Energy 263 (2023) 125490

Contents lists available at ScienceDirect

Energy
journal homepage: www.elsevier.com/locate/energy

Dynamic modeling and comprehensive analysis of direct air-cooling

coal-fired power plant integrated with carbon capture for reliable,
economic and flexible operation
Mingjuan Zhu a, Yudong Liu b, Xiao Wu a, **, Jiong Shen a, *
a
National Engineering Research Center of Power Generation Control and Safety, School of Energy and Environment, Southeast University, Nanjing, 210096, PR China
b
Huaneng Jiangsu Energy Development Co., Ltd., Nanjing, 210015, PR China

A R T I C L E I N F O A B S T R A C T

Keywords: Coal-fired power plants with direct air-cooling condensers (DACC-CFPP) are water-saving, eco-friendly and thus
Direct air-cooling coal-fired power plant widely installed in regions rich in coal but short of water. As such regions have better geological conditions for
Solvent-based carbon capture CO2 storage, retrofitting these plants with carbon capture techniques provides a cost-efficient way to reduce
Dynamic modeling
carbon emissions and retain dispatchable power. However, the integration of carbon capture poses significant
Reliability
Economics
challenges for CFPP and DACC. Therefore, this paper develops a plant-wide model of the 660MWe DACC-CFPP
Flexible operation integrated with post-combustion carbon capture (PCC). A connection system, including a steam extraction valve,
water spray and condensate returning, is designed, added to the Steam turbine-Feedwater heaters-Condenser
system and modeled to reflect dynamic interactions between the DACC-CFPP and PCC. This model is then used to
evaluate the reliability, economics and flexibility of the DACC-CFPP-PCC over wide operating conditions. Results
show that the integration of PCC causes changes in steam flowrate, pressure and temperature within the turbine,
creating reliability risks; brings 5%–10% extra exergy loss to the DACC-CFPP, but not necessarily economic
losses; reduces the minimum power load by 44MWe and improve the ramping speed by 3.75MWe/min. This
paper provides in-depth insights for decision-makers, designers, and operators to manage the DACC-CFPP-PCC
plant.

solvent-based post-combustion capture (PCC) is the most mature and

attractive option for the FFPP because it is suitable for dealing with
1. Introduction large-scale flue gas with low CO2 concentration and is easily retrofitted
for existing plants [7].
Rapid and far-reaching decarbonization of the energy sector is The main obstacle to applying PCC to the power plant is the high
crucial to achieving the Paris Agreement [1] target of limiting global energy penalty [8] because a large amount of steam is withdrawn from
warming to 1.5 ◦ C above pre-industrial levels, as the energy sector is the the steam turbine for solvent regeneration. Much effort has been
largest carbon emitter, accounting for nearly three-quarters of the total devoted to reducing reboiler heat duty by developing more efficient
emissions [2]. Carbon capture technology is irreplaceable in the tran­ solvents [9], upgrading PCC configuration [10], and optimizing process
sition to green energy [3]. First, considering the vast installed capacity parameters (reboiler temperature [11] and absorber/stripper pressure
of coal-fired power plants (CFPP) [4], retrofitting the plants with carbon [12]). Besides, many works analyze the energy penalty from the
capture technology can reduce CO2 emissions significantly at the source. perspective of integrated CFPP-PCC systems and seek more efficient
It is also cost-efficient because it can avoid the early retirement of integration schemes. Fernandez et al. [13] developed a steady-state
newly-built power plants, which are highly efficient [5]. Moreover, by model for the supercritical CFPP retrofitted with a monoethanolamine
operating flexibly over a wide range, the power plant-carbon capture (MEA)-based PCC system in Aspen Plus®. This work compared several
system can provide the grid with dispatchable and low-carbon elec­ integration options at different loads and reported that the advanced
tricity, which is beneficial for the reliable integration of intermittent integration option achieved higher efficiency over conventional
renewable power [6]. Among emerging carbon capture methods, amine

* Corresponding author.
** Corresponding author.
E-mail addresses: [email protected] (X. Wu), [email protected] (J. Shen).

https://doi.org/10.1016/j.energy.2022.125490
Received 12 May 2022; Received in revised form 10 September 2022; Accepted 16 September 2022
Available online 20 September 2022
0360-5442/© 2022 Published by Elsevier Ltd.
M. Zhu et al. Energy 263 (2023) 125490

Nomenclature (− )
ω velocity (m/s)
Cp specific heat capacity (J/kg K) ρ density (kg/m3)
CR CO2 capture rate (%) Φ diameter (m)
Ex and ex exergy (J) and exergy per unit mass (J/kg) η efficiency(%)
Ėx exergy flow (W)
g gravitational constant (9.8 m/s2) Subscripts
GV main steam governing valve opening (%) 0 and c turbine inlet and exhaust steam
h specific enthalpy (J/kg) air air
HHV higher heating value/gross calorific value/gross energy of ar coal composition on the as-received base
the fuel (J/kg) at reference thermodynamic state
HPFA power consumption of fans (MW) wi outlet water of ith feedwater heater
I exergy destruction/loss (J/kg) dryair dry air
İ exergy destruction/loss rate (J/kg s) dsw desuperheater water of boiler
km valve flow coefficients (m2) dwi drain water from ith feedwater heater
LHV lower heating value of the fuel (J/kg) fuel fuel
ṁ mass flowrate (kg/s) g and l gas-phase water and liquid-phase water
NBCR net benefit-cost ratio(%) gl mixture of gas-phase and liquid-phase water
n number (%) i and j reference number of turbine chest/feedwater heater
P power (W) ij turbine stage between the ith and jth chest
p pressure (Pa) sat saturated water
s specific entropy (J/kg) st main steam
SP setpoint wi the ith feedwater heater outlet water
t temperature (◦ C) wc condensate water
T temperature (K) Superscripts
u valve opening or ratio (%) in and out parameter at inlet and outlet
V volume (m3) ref reference value
Z height (m) min and max minimum or maximum values
Greek letters rated rated value
β excess air coefficient, the ratio of the total amount of air Lower lower limit
used in the combustion to the amount of stoichiometric air Upper upper limit

integration options. Zhai et al. [14] established integrated CFPP-PCC actual value of carbon capture from the perspective of the integrated
models under various capacities using Ebsilon® software. The eco­ system.
nomic indices of these plants, such as energy penalty, and thermal ef­ In addition, these studies only focused on the steady-state economics
ficiency, were then compared based on the grey relational analysis of the integrated CFPP-PCC system at designed operating conditions.
method. Results indicated that the 1000 MWe CFPP-PCC attained the However, it is unrealistic in the context of increasing penetration of
best economic performance, followed by the 660 MWe and the 330 MWe intermittent renewable power. The flexible operation has become a
plants. Olaleye et al. [15] carried out an advanced exergy analysis on a necessity for CFPPs rather than an exception. Flexible operation forces
550 MWe supercritical CFPP-PCC model. Compared with the conven­ the CFPPs to frequently shift the load over a wide range, resulting in
tional energy analysis, the exergy analysis concerns the quality of energy significant flue gas and steam variations. The decarbonization role of
in addition to the amount, thus can identify the location and magnitude PCC may never unfold if it fails to cope with these variations and
of irreversible exergy destructions and losses. Simulation results showed maintain a stable and reliable operation. However, the steady-state
that the steam turbine has lower exergy destruction but more fuel-saving models cannot reflect the dynamic behavior of the PCCs during the
potential than the boiler. Improved integration schemes are urgently transitions between possible operating scenarios. As a result, dynamic
needed to reduce the irreversibility of the steam turbine. Farajollahi and modeling and simulation studies have attracted much attention recently.
Hossainpour [16] developed a detailed model for a 350 MWe power Dutta et al. [18] simulated the PCC process model based on the
plant integrated with CO2 capture and compression processes. Three equilibrium-based approach in Aspen HYSYS®. This paper identified the
organic Rankine cycles were proposed to recover the waste heat from main reasons for high deviations of equilibrium-based models in mass
the flue gas cooler, CO2 compression intercoolers and extraction steam transfer calculation and presented effective modifications to improve
cooler, respectively, for energy penalty reduction. Simulations under the predictive performance. Lawal et al. [19] selected the rigorous
rated load and 90% capture rate condition indicated that plant efficiency rate-based model to describe CO2 mass transfer in the absorber/stripper,
can get increased by selecting the appropriate reboiler condensate which was shown to achieve good agreement with the experimental data
returning point and using an auxiliary steam turbine. Xu et al. [17] obtained from a pilot-scale PCC plant [20]. Dynamic simulations illus­
developed a CFPP-PCC model by the joint use of Ebsilon® and Aspen trated that the absorber is sensitive to the liquid-to-gas ratio and the
Plus®. They proposed a novel CFPP-PCC integration scheme, which used response of the stripper to the reboiler duty change was slow to 57 min
the un-reheated steam for PCC solvent regeneration after letting it due to the huge amount of solvent stored in the buffer tanks and column
expand in a supplementary steam turbine (SST). This scheme showed sumps. Flø et al. [21] modeled the Brindisi PCC pilot plant in the
superior performance at rated load and 90% capture rate because the K-Spice® simulation tool using the rate-based approach to evaluate
SST provided low-grade heat to the reboiler. However, The economic various flexible operating modes through dynamic simulations. They
indicators are not related to market conditions such as electricity price, reported that exhaust gas venting, varying solvent regeneration, and
coal price, and carbon tax benefit, which hinders the assessment of the solvent storage could help the PCC operate flexibly to maintain the CO2

