International Journal of Greenhouse Gas Control 55 (2016) 23–35

Contents lists available at ScienceDirect

International Journal of Greenhouse Gas Control

journal homepage: www.elsevier.com/locate/ijggc

Dynamic modelling of a commercial-scale CO2 capture plant

integrated with a natural gas combined cycle (NGCC) power plant
Zhenrong He, Luis A. Ricardez-Sandoval ∗
Department of Chemical Engineering, University of Waterloo, Waterloo, ON N2L 3G1, Canada

a r t i c l e i n f o a b s t r a c t

Article history: MEA-based post-combustion CO2 capture technology has been considered as one of the mature tech-
Received 12 June 2016 nologies to mitigate CO2 emissions since it can be easily integrated to existing fossil-fired power plants.
Received in revised form 26 October 2016 In this work, a model describing the dynamic operation of a 453MWe natural gas combined cycle (NGCC)
Accepted 1 November 2016
power plant integrated with a commercial-scale post-combustion CO2 capture plant has been developed.
Available online 11 November 2016
The proposed model has been used to evaluate the dynamic performance of the integrated process under
various scenarios, e.g. changes in the reboiler heat duty and power plant inputs. In addition, the transient
Keywords:
operation of the integrated plant using a pre-defined (scheduled) trajectory profile in the consumption of
Dynamic modelling
Process integration
steam in the reboiler unit has been compared to the case of constant withdrawal of steam from the power
Post-combustion CO2 capture plant. The results show that a coordinated effort between the two plants is needed to run the integrated
Natural gas combined cycle power plant plant efficiently and at near optimal economic points under changes in power demands. The insights
gained through this analysis will be instrumental to design basic and advanced control and scheduling
strategies for the integrated NGCC-CO2 capture plant.
© 2016 Elsevier Ltd. All rights reserved.

1. Introduction causes a drop in the power plant’s efficiency to deliver electricity

to the power grid. Significant interactions are therefore expected
Post-combustion technology is considered one of the most to occur during the normal operation of these two processes. In
mature approaches available to capture CO2 since it can be easily addition, sudden and unpredictable changes are expected to occur
integrated to existing fossil-fuel power stations; also, it provides during normal operation; thus, the transient behaviour of these
flexible operation for a wide range of operating conditions and integrated processes must be assessed to ensure that the oper-
can achieve high CO2 removal. Among the multiple solvents that ability of both the power plant and the CO2 capture plant remains
can be selected to absorb CO2 , e.g. piperazine (Gaspar et al., 2016), dynamically feasible. A detailed analysis on the dynamic interac-
methyldiethanolamine/piperazine (Closmann et al., 2009), ammo- tions between these plants can provide fundamental insight on
nia (Darde et al., 2010), and ionic liquids (Valencia-Marquez et al., the controllability and the design of suitable scheduling schemes
2015), monoethanolamine (MEA)-based post-combustion CO2 cap- for the integrated system. Early studies in this area were mainly
ture technology has been deeply explored and is considered as focused on coal-fired power plants integrated with CO2 capture
the benchmark to develop studies in post-combustion CO2 capture plants. Lawal et al. studied the integration of a commercial-scale
(Aroonwilas et al., 2003; Damartzis et al., 2014; Mac Dowell and coal-fired power plant (500 MWe sub-critical power plant) with a
Shah, 2014; Harun, 2012; He et al., 2015; Kvamsdal and Hillestad, post-combustion CO2 capture process (Lawal et al., 2012). In that
2012; Lin et al., 2011; Mac Dowell et al., 2013; Mahapatra et al., study, the dynamic performance of the integrated power plant/CO2
2014; Rabensteiner et al., 2015; Razi et al., 2013; Sahraei and capture plant was evaluated under set-point tracking changes in
Ricardez-Sandoval, 2014; Sharma et al., 2015). Despite their bene- CO2 capture rate. Zhang et al. performed a controllability analy-
fits, the energy dependence of the CO2 capture plant to regenerate sis on a 550 MWe coal-fired power plant model integrated with
the solvent used in the absorption process creates a direct depen- an MEA-based CO2 capture plant (Zhang et al., 2016). Lin et al.
dence of this process to the fossil-fired power plant, which in turn integrated a 580 MW bituminous coal-burning power plant model
with a CO2 capture plant using Aspen Dynamics; a decentralized
control scheme was proposed for this process (Lin et al., 2012).
GarÐarsdóttir et al. proposed a dynamic MEA-based CO2 capture
∗ Corresponding author. process using as a basis a steady state coal-fired power plant model
E-mail address: [email protected] (L.A. Ricardez-Sandoval).

http://dx.doi.org/10.1016/j.ijggc.2016.11.001
1750-5836/© 2016 Elsevier Ltd. All rights reserved.
24 Z. He, L.A. Ricardez-Sandoval / International Journal of Greenhouse Gas Control 55 (2016) 23–35

(GarÐarsdóttir et al., 2015). A decentralized control structure was Table 1

Main equipment parameters of the NGCC power plant.
designed in that work and compared its performance against open-
loop operation. Equipment Current model Process Data (Luo et al., 2015)
Although coal-fired power plants contribute in significant 1. HP Steam Turbine
proportions to global CO2 emissions, natural gas fired power pro- 1) Inlet steam Pressure(bar) 170.4 172.6
duction is often used in developed countries to accommodate a 2) Inlet steam Temperature(K) 874 874.85
significant amount of the electricity demands. Compared to coal- 3) Efficiencies (%) 92 92

