Energy 35 (2010) 5391e5399

Contents lists available at ScienceDirect

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

Operating pressure dependence of the pressurized oxy-fuel combustion

power cycle
Jongsup Hong a, Randall Field b, Marco Gazzino c, Ahmed F. Ghoniem a, *
a
Department of Mechanical Engineering, Massachusetts Institute of Technology, 77 Massachusetts Avenue, Cambridge, MA 02139-4307, USA
b
MIT Energy Initiative, Massachusetts Institute of Technology, 77 Massachusetts Avenue, Cambridge, MA 02139-4307, USA
c
ENEL Ingegneria e Innovazione S.p.A., 56122, Pisa, Via Andrea Pisano 120, Italy

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

Article history: Oxy-fuel combustion technology is an attractive option for capturing carbon dioxide (CO2) in power
Received 18 October 2009 generation systems utilizing hydrocarbon fuels. However, conventional atmospheric oxy-fuel combus-
Received in revised form tion systems require substantial parasitic energy in the compression step within the air separation unit
9 July 2010
(ASU), the flue gas recirculation system and the carbon dioxide purification and compression unit (CPU).
Accepted 12 July 2010
Available online 16 August 2010
Moreover, a large amount of flue gas latent enthalpy, which has high water concentration, is wasted. Both
lower the overall cycle efficiency. Pressurized oxy-fuel combustion power cycles have been investigated
as alternatives. Our previous study showed the importance of operating pressure for these cycles. In this
Keywords:
Oxy-fuel combustion
paper, as the extended work of our previous study, we perform a pressure sensitivity analysis to
Power cycle analysis determine the optimal combustor operating pressure for the pressurized oxy-fuel combustion power
CO2 capture and sequestration cycle. We calculate the energy requirements of the ASU and the CPU, which vary in opposite directions as
the combustor operating pressure is increased. We also determine the pressure dependence of the water-
condensing thermal energy recovery and its relation to the gross power output. The paper presents
a detailed study on the variation of the thermal energy recovery rate, the overall compression power
demand, the gross power output and the overall net efficiency.
Ó 2010 Elsevier Ltd. All rights reserved.

1. Introduction Current research progress suggests that pressurized oxy-coal

combustion power cycles utilize the higher heating value of a fuel,
Coal has been a primary energy source of electricity generation produce more gross power and have higher net efficiencies than
for many decades because of its availability and low costs. conventional atmospheric oxy-coal combustion power systems.
According to estimates by Energy Information Administration, coal Based on a patented pressurized oxy-coal combustion process and
accounts for over 50% of the electricity generated in the United technology, ISOTHERMÒ, ENEL has conducted a number of experi-
States [1]. Coal, which is cheap and abundant, provides primary mental studies that show the advantages arising from pressurized
energy at a price that is the half of the cost of oil and natural gas [2]. conditions [7e9]. It is shown that the pressurized systems have the
Many researchers and policy makers agree that coal will maintain increased heat transfer rates in the Heat Recovery Steam Generator
its important role in power generation in the near future [3e5]. (HRSG), possibilities to burn cheap coals and the reduced size of
Coal combustion emits large amounts of carbon dioxide, components. These benefits are demonstrated by using a refractory
believed to be a significant contributor to global warming. As lined combustor at 4 bars on a 5-MWth system [10e14]. CANMET
concerns over greenhouse gas emissions grow, scientists and and ThermoEnergy have examined the pressurized oxy-coal
investigators have endeavored to find a way to decrease carbon combustion systems as well, in terms of technical and economic
dioxide emissions from coal-fired power plants. Carbon dioxide studies. Their studies show that the high-pressure oxy-combustion
capture and sequestration (CCS) technologies should play a critical systems have benefits of low efficiency penalties and reduction in
role in this regard, enabling coal to continue to satisfy an increasing capital costs, compared to the conventional atmospheric oxy-fuel
fraction of worldwide energy needs [6]. Oxy-fuel combustion power cycles [15e21].
system is one of the promising CCS options. In our previous study, an oxy-fuel combustion power cycle
utilizing a pressurized coal combustor [22] was analyzed, starting
* Corresponding author. Tel.: þ1 617 253 2295; fax: þ1 617 253 5981. with the process design proposed by Gazzino et al. [13] for the
E-mail address: [email protected] (A.F. Ghoniem). integration of pressurized oxy-combustion with Rankine cycle. In

