Energy 61 (2013) 179e195

Contents lists available at ScienceDirect

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

Analysis and optimization of cascade Rankine cycle for liquefied

natural gas cold energy recovery
In-Hwan Choi, Sangick Lee, Yutaek Seo, Daejun Chang*
Division of Ocean Systems Engineering, Korea Advanced Institute of Science and Technology, 291 Daehak-ro, Yuseong-gu, Daejeon 305-701,
Republic of Korea

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

Article history: This study proposes a new concept called the cascade Rankine cycle, which recovers LNG (liquefied
Received 19 October 2012 natural gas) cold energy for power generation, optimizes the cycle by investigating the effects of key
Received in revised form parameters on its performance, and compares its thermal efficiency, exergy efficiency and economic
14 July 2013
evaluation to those of the conventional alternatives. The cascade Rankine cycle consists of multiple
Accepted 22 August 2013
Available online 2 October 2013
stages of the organic Rankine cycle in a layered structure in which the first stage encompasses the second
one that, in turn, encompasses the next. Due to its layered configuration, optimization of the cycle is
straightforward and involves sequentially optimizing the individual stages. Optimization of the subse-
Keywords:
Cascade Rankine cycle
quent stages, however, required process simulation considering the equipment efficiency and the ther-
LNG cold energy recovery modynamic properties of the working fluid. Process simulation indicated that the indicators such as net
Power generation cycle power output, thermal efficiency, and exergy efficiency generally increase as the number of stages in-
Organic Rankine cycle creases. These indicators were, however, significantly affected by the thermodynamic properties of the
Exergy efficiency working fluids. The proposed cycles demonstrated significantly better performance in these indicators
Economic evaluation than the conventional cycles. The three-stage cascade Rankine cycle with propane as the working fluid
exhibited the highest net power output, thermal efficiency and exergy efficiency within the set.
Crown Copyright Ó 2013 Published by Elsevier Ltd. All rights reserved.

1. Introduction
The supply chain of liquefied natural gas (LNG) generally con-
sists of four segments: natural gas production, LNG production, LNG
transport, and LNG regasification. Within this supply chain, the LNG
production segment consumes a significant amount of exergy due
to the required liquefaction processes. Numerous liquefaction
processes have been developed to reduce the energy required for
liquefaction. Of the commercialized processes, the C3 mixed
refrigerant cycle is known to exhibit the highest efficiency,
consuming approximately 300 kWh of electrical energy to produce
one ton of LNG, or 1080 kJ/kg-LNG [1].
In the LNG regasification segment, LNG is converted into natural
gas and releases its cold energy into the heating medium, usually
seawater, without any useful byproduct. In other words, the large
amount of energy absorbed during the LNG production segment is
discharged into the seawater. The cold energy contained in LNG is
approximately 864 kJ/kg at the ambient temperature (20
C), which
is equivalent to 240 kWh/t of LNG [2].
* Corresponding author. Tel.: þ82 42 350 1514/47; fax: þ82 42 350 1510.
E-mail address: djchang@kaist.edu (D. Chang).
The effective recovery of LNG cold energy has drawn significant
research attention. LNG cold energy can be directly utilized in the
air separation, ethylene separation, liquefied CO2 production,
freezing of food, hydrogen liquefaction, low-temperature crushing,
freeze drying, high purity ozone production, and many other
applications.
In addition to direct utilization, the LNG cold energy can be
recovered through power generation. While the direct expansion
cycle (DEC) directly utilizes LNG as a working fluid, the organic
Rankine cycle (ORC) uses seawater as the primary heat source and
LNG as the heat sink with other hydrocarbons as working fluids.
Since both the DEC and the ORC have relatively low first law effi-
ciency due to the utilization of the ambient heat sources, the
modified ORC and the Combined Cycle (CC) have been developed to
improve their thermal efficiency and performance using various
heat sources and working fluids.
Many studies have been dedicated to developing power cycles
from different temperature heat sources and working fluids and
improving the efficiency of their cycles. Table 1 presents a review of
previous studies on power generation cycles using cold energy
sources. They are classified into three groups according to the heat
source temperature: high-temperature heat sources, medium-
temperature heat sources and low-temperature heat sources.