fired power plants, gas-fired power plants are regarded to be 2. IP Steam Turbine
highly flexible and environmentally friendly (IEAGHG, 2012), i.e., 1) Inlet steam Pressure(bar) 41.5 41.5
2) Inlet steam Temperature(K) 873.75 874.15
they produce relatively low CO2 emissions. However, CO2 cap-
3) Efficiencies (%) 94 94
ture and storage (CCS) technology still plays an important role
to deeply reduce CO2 emissions from natural gas burning power 3. LP Steam Turbine
1) Inlet steam Pressure(bar) 5.8 5.8
plants (IEAGHG, 2012). Despite this fact, studies that evaluate the
2) Inlet steam Temperature(K) 622.15 566.25
performance of integrated natural gas power plants with CO2 cap- 3) Efficiencies (%) 90 90
ture process are very limited. Luo et al. presented a full-scale natural
4. Gas Turbine
gas combined cycle (NGCC) power plant integrated with an MEA- 1) Brand & Type GE 9371FB GE 9371FB
based CO2 capture plant (Luo et al., 2015). Both the power plant 2)Compressor Pressure Ratio 18.2 18.2
and the CO2 capture plant models were developed at steady-state 3)Net Efficiency(%, LHV) 38.7% –
using Aspen Plus. To the authors’ knowledge, the study reported
by Ceccarelli et al. is the only work that has provided insight on
the dynamic operation of the integrated gas-fired power plant and The flue gas coming from the HRSG unit is sent to the CO2 capture
CO2 capture plant during start-up and shut-down (Ceccarelli et al., plant for its processing and treatment.
2014). Tables 1 and 2 present the key equipment parameters and the
The aim of this study is to evaluate the dynamic performance nominal operating conditions for the NGCC power plant modelled
of the integrated NGCC-CO2 capture plant. In this work, a 453MWe in this study. Table 1 also presents the brand, type and net efficiency
NGCC dynamic power plant model integrated with a MEA-based (%, LHV) of the gas turbine and the compressor pressure ratio. As
post-combustion dynamic CO2 capture process has been developed shown in Table 1, the equipment parameters specified for the three
and validated using data reported in the literature. In the present steam turbines, i.e. temperature and pressure of the inlet steam, are
study, the dynamic performance of the integrated NGCC-CO2 cap- slightly different from that reported in a previous study (Luo et al.,
ture plant has been evaluated under different scenarios that are 2015); the efficiencies of the steam turbines are specified the same
expected to occur during normal operation. To illustrate the poten- as those reported by Luo et al. The differences observed were mostly
tial benefits of process scheduling and control for the integrated due to the transition from the Aspen Plus steady-state model to the
plant, a case study featuring a pre-defined scheduling of the heat corresponding Aspen Dynamics model, i.e. some model parame-
duty consumption in the reboiler unit during oscillatory changes ters and operating conditions were re-adjusted to reach a stable
in the electricity demands has been performed and compared with steady-state operation of the Aspen Dynamics model. As a result,
the case of constant withdrawal of steam from the power plant. as shown in Table 2, the current nominal operating conditions spec-
ified for the NGCC power plant were slightly different from those
reported in the previous study. For instance, the power output with-
2. Model development
out CO2 capture plant obtained with the present model is 0.4MWe
lower than that reported in by Luo et al., which only represents a
This section presents the procedure followed to set-up the
0.1% deviation with respect to that obtained by the present model.
dynamic model employed in this work to assess the dynamic per-
The net plant efficiency is lower by 8.5% when compared to the
formance of the integrated NGCC-CO2 capture plant. A dynamic
previous study. This may be partly because of the slightly differ-
integrated process model that considers a load-following NGCC
ences between the process flowsheets used here and that reported
power plant and an MEA-based post-combustion CO2 capture plant
by Luo et al. to model the NGCC plant; that is, due to the limited
was implemented in this study. Each of these models is described
information provided on the heat exchangers’ specifications, 10
next.
heat exchangers(Aspen-Tech, 2013a) were required by the current
model (see Fig. 1) to achieve similar performance and efficiency
2.1. Natural gas combined cycle (NGCC) power plant on the NGCC plant to those reported by Luo et al., which reported
14 heat exchangers in their flowsheet. Note that due to the inher-
In this study, a 453MWe NGCC plant has been developed ent characteristic of the fuel, i.e. natural gas, the concentration of
according the process data and information reported in the the CO2 in the flue gas from current NGCC power plant is approx-
literature(IEAGHG, 2012; Luo et al., 2015). As depicted in Fig. 1, imately 4.3 mol%, which is in good agreement with that reported
the NGCC power plant mainly consists of one gas turbine and a by Luo et al., i.e. 4.5 mol%. Table 2 also shows that the O2 concen-
high, intermediate and a low steam turbines operated at different tration in the flue gas is 11.84 mol%, which is in good agreement
steam pressures, i.e. 170.4 bar, 41.5 bar and 5.8 bar, respectively. with that reported in the literature. Accordingly, the ratio of natu-
The air stream with ambient temperature and pressure (i.e. 15 ◦ C ral gas to the air flowrate will be fixed for the present analysis to
and 1.013 bar) is compressed into 18 bar and then enters into the ensure complete combustion of the natural gas. The CO2 compres-
combustion chamber together with the natural gas stream. The sion electricity consumption and power consumption in the CO2
high temperature exhausted gas stream is first expanded in the capture plant were not considered in the current study and thus net
gas turbine to produce electricity. This section of the process is power output with CO2 capture plant was recorded approximately
known as the Gas Turbine unit. The hot gas leaving the chamber 40 MWe higher compared with process data reported by Luo et al.
then enters the Heat Recovery Steam Generator (HRSG) unit where The steady-state model for this process was initially developed on
the water supplied to this unit is heated with the exhausted gas to Aspen Plus and then transformed to its flow-driven dynamic ver-
generate steam, which is then passed through the high, interme- sion using Aspen Dynamics. To simplify the analysis, the off-design
diate and low pressure turbines to produce additional electricity. dynamic performance evaluation of the gas turbine and steam tur-
 Z. He, L.A. Ricardez-Sandoval / International Journal of Greenhouse Gas Control 55 (2016) 23–35 25

Fig. 1. NGCC power plant flow block diagram.

Table 2
Nominal operating conditions of NGCC power plant.

Operating Conditions Current model Process data (Luo et al., 2015)

Net power output with no capture(MWe) 453.5 453.9

Net power output with CO2 capture(MWe) 421.7 379.85
Net plant efficiency with CO2 capture (%, LHV) 45% 49.16%

Natural gas stream:

1) Temperature (K) 282 282
2) Flowrate (kg/s) 16.62 16.62

Composition (vol%)
CH4 89 89
C2 H6 7 7
C3 H8 1 1
C4 H10 0.1 0.1
C5 H12 0.01 0.01
CO2 2 2
N2 0.89 0.89

Air stream:
Temperature (K) 282 282
Pressure (bar) 1.013 1.013
Flowrate (kg/s) 656.9 656.9

Flue gas stream:

Flowrate to HRSG (kg/s) 673.56 673.58
CO2 Concentration to CO2 capture plant (mole%) 4.3 4.5
O2 Concentration to CO2 capture plant (mole%) 11.84 11.4

bines under transient operations were not considered in the current increasing the CO2 concentration in the treated flue gas stream. As
dynamic modelling; thus, the performance curves for the turbines depicted in Fig. 2, the treated flue gas stream enters the absorber
were not considered in present model. The PR-BM property method tower where it reacts with the amine solvent (30 wt% MEA) and
was employed to estimate the thermodynamic properties of the captures the CO2 contained in the flue gas stream. The refined flue
system (Neau et al., 2009a,b; Peng and Robinson, 1976). The results gas stream is vented from the top of the absorber column whereas
presented in Tables 1 and 2 indicate that there is a reasonable the rich solvent loaded with CO2 is collected at the bottom of the
agreement between the design and operational characteristics of absorber and sent to the stripper section where the solvent is regen-
the present NGCC power plant model and that previously reported erated. A reboiler unit is attached at the bottom of the stripper and
in the literature. is used to strip the CO2 from the amine solvent. The lean amine
solvent stream exiting from the reboiler unit is recycled to the
2.2. MEA-based post-combustion CO2 capture plant absorber tower after exchanging heat with the rich solvent stream
coming from the bottom of the absorber unit. The rich CO2 stream
To perform process integration with the NGCC power plant, a exiting from the top of the stripper is flashed to remove the water
commercial-scale MEA-based post-combustion CO2 capture plant from CO2 and to further purify the CO2 product stream. The puri-
was also developed and implemented on Aspen Dynamics. As fied CO2 product stream is then sent to a compressor unit (not
shown in Fig. 2, the post-combustion CO2 capture process mainly shown for brevity), which increases the pressure of this stream
consists of the absorber section and the regeneration section. In and makes it ready for storage. The following equilibrium reactions
addition, a pre-treatment unit for the CO2 capture process was
implemented to remove part of the water in the flue gas thus
26 Z. He, L.A. Ricardez-Sandoval / International Journal of Greenhouse Gas Control 55 (2016) 23–35

Fig. 2. CO2 capture process flowsheet.

were considered to describe the chemistry for the absorption and Table 3
Design specifications of the CO2 capture plant.
the desorption processes:
Equipment Current model Reference (Luo et al., 2015)
2H2 O ↔ H3 O+ + OH− (1)
1. Absorber
H2 O + HCO3 − ↔ CO3 −2 + H3 O+ (2) 1) Height(m) 25 25
2) Diameter(m) 19.8 19.8
2H2 O + CO2 ↔ HCO3 − + H3 O+ (3) 3) Pressure(bar) 1.1 1.07

− − 2. Stripper
H2 O + MEACOO ↔ MEA + HCO3 (4) 1) Height(m) 15 15
+ +
2) Diameter(m) 10.2 10.2
H2 O + MEAH ↔ MEA + H3 O (5) 3) Pressure(bar) 2.1 2.1

A pilot-scale MEA-based post-combustion CO2 capture model

has been validated in a previous study (Sahraei and Ricardez-
in Table 4, the reboiler temperature satisfies the realistic range of
Sandoval, 2014) based on pilot experimental data presented by
the reboiler temperature reported by Harun et al. (2012). How-
Dugas (2006). In order to accommodate fluctuations in the flue
ever, the reboiler heat duty obtained for the current model is higher
gas coming from a 453MWe NGCC power plant, the CO2 capture
than that reported from the reference model; this change also pro-
model was scaled up by adjusting the operating conditions and
duced changes in other parameters, e.g. lean loading, rich loading
design specifications. Accordingly, a commercial-scale plant model
and L/G ratio. These adjustments were necessary to obtain a stable
previously reported in the literature was used a basis to design
steady-state dynamic model for the CO2 capture plant.
the CO2 capture plant used in this study (Luo et al., 2015). The
Electrolyte NRTL thermal property package was employed in the
Aspen dynamics model to describe the liquid-phase non-ideality 2.3. Integration of the NGCC power plant with the
with respect to activity coefficient of MEA-solvent and aqueous post-combustion CO2 capture process
electrolyte system (Aspen-Tech, 2013b; Austgen et al., 1989; Chen
and Song, 2004). The post-combustion CO2 capture is an energy intensive process
Rate-based model has been regarded to be more suitable to since significant amounts of steam are needed for solvent regener-
describe the absorption and desorption processes since it can pre- ation. This energy, often provided by the power plant, results in
dict more accurate results when compared to equilibrium-based a reduction of the power output supplied to the electricity grid.
models. However, due to the unavailability of the dynamic rate- Thus, extraction of steam from the NGCC power plant to regener-
based model in Aspen dynamics, the equilibrium-based model was ate the solvent in the stripper section has been performed when
employed in this study. To compensate for the deviations between integrating the NGCC power plant and CO2 capture plant. The loca-
these two modelling approaches, the present study adopted a sim- tion at which the steam should be extracted is not trivial; previous
ilar approach to that presented in a previous study (Zhang et al., studies have suggested that the steam extraction between the low
2016). Accordingly, the Murphree efficiency of the stripper was set pressure turbine and intermediate turbine (LP and IP) is suitable
to the unit whereas the Murphree efficiency of absorption column and feasible to reduce the energy losses for the power plant (Lawal
was adjusted until a reasonable agreement between the rate-based et al., 2012; Lucquiaud and Gibbins, 2009). In the present study, the
and the equilibrium models was achieved. Note that the adjusted steam draw-off configuration was performed as depicted in Fig. 1,
component (CO2 ) Murphree efficiency varied from 0.15 to 0.25 i.e. the steam required to regenerate the MEA solvent in the strip-
throughout the absorber column, which shows a slight deviation per was extracted from crossover pipe between LP turbine and IP
from the constant Murphree efficiency (0.25) reported in a previous turbine.
study (Øi, 2007). The resulting operating conditions and specifica-
tions from the proposed CO2 capture model have been compared 3. Dynamic performance analysis
to that from the reference model obtained by using a rate-based
model proposed by Luo et al. (2015). This comparison is presented The integrated NGCC-CO2 plant dynamic model has been used
in Tables 3 and 4. As shown in Table 3, the key operating conditions to evaluate the dynamic performance of the integrated system
obtained from the plant model are in reasonable good agreement under scenarios that are expected to occur during operation. As
with the data reported by Luo et al. (2015). Note that as shown mentioned above, the design specifications corresponding to the
 Z. He, L.A. Ricardez-Sandoval / International Journal of Greenhouse Gas Control 55 (2016) 23–35 27

Table 4
Nominal operating conditions of the CO2 capture plant.

