 1.67E+06 1.61E-01 1.67E+06
Table 35 - Atmospheric emissions of AMP, PZ and other compounds at 45.5 oC
(Absorber without condenser, Wash water @ 35 oC)
Chemical
Emissions
Atmospheric Emissions (Stream VENT2)
Vapour
Phase
Droplet
Phase
Total Vapour
Phase
Droplet
Phase
Total
mg/Nm3
dry CO2
Lean gas
mg/Nm3
dry CO2
Lean gas
mg/Nm3
dry CO2
Lean gas
mg/tonne
CO2
mg/tonne
CO2
mg/tonne
CO2
AMP 9.10E+01 7.58E+00 9.86E+01 3.18E+05 2.64E+04 3.44E+05
PZ 3.54E+01 7.71E+00 4.31E+01 1.24E+05 2.69E+04 1.51E+05
SO2 2.83E+00 6.32E-05 2.83E+00 9.87E+03 2.20E-01 9.87E+03
NO2 1.06E-02 5.79E-07 1.06E-02 3.70E+01 2.02E-03 3.70E+01
NO 4.78E+02 5.27E-05 4.78E+02 1.67E+06 1.84E-01 1.67E+06
73
Table 36 - Atmospheric emissions of AMP, PZ and other compounds at 41.2 oC
(Absorber without condenser, Wash water @ 25 oC)
Chemical
Emissions
Atmospheric Emissions (Stream VENT2)
Vapour
Phase
Droplet
Phase
Total Vapour
Phase
Droplet
Phase
Total
mg/Nm3
dry CO2
Lean gas
mg/Nm3
dry CO2
Lean gas
mg/Nm3
dry CO2
Lean gas
mg/tonne
CO2
mg/tonne
CO2
mg/tonne
CO2
AMP 5.51E+01 7.28E+00 6.24E+01 1.92E+05 2.54E+04 2.17E+05
PZ 2.14E+01 7.40E+00 2.88E+01 7.44E+04 2.58E+04 1.00E+05
SO2 2.72E+00 6.73E-05 2.72E+00 9.48E+03 2.35E-01 9.48E+03
NO2 9.61E-03 6.16E-07 9.61E-03 3.35E+01 2.14E-03 3.35E+01
NO 4.78E+02 6.30E-05 4.78E+02 1.67E+06 2.19E-01 1.67E+06
Table 37 - Atmospheric emissions of AMP, PZ and other compounds at 36.3 oC
(Absorber without condenser, Wash water @ 15 oC)
Chemical
Emissions
Atmospheric Emissions (Stream VENT2)
Vapour
Phase
Droplet
Phase
Total Vapour
Phase
Droplet
Phase
Total
mg/Nm3
dry CO2
Lean gas
mg/Nm3
dry CO2
Lean gas
mg/Nm3
dry CO2
Lean gas
mg/tonne
CO2
mg/tonne
CO2
mg/tonne
CO2
AMP 3.11E+01 7.01E+00 3.81E+01 1.08E+05 2.44E+04 1.33E+05
PZ 1.21E+01 7.12E+00 1.92E+01 4.21E+04 2.48E+04 6.69E+04
SO2 2.59E+00 7.23E-05 2.59E+00 9.02E+03 2.52E-01 9.02E+03
NO2 8.54E-03 6.55E-07 8.54E-03 2.97E+01 2.28E-03 2.97E+01
NO 4.79E+02 7.67E-05 4.79E+02 1.66E+06 2.67E-01 1.66E+06
Figure 44 – Vapour phase emissions of AMP and PZ (With condenser)
0.00
0.50
1.00
1.50
2.00
2.50
3.00
0 10 20 30 40 50
PP
M
V
Wash Water Temperature °C
Atmospheric Emissions via Stream VENT2
AMP
PZ
Total
74
Figure 45 – Vapour phase emissions of AMP and PZ (Without condenser)
The Aspen-Plus results show that the use of a condenser in between the absorber and the
Wash Tower reduces the Wash Tower load and the atmospheric emissions. Taking into
account Australian inland summer and winter temperatures for the cooling water, Table 38
below shows the maximum and minimum levels of atmospheric emissions of AMP and PZ
for both cases of capture plant designs i.e. with and without a condenser upstream of the
Wash Tower.
Table 38 – Maximum and minimum atmospheric emissions of AMP/PZ
(mg per Nm3 of Dry CO2 Lean Gas Basis)
Chemical
Capture Plant With Condenser Capture Plant Without Condenser
Maximum Emissions
@ 43.8 oC
Minimum Emissions
@ 26.3 oC
Maximum Emissions
@48.7 oC
Minimum Emissions
@ 36.3 oC
Vapour Liquid Vapour Liquid Vapour Liquid Vapour Liquid
AMP 8.99E+00 1.15E+00 9.01E-01 1.03E+00 1.31E+02 7.84E+00 3.11E+01 7.01E+00
PZ 1.52E+00 4.35E-01 1.45E-01 3.92E-01 5.16E+01 7.97E+00 1.21E+01 7.12E+00
Total 10.5E+00 1.58E+00 1.05E+00 1.42E+00 1.83E+02 15.8E+00 4.32E+01 14.1E+00
8.4 Atmospheric Emissions of MDEA/MEA Solvent
A blend of 4 molar MDEA and 1 molar MEA is supposedly the preferred solvent for PSR
process for CO2 capture that originated from the University of Regina, Canada. This process
takes advantage of high absorption reaction rates with MEA and twice the CO2 carrying
capacity of MDEA without causing an excessive rise in the reboiler energy demand for the
regeneration of blended solvent. Moreover, MDEA has better oxidative and thermal
degradation resistance than un-inhibited aqueous MEA solution. Whilst the degradation
characteristics of MDEA/MEA blend have been studied by the research group from the
University of Regina, no degradation kinetics and no detailed degradation product
distribution have been quantified as yet. Moreover, MDEA seems to degrade preferentially
when blended with MEA as observed by Idem et al. Therefore in the present Aspen-Plus
process analysis, the degradation kinetics of MDEA and MEA have not been included.
However, as new information on degradation of MDEA and MEA blend in a post combustion
environment becomes available in future, the Aspen-Plus process models will be upgraded
and the atmospheric emissions of MDEA/MEA blend and its degradation products updated.
0.00
5.00
10.00
15.00
20.00
25.00
30.00
35.00
40.00
45.00
0 10 20 30 40 50
PP
M
V
Wash Water Temperature °C
Atmospheric Emissions via Stream VENT2
AMP
PZ
Total
75
Figure 46 shows the Aspen-Plus simulation based material and energy flow for a capture
plant that uses the aqueous solution of 25% w/w or 2.1 molar MDEA and 5% w/w or 0.82
molar MEA. In arriving at these results the same flue gas input data and other operating
process information (for example, the Direct contact cooler) as for the base case which is
described in detail in the previous milestone report were used. The CO2 loadings of lean
solvent at inlet to the absorber and of rich solvent at outlet of the absorber were set at 0.1
and 0.3 moles of CO2 per mole of total amine respectively for this exercise. The Aspen-Plus
simulation results show that the reboiler energy requirement in this case is 4.28 MJ per kg of
CO2, roughly same as in the case of 30% w/w un-inhibited aqueous MEA solvent and
temperature of the CO2 lean gas stream at outlet of the absorber is only 41.4 ̊C. This allows
for not having a condenser for removing the volatiles prior to water washing since the total
vapour phase concentration of MDEA and MEA in stream VENT1 is only 5.25 ppmv. Figure
47 shows the impact of varying temperature of wash water stream W4 and thereby the Wash
Tower temperature as the summer and winter period cooling water temperature changes in
inland Australia. As the wash water temperature decreases, the total amine content of
Stream W1 at outlet of the Wash Tower increases i.e., in other words, the atmospheric
emissions of the amines MDEA and MEA decrease. Figure 48 shows the vapour phase
atmospheric emissions of MDEA, MEA and total amines expressed as ppmv decreasing with
the dropping temperature of wash water stream W4. Tables 39 to 42 list the atmospheric
emissions of various chemical species leaving the capture plant with the CO2 lean gas
stream VENT2. In arriving at these results, the chemical reactions of SO2 and NO2 with
MDEA as well as MEA were not considered, though in the previous milestone report
degradation products of MEA including nitrosamines were included. In addition, the
degradation data available on individual amines can not be applied as sum total to predict
the degradation product distribution for the blend since MDEA is known to degrade
preferentially in the presence of MEA as observed by Idem et al. Nevertheless these results
clearly signify the role of Wash Tower temperature control in reducing the atmospheric
emissions. Taking into account Australian inland summer and winter temperatures for the
cooling water, Table 43 below shows the maximum and minimum levels of atmospheric
emissions of MDEA and MEA.
76
Figure 46 – MDEA/MEA blend based CO2 capture plant material and energy balance flow sheet
40.0
1.05
12761.97
0.00
LEAN
46.2
1.04
13094.47
0.00RICH1
41.4
1.03
1468.31
1.00
VENT1
47.0
1.90
13094.48
0.00
RICH2
HEAT1
106.3
1.88
13094.48
0.01RICH3
57.0
1.34
12761.79
0.00 L3
30.0
1.22
0.00
0.00
MAKEUP
120.3
1.85
12761.79
0.00L1
L4
40.1
1.05
1832.38
0.00
W2
40.1
1.05
1832.20
0.00
W3
40.1
1.05
0.18
0.00WWATER
40.1
1.02
1832.38
0.00W1
40.1
1.02
1832.20
0.00
W4
60.1
1.80
332.86
1.00GASCO2
57.0
1.05
12761.97
0.00
L7
40.1
1.02
1469.14
1.00
VENT2
110.0
1.00
1469.14
1.00
VENTGAS
HEAT2
Q
20.0
1.02
1.00
0.00
WATER
MJIKG
ABSORBER
QC=0.000
QR=0.000
HX4
HX5
Q=-865.680
PUMP4
W=0.358
STRIPPER
QC=-147.000
QR=375.666
HX6
Q=-229.255
B3
PUMP3
W=0.002
HX3
Q=0.000
B5
HX7
Q=30.378
B2
Q=0.000
Temperature (C)
Pressure (bar)
Mass Flow Rate (tonne/hr)
Vapor Fraction
Duty (MW)
Q Duty (MW)
W Power(MW)
4.28
45.0
1.09
1801.08
1.00
MEAFLUE
77
Figure 47 – Impact of wash water temperature on MDEA/MEA content of Stream
Figure 48 – Vapour phase emissions of MDEA and MEA
Table 39 - Atmospheric emissions of MDEA, MEA and other compounds at 40.1 oC
(Wash water @ 40 oC)
Chemical
Emissions
Atmospheric Emissions (Stream VENT2)
Vapour
Phase
Droplet
Phase
Total Vapour
Phase
Droplet
Phase
Total
mg/Nm3
dry CO2
Lean gas
mg/Nm3
dry CO2
Lean gas
mg/Nm3
dry CO2
Lean gas
mg/tonne
CO2
mg/tonne
CO2
mg/tonne
CO2
MEA 1.37E-01 5.98E-01 7.35E-01 4.80E+02 2.10E+03 2.58E+03
MDEA 9.79E-02 8.09E-01 9.07E-01 3.44E+02 2.84E+03 3.18E+03
SO2 1.63E+00 4.93E-05 1.63E+00 5.73E+03 1.73E-01 5.73E+03
NO2 2.86E-03 2.27E-07 2.86E-03 1.00E+01 7.96E-04 1.00E+01
NO 4.75E+02 7.79E-05 4.75E+02 1.67E+06 2.73E-01 1.67E+06
16391.75
16391.8
16391.85
16391.9
16391.95
16392
0 10 20 30 40 50
M
D
E
A
Pl us M
E
A
F
lo w R
at e (k g/ h) Wash Water Temperature °C
Total MDEA+MEA in Stream W1
0.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0 10 20 30 40 50
PP
M
V
Wash Water Temperature °C
Atmospheric emissions via Stream VENT2
MEA
MDEA
Total
78
Table 40 - Atmospheric emissions of MDEA, MEA and other compounds at 35.8 oC
(Wash water @ 30 oC)
Chemical
Emissions
Atmospheric Emissions (Stream VENT2)
Vapour
Phase
Droplet
Phase
Total Vapour
Phase
Droplet
Phase
Total
mg/Nm3
dry CO2
Lean gas
mg/Nm3
dry CO2
Lean gas
mg/Nm3
dry CO2
Lean gas
mg/tonne
CO2
mg/tonne
CO2
mg/tonne
CO2
MEA 7.34E-02 5.77E-01 6.50E-01 2.58E+02 2.02E+03 2.28E+03
MDEA 5.73E-02 7.80E-01 8.38E-01 2.01E+02 2.74E+03 2.94E+03
SO2 1.57E+00 5.31E-05 1.57E+00 5.51E+03 1.86E-01 5.52E+03
NO2 2.60E-03 2.43E-07 2.60E-03 9.12E+00 8.52E-04 9.12E+00
NO 4.75E+02 9.30E-05 4.75E+02 1.67E+06 3.26E-01 1.67E+06
Table 41 - Atmospheric emissions of MDEA, MEA and other compounds at 31 oC
(Wash water @ 20 oC)
Chemical
Emissions
Atmospheric Emissions (Stream VENT2)
Vapour
Phase
Droplet
Phase
Total Vapour
Phase
Droplet
Phase
Total
mg/Nm3
dry CO2
Lean gas
mg/Nm3
dry CO2
Lean gas
mg/Nm3
dry CO2
Lean gas
mg/tonne
CO2
mg/tonne
CO2
mg/tonne
CO2
MEA 3.59E-02 5.57E-01 5.93E-01 1.26E+02 1.95E+03 2.08E+03
MDEA 3.08E-02 7.53E-01 7.84E-01 1.08E+02 2.64E+03 2.75E+03
SO2 1.50E+00 5.76E-05 1.50E+00 5.26E+03 2.02E-01 5.26E+03
NO2 2.31E-03 2.60E-07 2.31E-03 8.09E+00 9.14E-04 8.09E+00
NO 4.75E+02 1.13E-04 4.75E+02 1.67E+06 3.98E-01 1.67E+06
Table 42 - Atmospheric emissions of MDEA, MEA and other compounds at 28.5 oC
(Wash water @ 15 oC)
Chemical
Emissions
Atmospheric Emissions (Stream VENT2)
Vapour
Phase
Droplet
Phase
Total Vapour
Phase
Droplet
Phase
Total
mg/Nm3
dry CO2
Lean gas
mg/Nm3
dry CO2
Lean gas
mg/Nm3
dry CO2
Lean gas
mg/tonne
CO2
mg/tonne
CO2
mg/tonne
CO2
MEA 2.41E-02 5.47E-01 5.71E-01 8.46E+01 1.92E+03 2.01E+03
MDEA 2.17E-02 7.40E-01 7.62E-01 7.62E+01 2.60E+03 2.68E+03
SO2 1.46E+00 6.02E-05 1.46E+00 5.12E+03 2.11E-01 5.12E+03
NO2 2.15E-03 2.69E-07 2.15E-03 7.55E+00 9.46E-04 7.55E+00
NO 4.75E+02 1.26E-04 4.75E+02 1.67E+06 4.42E-01 1.67E+06
79
Table 43 – Maximum and minimum atmospheric emissions of MDEA/MEA
(mg per Nm3 of Dry CO2 Lean Gas Basis)
Chemical
Capture Plant With Condenser
Maximum Emissions
@ 40.1 oC
Minimum Emissions
@ 28.5 oC
Vapour Liquid Vapour Liquid
MEA 1.37E-01 5.98E-01 2.41E-02 5.47E-01
MDEA 9.79E-02 8.09E-01 2.17E-02 7.40E-01
Total 2.35E-01 1.41E+00 4.58E-02 1.29E+00
8.5 Comparison of MEA, AMP, PZ and MDEA Emissions
Comparing the Aspen-Plus process simulation results given in Tables 32 and 41 for the
vapour phase atmospheric emissions of AMP/PZ and MDEA/MEA blends, it is clear
that at identical temperature the vapour phase emissions of AMP are highest, followed
by PZ, then MEA and finally MDEA. This is in accordance with the literature data for
the volatility of various amines when in aqueous solution as a single amine solute
except for MEA. The explanation for lower vapour phase emission of MEA than that of
PZ is its low concentration in the MDEA/MEA blend (only 5% w/w). At this low
concentration, most of MEA in solution could be already in the carbamate form due to
its higher reactivity with CO2 compared to MDEA and thereby suppressing its volatility.
