2414 Ind. Eng. Chem. Res.

2006, 45, 2414-2420

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 Plant and the Boundary Dam CO2 Capture Demonstration Plant
Raphael Idem,* Malcolm Wilson, Paitoon Tontiwachwuthikul, Amit Chakma,
Amornvadee Veawab, Adisorn Aroonwilas, and Don Gelowitz
International Test Centre for CO2 Capture (ITC), Faculty of Engineering, UniVersity of Regina,
Regina, Saskatchewan, Canada S4S 0A2, and Department of Chemical Engineering, UniVersity of Waterloo,
Waterloo, Ontario, Canada N2L 3G1

Evaluations of the benefits of using a mixed MEA/MDEA solvent for CO2 capture in terms of the heat
requirement for solvent regeneration, lean and rich loadings, CO2 production, and solvent stability were
performed by comparing the performance of aqueous 5 kmol/m3 MEA with that of an aqueous 4:1 molar
ratio MEA/MDEA blend of 5 kmol/m3 total amine concentration as a function of the operating time. The
tests were performed using two pilot CO2 capture plants of the International Test Centre for CO2 Capture
(ITC), which provided two different sources and compositions of flue gas. The University of Regina CO2
plant (UR unit) processes flue gas from the combustion of natural gas while the Boundary Dam CO2 plant
(BD unit) processes flue gas from a coal-fired electric power station. The results show that a huge heat-duty
reduction can be achieved by using a mixed MEA/MDEA solution instead of a single MEA solution in an
industrial environment of a CO2 capture plant. However, this benefit is dependent on whether the chemical
stability of the solvent can be maintained.

1. Introduction Mixed amines have been reported to maximize the desirable

qualities of the individual amines.3 Thus, the specific goal with
Postcombustion capture of carbon dioxide (CO2) is undoubt- respect to the use of mixed amines is to have a solution
edly the technology for mitigating greenhouse gas (GHG) consisting of tertiary and primary amines or tertiary plus
emissions from existing fossil fuel-fired electric power plants. secondary amines that, in comparison with single amine systems,
It is also one of the technologies that can supply huge amounts retains much of the reactivity of primary or secondary amines
of CO2, one of the flooding agents used for enhanced oil at similar or reduced circulation rates but offers low regeneration
recovery. One of the most attractive methods to achieve this costs similar to those of tertiary amines.2 Consequently, by
goal for such dilute and low-pressure CO2 sources is absorption blending a primary or secondary alkanolamine with a tertiary
with chemical reaction using aqueous alkanolamine solutions. alkanolamine, bulk CO2 removal is easily accomplished while
A wide variety of alkanolamines such as monoethanolamine regeneration energy costs are minimized. In addition, another
(MEA), diethanolamine (DEA), di-2-propanolamine (DIPA), degree of freedom (the amine concentration) is gained. The
and methyldiethanolamine (MDEA) exists, and some have been amine concentration can be altered to achieve precisely the
used industrially for a number of years.1 There are differences desired separation for a given process.
in their performance in CO2 absorption using packed columns. Substantial reductions in energy requirements and modest
The first difference pertains to their reactivities or rates of CO2 reduction in circulation rates have been reported for amine
absorption. Primary and secondary amines such as MEA and blends relative to the corresponding single amine system of
DEA are very reactive and thus are able to effect a high volume similar total amine concentration.4 Also, simulation studies have
of acid gas removal at a fast rate.2 The second is that primary shown that, for CO2 loadings below 0.5 mol/mol amine, MDEA
and secondary amines have the limitation that their maximum + MEA and MDEA + DEA blends containing 2 kmol/m3 of
CO2 loading capacity based on stoichiometry is at best 0.5 mol each amine produced an equilibrium partial pressure of the
CO2/mol amine, unlike tertiary amines such as MDEA, which amine blend that is intermediate between those of the corre-
have an equilibrium CO2 loading capacity that approaches 1.0 sponding single amine systems of equivalent total amine
mol CO2/mol amine. The third is that stripping of CO2 from concentration.5 For higher CO2 loadings, the equilibrium partial
MEA or DEA during solvent regeneration requires a large pressures in blended amine systems were less or comparable
amount of energy input as compared to MDEA. It is widely with those of single amine systems. Furthermore, experiments
known that the heat duty for solvent regeneration can constitute on CO2 solubility in aqueous blends of MEA + MDEA and
up to 70% of the total operating costs in a CO2 capture plant. DEA + MDEA have confirmed that equilibrium solution
Other operating concerns involve solvent corrosiveness and loading is influenced most by the blended compositions under
solvent chemical instability to which, studies have suggested, conditions that are typical of industrial regenerators.6,7
primary and secondary amines are more prone than tertiary Most of these studies have been performed using laboratory
amines. scale experimentation or software simulation with simulated flue
gas where the effects of other components of real flue gas
* To whom correspondence should be addressed. E-mail: including fly ash (SiO2, Al2O3, Fe2O3, CaO, MgO, Na2O, K2O,
raphael.idem@uregina.ca. Fax: (306) 585-4855. and P2O5), sulfur dioxide (SO2), and oxides of nitrogen (NOx),
such as from a coal-fired power plant, are not accounted for.
University of Waterloo.