2
M. Zhu et al. Energy 263 (2023) 125490

Fig. 1. The schematic diagram of the direct air-cooling coal-fired power plant retrofitted with the MEA-based post-combustion capture system.

capture rate at 90% over the 24-h simulation. Besides, the dynamic integrated system in 2.54 h. Gardarsdóttir et al. [28] established a dy­
model can also be developed by data-driven identification method for namic model for a 408 MWe CFPP-PCC in Dymola®. They concluded
the solvent-based PCC using the techniques of nonlinear autoregressive that the CO2 capture did not significantly affect the load-following
with exogenous input (NLARX) [22], piece-wise linearization [23], and capability of CFPP. More stable power generation could be achieved if
artificial neural network [24], which can achieve satisfactory accuracy the CO2 capture rate was not an operational constraint. Liao et al. [29]
without much knowledge of the process and design specifications. developed a lumped-parameter dynamic model for a 660 MWe super­
Nevertheless, the standalone PCC studies cannot thoroughly investigate critical CFPP-PCC model. A neural network inverse model was then
the dynamic interactions of the CFPP-PCC processes, thus is unable to identified from the simulation data and used to provide optimal
expand the understanding of flexibility potentials and challenges of the feed-forward signals for the plant to achieve a fast track of the
integrated plant. Researchers have turned their attention to modeling power-carbon instructions. Wu et al. [30] further investigated the dy­
the integrated DACC-CFPP and PCC system. Lawal et al. [25] scaled up a namic interactions between the CFPP and PCC and found that the impact
validated pilot-scale MEA-based PCC model to match the flue gas con­ of reboiler steam change on CFPP was 100 times faster than that on the
dition emitted from a 500 MWe coal-fired power plant (CFPP). PCC. Based on this, they proposed to use the PCC as an ‘energy storage’
First-principle dynamic models of the integrated CFPP and PCC were device to manage the fast power ramping of the CFPP. These models
then developed using gCCS® libraries. Dynamic simulations found that focused too much on the boiler dynamics; instead, the steam turbine,
the PCC process responds much slower than the CFPP. As a result, the feedwater regenerative heaters and condenser, which are directly con­
slow adjustment of the reboiler steam flowrate would cause continual nected to the PCC, were oversimplified. Moreover, the connecting sys­
fluctuations in power output if independent control systems were tem between PCC and CFPP, which regulates the reboiler steam
applied to the CFPP and PCC, respectively. Wellner et al. [26] presented temperature, pressure and flowrate, is often overlooked. As a result,
a detailed rate-based PCC model and linked it to a 600 MWe CFPP. The these models failed to represent the precise temperature, pressure and
integrated CFPP-PCC system simulation showed that the reboiler steam flowrate profiles within the thermodynamic system, which are signifi­
valve is a suitable actuator providing fast and reliable electrical power cantly changed after the integration of PCC and the existing CFPP [31].
control. Marx-Schubach and Schmitz [27] developed a simplified model Such issues are exacerbated when the flexible operation of the inte­
of a 920 MWe CFPP-PCC system to study the start-up process. A grated plant is required [32]. The unknown of these parameters makes it
sequence control scheme was tested, which could automatically start the impossible to investigate the reliability of the CFPP-PCC system and

3
M. Zhu et al. Energy 263 (2023) 125490

Table 1
Typical operating parameters of the existing DACC-CFPP under 100%, 75% and
50% loading conditions.
Variables 100% BRL 75% BRL 50% BRL

Active power Pe/MWe 660 495 330

Turbine inlet flow ṁst/(t/h) 1960 1448 958
Fuel flow ṁCM/(t/h) 314.66 239.42 163.06
Governing valve GV/% 68.7 55.22 34.79
Turbine inlet steam pressure pst/MPa 16.72 14.69 11.36
Steam temperature tst/◦ C 566 566 566
Exhaust flow ṁc/(t/h) 1279.13 979.08 688.39
DACC pressure (backpressure) pc/kPa 10 10 10
Air pressure pair/hPa 995.8 995.8 995.8
Fig. 2. The schematic diagram of the A-frame DACC [37]. Air temperature tair/◦ C 12 12 12
Excess air coefficient βair/- 1.15 1.15 1.15

cannot lead to accurate evaluations of the integrated CFPP-PCC plant *BRL stands for the boiler-rated load. Data is obtained from the real operating
efficiency or feasible operating range. data of the considered DACC-CFPP and the relevant manual [38].
In addition, no studies have been carried out on the CFPP-PCC with a
direct air-cooling condenser (DACC). As widely recognized, the DACC- 2. System description & model development of the integrated
CFPPs are significantly water-saving and eco-friendly [33] and thus DACC-CFPP–PCC process
have been extensively deployed in recent years [34]. In addition, most
DACC-CFPPs are built in regions short of water and rich in coal [35], 2.1. DACC-CFPP system
which provide more convenient and economic conditions for CO2
geological storage [36]. A comprehensive evaluation of DACC-CFPP for The CFPP under consideration is a 660 MWe unit located in Inner
carbon capture retrofit is thus critical for the large-scale commercial Mongolia, China. Because the location is rich in coal but short of water,
deployment of CCS in the power industry. On the other hand, the DACC is installed instead of the conventional water-cooling condenser.
DACC-CFPP also has inherent defects, such as higher environmental and The schematic diagram of the DACC-CFPP is shown in the upper part of
load sensitivity and higher operating costs due to poorer cooling per­ Fig. 1, which can be divided into the following four subsystems:
formance. These features deteriorate the stable, efficient, and flexible
operation of the DACC-CFPP system [37]. The integration of PCC ex­ 1) Furnace system, including the coal mill (CM), furnace (FN) and flue
acerbates this as the reboiler extracted large amounts of steam, causing gas treatment system (FGT), where the pulverized coal is burned to
the turbine and DACC to deviate from safe design conditions. Therefore, flue gas and releases the heat.
it is urgent to gain in-depth knowledge of the dynamic characteristics of 2) Boiler system, including economizer (EC), evaporator (EV), down-
the integrated DACC-CFPP-PCC system, which will help understand the comer (DC), drum (DR), reheater (RH), superheater (SH), desu­
potential effects of carbon capture retrofit in terms of economics, reli­ perheater (DSH), where the feedwater absorbs the heat generated in
ability, and flexibility. the furnace and becomes superheated.
To this end, this paper presents a plant-wide, full-condition dynamic 3) Steam turbine - regenerative feedwater heater system, including
model for a 660 MWe DACC-CFPP-PCC system by modifying an original high-pressure (HP), intermediate-pressure (IP), low-pressure (LP)
DACC-CFPP model developed in Ref. [37] and integrating it with a 30 wt cylinders, generator (GT), feedwater heater (FWH) system, deaerator
% MEA-based PCC process model [11]. A detailed connection system (DE), condensate pump (CP) and feedwater Pump (FP), where the
model, including the steam extraction valve, water spray, pressure superheated steam is expanded to produce shaft work for power
reduction, and water returning, is developed and added to the rigorous generation and the condensed water is preheated before sent into the
steam turbine-feedwater regenerative heater-DACC system model to boiler.
reflect the dynamic interactions between the DACC-CFPP and PCC. A 4) Cold end system, including DACC, fan array (FA) and hotwell (HW),
comprehensive analysis is then carried out based on the proposed model where the turbine exhaust steam is distributed to the lower A-frame
to fully assess the impact of PCC integration, ranging from the plant finned tube bundles through the upper duct and then condensed by
economy to the operating flexibility and reliability, over broad oper­ the airflow drawn up by fans (illustrated in Fig. 2).
ating conditions. The reliability analyses are carried out by investigating
the temperature, pressure, and flowrate profiles along the thermody­ A dynamic first-principle model of the existing 660 MWe DACC-CFPP
namic system of the steam turbine-feedwater regenerative has been successfully developed in our previous work [37] using
heaters-DACC. Exergy destructions of the DACC-CFPP due to carbon MATLAB Simulink®. Lumped parameter dynamic models are developed
capture are then calculated. A novel indicator, the net benefits rate of for the equipment involved in the furnace and boiler subsystems. A
cost (NBCR), is proposed to reflect the economics of the integrated one-dimensional multi-section DACC model encompassing the cross­
DACC-CFPP-PCC system under external market conditions. Feasible wind and ambient temperature disturbances is developed to study the
operating regions and power ramping performance of the plant under DACC performance under changing loading and environmental condi­
different operating strategies are simulated to demonstrate the flexi­ tions. In addition, a rigorous steam turbine-FWHs dynamic model has
bility of the DACC-CFPP-PCC plant. Further discussions are conducted to been established to reflect the couplings between the DACC and CFPP,
investigate the integrated system under freezing weather conditions and which leads to a plant-wide transient model for the integrated
different integration schemes. This paper points out the reliability, DACC-CFPP system. A hybrid modeling approach is exploited, fusing the
economics, and flexibility of the DACC-CFPP-PCC system, which pro­ physics-based understanding of the process and the information in the
vides in-depth insights for decision-makers, designers, and operators to actual operating data. The resulting DACC-CFPP dynamic model has
manage the integrated plant. been sufficiently validated against measured data under various oper­
ating conditions. Due to the limited space, the detailed model equations
are not repeated in this paper. Three typical operating conditions and
coal characteristics of the 660 MWe DACC-CFPP are listed in Tables 1
and 2, respectively.