0360-5442/$ e see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.energy.2010.07.016
5392 J. Hong et al. / Energy 35 (2010) 5391e5399

the pressurized oxy-fuel system, oxygen is pre-compressed in the paper [22]. Aspen PropertiesÒ provides thermodynamic and
ASU where the mass flow rate is smaller than that of the flue gases transport properties used to estimate the pressure drop through
into the compression unit (CPU). By pre-compressing the gas the flue gas recirculation path, as explained in Section 2.2.1
stream in the ASU, the combustion flue gases are at the high
pressure and the compression work duty of the CPU, which 2.1. Base case and design variables
compresses the flue gases from the combustor operating pressure
to 110 bars, decreases. Imposing more compression work duty on The pressure sensitivity analysis was performed using the base
the ASU and less on the CPU, the overall system lowers the parasitic case we developed in our previous study [13,22], which is revised to
power demand. Moreover, the energy penalty associated with the adopt the best available technologies especially in a steam cycle
oxygen compression work required to feed oxygen into the pres- and the CPU. The base case plant consists of five primary units, the
surized combustor is partially offset by removing the pre-heater ASU, a pressurized coal combustor, a steam generation unit,
from the oxygen compression line, hence avoiding the thermal a power island and the CPU. Each unit represents commercially
energy penalty related to oxygen pre-heating. In addition, the available technologies or processes at an advanced development
elevated dew point and higher available latent enthalpy in the flue stage. Fig. 1 shows the overall process diagram for the base case
gases lead to higher thermal energy recovery from the flue gases. cycle for the pressurized oxy-fuel combustion system. As shown in
The available latent enthalpy is defined as the recoverable amount Fig. 1, after being compressed by the first feedwater pump, the
of the latent enthalpy of the water in the flue gases (hot stream) condensate recuperates most of the latent enthalpy in the flue
based on the inlet and outlet temperatures of the feedwater (cold gases while flowing the acid condenser. The condensate stream is
stream) across the heat exchanger. The increased water-condensing heated further by cooling the combustor walls. The feedwater
thermal energy recovery enables the system to produce more gross heating system and the HRSG receive the condensate stream, which
power by eliminating the steam bleeding from the low-pressure is compressed to the supercritical state by the second feedwater
steam turbines. Compared to a conventional atmospheric oxy-fuel pump, and generates supercritical steam. On the gas side which is
combustion system, the pressurized system was shown to be more represented by the red lines, the oxygen stream is mixed with the
efficient, taking advantage of lower overall compression work re-circulated flue gases before entering the pressurized oxy-coal
demand and higher gross power output. combustor. The combustion flue gases are cooled down by the
Moreover, our study showed that the fan compression work has secondary recycled flue gases. Passing through the HRSG and
a significant impact on the overall performance. Most oxy-fuel the acid condenser, the flue gases transfer thermal energy to the
combustion systems have implemented recirculation of the flue feedwater. At the end, the flue gas stream is purified and
gases to decrease the combustion temperature to a reasonable compressed by the CPU.
level. Thus, a significant amount of the flue gases is re-circulated The proposed system has two important updates in the power
within the system, requiring a substantial amount of energy to island and the CPU, compared to the pressurized oxy-fuel power
compensate for the pressure drop across the steam generation unit cycle analyzed in our previous study. Because the double reheat
and the recirculation pipe. This feature is even more important for steam cycle is not economically beneficial, it is believed to be a non-
the combustion technology patented by ITEA [7e9], which adopts viable option in conventional coal-fired power plants. Thus, we
two flue gas recycling lines employed to control the combustion updated the previous base case by replacing it with a single reheat
temperature and temperature at the HRSG inlet. steam cycle. We also found that the turbines used in the previous
Based on the finding from the previous study, we have study represent 20-year-old technologies and have lower isen-
concluded that the combustor operating pressure has a strong tropic efficiencies. In this regard, the power island is modified to
impact on the overall performance of the oxy-fuel system. In this utilize the best available steam turbines based on a reference ENEL
paper, we perform a pressure sensitivity analysis to find the optimal power plant that has been recently commissioned. Data provided
combustor operating pressure. We calculate the pressure depen- by ENEL also accounts for the pressure drop along the steam
dence of the thermal energy recovery rate, the overall parasitic bleeding line to the feedwater heaters. In addition, for CO2
power demand, the gross power output and the overall efficiency. production or for Enhanced Oil Recovery (EOR) applications, it is
The analysis enables us to determine which parameters are common to pressurize CO2 by a pump above 80 bars, which is
important in operating the pressurized system and to examine how beyond the CO2 critical point [23]. The previous CPU utilizes only
to improve the overall performance. The methodology used in the compressors in the CO2 compression steps. Therefore, the updated
study is explained in Section 2, and the detailed results are dis- CPU includes a pump as well as compressors to pressurize the CO2
cussed in Section 3. Section 4 includes conclusions. stream up to the delivery pressure of 110 bars. Other detailed
information on the overall processes and coal analysis data can be
2. Methodology found in our previous paper [22].
Based on the base case processes, we fixed a set of important
A 300 MWe coal-fired power plant, which is based on the fixed design variables to focus on system’s operating pressure dependence.
coal flow rate of 30 kg/s corresponding to 874.6 MWth higher Table 1 shows the design variables used in the pressure sensitivity
heating value (HHV) or 839.1 MWth lower heating value (LHV), is analysis. While the gas side variables describe the characteristics
used as the base case for the pressure sensitivity analysis. In the of the available ASU, the combustor and the HRSG, parameters in
previous study, we proposed the pressurized oxy-fuel combustion the steam side represent the supercritical Rankine cycle.
power cycle and discussed its characteristics. The cycle has been
modified to include the best available technologies; all important 2.2. Pressure drop estimation method
design variables are shown in Section 2.1. The combustor operating
pressure was controlled by changing the oxygen delivery pressure As explained in the previous section, the fan compression works
to the combustor from 1.238 bars to 30 bars. The proposed for the flue gas recirculation system accounts for a significant
approach will be explained and discussed in the following sections.
Two commercial simulation packages, ThermoflexÒ and Aspen
PlusÒ, were used to conduct the study. An integration method 1
ThermoflexÒ, Aspen PlusÒ and Aspen PropertiesÒ are registered trademarks of
between these two simulation tools is mentioned in the previous Thermoflow LTD and Aspen Technology, Inc., respectively.
 J. Hong et al. / Energy 35 (2010) 5391e5399 5393