Operating Conditions Current model Reference model (Harun et al., 2012; Luo et al., 2015)

CO2 capture rate (%) 90 90

Heat duty of reboiler (MWth ) 199.3 186.8
Reboiler temperature(K) 386.5 383–393
CO2 mass flowrate produced (kg/s) 39.3 41.11
Lean loading(mol CO2 /mol MEA) 0.32 0.32
Rich loading(mol CO2 /mol MEA) 0.46 0.46
L/G(kg/kg) 2.57 1.79

Lean solvent stream:

1) Temperature(K) 314 303
2) Pressure(bar) 1.1 1
3) MEA concentration (wt%) 30 32.5

nominal operating condition have been adopted to perform the approximately 1.3 h. Fig. 3(g) and (h) shows the responses in rich
present analysis (see Tables 1–4). Each of the scenarios considered loading and lean loading, respectively; as depicted in these two
in this work are described next. panels, a decrease (increase) in the rich loading and lean loading
was observed when reboiler heat duty was increased (decreased).
Fig. 3(g) and (h) also shows that the rich loading and lean load-
3.1. Scenario 1: step changes in the reboiler heat duty
ing increased (decreased) by approximately the same magnitude
(0.085 mol CO2 /mol MEA).
The reboiler heat duty is a key factor that determines the pro-
cess performance and efficiency of the CO2 capture process and the
NGCC power plant. Step changes are common in chemical plants; 3.2. Scenario 2: ramp changes in the reboiler heat duty
hence, step changes in reboiler heat duty were considered in the
present study and expected to provide insight on the fundamen- In addition to step changes in reboiler heat duty, ramp changes
tal transient behaviour of NGCC-CO2 capture plant. As shown in in the reboiler heat consumption are also considered in this work;
Fig. 3(a), +/−10% step changes in the reboiler energy consumption it is expected that this type of change may be more realistic when
were selected to reflect realistic and acceptable changes in the operating an integrated NGCC + CO2 capture plant. As illustrated in
power plant’s efficiency while integrated with a CO2 capture unit, Fig. 4(a), two ramp tests were implemented in this scenario to lin-
e.g. higher withdrawals from the power plant may result in a sig- early increase (decrease) for a period of 1 h the reboiler heat duty
nificant decrease in the power plant’s output that may affect the resulting in a total +/−10% changes in the heat duty supplied to
consumer (plant not being able to meet the electricity demands). this unit. As in the previous scenario, +/−10% changes in the energy
As is shown in Fig. 3(b), a 2.3% increase in the CO2 capture rate consumption were selected to reflect a realistic change in the oper-
with a settling time of 1.6 h was observed when the reboiler energy ation of the plant, e.g. a realistic drop in the power plant’s efficiency
consumption was increased by 10%. On the contrary, the CO2 cap- while integrated with a CO2 capture unit. Fig. 4(b) indicates that
ture rate was reduced by 3.5% when a 10% decrease step change the CO2 capture rate increased from nominal condition (90%) to
in the reboiler heat duty was imposed. As shown in Fig. 3(c), an approximately 92% with a +10% ramp change in the reboiler duty.
overshoot (undershoot) response was observed for the CO2 mass Conversely, a 3.5% reduction in the CO2 capture rate was observed
flowrate in the production stream when the reboiler heat duty was when the reboiler heat duty was decreased by 10%. The settling
increased (decreased) by 10%. Fig. 3(d) shows that the power out- time for the CO2 capture rate (approximately 2 h) was longer than
put increased (decreased) by approximately 1% when the reboiler that observed for the other process variables reported in Fig. 4. This
heat duty was step changed by − (+) 10%. As shown in Fig. 3(e), an result suggests that a suitable control scheme with a fast control-
increase of 1% in the reboiler’s temperature was recorded when the loop for the CO2 capture rate may need to be designed for the
reboiler heat duty was increased by 10%. Conversely, the reboiler’s integrated system to recover rapidly from changes in the NGCC
temperature was reduced by 1% with a 10% step change decrease power plant. The responses in the CO2 mass flowrate in the pro-
in the reboiler heat duty. The regeneration heat duty, i.e. ratio duction stream are shown in Fig. 4(c). For +10% ramp change in
of the reboiler heat duty (GJ/s) to the CO2 mass flowrate in the the reboiler heat duty, once the reboiler heat duty reaches its sta-
product stream (ton/s) is presented in Fig. 3(f). Due to +10% step bility at around 1.3 h, the CO2 flowrate at the top of the stripper
change imposed in the reboiler duty at time 0.3 h, an overshoot reaches its maximum; at the same time, more CO2 entered recy-
was observed in the CO2 mass flowrate followed by an overdamped cled stream. This dynamic effect causes an initial increase in the CO2
non-oscillatory response (see Fig. 3(c)). As shown in Fig. 3(f), the mass flowrate followed by an overdamped response. A similar anal-
regeneration heat duty has a sudden increase in its value, which ysis can be performed when −10% ramp change in the reboiler heat
is due to the sudden change made in the heat duty supplied to the duty was introduced into the system. The settling time for the CO2
reboiler unit. After this initial and sudden change, the reboiler’s heat mass flowrate was the second longest recorded for the integrated
duty is maintained at its new operating point whereas the CO2 mass plant (approximately 1.9 h). As shown in Fig. 4(d), a linear increase
flowrate, which is inversely proportional to the regeneration heat (decrease) of approximately 1% in the power output was observed
duty, develops an overshoot due to the change made in the reboiler when the reboiler heat duty was decreased (increased) by 10%.
heat duty. Accordingly, the response observed in the regeneration Similarly, as depicted in Fig. 4(e), an approximately linear increase
heat duty shown in Fig. 3(f) is a contribution of these combined (decrease) by 1% in the reboiler temperature was observed when
effects, i.e. an initial change in its value followed by an undershoot the heat duty in this unit was linearly increased (decreased). Fig. 4(f)
(observed with respect to that new value in the regeneration heat illustrates that the regeneration heat duty increased by approx-
duty).An overall increase (decrease) in the regeneration heat duty imately 7.5% when the reboiler heat duty was ramped up by 10%
of 7.5% (6.7%) was recorded when the reboiler energy consump- whereas a 6.8% increase in the regeneration heat duty was recorded
tion was increased by + (−) 10%. Moreover, the settling time for when the reboiler heat duty was ramped down by 10%. Fig. 4(g)
both the regeneration heat duty and the CO2 mass flowrate was and (h) illustrates the responses of rich loading and lean loading in
28 Z. He, L.A. Ricardez-Sandoval / International Journal of Greenhouse Gas Control 55 (2016) 23–35