4
M. Zhu et al. Energy 263 (2023) 125490

Table 2
Coal characteristics of the 660 MWe DACC-CFPP.
Parameters as received Carbon Oxygen Hydrogen Nitrogen Sulfur Moisture Low heat value High heat value

Symbols Car Oar Har Nar Sar Mar LHV HHV

Units wt% wt% wt% wt% wt% wt% kJ/kg kJ/kg
Values 45.62 8.77 3.01 0.88 0.47 3.84 18981.0 19754.3

*Data is obtained from the real operating data of the considered DACC-CFPP and the relevant manual [38].

2.3. Connection system

Table 3
Design parameters of the 30 wt% MEA solvent-based CO2 capture plant.
A connection system is required to integrate the PCC and DACC-
Parameters Absorber Stripper Reboiler CFPP. On the flue gas side, pretreatment, including desulfurization, se­
Number 3 1 – lective catalytic NOx reduction, dust removal and cooling, is necessary
Packing type IMTP 50# Mellapak 250Y – before sending the flue gas into the PCC absorber to ensure a satisfactory
Specific area of packing/(m2/m3) 145 250 –
capture performance. Therefore, we assume that the flue gas entering
Height/m 20 22
the PCC contains CO2, N2 and water. The flue gas flowrate ṁPCC,in
–
Diameter/m 13 17 8 fluegas and
Capacity/m3 – – 2000 the corresponding composition (ṁPCC,in PCC,in
CO2 , ṁN2 , and ṁPCC,in
H2O ) are calcu­
Design pressure/bar 1.01 1.8 1.8 lated as:
Design capture rate/% 90 – – ⎧
Design temperature/K – – 392.2 ⎪
⎪
⎪
⎪ ṁPCC,in PCC,in
fluegas = ṁCO2 + ṁPCC,in
N2 + ṁPCC,in
H2 O
⎪
⎪
⎪ ṁ
⎪ PCC,in
= ṁCM (1 + 32/12)Car
⎪
⎪ CO2
⎨ PCC,in
Table 4 ṁN2 = 0.77βair ṁ0dryair (1)
Typical operating parameters of the 30 wt% MEA solvent-based CO2 capture ⎪
⎪ PCC,in
⎪ ṁH O = ṁCM (9Har + Mar ) + 0.01βair ṁ0dryair
⎪
plant.
⎪ 2
⎪ /
⎪
⎪
⎪
⎪ 0
Load Flue Capture Reboiler Lean solvent Reboiler ⎩ ṁdryair = ṁCM (32Car /12 + 8Har + Sar + 32Nar /14 − Oar ) 0.23
gas rate/(− ) temperature/ flow of three steam
flow/ K absorbers/ flow/(kg/ The connection at the water-steam side requires attention because
(kg/s) (kg/s) s)
the heat-power-carbon interaction between the CFPP and PCC is
100% 638.5 50% CR 392.2 1111.58 96.18 embodied there. Therefore, this section has developed a detailed dy­
BRL 90% CR 392.2 2071.54 186 namic model for the connection system on the water-steam side. The
95% CR 392.2 2223.72 200.30
75% 468.1 50% CR 392.2 816.89 69.19
schematic diagram of the connection system is illustrated in the bottom-
BRL 90% CR 392.2 1518.38 133.78 right part of Fig. 1.
95% CR 392.2 1628.20 144.15 The steam is supposed to be withdrawn from the IP-LP crossover to a
50% 318.7 50% CR 392.2 557.46 46.23 header (HD) first through the steam extraction valve (SEV). The
BRL 90% CR 392.2 1038.36 88.77
extracted steam flowrate (ṁSEV) is a function of SEV opening (uSEV),
95% CR 392.2 1114.37 95.61
valve flow coefficient (kmSEV), and the difference between extraction
*CR means capture rate. pressure (pi) and header pressure (pHD):
√̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
2.2. Solvent-based PCC system ṁSEV = kmSEV uSEV ρi (pi − pHD ) (2)

where ρi is the density of the extracted steam; here, kmSEV = 1.1 m2.
30 wt% MEA solvent-based PCC process is retrofitted to the DACC-
Modifications should also be made to the original steam turbine-
CFPP to reduce CO2 emission. Three standalone parallel absorbers are
FWH system model. The steam extraction from the IP-LP crossover to
applied instead of one with a large diameter to enhance the turndown
the reboiler reduces the IP-LP crossover pressure, thus will change the
ratio of the process. The flue gas and MEA solvent are equally distributed
steam distribution along the turbine. The pressure at the extraction point
to these absorbers, where they contact each other, and the CO2 is
(pi) is calculated considering the steam extraction:
chemically absorbed. A single stripper, which is used to separate CO2
⃒
from the solvent, is deployed in the process. In the reboiler, the solvent dpi 1 ( ) dρ⃒
rich in CO2 gets regenerated under the heating of steam withdrawn from = ṁ(i− 1)i − ṁi − ṁi(i+1) − ṁSEV , cTCi = VTCi ⃒⃒ (3)
dt cTCi dp ti
the steam turbine. The schematic diagram of the PCC process is shown in
the bottom-left part of Fig. 1. Main design specifications and typical where the turbine chest (TC) is the chamber located between the turbine
operating parameters of the PCC are listed in Tables 3 and 4. stages and provides steam for the FWH. From the turbine chest i (TCi),
The dynamic model of the PCC has been developed in our previous steam is drawn to FWHi. VTCi is the volume of TCi, here, it is the IP-LP
work [11] using the gCCS library in gPROMS®, a commercial software crossover volume (VTC4 = 0.5 m3); ṁij means the inlet steam flow of
specifically designed for carbon capture simulation. One-dimensional the turbine stage between TCi and TCj; ṁi represents the steam flow
rate-based models are developed for the absorber and stripper to accu­ withdrawn from TCi to FWHi.
rately describe the reactions between liquid and gas along the axial di­ Because the reboiler inlet steam temperature and pressure are rec­
rection, while other components such as the solvent sumps/tanks, heat ommended to be 410 K and 260 kPa to avoid solvent degradation, the
exchangers, and condenser are modeled through lumped parameter header is deployed in the connection system, where the extracted steam
approach. Although it is not practical to validate this industrial-scale is mixed with spray water withdrawn from the outlet of FWH1. The
PCC model due to the lack of operational data, the high fidelity of the specific enthalpy (hHD) and pressure (pHD) of the header outlet steam are
modeling methods applied has been validated in Ref. [19] against the calculated by:
actual experimental data of a pilot-scale capture plant [20]. Due to the
limited space, the model equations and parameters will not be repeated
here.

5
M. Zhu et al. Energy 263 (2023) 125490

Fig. 3. Cold end performances under different CO2 capture rates and loads: (a) is the volume flow of the last stage; (b) is the fan speed; (c) is DACC pressure.