Fig. 1. Overall process layout for the pressurized oxy-fuel system (modified from Ref. [10]).

amount of the parasitic power demand in the pressurized oxy-fuel L V2

DPpipe ¼ rfpipe ; (1b)
system. In this system, recirculation of 70e80% of the flue gases d 2
must be used in order to reduce the combustion temperature [24]. where fpipe ¼ FðRed ; 3=dÞ.
Because a considerable pressure drop occurs across the steam
generation unit and the flue gas recirculation pipe, a fan is required Using these correlations, we are able to evaluate the pressure
to re-circulate the flue gases. drop across the HRSG and the recirculation pipe, respectively. The
As expected, the pressure drop and the corresponding fan friction factors, f, are found from the following correlations [27,28],
compression work depend on the combustor operating pressure. To
estimate the operating pressure dependence of the fan compres- 0:181  104 0:792  108 0:165  1013
fHRSG ¼ c 0:162 þ þ 
sion work, we have to first estimate the pressure drop. From the Re Re2 Re3
pressure drop, DP, correlations [25,26] we know that: !
0:872  1016
þ ð2aÞ
r0 Vmax
2
Re4
DPHRSG ¼ Nc fHRSG ; (1a)
2
where fHRSG ¼ FðReD ; ST =DÞ ( " !#)2
ð23=dÞ 5:02 ð23=dÞ 13
fpipe ¼  2:0 log  log þ (2b)
7:4 Red 7:4 Red
Table 1
Design variables of the pressurized oxy-fuel system.
The heat exchanger and the recirculation pipe design assump-
Design variables Value tions include
Gas sidea
Oxygen purity 95% (mol%)
 The correction factor: c ¼ 1.076
Oxygen in the flue gas 3% (mol%)
Oxygen delivery temperature 200  C  The HRSG tube diameter: D ¼ 50 mm
CO2 compression pressure 110 bars  The HRSG tube length: Ltube ¼ 18.29 m
O2/CO2 compressor polytropic efficiency 85%  The HRSG tube transverse pitch: ST ¼ 100 mm
CO2 pump isentropic efficiency 75%  The HRSG tube longitudinal pitch: SL ¼ 100 mm
Steam side  The recirculation pipe length: L (explaining the pressure drop
HPT inlet pressure/temperature 250 bars/600  C across the valves and the curves within the recirculation path)
Reheat pressure/temperature 56 bars/610  C
 The recirculation pipe diameter: d (based on the volumetric
Turbine isentropic efficiency 85 w 96%
Deaerator pressure 10 bars flow rate and the velocity V)
Condenser pressure 0.0422 bars  The roughness: 3 ¼ 0.046 mm (commercial steel)
a
The combustion temperature and the inlet and outlet temperatures of the HRSG
 The maximum velocity: Vmax y 15.5 m/s
are set according to the process conditions set by ITEA and ENEL for their patented  The density of flue gases: r0 and r (based on the given pressure
combustion technology. and the temperature)
5394 J. Hong et al. / Energy 35 (2010) 5391e5399