Fig. 3. Plant responses during the step changes in the reboiler heat duty.

the CO2 capture plant, respectively. A similar dynamic behaviour air flowrates may need to be supplied during peak times to accom-
to that observed from the previous scenario was recorded for these modate additional requests on electricity demands. These changes
variables. Fig. 4 indicates that the settling time for the key variables in the NGCC power plant will also produce changes in the flue gas
in the integrated plant (e.g. CO2 capture rate, CO2 mass flowrate in flowrate thereby affecting the performance of the integrated sys-
the production stream and reboiler’s temperature) are larger than tem, especially the CO2 capture plant. In the present scenario, step
those obtained by the step changes in the reboiler heat duty. In changes were simultaneously imposed on the air and natural gas
addition to the responses observed in the rich loading and lean flowrates to reflect the expected load-following behaviour in the
loading, similar trends were also obtained for CO2 capture rate power plant. As shown in Fig. 5(a), a positive (negative) 10% step
and reboiler temperature than those recorded for Scenario 1. Note change in the natural gas flowrate was introduced into the inte-
that the dynamic behaviour of regeneration heat duty differs sig- grated model. To completely burn the natural gas in the chamber
nificantly between the two scenarios due to the different type of from the Gas Turbine unit (see Fig. 1), the air flowrate was also
changes considered. increased or decreased with a fixed natural gas to air flowrate ratio.
As in the previous scenario, +/−10% step changes in the natural gas
and air flowrates were considered. The step changes on the power
3.3. Scenario 3: step changes in the natural gas flowrate
plant inputs produce a change in the flue gas flowrate which affects
the operation of the CO2 capture plant. As shown in Fig. 5(b), a
Load-following power plants may typically reduce the elec-
−3.4% change in the L/G ratio was recorded for the positive step
tricity production during off-peak time periods by decreasing the
changes in the power plant’s natural gas and air flowrates. Simi-
natural gas and air flowrates; conversely, higher natural gas and
 Z. He, L.A. Ricardez-Sandoval / International Journal of Greenhouse Gas Control 55 (2016) 23–35 29

Fig. 4. Plant responses during the ramp changes in the reboiler heat duty.

larly, a 3% change in the same process variable was observed when imposed in the power plant’s input whereas a 10.2% decrease in
a negative 10% step changes were imposed. Similar responses were the power output was obtained when negative step changes were
recorded for the CO2 capture rate (see Fig. 5(c)), i.e. the CO2 capture introduced into the plant model. As shown in Fig. 5(f), the regen-
rate increased (decreased) by 3.4% (7.4%) when negative (positive) eration heat duty follows opposite trends to those observed for the
step changes in the natural gas and air flowrates were performed. CO2 mass flowrate. Fig. 5(g) shown that the rich loading increased
Note that the CO2 capture rate decreased (increased) due to the (decreased) by 0.7% (2.2%) when positive (negative) step changes in
fact that a smaller (larger) L/G ratio was obtained when the natural the natural gas and air flowrates were considered. This was mostly
gas and air flowrates were increased (decreased). This observation due to the changes in the CO2 mass flowrate from the flue gas
agrees with that reported in a previous study, i.e. larger L/G ratios stream, which were produced by the changes performed in the
are expected to improve the CO2 capture rate (GarÐarsdóttir and power plant inputs. As shown in Fig. 5(h), minor changes in the
Emilsdottir, 2012). The settling time recorded for the CO2 capture reboiler temperature were recorded for this scenario, i.e. positive
rate is approximately 0.7 h. Fig. 5(d) shows the response of the CO2 step changes in the power plant inputs resulted in a decrease in
mass flowrate in the product stream; this variable increased by 1.8% temperature of 0.1 K whereas a negative change of the same magni-
when positive step changes in the power plant’s inputs were intro- tude in the power plant inputs produced an increase in temperature
duced whereas an approximately 6.9% decrease on the CO2 mass of 0.5 K. Note that the gains obtained for the positive and negative
flowrate was obtained under step changes in the negative direc- changes in the air and natural gas flowrates in a few of the pro-
tion in the natural gas and air flowrates. The dynamic response of cess variables are different, e.g. the gains for the CO2 capture rate
the power plant output is shown in Fig. 5(e); a 9.5% increase in with respect to positive and negative step changes performed in
the power output was observed when positive step changes were the natural gas flowrate are −0.04 and −0.018, respectively. The
30 Z. He, L.A. Ricardez-Sandoval / International Journal of Greenhouse Gas Control 55 (2016) 23–35

Fig. 5. Plant responses during the step changes in the natural gas and air flowrates.

difference between the gains is a clear indication of the degree the order of minutes during the operation of an NGCC power plant.
of nonlinearity of the integrated plant model. Note that the gain As a result, four step-increments were performed (each consisting
is defined as the change in the process variable with respect to of a 0.5 kg/s increase in the natural gas flowrate every minute). The
changes in the input variables (at steady-state). Moreover, among 0.5 kg/s increase was selected to provide fundamental insight of the
the key variables illustrated in Fig. 5, the settling time for the power expected changes in the operation of the integrated plant. Fig. 6(a)
plant’s output was significantly smaller when compared with the illustrates that the L/G ratio decreased by 4.1% when the power
process variables from the CO2 capture plant. This result suggests plant’s input flowrates were increased. Fig. 6(b) shows the dynamic
that a slower response is expected from the CO2 capture plant when response of CO2 capture rate to these changes. As shown in this Fig-
compared with that of the NGCC power plant. This result agrees ure, an overall overdamped response in the CO2 capture rate was
with previous observations reported for coal-fired power plants observed for four step-increments in the natural gas flowrate. The
integrated with CO2 capture plants (Lawal et al., 2012). settling time for the CO2 capture rate was approximately 50 min,
which was long compared to the settling time observed for the
3.4. Scenario 4: step-increments in the natural gas flowrate other variables (see Fig. 6). Fig. 6(c) shows the response of the CO2
mass flowrate in the production stream. Step-increments in the
This scenario explores the response in the integrated system natural gas and air flowrates result in a series of increments in the
when continuous increments in the power outputs are required flue gas flowrate in the order of minutes, i.e. more CO2 and other
to meet the demands from the power grid. Accordingly, step-wise gases (particularly water in the form of steam) enter the CO2 cap-
increments in the natural gas and air flowrates were performed ture plant within the order of minutes. The sudden and continuous
on the integrated plant; the air flowrate was tuned accordingly to increase of water within the flue gas stream causes an imbalance
completely exhaust the natural gas with a fixed natural gas to air in the water flowing through the CO2 capture plant thus affecting
flowrate ratio. Natural gas and air flowrates may be increased in the efficiency of the absorber and stripping units to capture CO2
 Z. He, L.A. Ricardez-Sandoval / International Journal of Greenhouse Gas Control 55 (2016) 23–35 31