10.1021/ie050569e CCC: $33.50 2006 American Chemical Society

Published on Web 08/06/2005
 Ind. Eng. Chem. Res., Vol. 45, No. 8, 2006 2415

Also, a corrosion inhibitor is usually added to the solvent in reclaimer. Also, the pilot plant is equipped with the following
industrial operations in order to minimize equipment corrosion analytical facilities: (i) a continuous on-line SO2 and O2 analyzer
caused by the operating environment. It is necessary to evaluate located downstream of the Anderson 2000 SO2 scrubber, (ii)
how this industrial operating environment will affect the an on-line analyzer for CO2 concentration along the side of the
performance of mixed amines (in terms of heat requirement for CO2 absorption column, (iii) a data-logging system installed to
solvent regeneration, lean and rich loading, CO2 production, retrieve and transmit the process temperature and pressure data
and solvent stability) if the advantages of mixed amines are to to the control room, (iv) local gauges and routine manual logs
be exploited for field applications. In the present study, these used for the daily operation of plant facilities, and (v) a
evaluations were performed by comparing the performance of corrosometer and electrochemical probes installed to continu-
aqueous 5 kmol/m3 MEA with that of an aqueous 4:1 molar ously monitor the short-term and long-term corrosion rates at
ratio MEA/MDEA blend of 5 kmol/m3 total amine concentration various points in the plant. These points included downstream
as a function of the operating time. The tests were performed of the Anderson 2000 SO2 scrubber, the amine absorber bottom,
using two pilot CO2 capture plants of the International Test downstream of the feed gas cooler, the hot rich amine solution
Centre for CO2 Capture (ITC), which provided two different to regenerator, the amine reboiler, the regenerator overhead, the
sources and compositions of flue gas as well as two different amine reclaimer, and the regenerator bottom. The corrosometer
modes of solvent regeneration. These results are presented in (model RCS-8) and electrochemical probe (model 2500) were
this paper. obtained from Rohrback Csasco Systems, Santa Fe, CA. A
process flow diagram for the BD unit is given in Figure 1.
2. Experimental Section
2.2.2. Technology Development Pilot Plant at UR. The
2.1. Chemicals. Concentrated MEA and MDEA (commercial University of Regina technology development pilot plant
grade, 99% purity) were obtained from Prairie Petro-Chem, processes up to 4.8 103 m3/day of flue gas and captures up
Estevan, Saskatchewan, Canada. These solvents were diluted to 1 ton of CO2 per day. Figure 2 shows a process flow diagram
with distilled water to the desired concentrations. of the UR pilot plant. As seen in the figure, the UR pilot plant
2.2. Testing Facilities. The International Test Centre for CO2 consists of three main units connected in series as follows: (i)
Capture (ITC) has two major testing facilities: (i) a semicom- a flue gas generation/pretreatment unit, (ii) an absorption-based
mercial CO2 capture demonstration plant adjacent to the unit for CO2 capture, and (iii) a postconditioning unit for product
SaskPowers Boundary Dam power station (BDPS) and (ii) a purification. A steam boiler (250 kW) is the heart of the flue
technology development pilot plant located at the University gas generation/pretreatment unit, producing both flue gas and
of Regina. high-quality steam for the CO2 capture unit. Also in the same
2.2.1. Boundary Dam Subcommercial Technology Dem- unit is a 30 kW micro-gas-turbine connected to the steam boiler.
onstration Plant. The subcommercial technology demonstration The configuration is designed such that the boiler can operate
unit at the Boundary Dam (BD unit), first built in the summer on its own or in conjunction with the micro-gas-turbine. When