 The inlet and outlet temperatures of each heat exchanger: section. Under these cold stream temperature constraints,
TH,in ¼ the inlet temperature of the flue gas stream; increasing the dew point enables the latent enthalpy recovery at
TH,out ¼ the outlet temperature of the flue gas stream; higher temperatures. Consequently, we can recover more latent
TC,in ¼ the inlet temperature of the feedwater stream enthalpy from the flue gases at higher temperatures with
increasing the operating pressure. Fig. 2 shows the dependence of
To calculate DPHRSG, the following three equations [29e31] are the saturation temperature of the water (the blue line) and the
used to evaluate the Nusselt number,NuD ; the number of transfer available latent enthalpy (the red line) on the flue gas pressure. As
units, Ntu; and the number of tube rows, N, mentioned before, the available latent enthalpy is defined as the
recoverable amount of the latent enthalpy of the water in the flue
hD
NuD ¼ gases (hot stream) based on the inlet and outlet temperatures of
k 8 9
" the feedwater (cold stream) across the acid condenser.
   1=2 #=
2D <
1=2
0:62ReD Pr1=3 ReD The available latent enthalpy increases with the operating
¼ 1þ 0:3þ !2=3 #1=4 1þ 282;000
3SL : ; pressure and is saturated above 10 bars, as shown in Fig. 2. This is
0:4
½1þ because we recover almost all HHV of the fuel as we increase the
Pr combustor operating pressure. As shown in Fig. 3, the fraction of
(3) the water whose latent enthalpy is recovered through the acid
condenser rises with the operating pressure. Above 10 bars, we
where, h ¼ heat transfer coefficient, k ¼ thermal conductivity. reach nearly 100% latent enthalpy recovery rate.
 1=2 E  1 Because of a large amount of water in the oxy-fuel combustion
Ntu ¼  1 þ R2 ln (4) flue gases, 48% (molar) in our base case, the increase in the satu-
Eþ1 ration temperature of the water enables the recovery of a signifi-
cant amount of the available latent enthalpy from the flue gases.
where, 3 ¼ Q_ =Q_ max ¼ m _ H cP;H ðTH;in  TH;out Þ=Cmin ðTH;in  TC;in Þ ¼
As shown in Fig. 4, the total thermal energy (the blue line) recov-
_
mC cP;C ðTC;out  TC;in Þ=Cmin ðTH;in  TC;in Þ, E ¼ ½2ðF  RÞ=ðF  1Þ
ered from the flue gases increases by nearly 6%, as the combustor
ð1 þ RÞ=ð1 þ R2 Þ1=2 ; F ¼ ð3R  1=3  1Þ1=n , R ¼ Cmin =Cmax ; operating pressure is raised from 1.238 bars to 30 bars.
_ _
Cmin ¼ minfmH cP;H ; mC cP;C g; Cmax ¼ max fm _ H cP;H ; m
_ C cP;C g. The considerable increase in the thermal energy recovery lowers
the regeneration heat duty from the turbines. While the conven-
Ntu Cmin ST N C S tional atmospheric oxy-fuel system cannot recover the HHV of fuel
N¼ ¼ 8 tu min T 9
hc pDAc >
< S >
= through the steam generation unit, the pressurized oxy-fuel power
_H ST =2 cycle is able to recuperate most of the HHV of the fuel because of the
hc pDr m max T
; h   i
0 Vmax >
:ST D S2 þ ST 2 1=2 >
D;
increased dew point as discussed above. As a result, the former
L 2 needs a higher portion of the steam from steam turbines. As we
(5) recover more thermal energy from the flue gases, the regeneration
heat duty and the corresponding steam bleeding from the low
The number of tube rows N is calculated by combining Eqs. (3)e(5).
pressure and the intermediate pressure steam turbines decrease.
Adding (1a) and (1b), we estimate the overall pressure drops
Fig. 4 shows the variation in the overall steam bleeding (the red
across the HRSG and the recirculation pipe. The result is shown in
line) with increase in the combustor operating pressure.
Section 3.3.2
As shown in Fig. 5, the gross power output increases with
increasing the operating pressure, following the thermal energy
3. Pressure dependence of the system performance recovery curve in Fig. 4.

The performance of the pressurized oxy-fuel power system

depends on the parasitic power requirement, which includes the 3.2. Deaerator feedwater inlet temperature, acid condenser pinch
compression work of the ASU, the CPU and the recirculation fan, point and feedwater mass flow rate
and the thermal energy recovery. Thus, the pressure dependence of
these quantities must be carefully investigated. The fact that the deaerator feedwater inlet temperature should
be less than the saturation temperature determines the maximum
exit temperature of the feedwater out of the acid condenser.
3.1. Thermal energy recovery and gross power output

As the combustor operating pressure rises, so does the thermal

energy recovery from the flue gases that is used to generate steam.
At higher pressures, the saturation temperature of the water in the
flue gases rises as well. In this case, we can implement more
effective thermal integration between a hot stream (flue gases) and
a cold stream (feedwater). As shown in Fig. 1, the inlet temperature
of the cold stream to the acid condenser, 30  C, is determined by
the steam condenser, and the maximum exit temperature of
the feedwater out of the acid condenser is fixed by the deaerator
operating conditions which will be discussed in the following

2
Ntu does not change significantly with the operating pressure because the inlet
and outlet temperatures of the HRSG are fixed and the specific heat capacity of the
streams remains nearly constant as the pressure increases. However, as the cross- Fig. 2. Saturation temperature of the steam (the blue line) and the available latent
sectional area (the width) of the HRSG decreases, the number of tube rows enthalpy (the red line). (For interpretation of the references to colour in this figure
N needed to achieve the required heat transfer surface area becomes larger. legend, the reader is referred to the web version of this article.)
 J. Hong et al. / Energy 35 (2010) 5391e5399 5395

Fig. 5. Gross power output.