Nevertheless, Nguyen et al103-105 explain that when experimentally generating the
vapour-liquid-equilibrium data for aqueous solutions of different types of amines, they
observed that the vapour phase mole fraction of an amine was influenced by the
molecular groups in the amine structure and the shape of the structure. Amines having
one or more polar groups such as hydroxyl and ether, tended to be less volatile when
in solution due to favourable interactions with water. The presence of one or more
methyl groups in a structure seemed to contribute to non-polarity or greater volatility as
is observed in the present case for AMP. Nguyen et al concluded that as a second
order effect, the presence of an N-CH3 contribution in a straight chained amine, or a CCH3 contribution in a cyclic amine, correlates to lower volatility. Similarly, to a small
extent, the cyclic amines appear to be less volatile than the straight chain amines.
Thus, the amines with higher volatility invariably have high values for Henry’s constant.
Table 44 below shows the boiling points of a number of pure amines and their Henry’s
constant values when in aqueous solution at 40 ̊C. Thus, it is obvious that one amine
may be more volatile than another in its pure form but the opposite is true when each
amine is in water. Equation 48 below shows the Henry’s constant at 40 ̊C as a function
of various polar and non-polar group contributions according to Nguyen et al.
ln (H-amine, 40 oC) = [4.19 ± 0.09] – [1.65 ± 0.17](N) – [0.21 ± 0.07](NH) – [1.55 ± 0.17](R-O-R)
+ [0.7 ± 0.08](Non Cyclic C-CH3) + [2.63 ± 0.21](Cyclic N-CH3) ------ (48)
The analysis of Nguyen et al103-105 therefore shows what could be generally expected
for vapour phase atmospheric emissions of amines and their degradation products in
an industrial situation when aqueous solutions of individual amine or their blends are
used for post combustion CO2 capture, i.e. Irrespective of the normal boiling points, the
amines with higher values for Henry’s constant at a given stream temperature for the
CO2 lean gas leaving the wash tower will have greater vapour phase atmospheric
80
emissions or conversely these amines will require a larger wash water circulation in the
wash tower in order to meet a given atmospheric emission limit than the amines with
lower values, if these amines are either used as solvent for CO2 capture or they are
formed as degradation products of a different amine as solvent.
Table 44 – Comparison of boiling points and measured Henry’s constant
for amines
Amine Boiling Point (oC)
Henry’s Constant
(Pa)
Methyl-diethanolamine (MDEA) 245 12.7
Diglycolamine (DGA) 223 13.9
Piperazine (PZ) 146 43.4
2-Methyl-piperazine (2-MPZ) 155 48.2
Ethylenediamine (EDA) 117 62.7
Monoethanolamine (MEA) 170 70.7
1-Methyl-piperazine (1-MPZ) 119 114
2 –Amino-2-methyl-1-propanol (AMP) 166 288
9. Pilot Plant Based Solvent Degradation Observations
As mentioned in the introduction section, post combustion capture of CO2 from power
plant flue gas streams using the amine based gas separation technologies that were
originally developed for the gas processing and oil industries has been practised for the
last several decades. These plants have been built predominantly to supply CO2 for
enhanced oil recovery (EOR) and meet the demand for CO2 from food industry.
Currently, there are 4 large scale plants (200 to 800 TPD CO2) operating in the world
that are exclusively recovering CO2 from flue gas stream generated by the coal-fired
power plants. Table 45102 lists the post combustion capture plants currently operating
in the world at capacity more than 100 TPD CO2. These plants use inhibited 18 to 30%
w/w aqueous MEA or an aqueous solution of hindered amine with a rate promoter for
example KS-1. Lately, CANSOLV technology is being implemented at SaskPower’s
Boundry Dam based coal-fired power plant site to co-capture SO2 and CO2 from flue
gas.
Whilst the plants listed in Table 45 have been operating for several years and the
technology providers have concentrated over the years mainly on the corrosion,
solvent loss due to degradation, cost of equipment and energy consumption issues,
only lately environmental impact of the atmospheric emissions of solvent and its
degradation products particularly nitrosamines and nitramines has become a serious
issue. With regard to this latest issue, these process technologies are being reexamined at the process development and pilot plant scale. Table 46102 lists the pilot
plants currently operating in the world that are geared to investigate the solvent
degradation, corrosion and environmental impact (atmospheric emission) issues in
addition to assessing the performance of amine solutions in actual service.
81
Table 45 – Amine based CO2 capture plants
Capture Technology Location Size (TPD) Flue Gas CO2 Use
Kerr-MCGee/ABB Lumus Trona, California 600 Coal-fired Soda Ash
Kerr-MCGee/ABB Lumus Shady Point, Oklahoma 200 Coal-fired Food
Kerr-MCGee/ABB Lumus Botswana, Africa 300 Coal-fired Soda Ash
Kerr-MCGee/ABB Lumus Warrior Run, Maryland 200 Coal-fired Food
Fluor Econamine Uttar Pradesh, India 150 Gas Furnace Urea
Fluor Econamine Sechuan Province, PRC 160 Gas Furnace Urea
Fluor Econamine Bellingham, Massachusetts 320 Gas-fired Food
Fluor Econamine Chiba, Japan 160 Gas-fired Food
Mitsubishi Heavy
Industries
Kedah Darul Aman,
Malaysia 200 Gas Furnace Urea
Mitsubishi Heavy
Industries Fukoka, Japan 330 Gas Furnace General
Mitsubishi Heavy
Industries Aonla, India 450 Gas Furnace Urea
Mitsubishi Heavy
Industries Phulpur, India 450 Gas Furnace Urea
Mitsubishi Heavy
Industries Kakinada, India 450 Gas Furnace Urea
Mitsubishi Heavy
Industries Abu Dhabi, UAE 400 Gas Furnace Urea
Mitsubishi Heavy
Industries Bahrain 450 Gas Furnace Urea
Mitsubishi Heavy
Industries Ghotoki, Pakistan 340 Gas Furnace Urea
Mitsubishi Heavy
Industries Phu My, Vietnam 240 Gas Furnace Urea
CERI Shanghai, China 360 Coal-fired General
To the credit of technology providers, it must be said that the issue of solvent loss due
to degradation and the degradation of solvents causing corrosion particularly in the
equipment parts made out of carbon steel, stainless steel and copper alloys is well
understood by them from the point of keeping the fixed and operating costs of the
capture plants low. This is the reason why the use of oxygen scavengers such as
sodium sulphite (Na2SO3), hydrazine (N2H4), carbohydrazide (H6N4CO), erythorbate,
methylethylketoxime (MEKO), hydroquinone, dietylhydroxylamine and their mixtures is
listed in the public domain literature106 for either passivating metal surfaces by creating
an oxide layer or binding chemically the dissolved oxygen in amine solutions. Similarly,
proprietary corrosion inhibitors, for example Max-amine GT-741C made by GE107 are
82
used to prevent corrosion of process equipment caused by dissolved CO2 and organic
acids.
Table 46 – Pilot plant scale test facilities
Owner Location Size (kg/h) Flue Gas
SINTEF/NTNU Trondheim, Norway 10 Propane burner
Dong Energy/Vattenfall Esbjerg, Denmark 1000 Coal-fired power plant
RWE/BASF/Linde Niderauβem, Tsykland 300 Lignite-fired power plant
RWE Power Didcot, UK 42 Coal-fired power plant
Aberthaw Power Station UK 2083
Coal-fired power plant,
CANSOLV technology
University of Texas Texas, USA 200 Coal-fired boiler
DOE Wilsonville, Alabama Coal/Bio-fuel fired
ITC for CO2 Capture University of Regina 167
SaskPower Boundary Dam, Canada 4000 Lignite-fired power plant
Huaneng Beijing, China 500 Coal-fired power plant
ENEL Brindisi, Itali 2000 Coal-fired power plant
Loy-Yang Power/CSIRO Loy-Yang, Australia 100 Lignite-fired power plant
Tarong Energy/CSIRO Tarong, Australia 150 Coal-fired power plant
Hazelwood Power/CO2CRC Hazelwood, Australia 25 to 50 TPD Lignite-fired power plant
The following sections describe the solvent degradation studies conducted at some of
the plants listed in Table 46. Any atmospheric emissions measurements of either
solvents and/or their degradation products, where measured and put in public domain,
are also included in the discussion.
9.1 Solvent Degradation Emissions – Dong Energy Plant
Dong Energy pilot plant study108-112 is a joint venture under EU Project Castor where
Dong Energy and Vattenfall Nordic built a 1 ton/hr CO2 capture plant at Esbjergvaerket,
400 MW coal-fired power plant, site in Denmark. The plant uses bituminous coal and
roughly 0.5% of flue gas is used as the feed gas stream for the capture plant. The
power plant is equipped with state of the art de-NOX (high dust selective catalytic
reactor), cold-sided electrostatic precipitator (ESP) and wet limestone based flue gas
desulphurisation (FGD) plants.
A slip stream of flue gas is taken at a position immediately after the SO2 scrubber. The
flue gas does not undergo any pre-scrubbing or cooling before it enters the CO2
absorber. The absorber consists of four consecutive packed-beds for CO2 absorption
and one water wash bed at the top. The absorber has an internal diameter of 1.1
metre. Each bed for CO2 absorption is 4.25 metres in height and filled with IMTP50
random packing. The water wash bed is 3 metres in height and filled with structured
packing. The water wash section functions as a closed loop. The rich solvent from the
absorber is pumped through two mechanical filters in series and a plate heat
exchanger (Rich-Lean exchanger with 10 ̊C ∆T) before it is fed to the stripper. The
83
stripper has an internal diameter of 1.1 metre and consists of two 5 metre high beds
filled with IMTP50 random packing and a water wash bed at the top (3 metres of
IMTP50 random packing). The stripper is driven by 350 kPa (absolute) saturated steam
via a thermosiphon reboiler. The overhead vapours from the stripper are quenched in a
condenser and the condensate returned to the stripper wash section. Essentially pure
CO2 product gas stream is returned to the main power plant flue gas duct. The
regenerated solvent after Lean/Rich heat exchange is cooled to the absorber inlet
temperature in a water-cooled trim cooler before returning to the absorber. A slipstream
of approximately 10% of the solvent flow is passed through a carbon filter to remove
organic products. Also part of the pilot plant is a reclaimer vessel where soda ash is
used to remove impurities. Table 47 below gives the CO2 capture plant specifications.
Table 47 – Dong Energy Pilot Plant Specifications
Parameter Design Value
Flue gas capacity 5000 Nm3/h
CO2 removal (at 12% by volume CO2) 1000 kg/h
CO2 capture efficiency 90%
Maximum solvent flow 40 m3/h
Maximum stripper pressure 300 kPa (absolute)
Flue gas condition 47 ̊C @ inlet to absorber,
<10 ppmv SO2, <65 ppmv NOX,
<10 mg/Nm3 (wet basis)
With the above plant uninhibited 30% w/w MEA, 3M (or 27% w/w) AMP blended with
1M (or 9% w/w) PZ and several other solvents were tested for determining the solvent
effectiveness as well as solvent degradation. Sections 9.1.1 and 9.1.2 below present
the results from the MEA and the AMP/PZ campaigns respectively.
9.1.1 Uninhibited MEA Solvent Campaign
Knudsen et al109 found that under optimal operating conditions for MEA campaign, the
liquid to gas ratio (L/G) in the absorber is between 2.5 and 3. At this ratio, the steam
demand for 90% CO2 recovery was 3.7 GJ/ton of CO2 captured and MEA consumption
was 1.4 kg/ton of CO2 captured. For CO2 recovery from gaseous fuels with no SO2,
MEA consumption of 1.6 kg/ton CO2 has been reported by Chapel et al113. Knudsen et
al109 assert that the uninhibited MEA solvent campaign lasted for only 500 continuous
hours and the solvent used during the campaign was fresh, therefore the overall MEA
consumption was somewhat low despite the fact that flue gas (after FGD section)
contained 6 ppmv SO2. However, the authors detected 0.19 kg of heat stable salts per
ton of CO2 captured in the lean solvent (or 0.5% w/w) after 500 hours of operation. The
uptake of sulphur by MEA solvent as a result of chemical reactions (see reaction
equations 14 to 16) was 85%; i.e. 85% of sulphur coming into absorber as SO2 reacted
with amine. At the end of 500 hours, detailed emission measurements conducted on
flue gas leaving the absorber (after the wash section) and the CO2 gas from stripper
condenser revealed the presence of volatile degradation products, e.g. ammonia,
formaldehyde, acetaldehyde and acetone in both streams. In particular the emission of
ammonia (25 mg/Nm3) from absorber was significant, though no MEA (<0.01 mg/Nm3)
was detected. Since ammonia is formed mainly as a result of oxidative degradation of
MEA, assuming 1:1 mole basis between ammonia and MEA, 0.4 kg MEA per ton of
CO2 or roughly 30% of the total MEA loss (consumption) could be due to oxidative
84
degradation. Assuming irreversible reactions between SO2 and MEA, and 1:1 mole
basis, Knudsen et al109 calculated that 0.15 kg of MEA per ton of CO2 (or 10% of the
total MEA loss) occurs due to SO2 induced degradation. The remaining MEA loss (≈
60%) or 0.85 kg MEA per ton of CO2 may be due to thermal degradation and reaction
with NOX. Aas and da Silva111 state that the atmospheric emissions of formaldehyde
with CO2 lean gas (leaving the water wash tower) were 0.06 mg per Nm3 of lean gas
and oxygen concentration in the lean gas stream was 7.5% by volume on dry basis.
Similarly, the NOX content of CO2 lean gas leaving the wash section was 73 mg/Nm3.