⎧
⎪ dhHD 1 [ in in
( in
)] the drain water from FWH3 to DE, respectively; gl stands for the mixture
⎪
⎨ dt = c1,HD ṁSEV hi + ṁWEV hWEV − ṁRBS hRBS − d1,HD ṁSEV + ṁWEV − ṁRBS
⎪ of liquid-phase and gas-phase water; ṁ4 and h4 are the flow and specific
⎪ dp 1 [ ( )] enthalpy of the steam extracted from TC4 to DE respectively.
⎪
⎪ HD
⎩ = ṁSEV hi + ṁWEV hWEV − ṁin in in
RBS hRBS − d2,HD ṁSEV + ṁWEV − ṁRBS
dt c2,HD
3. Characteristic analysis of the integrated DACC-CFPP and PCC
(4)
system
where ṁWEV and ṁin RBS represent the spay water flow and reboiler inlet
steam flow; hi, hWEV and hin Based on the developed model, this section investigates the reli­
RBS are the specific enthalpies of the extraction
steam, spray water and reboiler inlet steam. We assume that hin ability, economics, and flexibility of the integrated DACC-CFPP-PCC
RBS equals
hHD and the reboiler inlet steam pressure pin process over a wide operating range. The system’s reliability will be
RBS equals pHD. The dynamic
parameters c1,HD, c2,HD, d1,HD, d2,HD can be calculated by the following assessed by comparing some critical parameters along the steam turbine
equations: and DACC after and before carbon capture under different operating
/ conditions. Exergy analysis is then carried out to identify the impact of
⎧ c1,HD = b21 − b11 b22 b12 carbon capture on the economics of the main subsystems of the DACC-
/
⎪
⎨ d1,HD = hHD + b22 b12 CFPP. A novel indicator incorporating market conditions is presented
/ (5) to evaluate the economics of the plant from the perspective of an inte­
⎪
⎩ c2,HD = b22 − b12 b21 b11 grated energy system. Feasible power generation and CO2 capture
/
d2,HD = hHD + b21 b11 operating regions subject to capacity and safety constraints are calcu­
lated. The transient behavior of the entire system corresponding to the
⎧ ⃒
⎪ ∂ρ⃒⃒ changes of main manipulated variables is analyzed, through which a
⎪ b11 = VHD
⎪
⎪
⎪
⎪ ∂h⃒pHD new operating strategy is proposed better to explore the flexibility po­
⎪
⎪
⎪
⎨ ⃒ tential of the DACC-CFPP-PCC system. The ambient temperature is set to
∂ρ⃒⃒
b12 = VHD (6) 20 ◦ C, and the condenser pressure setpoint is fixed at 10 kPa. We assume
⎪ ∂p⃒hHD
⎪
⎪
⎪ that there is no crosswind in all simulations given in this section.
⎪
⎪ = VHD ρHD
⎪ b21
⎪
⎪
⎩
b22 = − VHD 3.1. Reliability analysis

where ρHD is the density of the header outlet steam; VHD = 0.5 m3 is the Carbon capture requires steam extraction from the steam turbine and
volume of the header. causes more condensed water to return to the deaerator. As a result, it
Then the steam at the header outlet flows through the steam side of profoundly alters the mass-heat distribution within the DACC-CFPP,
the reboiler (RBS) to release heat and condense, providing heat for the making key thermodynamic parameters deviate from the design and
regeneration of the MEA solvent. Assuming that the pressure loss even from the safe operating conditions. The reliability risk is further
(ΔpRBS) at the steam side of the reboiler is 10 kPa, the specific enthalpy exacerbated by the fact that these power plants must operate flexibly to
(hout
RBS) of steam exiting the reboiler equals saturated water enthalpy (hsat, change the load and CO2 capture rate over wide ranges. Since reliability
out
water) at the reboiler steam outlet pressure (pRBS): is based on a complex mechanism that is difficult to quantify, this section
⃒ identifies possible reliability risks of the retrofitted DACC-CFPP by
hout ⃒
RBS = hsat,water pout (7)
RBS monitoring the steam temperature, pressure, and flowrate in the steam
turbine and judging whether they exceed the operating range of the non-
pout in
RBS = pRBS − ΔpRBS = pHD − ΔpRBS (8) retrofit power plant.
The water leaving the reboiler is then pumped back to the deaerator.
The dynamic model of the deaerator is also modified. The deaerator 3.1.1. Steam flowrate changes
pressure pDE is calculated as: The exhaust steam flowrate, fan speed, and DACC pressure are shown
in Fig. 3. Fig. 3 (a) shows that exhaust steam flowing into the LP turbine
dpDE ṁ4 h4 + ṁdw3 hdw3 + ṁw3 (hw3 − hw4 ) + ṁout is significantly reduced due to increased carbon capture rate. As re­
out
RBS hRBS
= (9)
dt cDE ported by Xu et al. [39], the decrease of the steam volume flow causes
⃒ the local high temperature at the last stage blade tips. As a result, the
∂ρgl hgl ⃒⃒ maximum equivalent stress at the blade tip and the maximum defor­
cDE = VDE − VDE (10)
∂p ⃒pDE mation at the leading edge increase while the elastic modulus and
stiffness decrease, reducing the natural frequency. Cao et al. [40] sug­
where VDE is the DE volume and can be calculated using structure pa­ gest that the inlet steam mass flow of the low-pressure cylinder must be
rameters in Ref. [37]; ṁdw3 and hdw3 are the flow and specific enthalpy of greater than 10% of the rated value considering safety-related factors

6
M. Zhu et al. Energy 263 (2023) 125490

Fig. 4. Pressure differences of HP, IP and LP cylinders under different CO2 capture rates and load conditions.

Fig. 5. Steam temperatures at each turbine stage inlet under different CO2 capture rates and load conditions (reboiler steam is extracted at the inlet of TB45, i.e., the
IP-LP crossover).

such as stress, strain, frequency, and vibration. For this DACC-CFPP, the regulating the fan speed. However, under lower load and higher CO2
minimum values of the last stage mass flow and volume flow are 35.5 capture rate conditions, the fan speed has already reached the lower
kg/s and 491 m3/s. As shown in Fig. 3(a), this requirement can be met allowable limitation, resulting in losing control over the DACC system.
within the normal operating range. However, attention should still be Further ambient temperature drops or crosswinds can cause the DACC
paid to avoiding excessive steam extraction during dynamic transitions. pressure to drop and even the condensate to freeze in this condition.
Since variable-speed fans are used in the DACC-CFPP to blow air to Some fan columns have to be turned off by the DACC system to improve
condense the turbine exhaust steam, fan speed is manipulated to control operating reliability (discussed later in Section 4.1).
the turbine backpressure. In case of the steam flowrate decreases, the fan
speed should be decreased accordingly to maintain the given DACC 3.1.2. Pressure changes
pressure. Fig. 3 (b) and (c) present the fan speed and DACC pressure Fig. 4 shows the pressure differences of HP, IP and LP cylinders with
under different CO2 capture rates and power loads. Under normal and without PCC, which can help evaluate the axial thrust on the turbine
environmental conditions (air temperature 20 ◦ C, no crosswind), the cylinders. With the CO2 capture rate rise, the IP pressure difference in­
DACC pressure can be well maintained at the designed 10 kpa value by creases while the LP pressure difference decreases; the HP cylinder

7
M. Zhu et al. Energy 263 (2023) 125490

pressure difference is almost unchanged. That is because, according to

dExsystem
Stodola’s Cone Law [41], the inlet pressure of each turbine stage is I˙ = Ėxin − Ėxout − (12)
dt
related to the flowrate and the outlet pressure. The steam extraction
from the IP-LP crossover reduces the LP flowrate, which then creates a where Ėxin and Ėxout indicate the exergy inflow and outflow; dExsystem is
pressure drop at the LP cylinder inlet (i.e., the IP cylinder outlet) and the exergy change of the system itself, which equals zero if the system is
downstream. However, it hardly changes the HP flowrate, with only a in a steady state.
slight pressure drop across the HP cylinder. The exergy analysis in this section aims at the thermodynamic cycle
It is demonstrated in Fig. 4 that at 100% BRL and 95% CR, the IP of the DACC-CFPP-PCC system. Thus the entire system is divided into the
pressure difference reaches 2728 kPa, 6.78% larger than the original following subsystems: furnace-boiler (FB), steam turbine stages-
maximum IP pressure difference of 2555 kPa, and the axial thrust may generator (TB-GT), DACC, hotwell (HW), regenerative feedwater
exceed the safe range of the bearing. Rebalancing the axial thrust of the heaters, connection system and carbon capture unit.
IP cylinder during the carbon capture retrofit is thus required to prevent
reliability risks. 3.2.1. Furnace & boiler
The exergy destruction of the furnace-boiler system İFB equals the
3.1.3. Temperature changes at different steam turbine stages total exergy inflow minus the total exergy outflow:
According to Ref. [32], thermal stress caused by temperature
changes is one of the essential factors leading to plastic strain damage of I˙FB = Ėxfuel + ṁair exair + ṁin in in in
EC exEC + ṁdsw exdsw + ṁRH exRH − ṁst exst
rotor metal materials. The variations of steam temperature at each − ṁout out
(13)
RH exRH − Ėxfluegas
turbine stage are calculated under different power load and CO2 capture
rate conditions, which are illustrated in Fig. 5. TBij means the turbine where ,in in out
EC, dsw, RH, st and RH are the economizer inlet water, desu­
stage between turbine chest i and turbine chest j. The temperature in perheating water, RH inlet steam, main steam and RH outlet steam,
turbine chest i is denoted by ti. respectively.
As shown in Fig. 5, after the integration of PCC, the steam temper­ The exergy inflow by fuel Ėxfuel is calculated by:
ature (t4) decreases due to the IP-LP crossover pressure decrease. Over
the 50%–100% BRL operating range, the average temperature drop of t4 Ėxfuel = ṁCM exfuel = ṁCM HHV (14)
reaches 25 ◦ C and 75 ◦ C under 50% and 95% carbon capture rate con­
ditions, respectively. Fig. 5 illustrates that the average reductions of t5, t6 where ṁCM and exfuel are the mass flowrate and exergy of the pulverized
and t7 are 60 ◦ C, 50 ◦ C and 25 ◦ C, respectively, when the carbon capture coal.
rate is 95%; while the reductions of t1, t2 and t3 are 2 ◦ C, 7 ◦ C and 11 ◦ C, The exergy outflow by flue gas Ėxfluegas is calculated by:
respectively, much smaller than the downstream stages. The closer the ( )
Tfluegas
stage is to the extraction point, the more significant the temperature Ėxfluegas = ṁfluegas Cp,fluegas Tfluegas − Tat − Tat ln (15)
Tat
drop.
According to Fig. 5, under 50%–100% BRL conditions, when the
where ṁfluegas and exfluegas are the mass flowrate and exergy of the flue
carbon capture rate is adjusted between 50% and 95%, 5 ◦ C to 50 ◦ C
gas, respectively. We ignore chemical exergy and assume that the flue
steam temperature fluctuations will appear on different turbine stages
gas temperature and pressure are the same as the environment since
and the most significant variations occur at TB45 inlet, where the steam
they will eventually be emitted into the atmosphere.
is extracted. From this point of view, cyclic and sudden variations of
capture rate should be avoided to alleviate the steam turbine’s thermal
3.2.2. Turbine & Generator
stress or thermal fatigue.
The exergy destruction of the steam turbine İTB is calculated by:
3.2. Economic analysis ∑
7
I˙TB = ṁst exst + ṁout out in in
RH exRH − ṁRH exRH − PTB − ṁi exi − ṁc exc − ṁSEV exSEV
The energy penalty imposed by the PCC on the CFPP is the major
i=1