Fig. 3. The fraction of the water whose latent enthalpy is recovered through the acid
condenser.

3.3. Pressure drop and fan compression work

As shown in Fig. 1, the condensate leaving the steam condenser
flows into the acid condenser to recover the latent enthalpy from The pressure drop and the fan compression work play important
the flue gases. Next, the heated water stream enters the deaerator roles in the overall performance of the pressurized oxy-fuel power
after recuperating more thermal energy by cooling the combustor system. Because the flue gas recirculation is a unique characteristic
walls. Therefore, the deaerator feedwater inlet temperature is of the oxy-fuel system, it is essential to examine the related power
strongly related to the amount of thermal energy recovery through consumption in the thermodynamic cycle analysis.
the acid condenser. However, because the deaerator controls the As explained in Section 2.2, the pressure drop is strongly
feedwater inlet temperature so that it does not allow the feedwater dependent on the operating pressure. To achieve higher heat
to become steam before the second feedwater pump, we fix the transfer coefficients, we kept the maximum velocity (Vmax)
maximum temperature we can reach at the deaerator operating constant within the HRSG and had denser flue gases at higher
pressure (10 bars in our study). Fig. 6(a) shows that the deaerator operating pressures. Given that the mass flow rate of the flue gases
feedwater inlet temperature reaches the maximum value as the was held constant, maintaining the same maximum velocity was
process operating pressure increases. Consequently, the deaerator possible by reducing the width of the HRSG and, hence its cross-
operating conditions have an impact on the amount of thermal sectional area. In addition, the diameter of the flue gas recirculation
energy recovery from the flue gases through the acid condenser. pipe was determined so as not to have a huge pipe when increasing
Fig. 6(b) shows another important parameter in the pressurized
oxy-fuel system. As we increase the combustor operating pressure,
the pinch point temperature within the acid condenser varies, as
shown in Fig. 6(b). While the pinch point occurs at the dew point of
the water at low pressures, it moves down to the exit temperature
of the flue gas stream above 8 bars. Because the condensate
temperature (inlet temperature of the cold stream to the acid
condenser) is fixed by the condenser, the flue gas temperature
cannot decrease any more if the pinch point occurs at the hot
stream exit, which is at the condensate inlet. Therefore, the acid
condenser pinch point temperature also plays an important role in
the thermal energy recovery from the flue gases.
According to the change in the deaerator feedwater inlet
temperature and the acid condenser pinch point, the increase in the
mass flow rate of the feedwater is also limited, as shown in Fig. 7.

Fig. 4. The variation in the thermal energy recovery (the blue line) and the overall
steam bleeding to feedwater heaters (the red line). (For interpretation of the references Fig. 6. (a) Deaerator feedwater inlet temperature and (b) acid condenser pinch point
to colour in this figure legend, the reader is referred to the web version of this article.) temperature.
5396 J. Hong et al. / Energy 35 (2010) 5391e5399

Table 2
Reference flue gas velocities within the flue gas recirculation pipe.

1.238 bars 10 bars

Recirculation to the combustor (m/s) 25 4
Recirculation to the HRSG (m/s) 30 14

pipe pressure drop varies, as shown in Fig. 8(a). This variation leads
to the pressure ratio across the flue gas recirculation pipe, as shown
in Fig. 9(a). Combining the pressure ratios across the HRSG and the
recirculation pipe, we have the overall pressure ratio across the fan,
as shown in Fig. 9(b).
Because the fan compression work is a strong function of the
pressure ratio, it follows the variation in the pressure ratio over the
Fig. 7. Deaerator feedwater inlet mass flow rate.
fan and has the minimum value at 10 bars, as shown in Fig. 10.