0360-5442/$ e see front matter Crown Copyright Ó 2013 Published by Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.energy.2013.08.047
180 I.-H. Choi et al. / Energy 61 (2013) 179e195

Nomenclature V_ volume flow rate, m3/s

T temperature, K
DEC direct expansion cycle _
Ex exergy transfer rate, kW or kJ/s
ORC organic Rankine cycle Dex exergy change per unit mass, kJ/kg
CC combined cycle DH enthalpy change, kJ
CRC cascade Rankine cycle h specific enthalpy, kJ/kg
C1 methane as a working fluid s specific entropy, kJ/kg K
C2 ethane as a working fluid
C3 propane as a working fluid Subscripts
LNG liquefied natural gas 0 ambient condition (1 atm, 298 K)
MSVEP maximum saturated vapor enthalpy pressure, bar net net output from a cycle
NG natural gas net,i net output from i-th-stage of cycle
LCC life cycle cost input input into a cycle
CAPEX capital expenditure output output from a cycle
OPEX operational expenditure in inlet
HFC hydrofluorocarbons out outlet
Pump_LNG boosting pump of LNG p polytropic efficiency
Pump_WF,i boosting pump of working fluid in i-th stage p_turbine polytropic efficiency of turbine
Pump_SW,i boosting i-th pump of seawater in i-th stage is isentropic efficiency
Turbine, i turbine in i-th stage is_pump isentropic efficiency of pump
Vaporizer,i vaporizer in i-th stage is_turbine isentropic efficiency of turbine
CondensereEvaporator,i condensereevaporator in i-th stage pump pump
PV, i pressure of working fluid at the i-th vaporizer, bar pump_LNG boosting pump of LNG
PC, i pressure of working fluid at the i-th condenser pump_WF,i boosting pump of working fluid in i-th stage
eevaporator, bar pump_SW,i boosting pump of seawater in i-th stage
w work per unit mass, kJ/kg turbine turbine
W _ power, kW or kJ/s turbine, i turbine in i-th stage
Q_ heat transfer rate, kW or kJ/s th thermal
m_ mass flow rate, kg/s ex exergy
m mass, kg loss loss
k ratio of Cp and Cv i integer number (1,2,3,...)
Cp constant pressure heat capacity, kJ/kg K j integer number (1,2,3,...)
Cv constant volume heat capacity, kJ/kg K
n polytropic coefficient Superscript
r liquid density, kg/m3 s isentropic process
h efficiency, % n polytropic coefficient
P pressure, Pa

Table 1
Summary of previous investigations on power generations combined with cold energy recovery.

Reference Cycles Cycle description Working fluids Heat source Heat source
temperature