In 2009, same pilot plant was operated for 3360 hours (20 weeks) under EU Project
CESAR to closely monitor oxidative and thermal degradation of uninhibited 30% w/w
MEA. After 11 weeks of operation, liquid samples were taken from different parts of the
plant: the lean amine in the absorber, the rich amine in the stripper and the water
washes (both for the absorber and stripper). In addition, final lean MEA solution
corresponding to the end of the campaign was analysed. Lepaumier et al42 state that
more degradation products were observed in the final lean MEA solvent than what their
own laboratory studies at equivalent conditions had determined. In the real plant
environment, Lepaumier et al42 found that the contribution of thermal degradation was
quite limited. Accordingly, N-(2-hydroxyethyl)imidazolidone (HEIA) was identified but in
a very small amount and N-(2-hydroxyethyl)ethylenediamine (HEEDA) was not
detected at all. This point is in agreement with Strazisar34 et al’s observation when they
analysed the degraded lean amine solution at Kerr-McGee/ABB Lumus technology
based 800 tons/day CO2 capture plant at Trona, California. Oxazolidone, N-(2hydroxyethyl)imidazole (HEI) and N-(2-hydroxyethyl)formamide (HEF) – the products of
oxidative degradation of MEA – were all detected with other amide derivatives N-(2hydroxyethyl)acetamide (HEA), 2-hydroxy-N(2-hydroxyethyl)acetamide (HHEA) and
N,N’-bis(2-hydroxyethyl)oxalamide (BHEOX). In addition, two major products 4-(2hydroxyethyl)piperazin-2-one (HEPO) and N-(2-hydroxyethyl)-2-(2hydroxyethylamino)acetamide (HHEA) were identified as major oxidative degradation
products. The last two products were also identified by Strazisar et al34 at CO2 capture
plant based in Trona, California. Lepaumier et al42 state that HEA, HHEA and BHEOX,
which were detected in reasonable good amounts, are formed by reaction of MEA with
respectively acetic, glycolic and oxalic acids. This observation confirms that in a real
plant environment, the heat stable salts are not really stable but they form amide
derivatives depending upon the plant operating conditions and how long the solvent
has been circulating between the absorber and stripper without reclamation step. Aas
and da Silva111 present the results of gas sample analysis conducted during this long
uninhibited MEA solvent campaign as given in Table 48. Figures 49 and 50 present the
measurements of atmospheric emissions of acids, dust and constituents of dust taken
during the 3360 hours of operation at absorber inlet and at outlet of the absorber wash
section.
Table 48 – Atmospheric emission measurements (mg/Nm3 of dry CO2 lean gas)
Compound Absorber Inlet Absorber Outlet Water Wash Outlet
MEA <0.1 0.7 <0.3
DEA <0.2 <0.3 <0.2
Formaldehyde <0.1 0.7 to 1.1 <0.1
Methylamine <0.2 <0.3 <0.2
Acetamide <0.6 <1.0 <1.0
Ammonia <0.1 23 20
85
It is interesting to note that neither Knudsen et al109 nor Lepaumier42 et al comment on
the formation and detection of nitrosamines or nitramines as a result of NOX induced
degradation of MEA, though DEA formation has been detected during the CESAR
campaign (Table 48). As observed in the lab-scale experiments by Pedersen et al26,
one would expect formation of nitrosamines during both CASTOR and CESAR
campaigns.
Figure 49 – Atmospheric of acids and dust at Dong Energy Pilot plant
Figure 50 – Atmospheric emissions of dust constituents at Dong energy’s pilot plant.
86
The results shown in Figures 49 and 50 clearly indicate the contribution of the water
wash section of CO2 capture plant in reducing the atmospheric emissions of volatile
organics and the particulate matter. It is interesting to note from Figure 50 that as the
dust particles get washed out in the absorber, the iron (Fe) present in the dust particles
must end up in the CO2 rich amine stream and catalyse both the oxidative and thermal
degradation of uninhibited MEA solvent as observed by Goff6 and Sexton7.
It should be noted that the atmospheric emission results of Dong Energy’s pilot plant
have been used for the ANLEC R&D Project, “Process Modelling for Amine-based
Post-Combustion Capture Plant” Milestone 3.1 Report99 and that has further been a
basis in this report for determining the atmospheric emissions as a function of the
Wash Tower operating conditions in Australian context (See section 8). In reality, the
ambient conditions at the capture plant site, the design of capture plant and the overall
plant operating practice will greatly influence the actual extent of degradation product
formation and its subsequent atmospheric emissions.
9.1.2 AMP/PZ Campaign
In a recent publication, Mertens et al112 describe how Dong Energy’s CO2 capture plant
at Esbjergvaerket has been utilised lately to understand the effectiveness of 3M
AMP/1M PZ blend as an alternative of 30% w/w MEA solvent. For this campaign, the
plant was equipped with on-line continuous measurement techniques to monitor the
impact of various process parameters such as the wash tower operating conditions,
lean amine solvent temperature at inlet to the absorber, CO2 content of flue gas etc on
the atmospheric emissions of AMP and PZ. These authors state that increasing CO2
content of flue gas left less “free” amine in the solvent and therefore, emissions of AMP
and PZ at outlet of the wash section decreased. Similarly, lowering the CO2 loading of
lean solvent increased the wash tower load for the same operating conditions and
resulted into higher emissions. Lowering the wash water temperature or increasing the
frequency of changing water (more makeup water) lowered emissions of both AMP and
PZ. In the case of AMP, 70% of total AMP loss was determined as vapour phase
emissions whereas for PZ, this figure was 3%. This is in accordance with the
observations of Nguyen et al103-105 (see Section 8.5), that AMP has a higher volatility
than PZ. The remaining 30% for AMP and 97% for PZ was determined to be lost
through several other mechanisms: formation of heat stable salts, gaseous emissions
of degradation products, AMP and PZ losses with waste water from the plant etc. This
suggests that volatility of amines rather than droplet entrainment is a dominant pathway
for atmospheric amine emissions with AMP/PZ blend. Volatile organics such as
formaldehyde, acetaldehyde and ammonia were not detected at the outlet of the wash
section due to the functional limits of on-line measurement equipment. Surprisingly,
Mertens et al112 give no gas or liquid sample analysis results for the AMP/PZ blend
campaign and do not present any evidence for detection or otherwise of either
oxidative or thermal degradation products of AMP and PZ including Mononitrosopiperazine (MNPZ) and Di-nitrosopiperazine (DNPZ) which are expected to be
formed due to nitrosation of PZ in the presence of NOX in the flue gas (see Sections 4
to 5.2).
9.2 University of Regina (ITC) & SaskPower’s Boundary Dam Plants
The International test Centre (ITC) for CO2 capture technologies at the University of
Regina, Canada, has been operating a CO2 capture pilot plant from around 2001 to
develop the formulated amine solvent (mostly a mixture of MEA and MDEA) as an
87
economically viable alternative of 30% w/w MEA114. It uses 4800 m3/day of natural gas
combustion flue gas (11.6% v/v CO2, 1.1% v/v O2 and 87.3% v/v N2) to produce 1
tonne/day of CO2. The absorber is composed of 0.3 m-diameter sections for a total
height of 10 m and equipped with a series of temperature sensors and gas sampling
points at a regular interval of 0.6 m to allow measurements of temperature and gasphase CO2 concentration during testing. The stripper is also of same dimensions as the
absorber. A natural gas fired, 250 kW, steam boiler supplies saturated steam to the
stripper for the solvent regeneration.
SaskPower, Canada, has research collaboration for carbon capture and sequestration
with the ITC at University of Regina. Its Boundary Dam based CO2 capture plant115-117
uses a slip stream (14160 m3/day) of flue gas from 150 MW lignite-fired power plant to
produce 4 tons of CO2 per day. The slip stream has composition of 15% v/v CO2, 5%
v/v O2, 15% v/v H2O, 380 ppmv SO2 and 350 ppmv NOX116, 117. It is first passed
through a high efficiency bag-house followed by an Anderson 2000 SO2 scrubbing
unit116 before entering the capture plant. The absorber has 0.45 m diameter and 19.5
m height whereas the stripper has 0.4 m diameter and 18 m height. The Anderson
2000 FGD unit improves the flue gas quality prior to entering the absorber to virtually
free of fly ash and <10 ppmv SO2 and NO2.
At the above plants, 30% w/w MEA and a blend of MEA/MDEA have been tested as
solvents. The MEA/MDEA blend had total concentration of MEA and MDEA as 30%
w/w but MEA/MDEA ratio was 4:1. It should be noted that though MDEA is known to
act as a corrosion inhibitor, a proprietary corrosion inhibitor was added to the solvents
used in the Boundary Dam plant but no corrosion inhibitor was used at the ITC plant117.
It should be noted that though the scope of work for this report restricts to dealing with
the solvent degradation where flue gas is only from a black coal-fired power plant, the
MEA and MEA/MDEA degradation data for treating the flue gas from a lignite-fired
power plant is considered here because the flue gas composition for Boundary Dam
plant is surprisingly looks similar to that of a black coal-fired plant in Australia, at least
in terms of CO2, O2, SOX and NOX concentrations. Therefore, the section 9.2.1 below
presents the solvent degradation results for MEA and MEA/MDEA solvents as
observed at ITC and Boundary Dam plants.
9.2.1 MEA and MEA/MDEA Blend Degradation at Pilot Plants
Idem et al117 describe that at both the ITC and Boundary Dam plants the amine
emissions increased when the CO2 loading of lean amine for both MEA and
MEA/MDEA blend solvents was reduced. With the decreasing lean loading, a
corresponding increase in the reboiler heat duty was observed at both plants for the
same CO2 capture capacity. Thus, in the industrial situation one will have to trade off
the energy efficiency of process with the atmospheric emissions.
In the case of Boundary Dam plant straight chain amines such as 1-propanamine,
cyclic compounds such as 1,2,3,6-tetrahydro-1-nitrosopyridine and 2-pyrrolidinone,
dialcholos such as 1,2-ethanediol, as well as sulphur compounds such as
isothiocynatoethane and 1,1-dioxid-tetrahydrothiophene were observed as major
degradation products of MEA and MEA/MDEA mixed solvents. Idem et al117 further
state that the sulphur compounds may have resulted from contact of aqueous MEA and
MDEA with trace amounts of SO2 that survived the scrubbing process in the Anderson
2000 SO2 unit. Since, the detection limit of the SO2 analyser was about 5 ppmv, SO2 in
88
the flue gas at absorber inlet must have been less than 5 ppmv. The wide variety of
degradation products observed in the Boundary Dam demonstration plant samples,
despite the use of proprietary corrosion inhibitors and oxygen scavenging additives,
illustrates the effect of a harsher environment brought about by a coal-fired power plant
flue gas. In the case of the ITC plant, lean amine sample analysis for both MEA and
MEA/MDEA solvents indicated insignificant degradation and the presence of only trace
amounts of degradation products, which included 1-propanamine and 2-pyrrolidinone,
and just more than trace amounts of 1,2-ethanediol and 1,2,3,6-tetrahydro-1nitrosopyridine. It should be noted that the ITC flue gas was natural gas combustion
based and with lower oxygen level compared to flue gas for the Boundary Dam plant.
Also, the lean amine solution for the ITC plant did not have either corrosion inhibitors or
oxygen scavenging additives in it117.
Idem et al117 further state that the degradation rate for the aqueous MEA used in the
Boundary Dam plant was 0.5 mole% per day. In the case of mixed MEA/MDEA solution
for the same plant, the MEA degradation rate was 2.3 mole% per day while the rate of
MDEA was 1.5 mole% per day. Thus, MEA degraded preferentially in the presence of
MDEA contrary to what was observed by these authors in their laboratory scale
degradation studies82, 83. Since, the reboiler temperatures (115 to 120 ̊C) and the lean
amine concentrations were identical for both the Boundary Dam and ITC capture
plants, the only reason for extensive solvent degradation in the case of Boundary Dam
plant could be the quality of flue gas (coal-fired v/s gas fired) and catalytic effect of
corrosion inhibitor for degradation). Unfortunately, Idem et al117 neither clarify the exact
nature of the corrosion inhibitor nor the quantification of degradation products for both
the pilot scale capture plants.
It is unfortunate that Idem et al117 neither state the characterisation and quantification
of heat stable salts nor they mention presence of any thermal degradation products in
the lean amine solvent samples for both the pilot scale capture plants. Comparison of
the laboratory based degradation studies (see Table 49) by Lawal et al82, 83 with the
pilot plant solvent degradation study117 implies that in an industrial situation, the solvent
degradation products formed are quite different from those detected in the laboratory
environment due to inability to replicate the exact flue gas quality, hydrodynamics and
degradation process conditions.
Table 49 – Solvent degradation products in laboratory environment
(55 to120 oC, 250 kPa O2 Pressure)
MEA-H2O-O2-CO2 System
(7 molar MEA, 0.27 mol CO2/mol MEA)
MEA-MDEA-H2O-O2-CO2 System
(7 molar MEA, 2 molar MDEA, 0.43 mol CO2/mole total amine)
Product Formula Product Formula
Methyl-pyrazine C5H6N2 3-Methyl pyridine C6H7N
7-Oxabicyclo-ketone C6H6O2 1,2-Propanediamine C3H10N2
Ethylamine C2H7N 2-Butanamine C4H11N
1-Propanamine C3H9N 1-Amino-2-propanol C3H9NO
1,3,5-Triazine C3H3N3 N-Hydrocarbaminic acid C7H14N2O5
12-Crown-4 C8H16O4 Dimethylamine C2H7N
2-(2-ethoxy)ethanol C6H14O3 15-Crown-5 C10H20O5
15-Crown-5 C10H20O5 2-Ethenoxy-ethanol C4H8O2
Oxacyclo-octadecane C12H24O6 Ethyl-urea C3H8N2O
Dimethyl-1,2-ethanediamine C4H12N2
89
9.3 RWE/LINDE/BASF Niderauβem Pilot Plant
In July 2009, RWE, Linde and BASF jointly commissioned 7.2 tons/day (300 kg/h) CO2
capture plant at Niderauβem, Germany, based 1000 MW lignite-fired power plant118121. This plant was operated for 5000 hours with 30% w/w uninhibited MEA solvent at
90% CO2 capture efficiency. A slip stream of 1550 Nm3/h flue gas containing 14.2% v/v
CO2 (dry basis) with 190 mg/Nm3 NOX (dry basis at 6% O2), 93 mg/ Nm3 SOX (dry basis
at 6% O2) and 5% v/v O2 (dry basis) was treated with roughly 4025 kg/h of aqueous
MEA solvent. Upstream of the capture plant a wet FGD and electrostatic precipitators
were used. The direct contact cooler of the capture plant reduced flue gas temperature
from 65 ̊C to 40 ̊C at the absorber inlet and since it was operated with dilute caustic
soda solution, the SOX content of flue gas at inlet to the absorber was less than 5
ppmv. Similarly the NOx content of flue gas at inlet to the absorber fluctuated between
110 and 130 mg/Nm3. The absorber was intercooled and equipped with water wash
tower at the top to minimise the atmospheric emissions of MEA and its volatile
degradation products. The temperature of CO2 lean flue gas leaving the water wash
tower was usually around 35 to 40 ̊C. The liquid to gas ratio (L/G) in the absorber was
3.1 on mass basis.
With the above background information, Moser et al121 state that the overall MEA loss
or specific MEA consumption during 5000 hours of plant operation was 290 g/ton of
CO2. The concentration of MEA in CO2 lean gas leaving the plant was only 0.02 to 0.03
mg/Nm3 (dry basis) but the product CO2 stream leaving the stripper condenser carried
between 8.1 and 11.8 mg of MEA per Nm3 of product gas (dry basis). Thus, the vapour
phase loss of MEA (combination of absorber and stripper) was roughly 6 g per ton of
CO2. The CO2 lean gas leaving the capture plant contained ammonia (26.9 to 46.6
mg/Nm3), acetaldehyde (0.5 to 1 mg/Nm3) and acetone (0.2 mg/Nm3). The product
CO2 gas contained 1.1 mg/Nm3 ammonia, 11.6 to 12.9 mg/Nm3 acetaldehyde and 0.3
to 1 mg/Nm3 acetone. Thus, the total specific loss of MEA was determined as: 323 to
563 g/ton of CO2 due to ammonia formation, 4 to 8 g/ton of CO2 due to acetaldehyde
formation and less than 1 g/ton of CO2 due to acetone formation. Comparing with the
data on specific consumption of MEA, the total specific loss of MEA does not match as
far as the plant material balance is considered and this could be attributed to incorrect
stoichiometry applied by Moser et al121 for the formation of various volatile degradation
products such as ammonia, acetaldehyde and acetone from MEA due to oxidative
degradation.