3.4. ASU and CPU compression work

the flue gas velocity. Because the volumetric flow rate grows as the
combustor operating pressure is decreased from 10 bars to The pressurized oxy-fuel power cycle needs more compression
1.238 bars, we had to linearly increase the flue gas velocity. On the work in the ASU than that of the atmospheric oxy-fuel system.
other hand, above 10 bars, we kept the flue gas velocity constant Because the operating pressure is governed by the oxygen delivery
because we have to keep the pipe diameter within a reasonable pressure to the combustor, the ASU requires more compression
value as the volumetric flow rate decreases. Table 2 shows the work as the operating pressure increases. As shown in Fig. 11, the
reference velocities used in the recirculation pipe pressure drop ASU power consumption (the blue line) grows significantly by
estimation. As the pressure changes from 1.238 bars to 30 bars the nearly 44%, as the combustor pressure is increased from 1.238 bars
diameter of the flue gas recirculation pipe is reduced as the volu- to 30 bars. The ASU is the biggest power consumer among the
metric flow rate and the reference velocity are varied.3 Throughout equipment implemented in the pressurized oxy-fuel system. At
the range of operating pressures, the flow in the pipe remains higher operating pressures, it is important to control the ASU
turbulent. Moreover, the flue gas recirculation pipe has the valves power consumption to improve the overall performance of the
and the curves within the recirculation path. This accounts for system.
a large amount of the pressure drop across the recirculation pipe.
Fig. 8(a) shows the pressure drop across the HRSG (the blue line)
and the pressure drop across the flue gas recirculation pipe (the red
line). The flue gas recirculation pipe pressure drop is strongly
dependent on the flue gas density, the diameter of the flue gas
recirculation pipe, and the square of the flue gas velocity, see the
correlation (1b). As the operating pressure increases, the flue gas
density becomes larger, and the diameter of the flue gas recircu-
lation pipe decreases as explained above. On the other hand, based
on the reference velocities given in Table 2, the flue gas velocity is
linearly reduced at the low-pressure range and is kept constant
above 10 bars. These changes lead to the flue gas recirculation pipe
pressure drop as shown in Fig. 8(a). As a result, a higher pressure
drop is encountered as the operating pressure increases, as shown
in Fig. 8(b).
Based on the pressure drop, the pressure ratios across the HRSG
(the blue line) and the recirculation pipe (the red line) vary, as
shown in Fig. 9(a). Pressure ratios are defined as the ratio of the
inlet pressure to the outlet pressure. While the HRSG pressure ratio
rises continuously like the HRSG pressure drop, the pressure ratio
across the flue gas recirculation pipe decreases below 10 bars and
increases above 10 bars. Because the recirculation pipe pressure
drop is governed by the pipe diameter, as shown in the correlation
(1b), which is controlled by the flue gas velocity, the recirculation

3
Given the constant mass flow rate of the flue gases, the volumetric flow rate is
governed by the operating pressure or density. If we hold the diameter of the flue
gas recirculation pipe constant at: (a) the diameter at low P: the flue gas velocity
reaches the order of 1 m/s at high pressures, and it is not appropriate to operate at
this low velocity. (b) The diameter at high P: the flue gas velocity increases
significantly as the operating pressure is lowered to 1.238 bars. The magnitude
of this high velocity is not reasonable. To avoid these extreme cases and operate Fig. 8. (a) The HRSG pressure drop (the blue line) and the flue gas recirculation pipe
the flue gas recirculation system at a reasonable flue gas velocity, we changed pressure drop (the red line), (b) the overall pressure drop. (For interpretation of the
the diameter of the flue gas recirculation pipe, maintaining it within references to colour in this figure legend, the reader is referred to the web version of
a reasonable range. this article.)
 J. Hong et al. / Energy 35 (2010) 5391e5399 5397

Fig. 11. Compression work demand of the ASU and the CPU.

gases from the operating pressure up to 110 bars. Therefore, the

increase in the operating pressure lowers the pressure ratio across
the CPU, requiring less auxiliary compression work (the red line), as
shown in Fig. 11.

3.5. Overall compression work

Based on the results in Sections 3.3 and 3.4, we now define the
pressure dependence of the overall compression work demand in
the pressurized oxy-fuel system. Fig. 12 shows the variation in the
overall compression work as the operating pressure increases. The
total parasitic power demand follows the trend of the fan
compression work shown in Fig. 10. When the ASU compression
work is combined with that of the CPU, their sum does not vary
Fig. 9. (a) The pressure ratio across the HRSG (the blue line) and the pressure ratio much with the operating pressure. As a result, the overall
across the flue gas recirculation pipe (the red line), (b) the pressure ratio across the fan. compression work which includes the power consumptions of the
(For interpretation of the references to colour in this figure legend, the reader is ASU, the CPU and the recirculation fan is mostly dependent on the
referred to the web version of this article.) fan compression work. We conclude that the overall parasitic
However, the larger oxygen compression work required to power demand of the pressurized oxy-fuel system is a strong
operate the combustor at higher pressures is partially offset by function of the flue gas recirculation power requirement. Again, the
eliminating the oxygen pre-heater from the oxygen compression total parasitic power demand has the minimum at 10 bars as the
line. In conventional atmospheric combustion systems, the gaseous fan compression work does.
oxygen is pre-heated by utilizing high temperature thermal energy The variation in the overall parasitic work demand points out
sources such as steam or flue gases. The oxygen pre-heating steps that, in order to improve the performance of the pressurized oxy-
result in thermal energy penalties and lower the overall efficiency. fuel system, we should focus on the pressure drop within the flue
In contrast, the pressurized oxy-fuel power cycle reduces or elim- gas recirculation path.
inates the thermal duty of oxygen pre-heating while using oxygen
compression. 3.6. Net efficiency
On the other hand, because the flue gases are already at high
pressures, the compression work in the CPU decreases as the oper- The pressure sensitivity analysis provides the optimal pres-
ating pressure increases. The CPU compresses the CO2-concentrated sure, as shown in Fig. 13. Based on the parasitic power