Liu et al. [3] ORC Conventional ORC 10 Working fluids High 300
C
Wei et al. [4] ORC Conventional ORC HFC-245fa High 370
C
Desai and Bandyopadhyay [5] Modified ORC ORC with regeneration and turbine bleeding 16 Organic High 370
C
Tsatsaronis and Morosuk [6] CC ORC with gaseturbine N2 High 1290
C
Hisazumi et al. [7] CC ORC with gas/steam turbines HFC High 1000e1300
C
Kim and Ro [8] CC ORC with gas turbine e High 1350
C
Miyazaki et al. [9] CC ORC with refuse incinerator Ammoniaewater High 950
C
Lu and Wang [10] CC ORC with combustion process Ammoniaewater High 990
C
Deng et al. [12] CC Combined cycle CO2 High 1250
C
Zhang and Lior [13] CC ORC with Brayton cycle CO2 High 1300
C
Zhang [14] CC ORC with Brayton cycle CO2 High 720
C
Vélez et al. [15] ORC Transcritical ORC CO2 Medium 60e150
C
Qiang et al. [16] CC ORC with DEC Propane Medium 60e90
C
Shi and Che [17] CC ORC with DEC Ammoniaewater Medium 157e197
C
Roy et al. [18] ORC Conventional ORC R-12, R-123, R-134a Medium 140
C
Chen et al. [19] ORC Supercritical Rankine cycle Zeotropic mixture Medium 120e200
C
Wang et al. [20] ORC Conventional ORC Various working fluids Medium 70e230
C
Baik et al. [21] ORC Subcritical ORC/Transcritical ORC Four R125-based HFC Medium 100
C
Wang et al. [22] ORC Conventional ORC Ammoniaewater Medium 200
C
Liu and Guo [23] Modified ORC Organic Rankine cycle with a vapor absorption Binary mixture (CF4 and C3H8) Low 20
C
process
Szargut and Szczygiel [24] CRC Cascade Rankine cycle 8 Organic compounds Low 15
C
 I.-H. Choi et al. / Energy 61 (2013) 179e195
The power generation cycles using high-temperature heat
sources (400e1300
C) mostly employed a modified ORC or CC.
Studies on the modified ORC have investigated the selection of an
appropriate working fluid and a simple modification of ORC to
optimize the utilization of LNG cold energy and available heat
sources. Liu et al. [3] reported that the presence of hydrogen bonds
in certain molecules, such as water, ammonia and ethanol, results
in wet fluids due to the negative slope of the saturation vapor curve,
which is regarded as inappropriate for an ORC system. Conse-
quently the outlets streams of the turbine contain a fair amount of
liquid droplets. Presence of liquid inside turbine may damage tur-
bine blades and it also reduces the isentropic efficiency of the
turbine. Wei et al. [4] presented an ORC using HFC-245fa (1,1,1,3,3-
pentafluoropropane) as the working fluid. The system performance
analysis and optimization of an ORC were analyzed. Their results
indicated that the use of exhaust heat is a good way to improve the
net power output and efficiency and that a high ambient temper-
ature deteriorates the system performances and the net power
output is decreased more than 30%. Desai and Bandyopadhyay [5]
studied the modified ORCs with 16 working fluids. The modified
ORCs consist of a conventional ORC with an additional regeneration
unit and turbine bleeding. They reported that simultaneous
regeneration and turbine bleeding improves the thermal efficiency
of the ORC. Tsatsaronis and Morosuk [6] presented an advanced
exergetic analysis on the novel cogeneration system for generating
electricity and vaporizing LNG. The overall system consists of LNG,
N2 and Open gaseturbine power subsystems under the same
conditions. They performed an advanced exergetic analysis to
compare with conventional exergy-based analyses and suggested
that the advanced analysis allows engineers to better understand
the interconnections among components and the potential for
improving such systems.
Studies on CC have investigated the combination with gas/steam
turbines, refuse incinerators, and Brayton cycles. Hisazumi et al. [7]
suggested an ORC that employed the hydrofluorocarbons (HFC) as a
working fluid and included a combined cycle with gas and steam
turbines. To optimize the system, the effects of the process vari-
ables such as the gas turbine combustion pressure, steam pressure,
condensation temperature in the combined cycle, composition of
HFC and vapor pressure of natural gas were evaluated. Kim and Ro
[8] investigated the use of low-temperature LNG as source of cold
energy for the gas turbine inlet air cooling process in a conventional
combined cycle power plant. Hence, there refrigeration system in
the normal inlet chilling process is replaced by the heat exchange
between air and LNG. Miyazaki and Kang [9] proposed a combined
power cycle composed of an ammoniaewater Rankine cycle and a
refuse incinerator for recovering LNG cold energy. Their study
indicated that the efficiency of the combined power cycle was
significantly higher than that of the simple Rankine cycle. Lu and
Wang [10] presented a combined cycle consisting of a Rankine cycle
with an ammoniaewater mixture, a Brayton power cycle using a
combustion gas, and an open LNG cycle. Their study analyzed the
influence of key parameters on the thermal and exergy efficiencies
and optimized the system using the maximum economic benefits
as the objective function.
Additional studies of CCs have also been performed on the
possibility of using Rankine cycles and zero-emission of CO2 cap-
ture system using CO2 as a working fluid with exhaust gases from
a gas turbine as a heat source and LNG as a cold sink and the
thermal performance of such systems has been analyzed [11].
Deng et al. [12] proposed a new cogeneration power system with
two energy sources, fuel chemical energy and LNG cold energy,
and two outputs, electrical power and cooling power. The cascade
configuration utilizing the cryogenic energy enabled the system
to achieve improved energy efficiency. Zhang and Lior [13]
181
investigated a combined cycle process employing a supercritical
CO2 Rankine-like cycle and a CO2 Brayton cycle interconnected
through a heat transfer process. By utilizing the LNG evaporation
system as the cycle’s cold sink, the condensation of the cycle
achieved a cooling temperature much lower than the ambient
temperature. Zhang [14] proposed a process in which LNG and a
gas turbine were combined to form a CO2 recovery cycle. This cycle
comprehensively recovered the LNG cold energy and generated
power with reduced carbon dioxide emissions. The influences of
certain key parameters on the cycle’s performance, including the
Brayton cycle mass flow rate ratio, the low pressure turbine inlet
temperature, and the pressure ratio, were investigated. Vélez et al.
[15] proposed a carbon dioxide transcritical power cycle with CO2
as working fluid in a low-temperature heat source for power
generation.
In the case of using a medium-temperature heat source (60e
200
C), ORC and CC power generation cycles were used, which is
ORC topping DEC. Qiang et al. [16] investigated a combined power
cycle based on the cold energy of LNG and a low-grade heat source.
This cycle not only effectively recovered the low-temperature
exergy and pressure exergy of LNG but also utilized low-grade
input energy. To increase the thermal and exergy efficiencies of
the power cycle, some of the effects of key parameters were
analyzed, including the temperature of the low-grade heat source
and the condensation temperature of the second medium. Shi and
Che [17] proposed a combined power system in which waste heat is
efficiently recovered and the LNG cold energy is fully utilized. Their
system consists of an ammoniaewater mixture Rankine cycle and
an LNG power generation cycle. Roy et al. [18] investigated a waste
heat recovery system based on ORC, using R-12, R-123 and R-134a
as working fluids. Their cycles used the waste heat of flue gas at
140
C as a heat source and investigated the optimization of the
turbine inlet pressure for maximum work and efficiency of the
cycles along the saturated vapor line and isobaric superheating at
different pressures. Chen et al. [19] proposed a supercritical
Rankine cycle (SRC) using a zeotropic mixture as a working fluid for
the conversion of low-grade heat. They investigated the perfor-
mance of the proposed cycle through a comparative study of SRC
and ORC under the same thermal conditions and demonstrated that
the SRC achieved higher thermal efficiency than the ORC. D. Wang
et al. [20] proposed a thermal efficiency model theoretically based
on an ideal ORC to analyze the influence of working fluid properties
on the thermal efficiency, the optimal operation condition and
exergy destruction for various heat source temperatures were also
evaluated utilizing pinch point analysis and exergy analysis. Baik
et al. [21] investigated by comparison between the power of an
R134a subcritical cycle and that of binary HFC mixture transcritical
cycles for a low-temperature geothermal heat source of about
100
C. J. Wang et al. [22] investigated an ammoniaewater power
system with LNG as its heat sink to recover the low-grade waste
heat and to find the best thermodynamic parameters for the system
optimization design from both thermodynamic and economic
aspects.
Power generation cycles using a low-temperature heat sources
such as seawater (20
C) and the surroundings (15
C) are applied
by ORC and modified ORC, which consists of multiple stages. Liu
and Guo [23] proposed a combined cycle with a vapor absorption
process to improve the energy recovery efficiency of LNG cold
energy. They used binary mixtures of tetrafluoromethane (CF4)
and propane (C3H8) as working fluids, seawater as the heat source
and LNG as the heat sink. CF4 works as the main expansion fluid in
the turbine for power generation, and C3H8 works mainly as the
absorbent for CF4. Their proposed cycle investigated the effects of
some key parameters and the optimized conditions and compared
them with the C3 ORC. It is found that the proposed cycle is
182 I.-H. Choi et al. / Energy 61 (2013) 179e195