Moser et al121 further claim that the lean solvent analysis after 5000 hours of operation
showed that the oxidative degradation of MEA contributed significantly towards the
formation of heat stable salts (203 g/ton CO2) which consisted of 1.8 % w/w acetate,
0.2% w/w formate and less than 0.1% w/w oxalates. In addition, traces of thermal
degradation product 1-(2-hydroxyethyl)imidazolidone-2 (HEIA) were detected.
However, other thermal degradation products that are related to the carbamate
polymerisation such as Oxazolidine, HEEDA, trimer and polymer of MEA and cyclic
urea were not detected in the circulating solvent even after 5000 hours of operation.
Moser et al121 are silent on the detection of any nitrosamines and nitramines either in
the gas or liquid samples after 5000 hours. These investigators seem to have not found
1,2,3,6-tetrahydro-1-nitrosopyridine or 2-pyrrolidinone which Idem et al117 found in the
liquid sample at the Boundary Dam plant which is expected since these plants are
lignite-fired and have roughly similar flue gas composition.
90
This shows that in an industrial situation, the extent of solvent degradation depends
upon the plant operating practice and impact of process scale-up factors on
degradation which are normally difficult to reproduce at the laboratory scale
degradation studies.
9.4 University of Texas Pilot Plant Studies
The University of Texas has been operating 200 kg/h CO2 capture plant for a number
of years where the flue gas is derived from a coal combustion based steam generator.
The pilot plant is based within the Pickle Research Centre of the university. Its
absorber is made out of carbon steel and has internal diameter of 0.43 m. It is
equipped with 6.1 m high packed bed made in two equal sections with a collector plate
and redistributor for liquid between the two packed beds. The facility has been used for
testing the novel solvents, corrosion, mass transfer, chemical kinetics, energy
efficiency and cost issues associated with the chemical solvent based post combustion
capture process. The solvent degradation aspects have been tested mostly in the
laboratory scale experiments and the results of these investigations have been
described in previous sections. The primary focus of these degradation studies has
been to develop cost effective corrosion inhibitors and additives to suppress the
oxidative degradation of amine solvents. Voice and Rochelle40 claim Diethylene
triamine Penta acetic acid (DTPA) as one of the best corrosion inhibitors for application
with MEA solvent. Among the sulphur containing inhibitors, Dimercaptothiadiazole
(DMTD) has also been considered to be a very effective corrosion inhibitor, though it is
relatively costly. In addition to DTPA and DMTD, Voice and Rochelle recommend
Diethylene triamine penta methylene phosphonic acid (DTPMP) and Hydroxy
ethylidene diphosphonic acid (HEDP). These additives suppress ammonia and formic
acid formation as a result of oxidative degradation by as much as 80% and only
required to be added to lean MEA solution at less than 1% w/w. Some recent results
from Prof. Rochelle’s research group suggest that a mixture of HEDP and DTPA, when
added to lean MEA solution at 1.5% w/w, suppresses ammonia formation as much as
97%.
9.5 CSIRO Pilot Plant Studies
CSIRO has been operating two CO2 capture pilot plants that use 30% w/w uninhibited
MEA as solvent to treat a slip stream of flue gas in one case, from a lignite coal fired
power plant, and in another case, from a bituminous coal fired power plant. The first
capture plant at Loy Yang in the La Trobe Valley of Victoria is based at Loy Yang
Power’s lignite coal-fired power plant whereas the second plant is based at Tarong,
Queensland, at the black coal-fired power plant site of Stanwell. Both power plants
have no de-NOX and FGD facilities but only bag filters for particulate removal.
The Loy Yang based capture plant123 has been designed to treat a maximum of 150
kg/h of flue gas at 90% CO2 capture efficiency. The plant has two absorbers (211 mm
internal diameter) made of stainless steel, each 9.4 m high with 2.7 m high packed bed
of Pall rings. The two absorbers can be run in series or in parallel as required. The
stripper is also made of stainless steel with 161 mm internal diameter and 6.9 m height
which is packed with 3.9 m high Pall ring bed.
The Tarong based capture plant122 is also all stainless steel construction but has 4
times bigger capacity for flue gas (600 kg/h). As a result, at 85% CO2 capture
91
efficiency, it is designed to produce 100 kg/hr of pure CO2. The absorber has internal
diameter of 0.35 m and total of 4 packed sections of Sulzer Mellapack M250X, each
1.78 m high and thus rendering a total packed height of 7.14 m. The stripper has an
internal diameter of 0.25 m. It contains two packed sections of Sulzer Mellapak 350X
packing in the stripping section, with a further 1.12 m in the condensate return. Whilst
the absorber is running close to atmospheric pressure, the stripper operates at around
1.8 atm pressure (absolute).
The uninhibited 30% w/w MEA campaign results122, 123 for the solvent degradation and
atmospheric emissions are as below:
a) It was observed that decreasing the CO2 loading of lean amine at Loy Yang plant
contributed towards higher emissions MEA to the atmosphere. Similarly, increasing
the liquid flow rate in the absorber or operating the absorber at higher L/G ratio
increased the atmospheric emissions of MEA. Thus, at L/G ratio of 2.2, MEA
emissions were recorded as 20 g/h which nearly doubled when L/G ratio was
increased to 4. Ammonia emissions from the absorber seemed to fluctuate
between 1 and 4 g/h. These emissions also tended to increase with increasing L/G
ratio and decreasing lean solvent loading. The high values for MEA and ammonia
emissions could be attributed to the design of the capture plant in the sense that
the Loy Yang pilot plant was not equipped with the water wash section downstream
of the absorber. Recently, the water wash process step has been installed at the
plant and a rigorous campaign with respect to detecting and quantifying the solvent
degradation products is underway.
b) The Tarong pilot plant MEA campaign lasted for 500 hours. Table 50 shows the
typical flue gas composition, flow rate and temperature recorded at various
locations in the capture plant during the campaign.
Table 50 – Flue gas composition during MEA campaign at Tarong capture plant
Component Unit Plant
Inlet
Absorber
Inlet
Absorber
Outlet
Wash Tower
Outlet
Product
CO2
CO2 Vol% 10.1 10.9 1 0.9 98.2
H2O Vol% 7.8 4.3 8.3 3.2 1.8
O2 Vol% 6.3 7.1 6.6 6.8 0
N2 Vol% 75.8 77.7 84 89.2 0
SO2 ppmv 197 <5 5.3 <5 <5
SO3 ppmv <5 <5 <5 <5 <5
NO ppmv 135 135 136 135 7.4
NO2 ppmv <5 <5 <5 <5 <5
N2O ppmv <5 <5 <5 <5 <5
MEA ppmv <5 <5 <5 <5 <5
HCL ppmv 7 <5 <5 <5 <5
HF ppmv 11 <5 <5 <5 <5
Gas Flow Rate Kg/h 500 496 428 410 71
Gas Temperature oC 104.5 28.1 41.4 27.6 23.2
The capture plant was operated at L/G ratio between 2.4 and 5.4. The CO2 loadings of
lean and rich amine solutions, in and out of the absorber, remained in the range 0.08 to
0.36 and 0.124 to 0.524 respectively. The minimum reboiler energy demand was
measured as 3.8 MJ/kg CO2 at the L/G ratio of 2.4. The lean MEA analysis showed that
the heat stable salts seemed to increase linearly with time on-stream for the capture
plant. At the end of 500 hrs, the lean MEA solvent carried 0.37% w/w heat stable salts.
92
This result is in line with the observation of Knudsen et al109 for the formation of heat
stable salts at Esbjergvaerket plant during uninhibited MEA campaign (see Section
9.1.1). The rich amine solution analysis also showed that over 500 hours, the iron
content of solution varied with a maximum level of 1.5 mg per litre of the solution
measured. This could be due to corrosion of the corrosion coupons installed at the pilot
plant during the operation with MEA. 128 coupons in total were installed at the pilot
plant and consisted of carbon and stainless steels.
The gas analysis involved measurements of 24 species including water vapours,
carbon dioxide, carbon monoxide (CO), nitrous oxide (N2O), nitric oxide (NO), nitrogen
dioxide (NO2), sulphur dioxide (SO2), sulphur trioxide (SO3), ammonia (NH3), hydrogen
chloride (HCl), hydrogen fluoride (HF), methane (CH4), ethane (C2H6), ethylene (C2H4),
n-propane (C3H8), n-hexane (C6H14), formaldehyde (CH2O), acetaldehyde (C2H4O),
ethanol (C2H6O), ethanolamine (MEA, C2H7NO), Piperazine (C4H10N2),
Ethylenediamine (C2H8N2), N-Formylpiperazine (C5H10N2O), NOX (as sum of NO and
NO2), Total organic carbon (TOC) and Oxygen. The detection limit for the industrial
Fourier Transform Infra Red spectroscopy (FTIR) based gas sample analyser was only
5 ppmv. Hence, the solvent degradation product components detected at the wash
tower outlet at less than 5 ppmv have very limited accuracy. Table 51 shows the values
for various chemicals detected in the gas samples at the beginning and at the end of
the 500 hr campaign.
Table 51 – Gas sample analysis at the start and end of 500 hr campaign
Compound Campaign Start (ppmv) Campaign End (ppmv)
Absorber Outlet
Wash Tower
Outlet
Stripper
Outlet
Absorber
Outlet
Wash Tower
Outlet
Stripper
Outlet
CO 1.2 6.5 <5 8.4 48.1 0
N2O <5 <5 <5 <5 <5 <5
NO 135 229.6 6.07 0 204.7 7.4
NO2 <5 <5 <5 <5 <5 <5
SO2 <5 <5 6 <5 <5 5.1
SO3 <5 <5 <5 <5 <5 <5
NH3 8.4 19.5 <5 43.6 31.7 <5
HCl <5 <5 <5 <5 <5 <5
HF <5 <5 <5 <5 <5 <5
CH4 <5 <5 <5 <5 <5 <5
C2H6 <5 <5 <5 <5 <5 <5
C2H4 <5 <5 <5 <5 <5 <5
C3H8 <5 <5 <5 <5 <5 <5
C6H14 <5 <5 <5 <5 <5 <5
CH2O <5 <5 <5 <5 <5 <5
C2H4O <5 <5 <5 <5 <5 <5
C2H6O <5 <5 <5 <5 <5 <5
MEA <5 <5 <5 123.7 <5 <5
C4H10N2 <5 <5 <5 <5 <5 <5
C2H8N2 <5 <5 <5 <5 <5 <5
C5H10N2O <5 <5 <5 <5 <5 <5
NOX 135 230.3 6.07 0 205.7 7.6
Whilst the accuracy of detection below 5 ppmv is questionable, the results presented in
Tables 51 are important for ammonia formation as a signpost for the extent of oxidative
degradation of MEA. Ammonia emission at the outlet of the absorber on the start of
campaign is seen as 8.4 ppmv which rises to roughly 44 ppmv by the end of 500 hr
93
campaign. This result points towards increasing degradation of MEA over time. At the
wash tower outlet the atmospheric emissions of ammonia after 500 hrs is 32 ppmv. NO
seems to simply pass through the absorber.
Currently, efforts are being made to improve the characterisation and quantification of
gaseous emissions of the MEA degradation products.
10. Commercial CO2 Capture Technologies – Solvent Degradation
As a result of the environmental regulatory bodies and other statutory authorities
becoming increasingly aware of the potential risk of adverse environmental impact of
amine based CO2 capture technologies if used for global Greenhouse gas mitigation
purposes, the commercial technology providers have started to open up recently and
provide in a limited way information on the solvent degradation and resulting emissions
from the capture plants based on their technology. This information has been given
below.
Kerr-McGee/ABB-Lumus Technology
Arnold et al12 and Strazisar et al34 describe Trona, California based 600 tons/day CO2
production plant and provide the atmospheric emissions data as well as the solvent
degradation information. According to these authors, the plant uses 10 to 20% w/w
MEA solution containing a proprietary corrosion inhibitor and other additives to treat the
coal-fired boiler flue gas stream that carries 10 to 15% CO2, 5 to 10% O2, 5 to 15%
H2O, approximately 80% N2, 10 to 100 ppmv SO2 300 to 500 ppmv NOX. In the plant,
condensate from stripper condenser carries roughly 0.04% w/w MEA and it is used as
wash water in the absorber wash tower. The wash water circulating in the wash tower
carries 0.5% w/w MEA. The CO2 lean gas leaving the wash tower carried aerosol of
monoethanol sulphate and ammonium sulphate which are the products of reaction of
MEA and ammonia with SOX respectively. The wash water entrained in the CO2 lean
gas had the droplet size of less than 5 microns with majority less than 3 microns. A
later date installation of a Brink mist eliminator downstream of the wire-mesh demister
in the wash tower has mostly eliminated atmospheric emissions of the aerosols and
fine droplets from the plant. Table 52 and 53 give the analysis of various liquid
samples. Arnold et al further state that the solvent reclamation operation at the plant
usually indicated insignificant solvent degradation. Un
Table 52 – Liquid sample analysis
Analysis Absorber wash water
Stripper
reflux water
MEA absorber exhaust line
condensate
pH 8.7 6.4 9.0
Cl- (%) <0.001 <0.001 <0.01
SO42- (%) 0.06 0.008 <0.01
SO32- (ppmw) <10 <10 <10
CO2 (%) 0.6 0.08 0.32
NH3, ppmw (%) 17 34 40
NO3- ppmw 38 <8 NO2-, ppmw 17 <1 MEA (%) 0.5 0.04% 0.5
94
Table 53 – Amine solution analysis
Analysis Rich Amine
Lean
Amine
Reclaimer
Sample “A”
Reclaimer
Sample “B”
pH 9.1 10.3 11.0 10.0
K+ (%) 0.0002 0.002 0.001 0.009
Na+ (%) 0.012 0.013 0.7 2.0
Cl- (%) 0.006 0.005 0.04 0.2
SO42- (%) 1.6 1.7 4.8 15.5
SO32- (ppmw) 800 700 80 140
CO2 (%) 3.2 0.5 1.7 1.0
NH3, ppmw (%) 180 160 1700 1100
NO3- ppmw 650 650 5300 7000
NO2-, ppmw 250 250 <1 <1
MEA (%) 10.7 11.2 46 33
Insoluble (%) 0.004 0.01 0.02 0.08
Whilst Arnold et al12 describe the Trona plant operation and various liquid stream
composition analysis under taken in 1982, Strazisar et al34 have studied the solvent
degradation issue in 2003 while analysing the plant performance separately. They
analysed virgin concentrated MEA solution from storage tank for presence of impurities
and lean MEA solution at inlet to the absorber and the reclaimer bottoms for solvent
degradation impurities. Table 54 lists the solvent degradation products identified by
these authors in the reclaimer bottoms sample.
Table 54 – Solvent degradation products identified at Trona plant
Compound Chemical Formula
N-Formylethanolamine C3H7NO2
N-Acetylethanolamine C4H9NO2
2-Oxazolidone C3H5NO2
N-(hydroxyethyl)succinimide C6H9NO3
N-(2-hydroxyethyl)-lanthamide C5H11NO3
1-Hydroxyethyl-3-homopiperazine C7H14N2O2
1-(2-hydroxyethyl)-2-imidazazolidnone C5H10N2O2
1-Hydroxyethyl-2-piperazinone C6H12N2O2
4-Hydroxyethyl-2-piperazinone C6H12N2O2
3-Hydroxyethylamino-N-hydroxyethyl propanamide C7H16N2O3
2-Hydroxyethylamino-N-hydroxyethyl acetamide C6H14N2O3
Ammonia NH3
Acetic acid C2H4O2
Propionic acid C3H6O2
n-Butyric acid C4H8O2
MEA C2H7NO
2,6-Dimethyl-4-pyridinamine C7H10N2
2-Imidazolecarboxaldehyde C4H4N2O
1-Methyl-2-imidazolecarboxaldehyde C5H6N2O
95
Strazisar et al34 further state that in the lean MEA sample nitrosamines were found to
be present at a concentration of 2.91 µmol/mL. This is believed to be due to the
reaction between MEA and nitrogen oxides, which are known constituents of flue gas.