(16)
obstacle to large-scale commercial utilization. This section aims to
quantitatively assess the impact of carbon capture retrofit on the energy where PTB is the turbine power; ṁi and exi are the mass flowrate and
utilization and economic performance of the DACC-CFPP system. The exergy of the steam extracted from ith turbine chest to ith feedwater
exergy analysis method is adopted to determine the irreversible loss of heater; ṁSEV and exSEV are the mass flowrate and exergy of the steam
individual components under different loads and capture rates. This extracted from the turbine to the PCC; ṁc and exc are the mass flowrate
exergy method integrates the first and second laws of thermodynamics. and exergy of the exhaust steam respectively.
Because the exergy method can specify the location, type, and magni­
The exergy destruction of the generator İGT equals the turbine power
tude of irreversibility [42], it can provide in-depth insight into the
minus the electrical power Pe:
economics of the thermodynamic cycle within the DACC-CFPP-PCC
system. I˙GT = PTB − Pe (17)
Exergy of the steady flow stream such as water or steam within CFPP
is defined as the summation of kinetic, potential and physical exergy, 3.2.3. Direct air cooling condenser
ignoring chemical exergy. The exergy per unit mass, ex, is calculated as: The exergy destruction of the DACC İCD is calculated by:
ω2 (18)
(11) I˙CD = HPFA + ṁc exc − ṁout out
ex = h − hat − Tat (s − sat ) + + gZ CD exCD
2
where HPFA is the power consumption of fans; ṁout out
CD and exCD are flow and
where temperature Tat = 20 ◦ C and pressure pat = 0.1 MPa is considered
exergy of the DACC outlet water, respectively.
as thermodynamic reference state; hat and sat are the reference specific
enthalpy and entropy respectively; ω and Z are the velocity and altitude
3.2.4. Hotwell
respectively; g is the gravitational constant, 9.8 m/s2.
The exergy destruction of the hotwell İHW is calculated by:
For an open system, the exergy destruction İ is identified by:
I˙HW = ṁout out
CD exCD + ṁdw7 exdw7 − ṁwc exwc (19)

8
M. Zhu et al. Energy 263 (2023) 125490

Fig. 6. Exergy destruction and loss of subsystems under different operating conditions.

where the subscripts dw7 and wc represent the drain water from FWH7 I˙PCC = ṁout out out out
(23)
FP exFP − ṁFP exFP
to HW and the HW outlet water, respectively.
Note that the exergy change of the flue gas is not included here since
3.2.5. Regenerative feedwater heaters & pumps the exergy destruction caused by the temperature drop has already been
The exergy destruction of the regenerative feedwater heater İFWHi is considered in the furnace-boiler system and the chemical exergy is
calculated by: ignored.
⎧
⎪
⎪
⎪ 3.2.8. The integrated DACC-CFPP-PCC system
⎪ ṁdwi− 1 exdwi− 1 + ṁwi+1 exwi+1 + ṁi exi − ṁwi exwi − ṁdwi exdwi i = 2, 6
⎪
⎪
⎪ From the perspective of the integrated DACC-CFPP-PCC system, the
⎪
⎨ ṁwi+1 exwi+1 + ṁi exi − ṁwi exwi − ṁdwi exdwi i = 1, 5
⎪ exergy input system consists of the fuel exergy and the power con­
I˙FWHi = ṁdw3 exdw3 + ṁw5 exw5 + ṁ4 ex4 + ṁout out
RBS exRBS − ṁw4 exw4 i=4 sumption of pumps and fans; The exergy product is the generated
⎪
⎪
⎪ electrical power, while the exergy loss is the summation of the exergy
out out
⎪
⎪ ṁdw2 exdw2 + ṁFP exFP + ṁ3 ex3 − ṁw3 exw3 − ṁdw3 exdw3 i = 3
⎪
⎪
⎪
⎪ ṁdw6 exdw6 + ṁout
⎩
out
CP exCP + ṁ7 ex7 − ṁw7 exw7 − ṁdw7 exdw7 i=7 losses in all the subsystems above:

(20)
in
Ėxtotal = Ėxfuel + PCP + PFP + HPFA (24)

where the subscripts dwi, wi, out out out

RBS FP and CP represent the drain water
out
Ėxtotal = Pe (25)
from FWHi to FWH(i + 1), FWHi outlet water, reboiler outlet water, ∑
feedwater pump outlet water and condensate pump outlet water, I˙total = I˙FB + I˙TB + I˙GT + I˙CD + I˙HW + I˙FP + I˙CP + I˙FWHi + I˙Connection
respectively.
+ I˙PCC
The exergy destruction of the feedwater pump İFP and condensate
(26)
water pump İCP are calculated as:
{ Fig. 6 shows the exergy destruction of the subsystems, with the share
I˙FP = ṁw4 (pFP − pw4 )/ηFP + ṁw4 exw4 − ṁw4 exout of the PCC in the total exergy loss marked. The main exergy loss still
FP
(21)
I˙CP = ṁwc (pCP − pHW )/ηCP + ṁwc exwc − ṁwc exout
CP occurs in the furnace and boiler; because thermal energy is required for
solvent generation, the addition of the PCC brings 5%–10% extra exergy
3.2.6. Connection system linking PCC and DACC-CFPP loss to the integrated plant. Besides, mixing high-temperature steam and
The exergy destruction of the steam-water connection between the desuperheating water in the connection system also causes exergy
PCC and the DACC-CFPP İConnection is calculated by: destruction, about 1/6 of the PCC.
It is worth noting that the operation of PCC significantly reduces the
I˙Connection = ṁSEV exSEV + ṁWEV exWEV − ṁin in
RB exRB (22) irreversibility of the DACC, and the higher the capture rate, the more
this reduction is. The reason is that the steam extracted to PCC is not
3.2.7. Carbon capture system condensed in the DACC, and less fan power is consumed. Nevertheless,
The exergy destruction of the PCC system İPCC equals the exergy loss the exergy loss of the integrated DACC-CFPP-PCC is still more significant
of the reboiler steam, which can be calculated by: than the original plant. Modifications to the PCC process and the CFPP-
PCC system integration scheme are required to enhance the economic
performance of the entire plant.

9
M. Zhu et al. Energy 263 (2023) 125490

Table 5 is under the existing markets in China, the EU and the US. It suggests that
Price of electricity, carbon tax and coal. the current carbon tax prices are not high enough to incentivize power
Price Series Electricity price Carbon tax Coal price plants to capture CO2. Implementing a higher carbon tax of 50$/ton in
the U.S. would allow NBCR to increase as the capture rate increases (see
China 0.090 $/(kW h) 12 $/(t CO2) 75 $/(t Coal)
Europe Union (EU) 0.30 $/(kW h) 19 $/(t CO2) 75 $/(t Coal) U.S.2 in Fig. 7). We conclude that carbon capture is more economical if
U.S. 0.16 $/(kW h) 15 $/(t CO2) 75 $/(t Coal) the ratio of the carbon tax to electricity price increases. Promoting the
U.S. 2 0.16 $/(kW h) 50 $/(t CO2) 75 $/(t Coal) CO2 utilization technique, increasing carbon tax and improving the
*Data source: Carbon tax prices are from the World Bank Group [43]. Electricity integration scheme are needed to partly compensate for the operating
prices are from Statista [44]. Coal prices are from Ref. [45]. cost of carbon capture.