considerably superior to the ORC. Szargut and Szczygiel [24]
proposed three variants of cascade cycles with working fluids
and a low-temperature heat source at ambient temperature
(15
C). They proposed three cycles to minimize the exergy los-
sesdtwo stages of expansion in the higher part of the cascade,
then a two-stage condenser of krypton in the lower part of the
cascade and heat regeneration by means of the interstage
extraction vapordand analyzed the economic optimization of the
minimum heat transfer temperature difference (pinch value) in
particular heat exchangers and then determined the optimum
heat transfer area.
Several other studies are also worth mentioning. Yamada et al.
[25] proposed a new pumpless Rankine-type cycle for power
generation from low-temperature heat sources. The new cycle
primarily consisted of an expander, two heat exchangers, and
switching valves for the expander and heat exchangers. Instead of
using a working fluid pump, the valves are switched to control the
cycle. Yuanwei et al. [26] proposed a TeUeH diagram method of
analyzing the performance of a two-cascade LNG cold power gen-
eration system based on the LNG gasification characteristics.
Despite these efforts, the efficiency of power generation cycles
relying on low-temperature cold sources can still be improved. One
possibility is to continue utilizing a cold source that has already
been used to cool a hot working fluid but is still cold enough for
further utilization. If the cold stream is repeatedly utilized, the
following benefits are expected: additional energy recovery, a
decrease in the required heating medium, and a reduced risk of the
heat exchanger freezing.
The objectives of this investigation are to propose a new cascade
Rankine cycle that uses low-temperature heat sources, to analyze
the influences of the key parameters on the performance of this
cascade Rankine cycle and to compare thermal efficiency, exergy
efficiency and economic evaluation of the new cycle with those of
conventional cycles. After the main features of the representative
conventional cycles are described, the new cycle is proposed. Per-
formance measures are introduced for analysis and comparison. A
specific case involving the LNG gasification process is simulated for
direct comparison.
Fig. 2 shows a schematic of the organic Rankine cycle. The ORC
consists of two stages: one for the LNG and the other for the
working fluid. In the latter stage, the working fluid enters the pump
and is boosted to the operating pressure of the vaporizer. After
being vaporized by the heating medium in the vaporizer, the
working fluid produces work as it expands through the turbine and
is condensed by the cryogenic LNG in the condensereevaporator,
completing the cycle.
A schematic of the ORC combined with DEC is presented in
Fig. 3. This combined cycle (CC) is a combination of the DEC and the
ORC. The working fluid stage and the LNG stage are connected to
each other at the condensereevaporator. In the natural gas turbine
(Turbine 2 in Fig. 3), the LNG should be boosted by the LNG pump
(Pump 1 in Fig. 3) to a higher pressure than that utilized in the DEC.
In general, the CC produces more net power output than the DEC or
the ORC alone.
2.2. A new cycle: cascade Rankine cycle (CRC)
This study proposes a new power generation cycle called the
cascade Rankine cycle. This cycle employs multiple organic Rankine
cycles in series. The LNG stream leaving the condensereevaporator
in Fig. 2 is still cold enough to generate additional power. If another
ORC utilizes this stream, the benefit is three-fold: more exergy is
recovered, the heating medium flow rate is reduced, and the risk of
the condensereevaporator freezing is mitigated. This concept is
presented in Fig. 4. Further improvement can be achieved by add-
ing additional organic Rankine cycles, and a three-stage version is
presented in Fig. 5.
3. Analysis and comparison
3.1. Comparison measures
Some performance measures are used for comparison among
the afore-mentioned power generation cycles. The power is
_ and is expressed as Eq. (1).
denoted by W
_ ¼ mw
W _ (1)