Fig. 6. Plant responses of step increments in the natural gas and air flowrates.

and remove CO2 from MEA, respectively. Accordingly, the sudden ity demands for a given day in Ontario, Canada (IESO, 2015).
imbalance of water in the CO2 plant results in an undershoot fol- As shown in this figure, the electricity demands varied from
lowed by overshoot in the CO2 mass flowrate of the production 15,810 to 19,562 MW during the day. The fluctuations observed in
stream; once these transients have decayed, the CO2 mass flowrate power demands during a typical day confirm that load-following
reached its new steady-state condition after 40 min with an overall NGCC power plants need to be dynamically operated to meet the
increase in the CO2 mass flowrate of 1.2% with respect to the nom- power demands and remain economically attractive. Moreover, the
inal condition, as shown in Fig. 6(c). Note that the pre-treatment changes in the flue gas flowrate due to the changes in the power
unit used to remove water from the flue gas stream entering into output will affect the process efficiency of the CO2 capture plant.
the CO2 capture plant was set at its nominal operating condition Based on the above, the aim of the present scenario is to explore
during this scenario. This result suggests that the operation in the the transient operation of the integrated plant under this condi-
pre-treatment unit needs to be re-adjusted when changes in the tion using a pre-defined (scheduled) time-trajectory for the steam
natural and air flowrates are being implemented to avoid exces- requirements in the reboiler unit of the CO2 capture plant. An
sively large amounts of components that can be effectively removed approximated 24-h period sinusoidal signal in the power output
before entering the CO2 capture plant, e.g. water. A similar analy- was designed and used to represent the actual oscillatory behaviour
sis can be made for the regeneration heat duty (see Fig. 6(e)). As observed in the fossil-fired power plant output. To approximate an
shown in Fig. 6(d), a total 11.3% increase in the power output was oscillatory behaviour in the power output, a sinusoidal change in
achieved due to the four step-increments performed in the natu- the natural gas was introduced into the power plant. Similar to pre-
ral gas and air flowrates. Fig. 6(e) shows that the regeneration heat vious scenario, the air flowrate changed as per the fixed ratio of the
duty decreased by 1.2% when step-changes in the natural gas and air natural gas to the air flowrate. The corresponding sinusoidal natural
flowrates were imposed into the integrated system. As depicted in gas and air flowrates, which produce the corresponding (oscilla-
Fig. 6(f), the rich loading exiting from the absorber tower increased tory) power output behaviour, is shown in Fig. 8. Two different
as the step-increments were performed in the process. This posi- cases were considered in this scenario, i.e. the integrated plant was
tive changes in the rich loading are due to the increase of the flue simulated using a constant reboiler heat duty and a pre-defined
gas flowrate, i.e. larger amounts of CO2 enter the CO2 capture plant scheduled time-trajectory in the steam supplied to the reboiler.
with a constant reboiler heat duty thus increasing the concentra- Note that the natural and air flowrate signals shown in Fig. 8 were
tion of CO2 in the rich loading stream. Small deviations in the order used for both tests. To meet the peak electricity demands and uti-
of +/−0.1 K were observed for the reboiler’s temperate (not shown lize the power at off-peak time, an arbitrary pre-defined scheduled
for brevity). As in the previous scenario, significantly small settling reboiler heat duty trajectory with a 24-h period was designed (see
times were observed for power plant’s output than those recorded Fig. 9). As shown in this figure, a 10% change in the reboiler heat
for the CO2 capture plant. duty was introduced at off-peak time (t = 2 to t = 10) whereas the
reboiler heat duty was deceased by 10% during the peak time (t = 14
to t = 22). When compared to the case of constant steam withdrawal
3.5. Scenario 5: scheduled steam consumption profile in the from the power plant, the scheduled (time-varying) reboiler heat
reboiler unit duty trajectory dropped the power output as much as 1% during
the off-peak time and increased the electricity output by as much
As discussed above, the power output from load-following as 0.8% during the peak time, as shown in Fig. 10. Fig. 11(a) shows
power plants are subject to changes throughout the day to meet that, under the scheduled steam consumption profile, the CO2 cap-
the varying electricity demands. Fig. 7 shows the total electric-
32 Z. He, L.A. Ricardez-Sandoval / International Journal of Greenhouse Gas Control 55 (2016) 23–35

Fig. 7. Ontario electricity demand on the Nov. 5 of 2015 (IESO, 2015).

Fig. 8. Sinusoidal signals for the power plant inputs.

Fig. 9. Proposed reboiler heat duty: constant reboiler heat duty and scheduled reboiler heat duty.
 Z. He, L.A. Ricardez-Sandoval / International Journal of Greenhouse Gas Control 55 (2016) 23–35 33

Fig. 10. Power output responses under different scenarios: constant reboiler heat duty and scheduled reboiler heat duty.

Fig. 11. CO2 capture rate and regeneration heat duty responses under different scenarios: constant reboiler heat duty and scheduled reboiler heat duty.