Assuming all of the nitrosamines present as nitrosodiethanolamine (NDELA as
proposed by Schallert31, 33), the measured concentration of nitrosamines is equivalent
to 390 mg of NDELA per litre of lean amine solvent which seems rather high. It should,
however, be noted that Strazisar et al’s quantification of nitrosamine concentration in
lean amine was based on a generalised functional group test for NO and no specific
nitrosamines were detected. These authors further state that nitrosamines were not
detected in the reclaimer bottoms perhaps due to their low boiling point.
Fluor Econamine FG PlusSM Technology
This technology uses 30% w/w MEA solvent with a corrosion inhibitor that tolerates
oxygen concentration in flue gas as much as 15% v/v. In fact, the corrosion inhibitor
requires oxygen in flue gas and dissolved in lean amine solution to remain active as
reported by Reddy et al13. This implies that the corrosion inhibitor works via passivating
the carbon steel surface with a film of oxides of iron (mostly Fe2O3 and Fe3O4). The
corrosion inhibition via surface passivation has design implications for the capture plant
equipment in the sense that ultimate design should be conducive of minimising the
sloshing and splashing of liquid within the absorber and flash vessels due to high gas
velocity. Sloshing and splashing causes erosion of oxide based protective film. Once, a
small portion of protective layer has eroded, it allows the corrosion reactions to
propagate underneath the remaining protective film. Published information12 on L/G
ratio and absorber diameter indicates that the superficial gas and liquid velocities in the
CO2 absorber and flash vessels need to be no more than 1.5 m/s and 15 cm/s
respectively to avoid surface erosion.
It should be noted that Fluor Econamine FG PlusSM technology incorporates recent
process improvements such as inter-cooling of absorber, split flow feed, vapour
compression and solvent flash prior to stripping in order to reduce the energy demand
of the process to the lowest level. A publication by Reddy et al implies that it could be
using either Piperazine type rate promoter to accelerate the reaction rate or MEA
concentration in the solution could be as much as 40% w/w. Most of the information on
this technology put in public domain by Fluor Ltd is sanitised and silent on the issue of
solvent degradation and atmospheric emissions of the degradation products. However,
Fluor Ltd admits that the capture plant does need a solvent reclaimer when treating a
coal-fired power plant flue gas and the flue gas should not have SOX/NOX more than 10
ppmv. Also the particular matter content of flue gas should be less than 3 mg per Nm3.
Mitsubishi Heavy Industries (MHI) Technology
This61 technology uses an aqueous solution of hindered amine blended with an amine
based rate promoter to enhance the kinetics of CO2 absorption. The solvent known as
KS-1 is claimed to have far better corrosion resistance characteristics and CO2 carrying
capacity than conventional MEA solvent. It is claimed to not require a corrosion
inhibitor at all. Unfortunately, KS-1 is at least 5 times more expensive than MEA12. A
recent publication by MHI63 states that at least nine post combustion CO2 capture
plants around the world have been operating that are built with their technology and
four more will be built by 2012 end. Most of the currently operating MHI plants have
CO2 capture capacity in the range 350 to 450 tonnes per day. Unfortunately none of
these plants are based on coal-fired power plant flue gas. With this technology only
96
one pilot plant at 10 tons per day CO2 capture capacity has been operating since 2006
at Matsushima power station in Southern Japan where over 6000 hours of cumulative
on-stream time recorded62.
Published information64 from MHI states that in comparison with conventional MEA
solvent, KS-1 requires 40% less solvent circulation rate for the same CO2 capture duty
and 20% less energy for solvent regeneration. In addition, the solvent loss due to
degradation is only 10% of what is encountered with conventional MEA. However to
achieve such superior performance when processing a coal-fired power plant flue gas,
the build of particulate matter in the lean solvent will need to be less than 10 ppmw and
the SOX content of flue gas at inlet of the absorber must not be more than 0.1 ppmv61.
Table 55 below compares the influence of SO3 content of flue gas on the atmospheric
emissions (total of vapour plus droplet phase) of KS-1 and conventional MEA solvents.
Table 55 – Influence of SOX content on atmospheric emissions of KS-1 and MEA
SO3 Concentration
@ Absorber Inlet
(ppmv)
KS-1 Emissions
(ppmv)
MEA Emissions
(ppmv)
3 23.2 67.5
1 9.1 29.8
0 0.4 0.8
MHI further state that in their 10 tons per day pilot plant, they observed that 1 to 3% of
flue gas NOX was consumed by solvent KS-1 resulting into the formation of heat stable
salts at low levels after 5000 hours of operation. The overall solvent loss (including loss
due to degradation) is claimed to be less than 1 kg per ton of CO2 by MHI15. No further
information is available from MHI on either the characterisation and quantification of
degradation products from KS-1 or their atmospheric emissions. However, a recent
publication by Svendsen94 implies that the MHI solvent KS-1 could be a blend of 3
molar AMP with 1 to 2 molar PZ. If that is true then the presence of PZ in solution is
most likely to produce mono and di-nitrosopiperazines upon reaction with NOX from
flue gas as described in the earlier sections on solvent degradation for AMP and PZ.
The formation of nitrosopiperazines and their potential for atmospheric emissions
points to adverse environmental impact with the use of AMP/PZ blends for CO2
capture.
11. ANTICIP ATED VOLATILE DEGRADATION EMISSIONS
Both the laboratory and plant based solvent degradation information presented in the
preceding sections points towards uncertainty of characterisation and questionable
accuracy with quantification of various solvent degradation products due to mismatch
of the operating process environment between the laboratory scale experiments and
the actual plant operations. In addition, other issues such as the materials of
construction of equipment for degradation studies acting as catalyst for solvent
degradation, flashing or otherwise of dissolved oxygen prior to the solvent being
regenerated and resubjected to the oxidative environment and the presence or
absence of corrosion inhibitors in the amine solution have made prediction of
atmospheric emissions of amines and their degradation products rather a very difficult
task. In addition to the above difficulties, however, the process simulations point
97
towards significance of the efficiency of wash tower as key equipment for minimising
the solvent loss via atmospheric emissions and thereby reduce the cost of makeup
solvent. Thus, the extent of atmospheric emissions of an amine based solvent and its
degradation products have not only a direct impact on environment but also on the cost
of capital involved in building a most efficient water wash tower and the cost of
consumables for operating a CO2 capture plant around the clock. A simple way to
screen any amine solvent for its propensity to be emitted from the process is to assess
its volatility and the volatility of its all possible degradation products. Whilst, this may
not take into account the effect of reaction kinetics, hydrodynamics, materials of
construction etc, it will provide some glimpse into what lays ahead, should a particular
amine or a blend of amines were to be chosen for the capture duty. In essence, it is a
preliminary check of comparative advantages or disadvantages of using a group of
amine solvents.
Nguyen et al104 quantify volatility of an amine or its degradation product via an
apparent activity coefficient, Ƴi, defined by the modified Rault’s law:
Ƴi = Pi ÷ [Xi × Poi] ------------ (48)
Where,
Pi = Partial pressure of amine or its degradation product i in the gas phase,
Xi = Liquid phase mole fraction of the amine or its degradation product i, and
Poi = Vapour pressure of the amine or its degradation product i at a given temperature.
According to the above expression, higher the mole fraction of component i, higher its
partial pressure for given activity co-efficient or volatility. All other things being equal,
this would result in a higher emission level. It should be noted that vapour pressure of a
chemical species i is a function of its free state in the liquid. Thus, the reactions that
bind up an amine as a carbamate or protonated amine will result in its lower
concentration as a free amine and lower vapour pressure.
According to equation 48, the volatility of the component i also depends on its vapour
pressure which can be predicted by using an empirical equation as follows:
Ln [Poi] = b1 + b2/T + b3 ln(T) + b4Tb5 ----------- (49)
Where Poi is in Pascal and T is in oK.
For MEA, PZ, MDEA and AMP the values of co-efficient b1 to b5 are available in the
literature and can be used to compare the volatility of these amines. Using this
concept, da Silva et al124 have recently estimated likely emissions of various amines
and their likely degradation products. These anticipated emission levels are reproduced
in Table 56 to 58 from the SINTEF report102. The underlying assumptions for arriving at
these numbers are:
i. Degradation product emission rate is proportional to its rate of formation.
ii. The rate of formation for a degradation product can be represented as a fraction
of the ammonia emission rate.
iii. Ammonia formation rate is an indicator of overall degradation rate of an amine.
98
iv. Ammonia emission for MEA is 1 ppmv, for MDEA 0.5 ppmv and 0.05 ppmv for
AMP and Piperazine.
v. Any degradation product that has a free energy of solvation >-5kcal/mol is a
volatile product. Volatile degradation products do not accumulate in the plant.
vi. For medium volatility degradation products (components with free energy of
solvation <-5kcal/mol and that are non-ionic) vapour emissions are proportional
to volatility and concentration in the liquid. Da Silva et al have used the
numbers for concentrations in liquid as informed guesses.
vii. Ionic degradation products are non-volatile.
The assumption (iv) could be interpreted as whatever is the rate of ammonia formation
during the degradation of MEA half of that is with MDEA and 1/20th is with either AMP
or PZ or their blend as a solvent. Accordingly, emission rate of ammonia for MDEA
degradation is half of what could be with MEA and 1/20th is for AMP and PZ. This same
logic should be applied when reading the numerical values for the anticipated emission
levels of other degradation products.
Table 56 – Emissions of non volatile degradation products
Degradation
Product CAS No.
Emissions
(ppmv)
Solvent Comment
Formic acid/formate 64-18-6
0.003 MEA
Oxidative degradation
product
0.002 PZ
0.003 MDEA
0.002 AMP
Acetic acid/acetate 64-19-7
0.03 MEA
Oxidative degradation
product
0.002 PZ
0.003 MDEA
0.002 AMP
Oxalic acid 144-62-7
0.003 MEA
Oxidative degradation
product
0.002 PZ
0.003 MDEA
0.002 AMP
Propanoic acid 79-09-4 0.0003 MEA
Glycolic
acid/glocolate 0.0003
MEA
Lactic acid/lactate 598-83-3 0.0003 MEA
Glycine 56-40-6 0.0003 MEA Possible oxidative degradation product
N-Glycylglycine 556-50-3 0.0002 Likely oxidative degradation product
99
Table 57 – Emissions of medium volatility degradation products
Degradation Product CAS No. Emission (ppmv) Solvent
Comments
Formamide
0.007 MEA
0.003 PZ
0.006 MDEA
0.003 AMP
Oxazolidin-2-one 0.004 MEA
1-(2-hydroxyethyl)-2 imidazoliinone 3699-54-5 0.003 MEA
N-(2-hydroxyethyl)-ethylenediamine 111-41-1 0.003 MEA
N-(2-hydroxyethyl)-acetamide 142-26-7 0.003 MEA
1-(2-hydroxyethyl)-2,5-Pyrrolidinone 18190-44-8 0.003 MEA
N-(2-Hydroxyethyl)lactamide 5422-34-4 0.003 MEA
N,N-di(2-hydroxyethyl)-Urea 15438-70-7 0.003 MEA
N-(2-hydroxyethyl)-3-[(2hydroxyethyl)amino]Propanamide 587876-41-3 0.003 MEA
N-(2-hydroxyethyl)-3-[(2hydroxyethyl)amino]acetamide 144236-39-5 0.003 MEA
1-(2-hydroxyethyl)-2-piperazinone 59702-23-7 0.003 MEA
4-(2-hydroxyethyl)-2-piperazinone 23936-04-1 0.003 MEA
2-((2-
[(hydroxyethyl)amino]ethyl)amino)ethanol 4439-20-7 0.003 MEA
2-methylaminoethanol 109-83-1 0.01 MEA
2,2’-[[2-[(2-
hydroxyethyl)amino)ethyl]imino]bisEthanol
60487-26-5 0.003 MEA
1,3-bis(2-hydroxyethyl)-2Imidazolidinone) 71298-49-2 0.003 MEA
Oxalamide 471-46-5 0.003 MEA
1-(2-hydroxyethyl)imidazole 1615-14-1 0.03 MEA
1H-imidazole-2-carboxaldehyde 10111-08-7 0.003 MEA
1-methyl-1H-imidazole-2-carboxaldehyde 13750-81-7 0.003 MEA
N,N-bis(hydroxyethyl)piperazine 122-96-3 0.003 MEA
Glycine 56-40-6 0.003 MEA
Hydroxy-acetaldehyde 141-46-8 0.003 MEA
100
Table 57- Emissions of medium volatility degradation products (continued)
Degradation Product CAS No. Emission (ppmv) Solvent
Comments
2-imidazolidinone 120-93-4 0.003 MEA
Morpholine 119-91-8 0.02 MEA
Diethanolamine 111-42-2 0.003 MEA
Methylnitramine 598-57-2 0.02 MEA
0.002 PZ
2-(Nitroamino)ethanol 74386-82-6 0.006 MEA
Extent of nitramines
formation unknown in
CO2 capture plant
N-nitrosodiethanolamine 1116-54-7 0.0005 MEA Detected in capture plants
0.001 MDEA
2-Oxopiperazine 5625-67-2 0.002 PZ
2,5-Piperazinedione 106-57-0 0.002 PZ
Ethylenediamine 107-15-3 0.002 PZ
1-Piperazinecarboxaldehyde 7755-92-2 0.003 PZ
1-Acetylpiperazine 13889-98-0 0.003 PZ
1,1’-Carbonyl-bis-piperazine 17159-16-9 0.002 PZ
1-Piperazine-ethanol 103-76-4 0.005 PZ
N-(hydroxymethyl)piperazine 90324-69-9 0.002 PZ
1-Nitrosopiperazine 5632-47-3 0.002 PZ Possible nitrosamine degradation product
1,4-dinitrosopiperazine 140-79-4 0.2 PZ Possible nitrosamine degradation product
1-Nitropiperazine 42499-41-2 0.003 PZ Possible nitramine degradation product
4-methyl-1-Piperazine-ethanol 5464-12-0 0.2 MDEA Observed experimentally
Triethanolamine 102-71-6 0.006 MDEA
N-nitroso-diethanolamine 1116-54-7 0.0008 MDEA
2-(Methylnitrosoamino)ethanol 26921-68-6 0.3 MDEA
4,4-dimethyl-2-Oxazolidinone 26654-39-7 0.01 AMP
Nitro-2-amino-2-methylpropanol 0.0003 AMP
101
Table 58 – Emissions of volatile degradation products
Degradation Product CAS No. Emission (ppmv) Solvent
Comments
Ammonia 7664-41-7
1 MEA Oxidative
degradation
product for all
solvents
0.05 PZ
0.5 MDEA
0.05 AMP
Methylamine 74-89-5 0.006 MEA
Possible
oxidative
degradation
product
Formaldehyde 50-00-0
0.08 MEA
0.004 PZ
0.004 MDEA
0.004 AMP
Acetaldehyde 75-07-0
0.02 MEA
0.001 PZ
0.009 MDEA
0.001 AMP
Acetone 67-64-1 0.002 MEA
Dimethylamine 124-40-3
0.01 MEA
0.00005 PZ
0.0005 MDEA
0.00005 AMP
Ethylamine 75-04-7
0.01 MEA
0.000035 PZ
0.00035 MDEA
0.000035 AMP
Diethylamine 109-89-7
0.01 MEA
0.00007 PZ
0.0007 MDEA
0.00007 AMP
N-nitrosodimethylamine 62-75-9
0.02 MEA
0.001 PZ
0.01 MDEA
0.001 AMP
4-nitroso-morpholine 59-89-2
0.01 MEA
0.0004 MDEA
2-Methyl-3-nitroso-oxazolidone 39884-53-2 0.003 MEA
1,4-dinitropiperazine 140-79-4 0.0001 PZ
102
Table 58 – Emissions of volatile degradation products (continued)
Degradation Product CAS No. Emission (ppmv) Solvent
Comments
Dimethylnitramine 4164-28-7
0.02 MEA
0.0001 PZ
0.001 MDEA
0.0001 AMP
1,4 Dimethylpiperazine 106-58-1 0.004 MDEA
2-methyl-2-(methylamino)-1-Propanol 27646-80-6 0.0003 AMP Observed by H. Lepaumier
3,4,4-trimethyl-oxazolidin-2-one 15833-17-7 0.0005 AMP Observed by H. Lepaumier
4,4 dimethyl-2-isopropyl-3nitrosooxazolidine 39884-58-7 0.015 AMP
da Silva et al124 clarify in their report that the above values of emissions are rather
conservative and on high side. They have tried to cover all possible degradation
compounds including those likely to be formed only in a laboratory environment and
those with some degree of uncertainty for their occurrence in an industrial environment.