2. Description and optimization of power generation cycles where m _ is the mass flow rate and w is the work per unit mass.
One of the measures that may be used is the net power output
2.1. Conventional cycles _ net , which is defined by Eq. (2).
W

There are three representative cycles that recover LNG cold W _ output  W
_ net ¼ W _ (2)
input
energy for power generation: the direct expansion cycle (DEC), the _ output and W
_
Here, W input are the power output from and input
organic Rankine cycle (ORC), and the ORC combined with DEC. The
into the power generation cycle, respectively.
first cycle directly vaporizes LNG into natural gas, which is
The thermal efficiency of power generation is defined by the
expanded through a turbine to recover pressure energy or more
ratio of the net power output to the input heat transfer rate.
appropriately exergy. The second cycle employs a secondary cycle
utilizing a foreign working fluid, and the last cycle combines the _ net =Q_
hth ¼ W input (3)
previous two to improve efficiency.
The DEC, as illustrated in Fig. 1, utilizes a pump, a vaporizer, and The power output from power generation cycles utilizing LNG
a turbine. The feed LNG enters the pump as a saturated liquid and is cold energy originates in the shaft work performed by expansion
boosted to the operating pressure of the vaporizer. The temperature engines or turbines.
of the LNG increases slightly during this boosting process as the X
specific volume decreases. The pressurized LNG is boiled in the _ output ¼
W _
W (4)
turbine;i
vaporizer by a heating source, which is usually seawater. The i
generated vapor is fed to the turbine and produces work by rotating
a shaft connected to an electric generator. The work produced by The power consumers are the pumps, which may serve to boost the
the turbine, which is proportional to the pressure ratio over the LNG or the heating medium, which is usually seawater.
turbine, originates from the work supplied by the pump and the X X X
_ _ _ _
heat added by the vaporizer. Though the DEC is usually less efficient W input ¼ W pump_LNG þ W pump_WF;i þ W pump_SW;i
than other cycles, it is simpler to configure, requires less capital i i i

investment, and is more spatially compact. (5)

 I.-H. Choi et al. / Energy 61 (2013) 179e195 183

Fig. 1. Schematic of the direct expansion cycle (DEC).

Pump isentropic pressurizing work W_

is_pump
The pump operation is similar to the compressor operation in his_pump ¼ ¼
_ pump
(6)
Pump actual pressurizing work W
that it increases the pressure. The pump operation assumes that the
inlet fluid is incompressible. The density is constant, and the liquid When the changes in potential energy and kinetic energies of
volume is independent of pressure. The isentropic efficiency of liquid are negligible, the isentropic pump pressuring work can be
pump is defined by Eq. (6). expressed by Eq. (7)

Fig. 2. Schematic of the organic Rankine cycle (ORC).

184 I.-H. Choi et al. / Energy 61 (2013) 179e195

Fig. 3. Schematic of the combined cycle (CC).

  _ out  Pin Þ
mðP due to friction or dissipative effects. Also, the isentropic efficiency
_ _ hsout  hin ¼ _ out  P Þ
W is_pump ¼ m ¼ VðP of turbine can be expressed in terms of inleteoutlet pressures,
r in
temperature and k (the ratio of Cp/Cv), as shown in Eq. (10).
(7)
Combining the above Eqs. (6) and 7 leads to the following Tout  Tin
his_turbine ¼ h i (10)
expression (Eq. (8)) for the actual pump pressurizing work. Tin ðPout =Pin Þðk1Þ=k  1

_  
W _ hsout  hin
m _ out  P Þ
VðP The change in enthalpy for an ideal gas can be calculated from
_ pump ¼
W
is_pump
¼ ¼ in
(8)
his_pump his_pump his_pump the energy balance equation as expressed in Eq. (11).