However, exergy loss and destruction does not necessarily mean 3.3. Flexibility analysis
economic losses as long as appropriate carbon taxes are given. A novel
metric, the net benefits rate of cost (NBCR), is defined to reflect the In the context of increasing renewable power penetration, CFPPs
economics of the integrated system under external market conditions. must be flexible enough to change power loads over a wide range
quickly. In addition, the downstream CO2 transport network also ex­
pects flexible CO2 sources to meet the changing CO2 product demand.
captured
Electricity price Pe Carbon tax ṁCO2
NBCR = + − 1
Coal price ṁCM Coal price ṁCM However, after the existing DACC-CFPP is retrofitted with PCC, the
( captured
) (27) flexibility of the integrated system is unclear due to the complex in­
Electricity price Pe Carbon tax ṁCO2
=
Coal price ṁCM
1+
Electricity price Pe
− 1 teractions between the power plant and the carbon capture plant. This
section investigates the feasible operating region and transient charac­
where ṁcaptured is mass flowrate of captured CO2. teristics of the integrated DACC-CFPP-PCC to evaluate its flexibility. The
CO2
The net benefit-cost ratio can reflect price changes in the electricity, flowchart of the flexibility analysis is illustrated in Fig. 8 uCM, GV and
carbon and coal markets. Different price scenarios are given in Table 5, SPCR are the coal mill valve opening, governing valve opening and
including recent coal/carbon tax/electricity prices in China, European capture rate setpoint. The superscripts Lower and Upper denotes the
Union and the United States. A new price scenario with U.S. carbon tax lower and upper limits.
of 50$ is also considered. The corresponding NBCRs are calculated and
shown in Fig. 7. We can see that all the NBCRs are larger than 0, indi­ 3.3.1. Feasible operating region of the DACC-CFPP-PCC
cating that DACC-CFPP-PCC system the can earn revenues under all The feasible operating region of the DACC-CFPP-PCC is subject to the
these market prices. NBCR is mainly determined by the ratio of elec­ plant capacity constraints first, which are determined by the charac­
tricity price to coal price, so the EU has the highest NBCR. At higher teristics of the CFPP, PCC, and the integration systems. The following
loads, CFPPs are more efficient and, therefore, more profitable. feasibility conditions should also be satisfied for safety reasons.
As shown in Fig. 7, the higher the capture rate, the less economical it

Fig. 7. Net benefit-cost ratios under various loads, CO2 capture rates and market conditions.

10
M. Zhu et al. Energy 263 (2023) 125490

Fig. 8. The flowchart of the feasibility analysis for the DACC-CFPP-PCC system.

3.3.1.1. Minimum steam flowrate in the LP cylinder. As mentioned where ΦAB and nAB are the diameter and number of the absorbers,
above, the steam flowrate in the LP cylinder should be greater than 10% respectively. According to the thermal properties of the lean solvent, we
of the rated value. Therefore, the following Feasibility Condition I should assume that the density ρLS = 1061.32 kg/m3. Thus the value of ṁminLS is
be satisfied: 586.67 kg/s and the minimum lean solvent flowrate per absorber col­
ṁLP ≥ ṁmin rated
(28) umn is 195.57 kg/s.
LP = 10%ṁLP
The reboiler temperature is fixed at 392 K for efficient operation of
where ṁLP and ṁrated
LP is the actual and rated steam flowrate in the LP the PCC unit.
cylinder. The minimum flowrate ṁmin
LP = 42 kg/s for this CFPP.
3.3.1.3. Limits of the DACC pressure. The following Feasibility Conditions
3.3.1.2. Minimum and maximum lean solvent flowrate of the absorber. III is required considering the DACC pressure safety limits [37]:
Packed columns require a minimum liquid load to ensure adequate mass
pLower ≤ pc ≤ pUpper (31)
transfer. Below this minimum value, only a small part of the packing c c

surface is wetted, so liquid and gas are no longer in intimate contact, The operating zone of the DACC-CFPP-PCC system subject to the
resulting in a considerable decrease in the absorption efficiency. [46] capacity constraint is shown in grey in Fig. 9, which is enveloped by five
suggests that the minimum liquid velocity ωminLS needs to be greater than lines: Line 0% CR (also 0% SEV), Line 99% CR, Line 100% SEV, Line 50%
5 m3/(m2⋅h). Such a minimum value should be guaranteed in all the BRL and line 100% BRL. Line 0% CR represents the operating zone of the
operating conditions, especially the low power load and CO2 capture original DACC-CFPP without carbon capture; Line 99% CR refers to the
rate conditions. maximum allowable carbon capture rates; Line 50% BRL and 100% BRL
In addition, the CO2 capture rate (CR) should be in the range of 50%– represent the minimum and maximum CFPP load conditions. The
95% to avoid too high or too low liquid to gas ratio (L/G), which may extraction steam is driven by the pressure difference between the
cause flooding in the absorber [46]. Thus, Feasibility Condition II needs to extraction point and the reboiler, which in turn reduces the steam
be considered as follows: pressure at the extraction point. Line 100% SEV shows that under low
π CFPP load conditions, even if the steam extraction valve is fully open,
ṁLS ≥ Φ2AB nAB ωmin
LS ρLS (29) the extraction pressure drops too low to supply enough steam for solvent
4
regeneration to achieve the maximum CO2 capture rate.
50% ≤ CR ≤ 95% (30) The feasible operating zone of the integrated DACC-CFPP-PCC

11
M. Zhu et al. Energy 263 (2023) 125490

of the original DACC-CFPP is 330 MWe-660 MWe. Due to the carbon

capture, the minimum power of the power plant is extended to about
285 MWe, and the maximum power of the power plant drops to about
609 MWe. The reboiler steam flow rate variation also changed the
amount of CO2 being captured. Given 100% BRL condition, the power
and carbon adjustable ranges are 562–609 MWe and 38–138 kg/s,
respectively. When the boiler condition is fixed at 50% BRL, the power
range is reduced to 285–305 MWe, and the range of CO2 flowrate is
37.5–69 kg/s.
However, if the legislation imposes strict CO2 capture rate re­
quirements in the future, the power-adjustable range of the DACC-CFPP-
PCC will become much smaller. As shown in Fig. 9, such a range is only
286–562 MWe and 305–609 MWe under 95% and 50% CR conditions,
both smaller than the original power-adjustable range 330–660 MWe.

3.3.2. Dynamic flexibility of the integrated DACC-CFPP-PCC system

Integration of PCC to the DACC-CFPP can expand the power-
adjustable range of the plant and offer an effective way to improve the
power ramping speed by timely regulating the steam flowrate extracted
from the turbine to the reboiler. In light of this, a “varying capture rate”
control strategy is proposed for the PCC: when the load ramps up, the
CO2 capture rate setpoint is reduced to 50% so that more steam is
allocated to generate additional power rather than heat the solvent;
when the load ramps down, the CO2 capture rate setpoint is increased to
95%, so that more energy storage in the CFPP is consumed for solvent
Fig. 9. The power-carbon feasible operating region of the DACC-CFPP-
regeneration; the capture rate setpoint returns to the design value (90%)
PCC system.
when the power load reference is stable. Two Proportional Integral (PI)
controllers are developed to track the capture rate setpoint and regulate
system considering the safety limitations is shown in green in Fig. 9. the reboiler temperature. The diagram of the control structure is shown
Feasibility Conditions I and III are already satisfied in the operating zone, in Fig. 10.
subject to capacity constraints. However, the minimum lean solvent The load ramping-up process of the DACC-CFPP-PCC system from the
flowrate and CO2 capture rate constraints further limit the feasible 50% BRL & 90% CR condition to 100% BRL & 90% CR condition is
operating region. According to Fig. 9, the basic CFPP-PCC integration simulated to verify the flexibility potential of PCC in the dynamic
scheme considered in this study is almost sufficient to achieve 50%–95% transient process. At t = 10 min, the fuel valve opening and turbine
CO2 capture rates under different boiler load conditions. governing valve opening increase, as shown in Fig. 11, to increase the
Retrofitting the DACC-CFPP with PCC results in a power penalty. The power load. The simulation results are in Figs. 12–15.
drawn-off steam used initially for electricity generation is now used to As shown in Fig. 12, the “varying capture rate” strategy can accel­
provide heat for solvent regeneration. Consequently, given the same erate the power ramping speed under the same turbine governing valve
boiler loads, the higher the capture rate or the captured CO2 flowrate, and fuel valve changes. It takes about 33 min for the CFPP to rise from
the smaller the power generation (See Lines for 50%, 75%, and 100% 285 MWe to 567.80 MWe (8.51 MWe/min) under the “varying capture
BRL in Fig. 9). rate” operating strategy, while the ramping time is 59 min for the “fixed
The retrofit provides a new viable option for operators to adjust the capture rate” strategy (4.76 MWe/min). The decrease of the capture rate
power output and capture rate by manipulating the reboiler steam setpoint at the beginning stage of power ramping up will reduce the lean
flowrate withdrawn to the PCC system. The power load variation range solvent flowrate first (Fig. 13). And then, reboiler temperature rise,

Fig. 10. Diagram of the proposed PCC control strategy.