In addition, da Silva et al state that there is a great deal of uncertainty about formation
of nitrosoamines and their emissions. Therefore, when using these tables caution
should be exercised. For example, with N-nitrosodiethanolamine (NDELA) listed in
Table 57, if ammonia emissions are 1 ppmv when using MEA solvent, then NDELA
emissions will be 0.5 ppb. If the emission level of ammonia drops to 0.2 ppm (due to
say process improvement) then NDELA emissions will reduce to 0.1 ppb. Similar
understanding should be used with other degradation products and solvents.
12. RANKING OF SOLVENTS
The amine solvents considered in this report are representative of primary (MEA,
AMP), secondary (PZ) and tertiary amines (MDEA). They exhibit a varied level of CO2
carrying capacity and reactivity to CO2. Whilst, it is easy to rank them in terms of their
capacity to carry CO2 or reactivity, in order to assess their environmental impact, they
should be ideally ranked for their potential to form and emit harmful compounds to
atmosphere, particularly when their usage for global mitigation of Greenhouse effect is
desired. This ranking should also consider the toxicity of the solvent degradation
products and their eventual fate in the environment. At present, there is no protocol or
weighting template available to CSIRO to account for all the parameters (tendency to
degrade, volatility of degradation products, toxicity of degradation products and fate of
these products in atmosphere) on a single basis. In such a situation, we know that the
compounds that have greatest health and environmental risk are the nitrosamines. The
risk for their formation is highest with secondary amines, lower with tertiary amines and
the lowest with primary amines. Based on this, one could rank the solvents from
greatest to lowest risk as:
103
PZ > MDEA > AMP = MEA
Another way to look at the solvents for their ranking is to consider the extent of
degradation to alkylamines, secondary amines (that may eventually form nitrosamines)
and volatile products. Therefore, in terms of degradation rates, the following ranking is
possible:
MEA > MDEA > PZ=AMP
Combining the above two rankings and giving more weight to the risk of nitrosamine
formation, the following ranking is likely:
PZ > MDEA > MEA > AMP
If we now superimpose on the above ranking the volatility of amines and consider that
most volatile amine will have maximum emissions in quantity for a worst case of
industrial accident then the following nominal ranking is obtained:
PZ > AMP > MEA > MDEA
Thus, the following generalisation can be derived for the above solvents:
a) Secondary amines have highest risk of nitrosamine formation, followed by
tertiary amines, while primary amines have the lowest risk of nitrosamine
formation.
b) All other things being equal, solvents with low vapour pressure are safer than
solvents with high vapour pressure.
c) All other things being equal, a more stable solvent that will resist degradation is
safer than a less stable one since the more stable solvent will have lower
emissions of degradation products.
13. CONCLUSIONS
Literature review of the laboratory studies, pilot plant scale experiments and public
domain technical information from various technology vendors on degradation of amino
solvents clearly indicates that in an industrial environment of post combustion CO2
capture, these solvents will most certainly undergo both oxidative and thermal
degradation. The extent of degradation and as a result, the type of degradation
products formed will, however, depend upon not only the structure of amine but also
the process operating conditions, such as the concentration of amine, its CO2 loading,
absorber reaction temperature, solvent regenerator temperature, content of oxygen,
SOX, NOX and particulate matter in flue gas, composition of particulate matter (Fe, Ni,
V, P, Cr, CO etc), catalytic effect of the material of construction of plant equipment
towards degradation etc. The plant operating practices, such as the process control
and how often a solvent is reclaimed, will also decide both the extent of solvent
degradation and type of degradation products formed.
In an industrial operating environment, potential degradation products of amine
solvents (combination of oxidative and thermal degradation) will be one or more of the
following:
104
Ammonia, primary amines / alkanolamines, secondary amines / alkanolamines, tertiary
amines / alkanolamines, aldehydes (formaldehyde, acetaldehyde), carboxylates,
amides, piperazines, piperazinones, oxazolidpnes, nitrosamines, imidazolidones, N,Ndistributed ureas and nitramines.
Exact chemical structure of these degradation products depends upon the chemical
structure of parent amine and the degradation reaction pathways it has followed.
The above list is by no means complete or entirely accurate since characterisation of
the degradation products is still ongoing internationally among various research groups
both at the laboratory and pilot plant scales. There are also differences in identification
of degradation products among various research groups due to either the differences in
analytical techniques employed or the kinetic and mass transfer limitations during
degradation have forced open different degradation pathways and formed different
products. Unfortunately, both the characterisation and quantification of degradation
products carried out in the laboratory environment are not entirely applicable to the
operating plant situation because the process environment and scale factors of
industrial plant are not duplicable or transferable at the laboratory scale. This is
evident, for example, between the findings of laboratory studies by Rochelle group at
University of Texas and the results of both ITC and Boundary Dam pilot plant results
from University of Regina. A similar conclusion could be drawn by comparing results of
pilot plant campaigns by various research groups.
Whilst there are a number of pilot plant scale CO2 capture campaigns ongoing around
the world, there has been so far not even a single study that has attempted to close the
material balance around plant where formation of degradation products within the plant
and the atmospheric emissions of these products are fully accounted. Nevertheless,
these campaigns do confirm the following:
a) A large number of degradation products are formed in an industrial plant
environment than what various research groups have so far determined in their
laboratory studies.
b) It is the oxidative degradation that is contributing more towards degradation than
thermal degradation in the industrial environment.
c) Vapour phase atmospheric emissions of heat stable salts and thermal degradation
products of amines are likely to be minimal to the point of no concern.
d) Formation of nitroso compounds in the industrial plant environment is a reality since
both Boundary Dam and ITC pilot plants in Canada confirm formation and detection
of 1,2,3,6-tetrahydro-1-nitrosopyridine during their MEA and MEA/MDEA
campaigns. It should be noted that the Boundary Dam pilot plant was originally built
by Union Carbide and later refurbished by Fluor Ltd for SaskPower116 as per the
Fluor Econamine technology and it used proprietary corrosion inhibitors for both
amine campaigns. Thus, Strazisar et al34 are correct in asserting that they detected
nitrosamines in the lean amine solution at Kerr-MCGee/ABB Lumus technology
based capture plant at Trona, California, and these compounds may have been
formed up to 2.91 µmol per mL of solution. It should be noted that Trona plant also
uses proprietary corrosion inhibitors despite the lean amine (MEA) concentration
being less than 20% w/w.
105
e) Corrosion inhibitors currently being recommended and perhaps used by
commercial technology vendors for post combustion CO2 capture may be acting as
catalyst towards the solvent degradation. Certainly, the corrosion inhibitors
containing Copper, Vanadium, Cobalt and other metals used in the reducing
environment of gas processing industry are catalyst for solvent degradation in the
oxidative environment of post combustion capture. Laboratory studies of solvent
degradation by Rochelle and other investigators confirm this.
f) The wash tower downstream of absorber plays critical role in controlling
atmospheric emissions of various amine solvents and their volatile degradation
products from a CO2 capture plant. The AspenPlus process simulation results
clearly indicate that the wash tower performance is affected by ambient conditions,
particularly the cooling water temperature. Operating this tower at temperature as
low as possible in practice can substantially reduce the emissions of volatile
degradation products. No doubt other process or equipment performance
improvement measures such as inter-cooling the absorber or separating the
condensable species downstream of the absorber (using reflux condenser) prior to
water washing the CO2 lean flue gas will certainly help in reducing atmospheric
emissions.
Published information on operating performance of the capture plant at Trona,
California, and the technology related information from MHI Ltd clearly states that the
flue gas impurities viz. the particulate matter, SOX and NOX contribute significantly
towards solvent degradation and atmospheric emissions of amine solvents. In fact, the
MHI data categorically shows that by reducing SOX content at inlet to the absorber from
3 ppmv to 1 ppmv reduces atmospheric emissions of both KS-1 and MEA solvents by
more than half. Recently published information from MHI further states that in order to
eliminate formation of aerosols in the CO2 lean exhaust stream, the SOX content of flue
gas at inlet to the absorber should be less than 0.1 ppmv. Similarly, to avoid the sludge
build up in absorber and resulting foaming and flooding of absorber as well as to
reduce the metal catalysed degradation of solvent, the particulate level in lean amine
solvent should not be allowed to exceed 1 ppm by weight. All of this means that for
stable performance of CO2 absorption/desorption system with high efficiency, minimum
solvent degradation and minimum atmospheric emissions, the direct contact cooler and
the water wash tower in post combustion CO2 capture plant must be performing
efficiently. This is particularly important in Australian context since Australian power
stations do not have de-SOX and de-NOX systems.
In terms of choosing an amine solvent for post combustion capture with minimal
environmental adverse impact, one should note the guidelines below:
i. Secondary amines have highest risk of nitrosamine formation, followed by tertiary
amines, while primary amines have the lowest risk of nitrosamine formation.
ii. All other things being equal, solvents with low vapour pressure are safer than
solvents with high vapour pressure.
iii. All other things being equal, a more stable solvent is safer than a less stable one
since the more stable solvent will have lower emissions of degradation products.
Using the above guidelines, the amine solvents considered in this report can be ranked
in order of their likely maximum adverse impact to minimal adverse impact as:
106
PZ > AMP > MEA > MDEA
14. RECOMMENDATIONS & FUTURE WORK
From the perspective of identifying accurately and anticipating amine solvent
degradation and resulting atmospheric emissions of both solvent and its degradation
products in an industrial scale post combustion CO2 capture plant, this report clearly
states that various laboratory based degradation studies have been less than helpful
due to several reasons: forced or accelerated degradation by using either very high
oxygen concentration in feed gas or high reactor temperatures, mismatch of operating
flow regimes, i.e. mass transfer v/s chemical kinetics control, lack of representative
feed gas including impurities in the feed gas (particulate matter, SOX and NOX), lack of
actual corrosion inhibitors and oxygen scavengers that are used in the industrial
environment and lack of hydrodynamics that are representative of the industrial
practice. Nevertheless, these studies have provided some useful insights of what could
be happening to an amine solvent circulating between the absorber operating at low
temperature and the solvent regenerator operating at high temperature (at least twice
of the absorber temperature). For example, ammonia formation rate is indicative of
oxidation degradation rate and the rate of formation of formates is indicative of heat
stable salts due to corrosion. These studies have also established clearly the possibility
of nitrosoamines and nitramines formation in an actual plant environment which has
been confirmed from the pilot plant results. In addition, the laboratory studies clearly
imply that overdesigning an industrial absorber for the purpose of high CO2 capture
rate or using a packing that increases the dynamic liquid holdup within both absorber
and stripper will contribute towards higher solvent degradation simply due to increased
residence time for oxygen dissolution in absorber and CO2 catalysis in stripper. These
studies have, of course, provided some understanding of the degradation reaction
pathways and the mathematical models of reaction kinetics which have been useful
towards calculating the likely atmospheric emissions of both solvent and its
degradation products using the process simulation tools such as AspenPlus. However,
validation of these estimates requires the pilot plant scale studies to characterise and
quantify the degradation products both in the gas phase and liquid phase accurately
and close the material balance around all the input and output streams of the plant.
Since degradation of amine solvent is an ongoing phenomenon within a capture plant
that is operating at steady state, the characterisation and quantification of degradation
products and the resulting reconciliation of plant material balance must be done
continually as on-stream time of the plant builds.
The pilot plant programs ongoing so far in different parts of the world give impression
from their publications that the characterisation and quantifications of gas and liquid
streams at inlet and outlet to the absorber, stripper and water wash towers downstream
of both absorber and stripper has not yet been complete. So far there has been only
one pilot plant based degradation study (RWE/LINDE/BASF Niderauβem Pilot Plant,
Section 9.3) that has attempted to close material balance with respect to amine
consumption and included formation of degradation products and their atmospheric
emissions in the material balance. Unfortunately, the material balance does not close
with acceptable accuracy and in addition, the liquid phase emissions have not been
quantified.
107
It must be pointed out that the above study has been performed at a lignite-fired power
plant in Europe where the coal based power plants are equipped with state of the art
de-SOX, de-NOX and particulate filtration systems which is not the case with Australian
coal-fired power plants. Thus, there is a strong parameter of flue gas quality that needs
to be considered before using a European solvent degradation study for anticipating
the solvent degradation and resulting atmospheric emissions from a post combustion
CO2 capture plant linked to an Australian coal-fired power plant. In addition, the
ambient conditions for both summer and winter seasons in Europe are markedly
different from Australian ambient conditions for both inland and coastal locations. Thus,
impact of ambient conditions on performance of the water wash tower must also be
accounted for when using a European study in Australian context.
Since, the Australian flue gas quality and ambient conditions are considerably different
from equivalent European or US/Canadian situation, both direct contact cooler (DCC)
and the water wash towers downstream of absorber and stripper will need to be
designed accordingly. It will be preferable, if the flue gas desulphurisation as well as
the removal of oxides of nitrogen can be carried out in the DCC tower itself for an
Australian post combustion capture plant. This has potential to reduce the cost of CO2
capture from an Australian coal-fired power plant and may even have a positive impact
on the net power plant efficiency.
Since post combustion capture of CO2 needs to meet not only the criteria of minimal
impact on the power plant efficiency and cost of electricity generation but also minimal
adverse environmental impact, the solvent development program currently ongoing in
Australia should follow the guidelines for solvent ranking given at Section 12.
In light of the above recommendations, the following work program is proposed:
I. Full characterisation and quantification of degradation products of amine
solvents both in the gas and liquid streams in an Australian pilot plant scale
CO2 capture plant accompanied by a closed material balance around the input
and output streams of the plant. It is expected that this pilot plant will be
operating at steady state in the gas-liquid flow regimes representative of
currently operating industrial scale post combustion CO2 capture plants.
II. Process improvements around the DCC tower and the water wash towers
downstream of both absorber and stripper in an Australian pilot plant scale to
minimise first the adverse impact of flue gas impurities on amine solvents and
then atmospheric emissions of amine solvents and their degradation products.
III. Development of mathematical models from first principle (for example, using
the molecular modelling principles) – bottom up approach – which account for
the solvent degradation kinetics and yields of the degradation products
observed through steps I and II above to improve predictive capability of
anticipated atmospheric emissions from a full scale amine solvent based post
combustion capture plant linked to an existing black coal-fired power plant in
Australia.