The turbine operation is used to decrease the pressure of a high DH ¼ mCp ðTout  Tin Þ (11)
pressure inlet gas stream to produce an outlet stream with low
The actual turbine expansion work for isentropic process can be
pressure and high velocity. An expansion process involves con-
calculated as shown in Eq. (12).
verting the internal energy of the gas to kinetic energy and finally to
shaft work. The isentropic efficiency of turbine can be expressed by
_ _
Eq. (9). W turbine ¼ his_turbine W is_turbine
k h i
_ ¼ his_turbine Pin V_ in ðPout =Pin Þðk1Þ=k  1 (12)
Turbine actual expansion work W ð1  kÞ
his_turbine ¼ ¼
_
turbine
Turbine isentropic expansion work W is_turbine Meanwhile, polytropic process is similar to the real expansion
ðh h Þ process considering a heat transfer and friction loss. In the relation
¼  sout in 
hout  hin of the isentropic efficiency upon the polytropic efficiency with the
pressure ratio (Pout/Pin), the isentropic efficiency for turbine is al-
(9)
ways greater than the polytropic one, while the isentropic effi-
The isentropic process means reversible adiabatic process. That ciency of pump exhibits an opposite tendency. The polytropic
is, no heat added to system, and no energy transformations occur efficiency (hp) is defined by Eq. (13).
 I.-H. Choi et al. / Energy 61 (2013) 179e195 185

Fig. 4. Schematic of the two-stage cascade Rankine cycle.

entropy. h0 and s0 are the specific enthalpy and entropy at ambient

klnðTout =Tin Þ
hp_turbine ¼ (13) temperature (T0) and pressure.
ðk  1ÞlnðPout =Pin Þ
The exergy loss rate in the each equipment of power generation
In the polytropic process, the polytropic coefficient (n) is is defined as follows:
expressed by Eq. (14).
 Pump
PV n ¼ constant (14)
_
DEx _ _
loss ¼ mðexin  exout Þ þ W pump (19)
The polytropic coefficient (n) can be defined in terms of k (Cp/Cv
ratio) and the polytropic efficiency (np) as in Eq. (15).
 Turbine
n1 k1
¼ hp (15) _ _
n k DEx _
loss ¼ mðexin  exout Þ  W turbine (20)
The actual expansion turbine work for polytropic process can be
calculated in terms of inleteoutlet pressures as shown in Eq. (16).
 Condensereevaporator
n h i
_ _ Xj  
W turbine ¼ hp_turbine Pin V in ðPout =Pin Þðn1Þ=n  1 (16) _
DEx _ i exi;in  exi;out
ð1  nÞ loss ¼ i
m (21)

The exergy efficiency of power generation system is defined by

the ratio of exergy output to exergy input.  Vaporizer
P Xj  
Exergy output _ _ i exi;in  exi;out þ Q_ ð1  T0 =TÞ
hex ¼ P DEx loss ¼ i
m (22)
Exergy input
P
Exergy loss rate in each equipment
¼ 1 P (17)
Exergy input
3.2. Comparison cases and cycle simulation
The exergy transfer rate of fluid stream is defined as
The processes are simulated to compare the performance of the
_ ¼ mð
Ex _ DexÞ ¼ m½ðh
_  h0 Þ  T0 ðs  s0 Þ (18) proposed cycle with those of the conventional cycles. LNG is a
multi-component mixture, and its composition varies from source
where Dex is the exergy change per unit mass and m
_ is the mass to source. Small variations in the composition, however, have a
flow rate. h and s refers to the specific enthalpy and specific negligible effect on the performance of the cycles. The composition
 186
I.-H. Choi et al. / Energy 61 (2013) 179e195
Fig. 5. Schematic of the three-stage cascade Rankine cycle.
 I.-H. Choi et al. / Energy 61 (2013) 179e195 187

Table 2 Table 4
Composition of LNG. Comparison of cycles.

Component Mole fraction Cycle no Notation Working fluid Cycle

CH4 0.9133 1 C1 DEC e Direct expansion cycle

C2H6 0.0536 2 C2 ORC C2H6 Organic Rankine cycle
C3H8 0.0214 3 C3 ORC C3H8 Organic Rankine cycle
i-C4H10 0.0047 4 C3 CC C3H8 Combined cycle
n-C4H10 0.0046 5 C2/C3 CRC C2H6eC3H8 Two-stage cascade Rankine cycle
i-C5H12 0.0001 6 C3/C2 CRC C3H8eC2H6 Two-stage cascade Rankine cycle
n-C5H12 0.0001 7 C3/C3 CRC C3H8eC3H8 Two-stage cascade Rankine cycle
N2 0.0022 8 C2/C2/C2 CRC C2H6eC2H6eC2H6 Three-stage cascade Rankine cycle
9 C2/C2/C3 CRC C2H6eC2H6eC3H8 Three-stage cascade Rankine cycle
10 C2/C3/C3 CRC C2H6eC3H8eC3H8 Three-stage cascade Rankine cycle
11 C3/C3/C3 CRC C3H8eC3H8eC3H8 Three-stage cascade Rankine cycle