of 1987 at the site of the Saskatchewan Power Corporations operating the microturbine, a low CO2 concentration feed is
(SaskPowers) Boundary Dam power station, the largest lignite produced; therefore, the boiler is operated as a duct burner to
coal-burning station in Canada, processes flue gas from a coal- enrich the CO2 concentration of the flue gas while consuming
fired electrical power plant. The BD unit was designed to the high excess O2 content from the turbines exhaust. The CO2
process 500 000 SCFD of flue gases from the Boundary Dam capture unit design is based on the gas absorption technology
power station and capture up to 4 tons of CO2 per day. The using amine solvent. Here, CO2 is removed from the precon-
CO2 demonstration pilot plant at BDPS is considered to be the ditioned flue gas in one of the three absorption columns and
only one of an appropriate size to develop the engineering data high quality CO2 stream is released from a solvent regenerator,
required for design of commercial facilities. operated at an elevated temperature. Each absorption column
The Boundary Dam pilot plant consists of three main is composed of three 0.3 m-diameter sections for a total height
components that are connected in series as follows: (i) a high- of 10 m and is also equipped with a series of temperature sensors
efficiency baghouse unit for fly ash removal, (ii) an Anderson and gas sampling points at a regular interval of 0.6 m to allow
2000 scrubbing unit for removal of SO2, and (iii) an amine- measurements of the temperature and gas-phase CO2 concentra-
based CO2 capture unit. The CO2 absorption is based on the tion during testing. In the postconditioning unit, the CO2 product
reaction of a weak base (alkanolamine) with a weak acid (CO2)
from the solvent regenerator is treated in a CO2-wash scrubber.
to produce a water soluble salt. The reaction is reversible and
An independent chiller is used to provide the cooling medium
temperature dependent. The removal of CO2 from flue gas is
in order to cool the CO2 gas to the desired temperature of 4 C.
achieved by contacting the feed flue gas in an amine absorber
From here, the CO2 product can either exit back to the
(18 in. in diameter) with an aqueous solution of alkanolamine
atmosphere or pass through a dryer and purification unit for
at a low temperature, whereby the CO2 chemically binds to the
further treatment to meet food-grade specifications.
amine and thus is removed from the flue gas stream. The CO2
is then liberated from the CO2-rich amine solution by elevated The UR technology development pilot plant is considered to
temperature to reverse the absorption reaction. The elevated be one of the best testing facilities, equipped with a state-of-
temperature is supplied by stripping steam in the amine the-art process control/instrumentation and data acquisition
regenerator (16 in. in diameter). The regenerated CO2-lean amine system called DeltaV. This system has allowed us to control,
solution is then cooled and recycled to the amine absorber for monitor, and record a complete spectrum of process operating
further CO2 removal. In contacting the flue gas, the amine conditions including temperature, flow rate, CO2 removal
solution may become contaminated by the formation of deg- efficiency, CO2 production rate, and, more importantly, energy
radation products generated from side reactions and the collec- consumption for CO2 capture. These data can be retrieved from
tion of insoluble material, such as pipe scale and fly ash. The the historical database and saved in a spreadsheet format for
conversion of degradation products back to amine and the detailed analysis. A control room was constructed in the building
removal of solid impurities are accomplished by the amine to house the control system computers.
2416 Ind. Eng. Chem. Res., Vol. 45, No. 8, 2006

Figure 1. Process flow diagram of the Boundary Dam CO2 capture demonstration plant.