ture rate increased by a maximum of 1.4% at off-peak time when respectively. Moreover, the changes observed in these variables are
compared to that obtained with the constant steam consumption also correlated with the changes in the flue gas flowrate, which
profile. Similarly, a decrease of 6.2% in the CO2 capture rate at peak follow the sinusoidal profiles imposed on the power plant’s input
time was recorded when the scheduled steam consumption tra- flowrates (see Fig. 8). In the case of the CO2 capture rate, the devi-
jectory was employed when compared to the case of a constant ations observed in the positive direction are smaller than those
steam consumption profile. Hence, the scheduling approach that observed in the negative direction. This is a clear indication of the
supplies steam to the reboiler unit results in a feasible scheme since nonlinearity of the integrated system and that the selection of the
it can meet the varying electricity demands at either peak time or nominal operating point is key for the efficient operation of the
off-peak time while keeping the CO2 capture within acceptable lim- integrated system. Similarly, the variability observed in the regen-
its. Fig. 11(b) shows that the regeneration heat duty increased by eration heat duty also presents a nonlinear behaviour but this is not
approximately 8.5% at off-peak time when the scheduled reboiler as significant as that observed for the CO2 capture rate. These results
heat duty was considered. During the peak time, this variable suggest that changing of the steam consumption rate supplied to
decreased by approximately 4.3%. As shown in Figs. 10–11, the the reboiler unit during the course of a day produce significant vari-
response in the power output shows minor deviations between ability in the CO2 capture plant, even for small changes in the power
the two steam consumption trajectories; on the other hand, large plant’s load. Accordingly, a coordinated effort between the NGCC
variability in the CO2 capture rate and the regeneration heat duty plant and the CO2 capture plant is needed to maintain the dynamic
were observed when the schedule (time-varying) trajectory was operability of both plants within feasible limits and at near-optimal
employed. Note that the variability observed in these variables economic operating conditions. On the one hand, a fast and agile
are directly correlated with the changes imposed on the sched- control strategy is needed to meet the daily and seasonal changes
uled (time-varying) steam consumption case. That is, an increase in the power plant demands; on the other hand, these changes
and decrease in the CO2 capture rate and regeneration heat duty may produce significant and fast changes in the steam consumption
occur during the off-peak and peak operation of the power plant, supplied to the reboiler unit, which will introduce significant vari-
34 Z. He, L.A. Ricardez-Sandoval / International Journal of Greenhouse Gas Control 55 (2016) 23–35

ability in the CO2 capture plant. Accordingly, optimal scheduling Acknowledgement

trajectories in the steam supplied to the reboiler may be beneficial
and essential to balance both the power demands required by the The financial support provided by Wison Engineering Ltd. is
electricity grid and the CO2 capture performance specifications, e.g. gratefully acknowledged.
maintain a specific CO2 capture rate target during the course of a
day or a week. While the results presented from this scenario sug-
gest that a constant steam consumption strategy may reduce the References
variability in the CO2 capture plant, the time-varying scheduling
trajectory proposed in this work for the steam consumption was Aroonwilas, A., Chakma, A., Tontiwachwuthikul, P., Veawab, A., 2003.
Mathematical modelling of mass-transfer and hydrodynamics in CO2
arbitrarily designed. In addition, suitable control schemes that can
absorbers packed with structured packings. Chem. Eng. Sci. 58, 4037–4053.
maintain the dynamic operability of the plants under their safe and Aspen-Tech, 2013a. Aspen Plus IGCC Model. Aspen Technology, Inc., Burlington,
environmental restrictions need to be specified for the integrated USA.
Aspen-Tech, 2013b. Aspen Physical Property System:Physical Property Methods.
system. Therefore, optimal time-varying steam consumption tra-
Aspen Technology, Inc., Burlington, USA.
jectories that can efficiently operate both plants in closed-loop near Austgen, D.M., Rochelle, G.T., Peng, X., Chen, C., 1989. Model of vapor-liquid
their corresponding economic targets and under changes in the equilibria for aqueous acid gas-alkanolamine systems using the
power demands need to be specified and used to dynamically assess electrolyte-NRTL equation. Ind. Eng. Chem. Res. 28, 1060–1073.
Ceccarelli, N., Van Leeuwen, M., Wolf, T., Van Leeuwen, P., Van Der Vaart, R., Maas,
the technical viability and performance of the integrated NGCC-CO2 W., Ramos, A., 2014. Flexibility of low-CO2 gas power plants: Integration of the
plants. CO2 capture unit with CCGT operation. Energy Procedia 63, 1703–1726.
Chen, C.C., Song, Y., 2004. Generalized electrolyte-NRTL model for mixed-solvent
electrolyte systems. AIChE J. 50, 1928–1941.
Closmann, F., Nguyen, T., Rochelle, G.T., 2009. MDEA/Piperazine as a solvent for
CO2 capture. Energy Procedia 1, 1351–1357.
Damartzis, T., Papadopoulos, A.I., Seferlis, P., 2014. Optimum synthesis of
4. Conclusion solvent-based post-combustion CO2 capture flowsheets through a generalized
modeling framework. Clean Technol. Environ. Policy 16, 1363–1380.
Darde, V., Thomsen, K., Van Well, W.J.M., Stenby, E.H., 2010. Chilled ammonia
In this study, a 453MWe NGCC power plant integrated with process for CO2 capture. Int. J. Greenh. Gas Control 4, 131–136.
MEA-based CO2 capture process was developed using Aspen Dugas, E.R., 2006. Pilot Plant Study of Carbon Dioxide Capture by Aqueous
Monoethanolamine (M.S.E. Thesis). University of Texas at Austin.
Dynamics on the basis of the steady state model obtained from GarÐarsdóttir, S., Emilsdottir, S., 2012. Transient Behaviour of Post-Combustion
Aspen Plus. The dynamic performance of the integrated model CO2 Capture with MEA in Coal Fired Power Plants (Master of Science Thesis).
based on the nominal operating condition was evaluated under Chalmers University of Technology.
GarÐarsdóttir, S.Ó., Normann, F., Andersson, K., Prölß, K., Emilsdóttir, S., Johnsson,
different scenarios. The insight on the basic operation of plant pro- F., 2015. Post-combustion CO2 capture applied to a state-of-the-art coal-fired
vides a fundamental understanding of the dynamic performance power plant—The influence of dynamic process conditions. Int. J. Greenh. Gas
of the integrated plant and will be essential to design feasible and Control 33, 51–62.
Gaspar, J., von Solms, N., Thomsen, K., Fosbøl, P.L., 2016. Multivariable Optimization
efficient control schemes for the integrated process. For example,
of the Piperazine CO2 Post-Combustion Process. Energy Procedia 86.
results from this analysis suggest that the operating conditions Harun, N., Nittaya, T., Douglas, P.L., Croiset, E., Ricardez-Sandoval, L., 2012.
in the pre-treatment unit need to be adjusted when significant Dynamic simulation of MEA absorption process for CO2 capture from power
plants. Int. J. Greenh. Gas Control 10, 295–309.
changes in the natural gas and air flowrate are experienced; these
Harun, N., 2012. Dynamic Simulation of MEA Absorption Process for CO2 Capture
adjustments will be key to reduce the water content in the flue from Power Plants. PhD Thesis. Univ. Waterloo.
gas stream and therefore avoid drops in the efficiency of the He, Z., Sahraei, M.H., Ricardez-Sandoval, L.a., 2015. Flexible operation and
absorber and stripping units. Also, fast controllers may be required simultaneous scheduling and control of a CO2 capture plant using model
predictive control. Int. J. Greenh. Gas Control 48, 300–311.
to maintain the CO2 capture rate and the CO2 mass flowrate in the IEAGHG, 2012. CO2 Capture at Gas Fired Power Plants. International Energy
production stream on target and under fast and sudden changes Agency, Available at: http://www.globalccsinstitute.com/publications/CO2-
in load-following NGCC plants. More challenging changes in the capture-gas-fired-power-plants (assessed February 2015).
IESO, 2015. Ontario and Market Demand, Available at: http://www.ieso.ca/Pages/
operation of the plant will be considered in the future when the Power-Data.
integrated plant model is simulated in closed-loop, i.e. when a Kvamsdal, H.M., Hillestad, M., 2012. Selection of model parameter correlations in a
control system is engaged and used to maintain the dynamic oper- rate-based CO2 absorber model aimed for process simulation. Int. J. Greenh.
Gas Control 11, 11–20.
ation of the plant within its target limits when changes in the Lawal, A., Wang, M., Stephenson, P., Obi, O., 2012. Demonstrating full-scale
power plant’s load or in the CO2 capture plant set points are con- post-combustion CO2 capture for coal-fired power plants through dynamic
sidered. Also, considering the operation of load-following NGCC modelling and simulation. Fuel 101, 115–128.
Lin, Y.J., Pan, T.H., Wong, D.S.H., Jang, S.S., Chi, Y.W., Yeh, C.H., 2011. Plantwide
power plants, the dynamic operation of the integrated plant was
control of CO2 capture by absorption and stripping using monoethanolamine
evaluated using a constant and a pre-defined scheduled trajec- solution. Ind. Eng. Chem. Res. 50, 1338–1345.
tory for the steam supplied to the reboiler unit through a 24-h Lin, Y.-J., Wong, D.S.-H., Jang, S.-S., Ou, J.-J., 2012. Control strategies for flexible
operation of power plant with CO2 capture plant. AIChE J. 58, 2697–2704.
period. The results showed that the pre-defined scheduled steam
Lucquiaud, M., Gibbins, J., 2009. Retrofitting CO2 capture ready fossil plants with
consumption trajectory provided a viable and feasible scenario to post-combustion capture. Part 1: requirements for supercritical pulverized
operate a load-following NGCC plant integrated with a commercial- coal plants using solvent-based flue gas scrubbing. Proc. Inst. Mech. Eng. Part A
scale CO2 capture plant. The insight provided by the pre-defined J. Power Energy 223, 213–226.
Luo, X., Wang, M., Chen, J., 2015. Heat integration of natural gas combined cycle
scheduled trajectory is instrumental to develop new studies that power plant integrated with post-combustion CO2 capture and compression.
can evaluate the dynamic performance and technical viability of Fuel 151, 110–117.
the integrated plant in closed-loop under scheduled and sudden Mac Dowell, N., Shah, N., 2014. Optimisation of post-combustion CO2 capture for
flexible operation. Energy Procedia 63, 1525–1535.
changes in the NGCC plant’s demands. Future work in this research Mac Dowell, N., Samsatli, N.J., Shah, N., 2013. Dynamic modelling and analysis of
is therefore aimed at developing suitable control strategies for an amine-based post-combustion CO2 capture absorption column. Int. J.
the integrated system and to study the dynamic operability of Greenh. Gas Control 12, 247–258.
Mahapatra, P., Ma, J., Ng, B., Bhattacharyya, D., Zitney, S.E., Miller, D.C., 2014.
the closed-loop under changes in the power plant. Furthermore, Integrated Dynamic Modeling and Advanced Process Control of Carbon
model-based control strategies using single and multiple (coordi- Capture Systems. Energy Procedia 63, 1354–1367.
nated) model predictive controllers can also be explored to study Neau, E., Hernández-Garduza, O., Escandell, J., Nicolas, C., Raspo, I., 2009a. The
Soave, Twu and Boston-Mathias alpha functions in cubic equations of state.
the potential implementation of these control techniques to the
Part I. Theoretical analysis of their variations according to temperature. Fluid
integrated NGCC-CO2 capture system. Phase Equilib. 276, 87–93.
 Z. He, L.A. Ricardez-Sandoval / International Journal of Greenhouse Gas Control 55 (2016) 23–35 35