of the LNG considered in this study is shown in Table 2. The three

lightest hydrocarbons are considered as working fluids and are
denoted by C1, C2, and C3. Their thermodynamic properties are
presented in Table 3. Table 4 presents the eleven cycles to be
compared: one DEC, two ORCs, one CC, three two-stage CRCs, and
four three-stage CRCs.
The assumptions employed in the process simulation are sum-
marized in Table 5. The efficiencies of the pumps and turbines are
chosen to be within commercially viable ranges. Considering the
temperature of the seawater entering the seawater pumps and the
minimum approach temperature of the vaporizers, the maximum
temperature of the working fluid in the vaporizers is 10
C (¼15
Ce
5
C). Analogously, the minimum temperature of the working fluid
in the condenser can be derived from the temperature of the
cooling medium and the minimum approach temperature in the
condenser. For example, the condensation temperature of the
working fluid of the ORC in Fig. 2 is 158
C (¼163
C þ 5
C).
The minimum pressure in the condenser is limited to prevent
leakage of air into the vacuum condenser. This leakage can cause
two problems. Air leakage deteriorates the performance of the
condenser due to the presence of non-condensable components
such as nitrogen and oxygen. There is also a risk of fire and ex-
plosion when the working fluid is a hydrocarbon, as in this study.
This limitation is ameliorated if a vacuum pump or an ejector is
used to eject a portion of the vapor from the condenser. This
approach, however, results in additional problems. In practical
design, when using a flammable hydrocarbon as the working fluid,
it is desirable to keep the condenser pressure above the ambient
pressure.
4. Results and discussion
4.1. Conventional ORC
Fig. 6 presents the specific enthalpyepressure diagram of the C3
ORC (Cycle 3) with two isentropic lines. The diagram was generated
using Refrigeration Utilities Ver. 2.84 developed by DTU (Denmark
Technical University). The working fluid in the liquid phase un-
dergoes Step A / B, which represents the pressure boosting per-
formed by the pump. Because this step is irreversible, it deviates
from the isentropic line. Following Step B / C, the working fluid is
Table 3
Properties of the studied working fluids.
Name Notation Normal boiling Liquid density Critical Critical
point (NBP),
C at NBP, kg/m3 pressure, temperature,
C
bar
Methane C1 164 0.425 46.0 82.7
Ethane C2 89 0.548 48.8 32.2
Propane C3 42 0.582 42.5 96.6
vaporized in the condensereevaporator, which causes no pressure
drop in the working fluid in accordance with the simulation’s as-
sumptions. Step C / D is the expansion of the working fluid
through the turbine. Like Step A / B, this step is irreversible and
thus deviates from the isentropic line. Finally, the working fluid is
condensed in Step D / A without any pressure loss in the
condensereevaporator.
The fact that points C and D are closely located to the saturated
vapor line helps provide the first guess for the optimal process
conditions that produce the maximum net power output. Consid-
ering the shape of the enthalpyepressure curve, the conditions for
the optimal pressure are as follows.
Condition 1*
The inlet pressure of the turbine, which is equal to the pressure of
the vaporizer, should be the pressure that corresponds to the
maximum enthalpy of the vapor phase. This pressure is called the
maximum saturated vapor enthalpy pressure (MSVEP).
Condition 2*
The outlet pressure of the turbine, which is equal to the pressure of
the condensereevaporator, should be as small as possible.
These conditions are ideal, and both pressures maybe infeasible
depending on the surrounding conditions and the properties of the
working fluids. Condition 1* cannot be satisfied if the pressure is
not achievable by the heating medium. The saturated vapor pres-
sure of the working fluid is proportional to its temperature, which
is determined by the heating medium, seawater in the ORC. In a real
design, the vaporization temperature of the working fluid is lower
than the temperature of the heating medium; this difference is
termed the approach temperature and is required for heat transfer.
Table 5
Assumptions used in process simulation.
Variables Values
Adiabatic efficiency of the pumps 80%
Adiabatic efficiency of the turbines 80%
Pressure drop in the heat exchangers 0
Minimum approach temperature in the condensereevaporators 5
C
Minimum approach temperature in the vaporizers 5
C
Minimum pressure in the condensereevaporators 1 bar
Temperature of seawater entering the seawater pumps 15
C
Pressure of seawater discharged from the pumps 3 bar
Temperature of the LNG entering the LNG pumps 162
C
Pressure of the LNG entering the LNG pumps 1 bar
Pressure of the product natural gas 60 bar
Temperature of the product natural gas 10
C
188 I.-H. Choi et al. / Energy 61 (2013) 179e195

Fig. 6. Specific enthalpyepressure diagram for Cycle 3.