2.3. Feed to the Absorption Plants. 2.3.1. Boundary Dam hydroxide (NaOH) is added to the system as a makeup reagent,
Subcommercial Technology Demonstration Plant. Flue gas it converts the NaHSO3 product back to Na2SO3 for further SO2
from the coal-fired power station provides a plentiful supply of removal. The amine extraction unit treats the preconditioned
CO2 but at the same time contains fly ash, oxygen (O2), and flue gas from the Anderson 2000 unit. The flue gas is virtually
trace contaminants such as sulfur dioxide (SO2) and nitrogen free of fly ash but may contain some (<10 ppm) SO2 and NO2.
dioxide (NO2) which are undesirable to the amine treating unit. 2.3.2. Technology Development Pilot Plant at UR. The flue
Flyash is the fine solid component in the flue gas typically gas to this capture unit is obtained by burning natural gas. There
composed of SiO2, Al2O3, Fe2O3, CaO, MgO, Na2O, K2O, and is no fly ash, SO2, or NOx in this flue gas. Thus, unlike in the
P2O5. The presence of fly ash in a gas stream results in a number case of the BD unit, the only conditioning performed on the
of operational difficulties. It can deposit or cake on the process raw flue gas before entering the CO2 capture unit is to clean
equipment, thereby blocking flow, wrenching pumps and pump and cool it to a specific temperature in an inlet gas scrubber.
seals, and clogging equipment and instrumentation. Most of the 2.4. Operating Conditions. The MEA concentration used
fly ash is usually removed from the flue gas before it is for both the BD and UR plants was 5 kmol/m3 for MEA. In the
discharged to the atmosphere through the flue stack. For the case of the mixed MEA/MDEA, the total amine concentration
contaminated gases, oxygen in the flue gas originates as excess was 5 kmol/m3 and the MEA/MDEA ratio was 4:1. A
combustion air in the boiler and SO2 and NO2 are the proprietary corrosion inhibitor was added to the solvents used
combustion products of sulfurous and nitrogen compounds in in the BD plant, while the solvents used in the UR plant had
the coal. no corrosion inhibitor. These solvents were used at a circulation
The residual fly ash is removed in a high efficiency baghouse, rate of 8 L/min. The CO2 concentrations used for the flue gas
which consists of 36 composite filter bags housed by a structure of the UR plant were 11 and 15 mol %, while that for the BD
complete with a hopper bottom for dust collection and removal. plant was only 15 mol %.
The structure is heat traced and insulated to withstand extreme 2.5. Analysis of Samples. Analysis of the samples (fresh and
climate variances and to maintain a temperature above the sulfur used) was carried out using a gas chromatograph/mass spec-
dew point on the baghouse walls. The bags are cleaned using trometer (GC/MS model HP 6890/5073 supplied by Hewlett-
blasts of air via solenoids controlled by an electronic control Packard Canada Ltd., Montreal Quebec, Canada). An HP
panel that measures pressure differential across the bags. SO2 Innowax column (length ) 30 m, internal diameter ) 250 m,
is removed by scrubbing in an Anderson 2000 unit, which is a thickness ) 0.25 m) packed with cross-linked polyethylene
two-stage wet-scrubbing system that uses sodium sulfite (Na2- glycol was used in the GC for the separation of components.
SO3) as an aqueous absorbent for SO2 removal. The primary These components were identified by their mass spectra. Prior
reaction product is sodium bisulfite (NaHSO3). When sodium to GC/MS analysis, each sample was diluted with deionized
 Ind. Eng. Chem. Res., Vol. 45, No. 8, 2006 2417

Figure 2. Process flow diagram of the technology development pilot plant at the University of Regina.