108
15. REFERENCES
1. Newman, S. A., Edit., “ Acid and Sour Gas Treating Processes”, Gulf Publishing
Co., ISBN – 10: 0872018393
2. Mimura, T., Suda, T., Iwaki, I., Honda, A. and Kumazawa, H., “Kinetics of reaction
between CO2 and sterically hindered amines for CO2 recovery from power plant flue
gases”, Chem. Enng. Comm., (1998), Vol. 170, pp. 245-260
3. Kohl, A. And Neilsen, R. B., Edit., “Gas Purification”, 5th Edition, Gulf Professional
Publishing Co., ISBN – 10: 0884152200
4. Freeman, s. A., “Thermal Degradation and Oxidation of Aqueous Piperazine for
Carbon Dioxide Capture”, Ph. D. Dissertation, The University of Texas at Austin,
May 2011
5. Dave, N., Duffy, G. J., Edwards, J. H. and Lowe, A., “Evaluation of the Options for
Recovery and Disposal/Utilisation of CO2 from Australian Black Coal-fired Power
Stations”, ACARP Project Report No: C7051, August 1999.
6. Goff, G. S., “Oxidative Degradation of Aqueous MEA in CO2 Capture Processes:
Iron and Copper Catalysis, Inhibition and O2 Mass Transfer”, Ph. D. Dissertation,
The University of Texas at Austin, May 2005
7. Sexton, A. J., “Amine Oxidation in CO2 Capture Processes”, Ph. D. Dissertation,
The University of Texas at Austin, December 2008
8. Chanchey, A., Saiwan, C, Supap, T., Idem, R. and Tontowachwuthikil, P., “Off-Gas
Emission in CO2 Capture Process Using Aqueous MEA Solution”, Energy
Procedia, Proceedings of GHGT-10 Conference, 19-23 September, 2010,
Amsterdam, The Netherlands.
9. Veltman, K., Singh, B. and Hertwich, E. G., “Human and environmental Impact of
Post Combustion CO2 Capture Focussing on Emissions from Amine-based
Scrubbing Solvents to Air”, Environ. Sci. Technol., (2010), Vol, 44, No. 4, pp. 14961502
10. Thitakamol, B., Veawab, A. and Aroonwilas, “Foaming in Amine-based CO2
Capture Process: Experiment, Modelling and Simulation”, Energy Procedia,
Proceedings of GHGT- 9 Conference, 16-20 November, 2008, Washington DC,
USA
11. Davis, J. D., “Thermal Degradation of Aqueous Amines Used for Carbon Dioxide
Capture”, Ph. D. Dissertation, The University of Texas at Austin, August 2009
12. Arnold, D. S., Barrett, D. A. and Isom, R. H., “CO2 Production from Coal-fired Boiler
Flue Gas by MEA Process”, Proceedings of Gas Conditioning Conference, 1982
13. Reddy, S., Johnson, D. and Gilmartin, J., “Fluor’s Econamine FG PlusSM
Technology for CO2 Capture at Coal-fired Power Plants”, Proceedings of “Mega”
Symposium on Power Plant Air Pollution Control, 25-28 August, 2008, Baltimore,
USA.
14. Fredriksen, S. B., Jens, Claus and Eimer, D., “Theoretical Evaluation of the
Probability to Form and Emit Harmful Components from the Aqueous Amine
Solutions of MEA, MDEa, AMP, PZ and Some of Their Mixtures”, A Literature
Report Submitted to Gassnova Inc, Report No: 22100-2, October 2010
109
15. Mimura, T., Nojo, T., Ijima, M. and Mitsuoka, S., “Development and Application of
Flue Gas CO2 Recovery Technology”, Proceedings of GHGT-5 Conference, 13-16
august, 2000, Cairns, Australia.
16. Yoshida, K., Mimura, T., Shimojo, S. Karasaki, M., Ijima, M., Touru, S. and
Mitsuoka, S., “Method for removing CO2 from Combustion Exhaust Gas”, US Patent
No: 6500397, Year 2002
17. Rooney, P. C., DuPart, M. S. and Bacon, T. R., “Oxygen’s Role in Alkanolamine
Degradation”, Hydrocarbon Processing, Int. Ed. (1998), Vol 77, No. 7, pp.109-133
18. Vevelstad, S. J., Eide-Haugmo, I., da Dilva, E. F. and Svendsen, H., “Degradation
of MEA; A Theoretical Study”, Energy Procedia, Proceedings of GHGT-10
Conference, 19-23 September, 2010, Amsterdam, The Netherlands.
19. Goff, G. S. and Rochelle, G. T., “Monoethanolamine Degradation: O2 Mass Transfer
Effects under CO2 Capture Conditions”, Ind. Eng. Chem. Res. (2004), Vol. 43, No.
20, pp. 6400-6408
20. Blachy, C. H. and Ravner, H., “The Stabilisation of Monoethanolamine Solutions for
submarine Carbon Dioxide Scrubbers”, AD609888, NRL-6189, NRL-FR-6189,
Naval Research Laboratory, Washington DC, (1964)
21. Girdler Corporation, Contract No: 50023, “Carbon Dioxide Absorbants”, (1950)
22. Hofmeyer, B. G., Scholten, H. G. and Loyd, W.G., “Contamination and Corrosion in
MEA Gas Treating Solutions”, Am. Chem. Soc., Div. Pet. Chem., Prepr. - Symp.,
(1956), Vol. 1, No. 2, pp. 91-99
23. Chi, S. and Rochelle, G. T., “Oxidative Degradation of MEA”, Ind. Eng. Chem. Res.,
(2002), Vol. 41, No. 17, pp. 4178-4186
24. Goff, G. S. and Rochelle, G. T., “Oxidative Degradation of Aqueous
Monoethanolamine in CO2 capture Systems under Absorber conditions”, Energy
Procedia, Proceedings of GHGT-6 Conference, Kyoto, Japan, 2003
25. Wen, H. and Narula, R, “Impacts of Carbon Capture on Power Plant Emissions”,
International post combustion capture network, 28th September – 1st October, 2009,
Regina, Canada
26. Pedersen, S., SjØvoll, M. and Fostås, B. F., “Flue Gas Degradation of Amines”, IEA
GHG Workshop, 16th February, 2010, Oslo, Norway
27. Fostås, B. F. and SjØvoll, M., “Flue Gas Degradation of Amines”, Climit Project
196051, 6th October, 2010 presentation
28. Fostås, B. F., Gangstad, A. and Nenseter, B., “StatOil’s Approach to Nitrosamines”,
IEA GHG Seminar, 1-2 February, 2011, Essen, Germany
29. Fostås, B. F., Gangstad, A., Nenseter, B., Pedersen, S., SjØvoll, M. and SØrensen,
A. L., “Effects of NOX in the Flue Gas Degradation of MEA”, Energy Procedia,
Proceedings of GHGT-10 Conference, 19-23 September, 2010, Amsterdam, The
Netherlands
30. Jackson, P. And Attala, M., “Environmental Impacts of Post Combustion Capture –
New Insights”, IEA GHG Seminar, 1-2 February, 2011, Essen, Germany
31. Schallert, B., “Expected NO2 Levels at the Inlet of CO2 Capture Plants”, IEA GHG
Seminar, 1-2 February, 2011, Essen, Germany
32. Frischmann, M., “Plausibilityof Total and Individual Nitrosamine Measurements and
Improvements”, IEA GHG Seminar, 1-2 February, 2011, Essen, Germany
110
33. Schallert, B., “Measures to Control Nitrosamine Concentrations”, IEA GHG
Seminar, 1-2 February, 2011, Essen, Germany
34. Strazisar, B. R., Anderson, R. R. and White, C. M., “Degradation Pathways for MEA
in a CO2 Capture Facility”, Energy & Fuels, (2003), Vol. 17, No. 4, pp.1034-1039
35. Idem, R.O., Uyanga, I. J, Aboudheir, A. and Tontiwachwuthikul, P., “Kinetic Study of
SO2 and O2 Induced Degradation of Aqueous MEA”, Carbon Capture and Utilisation
Workshop, HTC Energy, 2007
36. Bello, A. and Idem R. O., “Comprehensive Study of the Kinetics of the Oxidative
Degradation of CO2 Loaded and Concentrated Aqueous MEA With and Without
Sodium Metavanadate During CO2 Absorption from Flue Gases”, Ind. Eng. Chem.
Res. (2006), Vol. 45, No. 8, pp.2569-2579
37. Uyanga, I. J. and Idem, R. O., “Studies of SO2 and O2 Induced Degradation of
Aqueous MEA During CO2 Capture from Power Plant Flue Gas Streams”, Ind. Eng.
Chem. Res., (2007), Vol. 46, No. 8, pp.2558-2566
38. Supap, T., Idem, R., Tontiwachwuthikul, P. and Saiwan, C., “Kinetics of SO2 and O2
Induced Degradation of Aqueous MEA Solution During CO2 Absorption from Power
Plant Flue Gas Streams”, Int. J. of Greenhouse Gas Control, (2009), Vol, 3, pp.133142
39. Sexton, A. J. and Rochelle, G. T., “Catalysts and Inhibitors for MEA Oxidation”,
Energy Procedia, Proceedings of GHGT- 9 Conference, 16-20 November, 2008,
Washington DC, USA
40. Voice, A. and Rochelle, G. T., “MEA Oxidation in CO2 Capture – Inhibitor Screening
with Hot-gas FTIR”, Presentation at Luminant Carbon Management Program’s
Research Review Meeting, June 14, 2011, Trondheim, Norway
41. Davis, J. and Rochelle, G., “Thermal Degradation of MEA at Stripper Conditions”,
Energy Procedia, Proceedings of GHGT- 9 Conference, 16-20 November, 2008,
Washington DC, USA
42. Lepaumier, H., da Silva, E., Einbu, A., Grimstvedt, A., Knudsen, J., Zahlsen, K. And
Svendsen, H. F., “Comparison of MEA Degradation in Pilot-scale with Lab-scale
Experiments”, Energy Procedia, Proceedings of GHGT- 10 Conference, 19-23
September, 2010, Amsterdam, The Netherlands
43. Thomas, D. and Vanderschuren, J., “Modelling of NOX Absorption into Nitric Acid
Solutions Containing Hydrogen Peroxide”, Ind. Eng. Chem. Res., (1997), Vol. 36,
No. 8, pp.3315-3322
44. Sun, W., Yong, C., Li, M., “Kinetics of the Absorption of CO2 into Mixed Aqueous
Solutions of AMP and Piperazine”, Chem. Eng. Sci., (2005), Vol.60, pp. 503-516
45. Choi, W., Cho, K., Lee, S., Shim, J., Hwang, H. Park, S. and Oh, K., “Removal of
CO2 by Absorption into Blended amines: Kinetics of Absorption into Aqueous
AMP/HMDA, AMP/MDEA and AMP/PZ Solutions”, Green Chemistry, (2007), Vol. 9,
pp.594-598
46. Lensen, R., “The Promoter Effect of Piperazine on the Removal of CO2”,
http://www.bsdfreaks.nl/files/Literatuurscriptie_def.pdf
47. Yeh, J. T., Pennline, H. W. and Resnik, K. P. “Study of CO2 Absorption and
Desorption in a Packed column”, Energy & Fuels, (2001), Vol. 15, No. 2, pp.274278
111
48. Vaidya, P. D. and Kenig, E. Y., “CO2-Alkanolamine Reaction Kinetics: A Review of
Recent Studies”, Chem. Eng. Technol., (2007), Vol. 30, No. 11, pp.1467-1474
49. Veawab, A., “Corrosion and Corrosion Control in CO2 Absorption Process Using
Aqueous Amine Solutions”, Ph. D. Thesis, University of Regina, (2000)
50. Knudsen, J., “Results from Test Campaigns at the 1 t/h CO2 Post-Combustion
Capture Pilot-Plant in Esbjerg Under the EU FP7 CESAR Project”, 1st Post
Combustion Capture Conference, 17-19 May, 2011, Abu-Dhabi, UAE.
51. Wang, T. and Jens K., “Oxidative Degradation of AMP”, 1st Post Combustion
Capture Conference, 17-19 May, 2011, Abu-Dhabi, UAE.
52. Veawab, A. and Aroonwilas, A., “Identification of Oxidising Agents in Aqueous
Amine-CO2 Systems”, Corrosion Science, (2002), Vol. 44, pp.967-987
53. Wang, T., Chen, C. And Jens, K., “A Study of Oxidative Degradation of AMP for
Post-combustion CO2 Capture”, Presentation at Luminant Carbon Management
Program’s Research Review Meeting, June 15, 2011, Trondheim, Norway
54. Voice, A. and Rochelle, G., “Oxidation of Amines at Absorber Conditions for CO2
Capture from Flue Gas”, Energy Procedia, Proceedings of GHGT- 10 Conference,
19-23 September, 2010, Amsterdam, The Netherlands
55. Lepaumier, H., Picq, D. and Carrette, P., “New Amines for CO2 Capture :
Mechanisms of Amine Degradation in the Presence of CO2”, Ind. Eng. Chem. Res.,
(2009), Vol. 48, No. 20, pp.9061-9067
56. Lepaumier, H., Picq, D. and Carrette, P., “Degradation Study of New Solvents for
CO2 Capture in Post-combustion”, Energy Procedia, Proceedings of GHGT- 9
Conference, 16-20 November, 2008, Washington DC, USA
57. Oh, K., Kim, S. and Park, S., “Simultaneous Absorption of CO2, SO2 and NO2 into
Aqueous AMP”, Korean J. Chem. Eng., (2011), Vol. 28, No. 6, pp.1444-1450
58. Choi, W., Seo, J., Cho, S., Park, S. and Oh, K., “Simultaneous Absorption of CO2
and SO2 into Aqueous AMP/NH3 Solutions in Binary Composite Absorption
System”, Korean J. Chem. Eng., (2009), Vol. 26, No. 3, pp.705-710
59. Seo, J., Jeon, S., Choi, W., Kim, J. Lee, G. And Oh, K., “The Absorption Rate of
CO2/SO2/NO2 into a blended AMP/NH3 Solution”, Korean J. Chem. Eng., (2011),
Vol. 28, No. 1, pp.170-177
60. Mimura, T., Shimojo, S., Ijima, M. and Mitsuoka, S., “Process for RemovingCO2 and
NOX from Combustion Gases”, U.S. Patent 5,648,053 (July 15, 1997)
61. Kishimoto, S., Hirata, T., Iijima, M., Ohishi, T., Higaki, K. and Mitchell, R., “Current
Status of MHI’s CO2 Recovery Technology and Optimisation of CO2 Recovery Plant
with a PC Fired Power Plant”, Energy Procedia, Proceedings of GHGT- 9
Conference, 16-20 November, 2008, Washington DC, USA
62. Endo, T., Kajiya, Y., Nagayasu, H, Iijima, M., Ohishi, T., Tanaka, H. and Mitchell,
R., “Current Status of MHI CO2 Capture Plant Technology, Large Scale
Demonstration Project and Road Map to Commercialization for Coal-fired Flue Gas
Application”, Energy Procedia, Proceedings of GHGT- 10 Conference, 19-23
September, 2010, Amsterdam, The Netherlands
63. Kamijo, T., “Amine Emission Control Technology of KM CDR Process”, Amine
Workshop: EPRI Meeting, 16th August 2011, Palo Alto, USA.
112
64. MHI Ltd: Technology Publication, “Flue Gas CO2 Capture Technology of KS-1”,
Year 2005.