Fig. 10. Fan compression work demand. Fig. 12. Overall parasitic compression work demand.
5398 J. Hong et al. / Energy 35 (2010) 5391e5399

PropertiesÒ. Hussam Zebian’s support for the evaluation of the

pressure drop estimates is acknowledged.

References

[1] Energy Information Administration. Annual Energy Outlook 2008, report

no. DOE/EIA-0383 (2008). Washington, D.C.: U.S. Department of Energy.
Available from: http://www.eia.doe.gov/oiaf/aeo/pdf/0383%282008%29.pdf;
2008.
[2] MIT. The future of coal, in an interdisciplinary MIT study. Cambridge, MA:
MIT; 2007.
[3] Saha PC. Sustainable energy development: a challenge for Asia and the Pacific
region in the 21st century. Energy Policy 2003;31(11):1051e9.
[4] Zhou F, Zhou D. Study on long-term energy development strategies of China.
Beijing, China: China Planning Press; 1999.
[5] International Energy Agency. Roadmapping coal’s future. Paris, France: OECD/
IEA Coal Industry Advisory Board. Available from: http://www.iea.org/papers/
2005/roadmapping.pdf; 2005.
Fig. 13. Overall net efficiency based on HHV (the blue line) and LHV (the red line). (For
[6] Gielen D. The future role of CO2 capture and storage: results of the IEA-ETP
interpretation of the references to colour in this figure legend, the reader is referred to
model, report no. EET/2003/04. Paris, France: International Energy Agency.
the web version of this article.)
Available from: http://www.iea.org/papers/2003/eet04.pdf; 2003.
[7] Malavasi M, Rossetti E. High-efficiency combustors with reduced
requirement and the thermal energy recovery, the overall net environmental impact and processes for power generation derivable
efficiency varies as the operating pressure increases. At low therefrom. International patent, international publication no. WO 2005/
pressures, gross power output, shown in Fig. 5, increases and the 108867; 2005.
[8] Rossetti E, Malavasi M. Method and plant for the treatment of materials, in
overall compression work, shown in Fig. 12, decreases, the net particular waste materials and refuse. International patent, international
efficiency grows by 3% point when the combustor pressure is publication no. WO 2004/094904; 2004.
raised from 1.238 bars to 10 bars. On the other hand, at high [9] Malavasi M, Di Salvia G, Rossetti E. Combustion process. International patent,
international publication no. WO 2009/071239; 2009.
pressures, the efficiency decreases slowly as the parasitic [10] Benelli G, Malavasi M, Girardi G. Oxy-coal combustion process suitable for
compression power dominates the overall performance. As future and more efficient zero emission power plants. PowerGen Europe
a result, the maximum efficiency is achieved in the vicinity of conference and exhibition. Madrid, Spain; 2007.
[11] Benelli G, Girardi G, Malavasi M, Saponaro A. ISOTHERM: a new oxy-
the 10-bar operating pressure. combustion process to match the zero emission challenge in power genera-
tion. The 7th high temperature air combustion and gasification international
symposium. Phuket, Thailand; 2008.
4. Conclusions [12] Benelli G, Cumbo D, Gazzino M, Morgani E. Pressurized oxy-combustion of
coal with flue gas recirculation: pilot scale demonstration. PowerGen Europe
conference and exhibition. Milan, Italy; 2008.
The combustor operating pressure has a critical impact on the
[13] Gazzino M, Benelli G. Pressurized oxy-coal combustion Rankine-cycle for
overall performance of the pressurized oxy-fuel combustion power future zero emission power plants: process design and energy analysis.
cycle. With increasing the operating pressure, the thermal energy The 2nd international conference on energy sustainability. Jacksonville, FL,
recovery from the flue gases grows as we recuperate more available USA; 2008.
[14] Gazzino M, Riccio G, Rossi N, Benelli G. Pressurized oxy-coal combustion
latent enthalpy at higher dew point. Increasing the energy recovery Rankine-cycle for future zero emission power plants: technological issues.
enables the pressurized oxy-fuel system to save some of the steam The third international conference on energy sustainability. San Francisco, CA,
bleeding needed in the regeneration steps, and it subsequently USA; 2009.
[15] Zheng L, Pomalis R, Clements B. Technical feasibility study of TIPS process and
produces more gross power. In addition, the fan compression work comparison with other CO2 capture power generation processes. The 32nd