water to five times its original volume to avoid column overload CO2 loading in the lean and rich amine samples were determined
and to improve separation of the components. Sample injection using a Chittick CO2 analyzer.
into the GC column was done using an autoinjector (model 2.6. Measurements of Heat Duty for Regeneration and
7683) supplied by Hewlett-Packard Ltd. The exact concentration CO2 Production. The heat duty for regeneration was evaluated
of fresh aqueous MEA and MDEA was obtained by titrating by making an energy balance around the reboiler and/or the
known volumes of MEA/MDEA with 1 N HCl using 0.1 wt % stripping column. In the BD column, the heat used in the reboiler
methyl orange solution as indicator. Several trials were done for stripping was supplied by a glycol heater, whereas that for
to select the optimum operating conditions for the GC/MS. The the UR plant was supplied by a steam boiler.
conditions used are summarized as follows. An autoinjector
(model 7683, supplied by Hewlett-Packard) was used to 3. Results and Discussion
automatically introduce samples into the GC column to give
better reproducibility. A 10 L syringe with an injection volume The CO2 capture performance for the single amine (MEA)
of 0.2 L was used, and a split mode was selected for the inlet and the mixed amine (MEA/MDEA ) 4:1) solvents were
with a split ratio of 10:1, split flow of 10.3 mL/min, and total evaluated in terms of CO2 production, lean and rich amine
flow of 13.9 mL/min. The inlet temperature and pressure were loading, heat duty for regeneration, and solvent stability as a
70 C and 9.18 psi, respectively. The initial temperature of the function of time.
oven was 100 C with a hold time of 0 min, while the final 3.1. CO2 Production Rate. Figures 3, 4, and 5 show
temperature was 240 C with a hold time of 10 min with an comparisons between the single amine (MEA) and mixed amine
oven ramp of 10 C/min for a total run time of 27 min. The (MEA/MDEA) of the relationship of CO2 production rate with
column flow rate was 1 mL/min, while the pressure and average reboiler heat duty for the UR plant 11 mol % CO2 feed
velocity were 9.18 psi and 37 cm/sec. For the MS parameters, composition, the UR plant 15 mol % CO2 feed composition,
the interface, quadruple, and source temperatures were 250, 150, and the BD plant 15 mol % CO2 feed composition, respectively.
and 230 C, respectively, and the electron multiplier (EM) In the two UR cases (Figures 3 and 4), the CO2 production
voltage was 1200 V. The error of the GC/MS measurement was rates for the mixed MEA/MDEA (4:1) were higher than that
estimated to be less than (3%. The majority of the samples for the single MEA of the same total molar concentration and
were analyzed two or more times to check the reproducibility the same heat duty for regeneration. This implies that the heat
of analysis. The error was also less than (2%. MEA and MDEA penalty for regeneration can be reduced significantly by
concentrations were based on peak area % only, and thus, employing a mixed amine instead of a single amine. In the BD
represented relative concentrations. The absolute molar con- plant case (Figure 5), the opposite was the case. The CO2
centrations were obtained with HPLC with a refractive index production rate for the mixed MEA/MDEA (4:1) is lower than
detector (Agilent Technologies Canada, Mississauga, Ontario, that for the single MEA of the same total molar concentration
Canada) using a nucleosil column with a phosphate buffer. The and the same heat duty for regeneration, implying that we could
2418 Ind. Eng. Chem. Res., Vol. 45, No. 8, 2006

Figure 6. Lean-loading effect of 4:1 MEA/MDEA mixed amine at 11%

Figure 3. CO2 production at 4:1 MEA/MDEA mixed amine and 11% inlet inlet CO2 concentration compared to 5 M MEA and 11% inlet CO2
CO2 concentration compared to 5 M MEA and 11% inlet CO2 concentration concentration (UR unit).
(UR unit).

Figure 4. CO2 production at 4:1 MEA/MDEA mixed amine and 15% inlet Figure 7. Lean-loading effect of 4:1 MEA/MDEA mixed amine at 15%
CO2 concentration compared to 5 M MEA and 11% inlet CO2 (UR unit). inlet CO2 concentration compared to 5 M MEA and 11% inlet CO2
concentration (UR unit).

Figure 5. CO2 production at 4:1 MEA/MDEA mixed amine compared to

baseline conditions of 5 M MEA for 15% inlet CO2 concentration (BD
unit).
Figure 8. Lean-loading effect of 4:1 MEA/MDEA mixed amine compared
not exploit the benefits of the mixed amine formulation in the to baseline conditions of 5 M MEA (BD unit).
BD plant environment. To determine the circumstances that
resulted in this anomalous behavior, we evaluated parameters the relationship of lean amine CO2 loading for the UR plant
such as lean amine loading, rich amine loading, and the solvent are given in Figures 6 and 7 for CO2 compositions of flue gas
stability for possible effects on CO2 production rate in the two of 11 and 15 mol %, respectively, while that for 15 mol % CO2
plants. composition for the BD unit is given in Figure 8. These figures
3.2. Lean Amine CO2 Loading. The comparisons between show that the lean loadings of the mixed amine (MEA/MDEA
the single amine (MEA) and mixed amine (MEA/MDEA) of ) 4:1) were either the same or lower than those for the single
 Ind. Eng. Chem. Res., Vol. 45, No. 8, 2006 2419

Figure 11. Rich-loading effect of 4:1 MEA/MDEA mixed amine compared

to baseline conditions of 5 M MEA (BD unit).