65. Freeman, S., Davis, J. and Rochelle, G., “Degradation of Aqueous Piperazine in
CO2 capture”, Int. J. of Greenhouse Gas Cont., (2010),
doi:10.1016/j.ijggc.2010.03.09
66. Freeman, S., Dugas, R., Wagener, D. V., Nguyen, T. and Rochelle, G., “CO2
Capture with Concentrated Aqueous Piperazine”, Energy Procedia, Proceedings of
GHGT- 9 Conference, 16-20 November, 2008, Washington DC, USA
67. Bishnoi, S., “CO2 Absorption and Solution Equilibrium in Piperazine Activated
Methyldiethanolamine”, Ph. D. Dissertation, The University of Texas at Austin, 2000
68. The Dow Chemical Company, “Ethyleneamines”, Technical Publication, Year 2001
69. Hillard, M. D., “A Predictive Thermodynamic Model for an Aqueous blend of
Potassium Carbonate, Piperazine and MEA for CO2 Capture from Flue Gas”, Ph. D.
Dissertation, The University of Texas at Austin, 2008
70. Merk Corp., “Material Safety Data Sheet (MSDS) for Anhydrous Piperazine”, Year
2009
71. Muhammad, A., Mutalib, M, Murugesan, T. and Shafeeq, A., “Thermophysical
Properties of Aqueous Piperazine and Aqueous MDEA/PZ Solutions at
Temperatures 298.15 to 338.15 K”, J. of Chem. & Eng. Data, (2009), Vol. 54, No. 8,
pp.2317-2321
72. Closmann, F., “Oxidation and Thermal degradation of MDEA/PZ in CO2 Capture”,
Ph. D. Dissertation, The University of Texas at Austin, December 2011
73. Ashouripashaki, M. and Rochelle, G., “Nitrosamine Formation in CO2 Capture”,
Presentation at Luminant Carbon Management Program’s Research Review
Meeting, June 14, 2011, Trondheim, Norway
74. Jackson, P. and Attala, M., “Environmental Impact of Post-combustion Capture –
New Insights”, Energy Procedia, Proceedings of GHGT- 10 Conference, 19-23
September, 2010, Amsterdam, The Netherlands
75. Freeman, S., Closmann, F. B. and Rochelle, G., “Oxidation of Aqueous Piperazine:
Oxidation Rates, Products and High-temperature Oxidation”, CO2 Capture by
Aqueous Absorption – Summary of 3rd Quarterly Progress Report, 2011
76. Freeman, S. and Rochelle, G., “Thermal Degradation of Aqueous Piperazine for
CO2 Capture – Part 1, Effects of Process Conditions and Comparison of Thermal
Stability of CO2 Capture Amines”, CO2 Capture by Aqueous Absorption – Summary
of 3rd Quarterly Progress Report, 2011
77. Freeman, S. and Rochelle, G., “Thermal Degradation of Aqueous Piperazine for
CO2 Capture – Part 2, Product Types and Generation Rates”, CO2 Capture by
Aqueous Absorption – Summary of 3rd Quarterly Progress Report, 2011
78. Chen, X. and Rochelle, G., “
79. Chakma, A. and Meisen, A., “Methyl-Diethanolamine Degradation – Mechanism
and Kinetics”, The Canadian J. of Chem. Eng., (1997), Vol.75, pp.861-871
80. Closmann, F., Nguyen, T. and Rochelle, G., “MDEA/Piperazine as a Solvent for
CO2 Capture”, Energy Procedia, Proceedings of GHGT- 9 Conference, 16-20
November, 2008, Washington DC, USA
113
81. Closmann, F. and Rochelle, G., “Degradation of Aqueous MDEA by Temperature
and Oxygen Cycling”, Energy Procedia, Proceedings of GHGT- 10 Conference, 1923 September, 2010, Amsterdam, The Netherlands
82. Lawal, A. and Idem, R., “Effects of Operating Variables on the Product Distribution
and Reaction Pathways in the Oxidative Degradation of CO2-Loaded Aqueous
MEA-MDEA Blends during CO2 Absorption from Flue Gas Streams”, Ind. Eng.
Chem. Res., (2005), Vol. 44, No. 4, pp. 986-1003
83. Lawal, A., Bello, A., and Idem, R., “The Role of MDEA in Preventing the Oxidative
Degradation of CO2 Loaded and Concentrated Aqueous MEA/MDEA Blends During
CO2 Absorption from Flue Gas”, Ind. Eng. Chem. Res., (2005), Vol. 44, No. 6,
pp.1874-1896
84. Bello, A. and Idem, R., “Pathways for the Formation of Products of the Oxidative
Degradation of CO2-Loaded Concentrated Aqueous MEA Solutions during CO2
Absorption from Flue Gas”, Ind. Eng. Chem. Res., (2005), Vol. 44, No. 4, pp. 945969
85. Lawal, A., and Idem, R., “Kinetics of the Oxidative Degradation of CO2-Loaded and
Concentrated Aqueous MEA/MDEA Blends during CO2 Absorption from Flue Gas”,
Ind. Eng. Chem. Res., (2006), Vol. 45, No. 8, pp.2601-2607
86. Ramachandran, N., Aboudheir, A., Idem, R. and Tontiwachwuthikul, P., “Kinetics of
the Absorption of CO2 into Mixed Aqueous Loaded Solutions of MEA and MDEA”,
Ind. Eng. Chem. Res., (2006), Vol. 45, No. 8, pp.2608-2616
87. CANSOLV Integrated SO2/CO2 Removal Process http://www.cansolv.com/en/multipollutant.ch2
88. Huttenhuis, P.J.G., van Elk, E. P. and Versteeg, G. F., “Performance of Aqueous
MDEA Blends for CO2 Removal from Flue Gas”, Presentation at 10th Meeting of the
International Post-combustion CO2 Capture Network, 24th May 2007, Lyon, France
89. Dubois, L. and Thomas, D., “CO2 Absorption into Aqueous Amine Based Solvents:
Modelling and Absorption Tests”, Energy Procedia, Proceedings of GHGT- 10
Conference, 19-23 September, 2010, Amsterdam, The Netherlands
90. Arcis, H., Rodier, L. and Coxam, J., “Enthalpy of Solution of CO2 in Aqueous
Solutions of AMP”, J. Chem. Thermodynamics, (2007), Vol. 39, pp.878-887
91. Dey, A. and Aroonwilas, A., “CO2 Absorption Characteristics of Blended MEA and
AMP”, EIC Climate Change Technology Conference, IEEE, Ottawa, Canada, 10-12
May, 2006, pp.1-5
92. Dubois, L. and Thomas, D., “CO2 absorption into Aqueous Solutions of MEA,
MDEA, PZ and Their Blends”, Chem. Eng. Technol., (2009), Vol. 32, No. 5, pp.710718
93. Choi, W., Seo, J., Jang, S., Jung, J. and Oh, K., “Removal Characteristics of CO2
Using Aqueous MEA/AMP Solutions in the Absorption and Regeneration Process”,
J. of Environmental Sciences, (2009), Vol. 21, pp.907-913
94. Aronu, U., Svendsen, H., Hoff, K. and Juliussen, O., “Solvent Selection for CO2
Absorption”, Energy Procedia, Proceedings of GHGT- 9 Conference, 16-20
November, 2008, Washington DC, USA
95. Derks, P. W. J., “CO2 Absorption in Piperazine Activated N-Methyldiethanolamine”,
Ph.D. Thesis, University of Twente, (2006)
114
96. Chakma, A., “Formulated Solvents: New Opportunities for Energy Efficient
Separation of Acid Gases”, Energy Sources, (1999), vol. 21, pp. 51-62
97. Chakma, A., “An Energy Efficient Mixed Solvent for the Separation of CO2”, Energy
Conv. Mgmt., (1995), Vol. 36, No. 6-9, pp.427-430
98. Chakma, A., and Tontiwachwuthikul, P., “Designer Solvents for Energy Efficient
CO2 Separation from Flue Gas”, Proceedings of Greenhouse Gas Control
Technologies, 30th Aug – 2nd Sep, 1998, Interlaken Switzerland, Elsevier
Publication, ISBN 008-043018-X, pp.35-42
99. Do, T., Dave, N., Azzi, M. and Feron, P., “Environmental Impact of Amine based
Post-combustion CO2 Capture Processes : Estimated emissions to the atmosphere
from amine based PCC processes for a black coal-fired power station based on
literature and modelling”, Task 3.1 Report, ANLEC R&D, January 2012.
100. Bruder, P., Grimstvedt, A., Mejdell, T. and Svendsen, “CO2 Capture into Aqueous
Solutions of Piperazine Activated AMP”, Chem. Eng. Sci., (2011), Vol. 66, pp.61936198
101. Yang, Z., Soriano, A., Caparanga, A. and Li, M., “Equilibrium Solubility of CO2 in
AMP + PZ + Water”, J. Chem. Thermodynamics, (2010), Vol. 42, pp.659-665
102. SINTEF Report, “Emission Reducing Technologies H&ETQP Amine 6”, SINTEF
A 18095, 14th Feb. 2011.
103. Nguyen, T., Hilliard, M. and Rochelle, G., “Volatility of Aqueous Amines in CO2
Capture”, Energy Procedia, Proceedings of GHGT- 10 Conference, 19-23
September, 2010, Amsterdam, The Netherlands
104. Nguyen, T., Hilliard, M. and Rochelle, G., “Amine Volatility in CO2 Capture”, Int. J.
of Greenhouse Gas Control, (2010), Vol. 4, pp.707-715
105. Nguyen, T., Frailie, P. and Rochelle, G., “Amine Volatility in CO2 Capture”, 1st
Post-combustion CO2 Capture Conference, 17-19 May, 2011, Abu Dhabi, UAE.
106. Arkema Inc., “Oxygen Scavengers”, Technical Brochure No: QC173 7/01
107. GE Water & Process Technology, “Max-Amine GT-741C”, Technical Brochure
No: HPG710EN 0604
108. Knudsen, J., Vilhelmsen, P., Jensen, J. and Biede, O., “CASTOR SP2:
Experiments on Pilot Plant”, Presentation at Common Technical Training
Workshop, 22-24 January, 2008, Lyon, France
109. Knudsen, J., Jensen, J., Vilhelmsen, P. and Biede, O., “Experience with CO2
Capture from Coal Flue Gas in Pilot –scale: Testing of Different Amine Solvents”,
Energy Procedia, Proceedings of GHGT- 9 Conference, 16-20 November, 2008,
Washington DC, USA
110. Knudsen, J., Andersen, J., Jensen, J. and Biede, O., “Evaluation of Process
Upgrades and Novel Solvents for the Post-combustion CO2 Capture in Pilot-scale”,
Energy Procedia, Proceedings of GHGT- 10 Conference, 19-23 September, 2010,
Amsterdam, The Netherlands
111. Aas, N. and da Silva, E., “Emission Measurements at Dong’s Pilot Plant for CO2
Capture in Esbjerg”, EU Project CESAR Presentation
112. Mertens, J., Knudsen, J., Thielens, M. and Andersen, J., “On-line Monitoring and
Controlling Emissions in Amine Post-combustion Carbon Capture: A Field Test”, Int.
J. of Greenhouse Gas Control, (2012), Vol. 6, pp.2-11
115
113. Chapel, D., Ernest, J. and Mariz, C., “Recovery of CO2 from Flue Gases:
Commercial Trends”, Canadian Soc. Chem. Engrs. Annual Meeting”, 4-6 Oct.,
1999, Saskatoon, Saskatchewan, Canada
114. Wang, M., Lawal, A., Stephenson, P., Sidders, J. and Ramshaw, C., “Postcombustion CO2 Capture with Chemical Absorption: A State-of-the-art Review”,
Chem. Eng. Res. & Des., (2011), Vol. 89, pp.1609-1624
115. http://www.cslforum.org/publications/documents/ITC_CO2_Capture_with_Chemic
al_Solvents.pdf
116. Amoco Canada Petroleum Co Ltd, “Boundary Dam CO2 Extraction Pilot Plant”,
December, 1989
117. Idem, R., Wilson, M., Tontiwachwuthikul, P., Chakma, A., Veawab, A,
Aroonwilas, A and Gelowitz, D., “Pilot Plant Studies of the CO2 Capture
Performance of Aqueous MEA and Mixed MEA/MDEA Solvents at the University of
Regina CO2 Capture Technology development Planand the Boundary Dam CO2
Capture Demonstration Plant”, Ind. Eng. Chem. Res., (2006) Vol. 45, No.8,
pp.2414-2420
118. Moser, P., Schmidt, S., Sieder, G., Garcia, H., Ciattaglia, I. And Klein, H.,
“Enabling Post-combustion Capture Optimization – The Pilot Plant Project at
Niederaussem”, Energy Procedia, Proceedings of GHGT- 9 Conference, 16-20
November, 2008, Washington DC, USA
119. Moser, P., Schmidt, S., Sieder, G., Garcia, H., Stoffregen, T. and Stamatov, V.,
“The Post-combustion Capture Pilot Plant Niederaussem – Results of the First Half
of the Testing Programme”, Energy Procedia, Proceedings of GHGT- 10
Conference, 19-23 September, 2010, Amsterdam, The Netherlands
120. Moser, P., Schmidt, S., Uerlings, R., Sieder, G., Titz, J., Hahn, A. and Stoffregen,
T., “Materials Testing for future commercial Post-combustion Capture Plants –
Results of the Testing Programme Conducted at the Niederauseem Pilot Plant”,
Energy Procedia, Proceedings of GHGT- 10 Conference, 19-23 September, 2010,
Amsterdam, The Netherlands
121. Moser, P., Schmidt, S. and Stahl, K., “Investigation of Trace Elements in the Inlet
and Outlet Streams of a MEA-based Post-combustion Process – Results from the
Test Programme at the Niederaussen pilot Plant”, Energy Procedia, Proceedings of
GHGT- 10 Conference, 19-23 September, 2010, Amsterdam, The Netherlands
122. Cousins, A., Cottrell, A., Huang, S., Beavis, P. Chow, T, Lawson, A. and Feron,
P., “Tarong Post Combustion Capture Pilot Plant Project – Final Report”, Submitted
to Department of Resources, Energy and Tourism, Australian Federal Government,
2011.
123. Leusden, I, Artanto, Y. And Meuleman, E., “Emissions from CSIRO’s PCC Pilot
Plant at Loy Yang – MEA Base Case”, CSIRO Report: ET/IT XXXR, January 2010
124. da Silva, E., Booth, A., Loeppky, R., Ma’Mun, S., Karl, M., Heimstad, E.,
Johnson, M. and Feilberg, K., “Protocol For Evaluation of Solvents – Process &
Atmospheric Chemistry”, SINTEF Report, 19th June 2010
125. Closmann, F., “Solvent Degradation – MEA and MDEA/PZ Blended Systems”,
Luminant Carbon Management Program presentation, June 15. 2009
126.
116

PDF Document reader online

This website is focused on providing document in readable format, online without need to install any type of software on your computer. If you are using thin client, or are not allowed to install document reader of particular type, this application may come in hand for you. Simply upload your document, and Docureader.top will transform it into readable format in a few seconds. Why choose Docureader.top?

  1. Unlimited sharing - you can upload document of any size. If we are able to convert it into readable format, you have it here - saved for later or immediate reading
  2. Cross-platform - no compromised when reading your document. We support most of modern browers without the need of installing any of external plugins. If your device can oper a browser - then you can read any document on it
  3. Simple uploading - no need to register. Just enter your email, title of document and select the file, we do the rest. Once the document is ready for you, you will receive automatic email from us.

Previous 10

Next 10