12
M. Zhu et al. Energy 263 (2023) 125490

Fig. 11. Changes of the coal valve and turbine governing valve during the load ramp-up process.

Fig. 12. Generator Power during the load ramp-up process.

Fig. 13. CO2 capture rate and lean solvent flowrate during the load ramp-up process.

resulting in a reduction in reboiler steam flowrate (Fig. 14). Since more stronger variations for the turbine exhaust steam. Although much con­
steam enters the LP steam turbine, the generator power ramps faster in trol effort has been made by changing the fan array speed, due to the
Fig. 12. The simulation results also illustrate that the CO2 capture level limited cooling performance of air, fluctuations still appear in the DACC
can smoothly follow the given setpoint. The fluctuations of the reboiler pressure, as can be observed in Fig. 15. Long-term deviation of the DACC
temperature are trivial due to the vast amount of solvent stored in the pressure from the design value will lead to economic losses and safety
PCC. risks. Applying a feed-forward fan array speed controller that directly
Nevertheless, the “varying capture rate” operating strategy leads to percepts the CO2 capture rate setpoint variations may help stabilize the

13
M. Zhu et al. Energy 263 (2023) 125490

Fig. 14. Reboiler temperature and reboiler inlet steam flowrate during the load ramp-up process.

Fig. 15. DACC pressure and fan array speed during the load ramp-up process.

DACC pressure. saturation pressure is 3 kPa). The operating region of the DACC-CFPP-
PCC system for power generation-carbon capture with an anti-freezing
4. Further discussion safety margin of 25 ◦ C is shown in Fig. 16.
The steam flowrate extracted from the steam turbine is forced to
4.1. Effects of the cold weather on the integrated DACC-CFPP-PCC reduce to meet the Anti-freezing Condition. As illustrated in Fig. 16,
system keeping all nine DACC fan columns on leads to no feasible operating
region for the DACC-CFPP-PCC system, which indicates that the PCC
The above analyses are all carried out under conventional ambient cannot operate. It is thus necessary to shut down some fan columns to
conditions. However, in many areas where DACC-CFPPs are deployed, improve the regulation capability of the cold end to increase the feasible
the ambient temperature reaches − 20 ◦ C or even − 30 ◦ C in winter. The region of the entire system. It can be observed from Fig. 16 that the
cold weather brings a significant impact on the cooling process of DACC, feasible operating region gets enlarged when fewer fan columns are put
which can sharply drop the DACC pressure below the reference value into operation. After installing PCC equipment, operators can only keep
(10 kPa) and cause the DACC outlet water temperature to fall below the five DACC fan columns on to maintain the same operating zone as before
freezing point, even if the fan array speed has been reduced to the when the temperature is - 30 ◦ C.
minimum value (30%).
The exhaust steam flowrate of the DACC-CFPP becomes much
smaller after integrating the PCC equipment; consequently, the DACC is 4.2. Effects of the integration schemes on the integrated DACC-CFPP and
more likely to freeze in cold weather conditions. Therefore, it is neces­ PCC system
sary to investigate the impact of cold weather on the feasible operating
zone of the DACC-CFPP-PCC plant. The selection of an integration scheme between the CFPP and PCC
The workflow shown in Fig. 8 is re-performed to identify the feasible processes is essential for the economics and flexibility of the entire
operating region by changing the ambient temperature to − 30 ◦ C. In system. Extracting steam from the turbine stage with higher pressure
addition, an Anti-freezing Condition is added to the Feasibility Condition increases the exergy loss because more feedwater is needed to cool the
that the DACC outlet temperature toutCD should be larger than 25 C (its
◦ extraction steam to a suitable temperature for solvent regeneration.
However, extracting steam from the turbine stage with lower pressure

14
M. Zhu et al. Energy 263 (2023) 125490

condensate water first heats the outlet water of the condensate pump
and then returns to the hotwell.
In all the integration schemes, the cooling water is from the outlet of
FWH1. The Feasible operating regions of the DACC-CFPP-PCC under
these integration schemes are illustrated in Fig. 18.
Extracting the steam from TC5, which is of lower pressure, Integra­
tion Scheme 3 shows the smallest operating range. The TC5 pressure
easily falls below the required level (260 kPa) in case of extensive steam
extraction and cannot provide sufficient pressure difference to drive the
steam flow. In contrast, Integration Scheme 1 has the largest operating

Fig. 16. The power-carbon feasible operating region of the DACC-CFPP-PCC

system (air temperature − 30 ◦ C, no crosswind).

may narrow the feasible operating range because the pressure at the
extraction point can fall below the required level (260 kPa) in the case of
extensive steam extraction. Therefore, the following four integration
schemes, as shown in Fig. 17, are compared in this section:
Integration Scheme 1: Steam is extracted from TC4; the reboiler
condensate water returns to the deaerator;
Integration Scheme 2: Steam is extracted from TC3; the reboiler
condensate water returns to the deaerator.
Integration Scheme 3: Steam is extracted from TC5; the reboiler
condensate water returns to the deaerator. Fig. 18. The power-carbon feasible operating region of the DACC-CFPP-PCC
Integration Scheme 4: Steam is extracted from TC4; the reboiler system under different integration schemes.

Fig. 17. The flow diagram of several integration schemes.