Figure 9. Rich-loading effect of 4:1 MEA/MDEA mixed amine at 11% CO2 in the absorption column. This is usually expressed in terms
inlet CO2 concentration compared to 5 M MEA and 11% inlet CO2
concentration (UR unit). of mol CO2/mol of total amine. A lower rich amine loading for
the mixed amine in the BD case means that the capacity of this
solvent to absorb CO2 has been significantly reduced. When
this result is compared with those in the UR plant, where the
rich amine loadings for the mixed and single amines were almost
the same, it may mean that the loss in capacity could be
attributed to parameters other than the solvent itself. This
involves the chemical stability of the solvent.
3.4. Solvent Chemical Stability. Liquid samples of lean
MEA were collected and analyzed to verify the chemical
stability of aqueous MEA and mixed MEA/MDEA when
exposed to various operating conditions in a typical amine-based
CO2 capture environment. These tests were done for the
Boundary Dam demonstration plant as well as for the UR
technology development plant for 5 kmol/m3 aqueous MEA and
5 kmol/m3 total aqueous mixed MEA/MDEA solvent of MEA/
MDEA ratio of 4:1. Analysis was performed by gas chroma-
tography, and the components were identified by their mass
Figure 10. Rich-loading effect of 4:1 MEA/MDEA mixed amine at 15% spectra. The results showed that components other than MEA
inlet CO2 concentration compared to 5 M MEA and 11% inlet CO2 and MDEA were present in the lean MEA samples. The major
concentration (UR unit). compounds observed included straight chain amines such as
1-propanamine, cyclic compounds such as 1,2,3,6-tetrahydro-
amine of the same total amine concentration for the same heat 1-nitrosopyridine and 2-pyrrolidinone, dialcohols such as 1,2-
duty for regeneration. In fact, the lean loading of the mixed ethanediol, as well as sulfur compounds such as isothiocyna-
amine (MEA/MDEA ) 4:1) in the BD case was clearly lower toethane and 1,1-dioxid-tetrahydrothiophene. The sulfur com-
than that for the single amine. This shows that for the same pounds may have resulted from contact of aqueous MEA and
heat duty for stripping, more CO2 stripping is actually achieved MDEA with trace amounts of sulfur dioxide (SO2) that survived
from the mixed amine system than from the single amine. Again, the scrubbing process in the SO2 unit. The detection limit of
this confirms that the benefits of using mixed amines to achieve the SO2 analyzer is about 5 ppm. Thus, the trace amount of
a lower heat duty for regeneration can be achieved in an SO2 in the flue gas stream into the absorber is <5 ppm. Also,
industrial environment. It therefore means that the anomaly the presence of these sulfur compounds indicates that it was
observed in the BD plant of a lower CO2 production rate of the not possible to regenerate MEA and MDEA from these
mixed amine as compared to that of the single amine cannot be compounds in the regeneration unit.
attributed to the stripping performance. The wide variety of degradation products observed in the
3.3. Rich Amine CO2 Loading. Figures 9, 10, and 11 show Boundary Dam demonstration plant samples illustrates the effect
comparisons between the single amine (MEA) and mixed amine of a harsher environment brought about by a coal-fired power
(MEA/MDEA) of the relationship of rich amine CO2 loading plant flue gas. In the case of the UR lean amine, results indicate
with reboiler heat duty for the UR plant 11 mol % CO2 feed insignificant degradation and the presence of only trace amounts
composition, UR plant 15 mol % CO2 feed composition, and of degradation products, which included 1-propanamine and
the BD plant 15 mol % CO2 feed composition, respectively. In 2-pyrrolidinone, and just more than trace amounts of 1,2-
the two UR cases (Figures 9 and 10), the rich amine CO2 ethanediol and 1,2,3,6-tetrahydro-1-nitrosopyridine. On the other
loadings for the mixed MEA/MDEA (4:1) were about the same hand, our results show a more substantial degradation of the
or just slightly lower than those for the single MEA of the same MEA and mixed MEA/MDEA solvents from the Boundary Dam
total molar concentration and the same heat duty for regenera- demonstration plant as compared with those from the UR
tion. In the case of the BD plant, the rich amine loading for the technology development plant. For example, HPLC analysis
mixed amine was clearly below that for the single amine. The showed that the degradation rate for the aqueous MEA used in
CO2 loading is a test of the capacity of the solvent to absorb the BD unit was 0.5 mol %/day. In the mixed MEA/MDEA
2420 Ind. Eng. Chem. Res., Vol. 45, No. 8, 2006