varies significantly with the operating pressure as the overall pressure international technical conference on coal utilization and fuel systems.
drop across the steam generation unit and the recirculation pipe Clearwater, FL, USA; 2007.
[16] Pomalis R, Zheng L, Clements B. ThermoEnergy integrated power system
increase. While the ASU compression consumes more energy at economics. The 32nd international technical conference on coal utilization
higher pressures, the CPU requires less auxiliary energy. Conse- and fuel systems. Clearwater, FL, USA; 2007.
quently, the overall parasitic compression work demand is strongly [17] Zheng C, Zheng L, Pomalis R, Turner R, Clements B. Conceptual design and
experimental study overview: flue gas treatment and CO2 recovery experi-
dependent on the recirculation fan compression work. The maximum
mental system for high-pressure oxygen fired coal combustion. The 33rd
efficiency can be achieved in the vicinity of the 10-bar operating international technical conference on coal utilization and fuel systems.
pressure. Given that the high-pressure system raises the flue gas Clearwater, FL, USA; 2008.
[18] Fassbender A. Pressurized oxy-fuel combustion for multi-pollutant capture.
density and reduces equipment size, the best system from an
The 30th international technical conference on coal utilization and fuel
economic point of view might be found at higher operating pressures. systems. Clearwater, FL, USA; 2005.
The overall performance of the pressurized oxy-fuel combustion [19] Fassbender A. Simplification of carbon capture power plants using pressurized
power cycle is a strong function of the thermal energy recovery oxyfuel. The 32nd international technical conference on coal utilization and
fuel systems. Clearwater, FL, USA; 2007.
from the flue gases and the energy requirement in the flue gas [20] Fassbender A, Tao L, Henry R. Physical properties and liquid vapor equilibrium
recirculation system. As the operating pressure increases, the of pressurized CO2 rich gases from pressurized oxy-fuel combustion of coal.
variation in the net efficiency shows that the overall performance is The 33rd international technical conference on coal utilization and fuel
systems. Clearwater, FL, USA; 2008.
sensitive to the change in these two parameters. Therefore, it is [21] Henry R. Modeling of pressurized oxy-fuel for recovery of latent heat and
crucial to control the flue gas recirculation system and thermal carbon capture using usibelli coal. The 32nd international technical
energy recovery rate in order to improve the overall performance of conference on coal utilization and fuel systems. Clearwater, FL, USA;
2007.
the pressurized oxy-fuel combustion power cycle with increasing [22] Hong J, Chaudhry G, Brisson JG, Field R, Gazzino M, Ghoniem AF. Analysis of
the operating pressure. oxy-fuel combustion power cycle utilizing a pressurized coal combustor.
Energy 2009;34(9):1332e40.
[23] Fradet A, Saysset S, Odru P, Broutin P, Ruer J, Bonnissel M. Technical and
Acknowledgements economic assessment of CO2 transportation for CCS purposes. Pipeline Engi-
neering 2007;3rd quarter:173e9.
[24] Buhre BJP, Elliott LK, Sheng CD, Gupta RP, Wall TF. Oxy-fuel combustion
This research is financially supported by ENEL. Thermoflow and technology for coal-fired power generation. Progress in Energy and
Aspen Technology provided ThermoflexÒ, Aspen PlusÒ and Aspen Combustion Science 2005;31(4):283e307.
 J. Hong et al. / Energy 35 (2010) 5391e5399 5399

[25] Zukauskas A, Ulinskas R. Efficiency parameters for heat transfer in tube banks. [29] Churchill SW, Bernstein M. A correlating equation for forced convection from
Heat Transfer Engineering 1985;6(2):19e25. gases and liquids to a circular cylinder in crossflow. Heat Transfer 1977;99:300e6.
[26] White FM. Fluid mechanics. 5th ed. New York: McGraw-Hill; 2003. [30] Shah RK, Mueller AC. Heat exchanger basic thermal design methods. 2nd ed.
[27] Hewitt GF. Heat exchanger design handbook. New York/Wallingford (UK): New York: McGraw-Hill; 1985.
Begell House Inc; 1998. [31] Gnielinski V, Zukauskas A, Skrinska A. Banks of plain and finned tubes, in
[28] Zigrang DJ, Sylvester ND. Explicit approximations to the solution of Cole- hemisphere heat exchanger design handbook. Washington, DC: Begell House;
brook’s friction factor equation. AIChE 1982;28:514e5. 1990.