15
M. Zhu et al. Energy 263 (2023) 125490

region since the reboiler steam is extracted from a higher-pressure tur­ Data availability
bine chest. The adjustable power ranges of Integration Schemes 1–4 are
285-609MWe, 270-596MWe, 467.8–616MWe, and 284-607MWe, Data will be made available on request.
respectively. Integration Scheme 3 cannot afford the load changes
comparable to other Schemes, and it also fails to achieve higher CO2 Acknowledgments
capture rates.
However, given the same fuel consumption and CO2 captured The authors would like to acknowledge the National Natural Science
amount, Integration Scheme 3 has the highest generator power, indi­ Foundation of China under Grants 51576041 and 51976030, Natural
cating that the economics of the entire DACC-CFPP-PCC plant is best Science Foundation of Jiangsu Province for Outstanding Young Scholars
under this scheme. As the steam of TC5 is of lower pressure and tem­ under Grant BK20190063 and Fundamental Research Funds for the
perature, less feedwater is needed to cool the extraction steam to a Central Universities for funding this work.
suitable temperature for solvent regeneration, which results in a smaller
exergy loss. For the same reason, Integration Scheme 2 is observed to References
have the worst operating economics.
Integration Schemes 1 and 4 show similar economic performance [1] Delbeke J, Runge-Metzger A, Slingenberg Y, Werksman J. The paris agreement.
2019. p. 24–45. https://unfccc.int/process-and-meetings/the-paris-agreement/th
and operating regions. When the load is 100% BRL and the CO2 capture e-paris-agreement.
rate is 90%, the generator power of Integration Scheme 4 is only 2 MWe [2] Ritchie Hannah, Roser Max. CO2 and greenhouse gas emissions. 2020. https://
less than that under Integration Scheme 1. Although Integration Scheme ourworldindata.org/emissions-by-sector.
[3] IPCC. Global Warming of 1.5◦ C.An IPCC Special Report on the impacts of global
4 does not require the pump to transport the return water to the high- warming of 1.5◦ C above pre-industrial levels and related global greenhouse gas
pressure deaerator, it requires a heat exchanger between the FWH7 emission pathways. In: The context of strengthening the global response to the
and the condensate pump to avoid flashing in the HW. threat of clim; 2018. n.d.
[4] International Energy Agency. World energy outlook 2021. 2021.
[5] Kefford BM, Ballinger B, Schmeda-Lopez DR, Greig C, Smart S. The early retirement
5. Conclusion challenge for fossil fuel power plants in deep decarbonisation scenarios. Energy Pol
2018;119:294–306.
[6] International Energy Agency. The role of CCUS in low-carbon power systems.
Coal-fired power plants with direct air-cooling condensers can
2020. https://www.iea.org/reports/the-role-of-ccus-in-low-carbon-power-systems.
effectively reduce water consumption and thus have recently been [7] Wu X, Wang M, Liao P, Shen J, Li Y. Solvent-based post-combustion CO2 capture
widely installed in regions rich in coal but short of water. Because such a for power plants : a critical review and perspective on dynamic modelling , system
region provides better conditions for CO2 geological storage, it is of great identification , process control and flexible operation. Appl Energy 2020;257:
113941.
significance to study the interactions between the DACC-CFPP and PCC [8] Goto K, Yogo K, Higashii T. A review of efficiency penalty in a coal-fired power
to guide the safe, economical, flexible and low-carbon operation of the plant with post-combustion CO2 capture. Appl Energy 2013;111:710–20.
entire plant. This paper develops a plant-wide dynamic model for an [9] Borhani T N, Wang M. Role of solvents in CO2 capture processes: the review of
selection and design methods. Renew Sustain Energy Rev 2019;114:109299.
integrated DACC-CFPP-PCC (Direct air-cooling condensers - Coal-fired [10] Oh SY, Yun S, Kim JK. Process integration and design for maximizing energy
power plants - Post-combustion carbon capture) system. To connect efficiency of a coal-fired power plant integrated with amine-based CO2 capture
the DACC-CFPP and solvent-based PCC plant, the rigorous steam process. Appl Energy 2018;216:311–22.
[11] Liao P. The dynamic modelling and flexible operation of coal-fired power plant
turbine-feedwater heater process model is modified by considering the integrated with post-combustion carbon capture process. In: Chinese). Ph. D thesis.
dynamics of reboiler steam extraction, water spray, pressure reduction, Southeast University; 2020.
and water returning processes. The proposed model can well reflect the [12] Cifre PG, Brechtel K, Hoch S, García H, Asprion N, Hasse H, et al. Integration of a
chemical process model in a power plant modelling tool for the simulation of an
distributions of steam/water temperature, pressure, and flowrate along amine based CO2 scrubber. Fuel 2009;88:2481–8.
the steam turbine and thus is used to evaluate the performance of the [13] Sanchez Fernandez E, Sanchez del Rio M, Chalmers H, Khakharia P, Goetheer ELV,
DACC-CFPP-PCC in terms of operating reliability, economics, and flex­ Gibbins J, et al. Operational flexibility options in power plants with integrated
post-combustion capture. Int J Greenh Gas Control 2016;48:275–89.
ibility over wide operating conditions.
[14] Zhai R, Feng L, Yu H, Wang Y, Yang Y. Integration of power plants with different
The results show that: 1) Carbon capture can reduce the temperature capacities with aqueous ammonia-based CO2 Capture. Energy Fuel 2019;33:
of the IP-LP crossover by 75 ◦ C, increase the axial thrust of the IP cyl­ 9040–54.
inder by 6.78%, and increase control challenges for the DACC. To alle­ [15] Olaleye AK, Wang M, Kelsall G. Steady state simulation and exergy analysis of
supercritical coal-fired power plant with CO2 capture. Fuel 2015;151:57–72.
viate the corresponding risks, axial thrust rebalancing, avoidance of [16] Farajollahi H, Hossainpour S. Application of organic Rankine cycle in integration of
sudden and cyclic capture rate changes, and antifreeze strategies are in thermal power plant with post-combustion CO2 capture and compression. Energy
urgent need. 2) PCC itself brings 5%–10% additional exergy loss for the 2017;118:927–36.
[17] Xu C, Li X, Liu X, Li J. An integrated de-carbonization supercritical coal-fired power
DACC-CFPP, but not necessarily economic losses, depending on the plant incorporating a supplementary steam turbine , process heat recovery and a
electricity price, coal price, and carbon tax. 3) Flexible operation of PCC modified boiler structure. Appl Therm Eng 2020;178:115532.
reduces the minimum power load of the DACC-CFPP from 330MWe to [18] Dutta R, Nord LO, Bolland O. Prospects of using equilibrium-based column models
in dynamic process simulation of post-combustion CO2 capture for coal-fired power
286MWe and can increase the power ramping rate from 4.76 MWe/min plant. Fuel 2017;202:85–97.
to 8.51 MWe/min. 4) Both economy and feasibility should be balanced [19] Lawal A, Wang M, Stephenson P, Koumpouras G, Yeung H. Dynamic modelling and
when PCC is integrated into the DACC-CFPP. analysis of post-combustion CO2 chemical absorption process for coal-fired power
plants. Fuel 2010;89:2791–801.
[20] Dugas RE. Pilot plant study of carbon dioxide capture by aqueous
Credit author statement monoethanolamine. The University of Texas at Austin; 2006.
[21] Flø NE, Kvamsdal HM, Hillestad M. Dynamic simulation of post-combustion CO2
capture for flexible operation of the Brindisi pilot plant. Int J Greenh Gas Control
Mingjuan Zhu: Writing, Software, Methodology, Data curation.
2016;48:204–15.
Yudong Liu: Writing - Review & Editing, Visualization, Validation. Xiao [22] Abdul Manaf N, Cousins A, Feron P, Abbas A. Dynamic modelling, identification
Wu: Writing - Review & Editing, Conceptualization, Investigation. and preliminary control analysis of an amine-based post-combustion CO2 capture
pilot plant. J Clean Prod 2016;113:635–53.
Jiong Shen: Supervision, Funding acquisition, Project administration.
[23] Liao P, Wu X, Li Y, Wang M, Shen J, Lawal A, et al. Application of piece-wise linear
system identification to solvent-based post-combustion carbon capture. Fuel 2018;
Declaration of competing interest 234:526–37.
[24] Xi H, Wu X, Chen X, Sha P. Artificial intelligent based energy scheduling of steel
mill gas utilization system towards carbon neutrality. Appl Energy 2021;295:
The authors declare that they have no known competing financial 117069.
interests or personal relationships that could have appeared to influence
the work reported in this paper.

16
M. Zhu et al. Energy 263 (2023) 125490

[25] Lawal A, Wang M, Stephenson P, Obi O. Demonstrating full-scale post-combustion [35] Shang Y, Wang J, Liu J, Jiang D, Zhai J, Jiang S. Suitability analysis of China’s
CO 2 capture for coal-fired power plants through dynamic modelling and energy development strategy in the context of water resource management. Energy
simulation. Fuel 2012;101:115–28. 2016;96:286–93.
[26] Wellner K, Marx-Schubach T, Schmitz G. Dynamic behavior of coal-fired power [36] Li Q, Wei Y, Liu G, Jing M, Zhang M, Fei W. Feasibility of the combination of CO 2
plants with post combustion CO2 capture. Ind Eng Chem Res 2016;55:12038–45. geological storage and saline water development in sedimentary basins of China.
[27] Marx-Schubach T, Schmitz G. Modeling and simulation of the start-up process of Energy Environ 2013;37:4511–7.
coal fired power plants with post-combustion CO2 capture. Int J Greenh Gas [37] Zhu M, Wu X, Shen J, Lee K. Dynamic modeling, validation and analysis of direct
Control 2019;87:44–57. air-cooling condenser with integration to the coal-fired power plant for flexible
[28] Gardarsdóttir S, Montañés RM, Normann F, Nord LO, Johnsson F. Effects of CO2- operation. Energy Convers Manag 2021;245:114601.
absorption control strategies on the dynamic performance of a supercritical [38] Shanghai Steam Turbine Co. Ltd. Steam turbine manual. 2016.
pulverized-coal-fired power plant. Ind Eng Chem Res 2017;56:4415–30. [39] Xu M, Cao L, Song L. Structure and modal analysis of the last stage blade in steam
[29] Liao P, Li Y, Wu X, Wang M, Oko E. Flexible operation of large-scale coal-fired turbine under low volume flow conditions. Eng Fail Anal 2022;131:105820.
power plant integrated with solvent-based post-combustion CO2 capture based on [40] Cao L, Wang W, Luo H, Hu P, Wang Y, Sun L. Determination of minimum flow rate
neural network inverse control. Int J Greenh Gas Control 2020;95:102985. of low pressure cylinder of steam turbine under deep peak load regulation
[30] Wu X, Wang M, Lee KY. Flexible operation of supercritical coal-fired power plant conditions. Jixie Gongcheng Xuebao/J Mech Eng 2020;56:98–108.
integrated with solvent-based CO2 capture through collaborative predictive [41] Chaibakhsh A, Ghaffari A. Steam turbine model. Simulat Model Pract Theor 2008;
control. Energy 2020;206:118105. 16:1145–62.
[31] Lucquiaud M, Gibbins J. Effective retrofitting of post-combustion CO2 capture to [42] Moran MJ, Shapiro HN. Fundamentals of engineering thermodynamics. eighth ed.
coal-fired power plants and insensitivity of CO2 abatement costs to base plant John Wiley & Sons; 2011.
efficiency. Int J Greenh Gas Control 2011;5:427–38. [43] World bank group. State and trends of carbon pricing 2020. 2020. https://openk
[32] Sun YJ, Liu XQ, Hu LS, Tang XY. Online life estimation for steam turbine rotor. nowledge.worldbank.org/bitstream/handle/10986/33809/9781464815867.pdf?
J Loss Prev Process Ind 2013;26:272–9. sequence=4.
[33] Pan S-Y, Snyder SW, Packman AI, Lin YJ, Chiang P-C. Cooling water use in [44] Statista. Household electricity prices worldwide in September 2021. 2022. https
thermoelectric power generation and its associated challenges for addressing ://www.statista.com/statistics/263492/electricity-prices-in-selected-countries/.
water-energy nexus. Water-Energy Nexus 2018;1:26–41. [45] IEA. Coal 2020. 2020. https://www.iea.org/reports/coal-2020.
[34] China Electricity Council. Annual compilation of statistics of the power industry [46] Stichlmair JG, Klein H, Rehfeldt S. Distillation: principles and practice. second ed.
2013. 2013. Beijing, China. Hoboken: New Jersy; 1998.

17