Figure 12. Effect of degradation for 4:1 MEA/MDEA mixed amine (BD unit).

solution at the BD unit, the MEA degradation rate was 2.3 mol CO2 production rate. The declining CO2 capacity of the solvent
%/day while the rate for MDEA was 1.5 mol %/day. Since the with time can thus be attributed to the declining concentrations
reboiler temperatures were very similar in both the Boundary of both MEA and MDEA, especially MEA.
Dam demonstration plant and the UR technology plant (115-
120 C), it does not appear that temperature is a strong factor 4. Conclusions
that can account for the variability of the results from the two A huge heat-duty reduction can be achieved by using a mixed
plants. Also, the UR plants operated with aqueous MEA or MEA/MDEA solution instead of a single MEA solution in an
mixed MEA/MDEA solutions of the same concentrations (5 industrial environment of a CO2 capture plant. However, this
kmol/m3) as those used at the Boundary Dam plant (5 kmol/ benefit is dependent on whether the chemical stability of the
m3). This implies that concentration is also not an issue when solvent can be maintained.
used under the temperatures of 115-120 C and should not
contribute significantly to MEA degradation. As mentioned Acknowledgment
earlier, the major difference between the Boundary Dam
demonstration plant and the UR technology development plant Funding provided by the Consortium members of the
is primarily that the former is fed with flue gas from a coal- International Test Centre for CO2 Capture (ITC) and Saskatche-
wan Industry and Resources, Regina, Canada is gratefully
fired power plant (which is composed of CO2, O2, SO2, N2,
acknowledged.
particulates, Hg, etc.) whereas the latter is fed with flue gas
derived from the combustion of natural gas (composed mainly Literature Cited
of CO2, O2, and N2). This difference in flue gas feed components
is considered to be one of the factors responsible for the (1) Kohl, A. L.; Riesenfield, F. C. Gas Purification, 4th ed.; Gulf
Publishing: Houston, TX, 1985.
difference in the degradation product slate between the two (2) Dawodu, O. F.; Meisen, A. Degradation of alkanolamine blends by
plants. The other factor may involve the inhibitor. It is possible carbon dioxide. J. Chem. Eng. 1996, 74, 960-962.
that the inhibitor may also be acting as catalyst to facilitate (3) Charkravarty, T.; Phukan, U. K.; Weiland, R. H. Reaction of acid
degradation. gases with mixtures of amines. Chem. Eng. Prog. 1985, 81, 32-36.
(4) Campbell, S. W.; Weiland, R. H. Modeling of CO2 removal by amine
The degradation rate results indicate that the flue gas blends. Presented at the AICHE Spring National Meeting, Houston, TX,
composition and the presence of an inhibitor in the solvents in April 2-6, 1989.
the BD case affected the mixed amine chemical stability more (5) Austgen, D. M.; Rochelle, G. T.; Chen, C. C. A model of vapor
than that of the single amine. Consequently, this would explain liquid equilibria for aqueous acid gas alkanolamine systems II representative
of H2S and CO2 solubility in aqueous MDEA and CO2 solubility in aqueous
why the rich amine loading was higher for the single amine mixture of MDEA and MEA or DEA. Ind. Eng. Chem. Res. 1991, 30 (3)
than the mixed amine. Figure 12 shows a plot of the MEA and 543-555.
MDEA mole fractions in the mixed solvent (in terms of area (6) Li, M. H.; Shen, K. P. Densities of solutions of carbon dioxide in
%), CO2 production with the mixed amine, and rich amine water and monoethanolamine + N-methyldiethanolamine. J. Chem. Eng.
Data 1992, 37, 288-290.
loading of the mixed amine for test runs conducted between
(7) Dawodu, O. F.; Meisen, A. Mechanism and kinetics of COS-induced
February and April 2005. The figure shows a strong correlation diethanolamine degradation. Ind. Eng. Chem. Res. 1994, 33, 480-487.
between these four parameters. The implication is that the rich
loading, which represents the capacity to store or absorb CO2, ReceiVed for reView May 15, 2005
started high in February but started to decline as a function of ReVised manuscript receiVed July 2, 2005
Accepted July 8, 2005
time. With the lean amine being more or less constant for the
same testing period, this led to a corresponding decline in the IE050569E