Neau, E., Raspo, I., Escandell, J., Nicolas, C., Hernández-Garduza, O., 2009b. The Sahraei, M.H., Ricardez-Sandoval, L.A., 2014. Controllability and optimal
Soave, Twu and Boston-Mathias alpha functions in cubic equations of state. scheduling of a CO2 capture plant using model predictive control. Int. J.
Part II. Modeling of thermodynamic properties of pure compounds. Fluid Phase Greenh. Gas Control 30, 58–71.
Equilib. 276, 156–164. Sharma, M., Qadir, A., Khalilpour, R., Abbas, A., 2015. Modeling and analysis of
Øi, L.E., 2007. Aspen HYSYS simulation of CO2 removal by amine absorption from a process configurations for solvent-based post-combustion carbon capture.
gas based power plant. SIMS 2007 Conf., 73–81. Asia- Pac. J. Chem. Eng 10, 764–780.
Peng, D.-Y., Robinson, D.B., 1976. A new two-constant equation of state. Ind. Eng. Valencia-Marquez, D., Flores-Tlacuahuac, A., Ricardez-Sandoval, L., 2015.
Chem. Fundam. 15, 59–64. Technoeconomic and Dynamical Analysis of a CO2 Capture Pilot-Scale Plant
Rabensteiner, M., Kinger, G., Koller, M., Gronald, G., Hochenauer, C., 2015. Using Ionic Liquids. Ind. Eng. Chem. Res. 54, 11360–11370.
Investigation of carbon dioxide capture with aqueous piperazine on a post Zhang, Q., Turton, R., Bhattacharyya, D., 2016. Development of Model and
combustion pilot plant-Part I: Energetic review of the process. Int. J. Greenh. Model-Predictive Control of a MEA-Based Postcombustion CO2 Capture
Gas Control 39, 79–90. Process. Ind. Eng. Chem. Res. 55, 1292–1308.
Razi, N., Svendsen, H.F., Bolland, O., 2013. Validation of mass transfer correlations
for CO2 absorption with MEA using pilot data. Int. J. Greenh. Gas Control 19,
478–491.