Condition 2* is limited by two factors: the temperature of the medium, or 6.3 bar, while Condition 2 implies that the condensere
cooling medium and the minimum allowable condensereevapo- evaporator pressure should be 1 bar.
rator pressure to prevent leakage (1 bar). The former limitation is Fig. 7 illustrates the turbine’s power output against the pressure
imposed by the temperature of the cooling medium, which is LNG of the vaporizer. The turbine’s power output increases with an in-
in Cycle 3, and plays a role similar to that on the temperature of the crease in the pressure of vaporizer up to the MSVEP, and decreases
heating medium in the vaporizer. The temperature of the
condensed liquid cannot be reduced below that of the cooling
medium; the difference between the two temperatures should be
greater than the approach temperature of the condensere
evaporator.
Considering these limitations, the ideal conditions on the
optimal pressure should be revised for the actual cycle design as
follows.

Condition 1
The inlet pressure of the turbine, which is equal to the pressure of
the vaporizer, should be the lower of the MSVEP and the maximum
saturated vapor pressure achievable by the heating medium.

Condition 2
The outlet pressure of the turbine, which is equal to the pressure of
the condensereevaporator, should be the larger of the atmospheric
pressure and the minimum saturated liquid pressure achievable by
the cooling medium.

In Cycle 3, Condition 1 indicates that the inlet pressure of the

turbine, which equal to the pressure of the vaporizer, should be the Fig. 7. Power output from the turbine against the pressure of the inlet stream in Cycle
maximum saturated vapor pressure achievable by the heating 3.
 I.-H. Choi et al. / Energy 61 (2013) 179e195
_
Fig. 8. Power output and input of the turbine and the pump in Cycle 3 (W turbine : Power
_ pump_WF : Power input into working fluid pump (kJ/s),
output from the turbine (kJ/s), W
W_ _ Analogous to Cycle 3, the power output increases more rapidly than
pump_SW : Power input into seawater pump (kJ/s), W net : Net power output from a
cycle (kJ/s)).
thereafter. Note that an inlet stream pressure greater than 6.3 bars
is infeasible because it is assumed temperature of the seawater
(15
C) as a heating medium. In considering the minimum approach
temperature of the vaporizers is 5
C, the maximum temperature of
the working fluid in the vaporizers is 10
C.
Fig. 9. Specific enthalpyepressure diagram for Cycle 2.
189
Fig. 8 depicts the turbine’s power output and the pump’s power
input together over the range of feasible vaporizer pressures. Both
the power output and input increase as the vaporizer pressure in-
creases However, the power output increases far more rapidly than
the input, meaning that the net power output increases as the
vaporizer pressure increases.
Fig. 9 shows the specific enthalpyepressure diagram of the C2
ORC (Cycle 2) with ethane as the working fluid (Cycle 2). Compared
with Cycle 3, the point D, which is the state of the working fluid at
the turbine outlet, is located on the left of the saturated vapor line,
implying that the stream is a two-phase flow containing some
droplets in the main vapor stream.
Fig. 10 shows that the dependence of the turbine’s power output
on the vaporizer pressure of Cycle 2 is similar to that of Cycle 3. The
maximum allowable vaporizer pressure for Cycle 2 is approxi-
mately 30 bar, and the minimum allowable condensereevaporator
pressure is 1 bar. The turbine’s power output increases steadily as
the vaporizer pressure increases within the feasible range.
Fig. 11 illustrates the power output from the turbine and input
into the pumps within the feasible vaporizer pressure range.
the power input, indicating that the net power output increases as
the vaporizer pressure increases.
4.2. Two-stage cascade Rankine cycle (CRC)
When another stage is added to an ORC cycle to form the second
stage of a two-stage CRC, it should be noted that the optimal con-
ditions of the first stage are not affected by the presence of the
second stage. Thus, the optimization of the two-stage CRC can be
190
Fig. 10. Power output from the turbine in Cycle 2.
separated into the optimization of the first stage followed by the
optimization of the second stage. Because the first stage is already
optimized in terms of the vaporizer pressure, the remaining task is
to optimize the second stage.
Optimization of the second stage is not as straightforward as
that of the first stage. The specific enthalpy difference, or the work
per unit mass and mass flow rate of the working fluid from the
second turbine naturally decreases as the pressure of the second
condensereevaporator increases (CondensereEvaporator 2 in
Fig. 4); this pressure is denoted PC,2. Fig. 12 presents the specific
enthalpy difference in turbine 2. For the second stage, the specific
enthalpy difference decreases as the condensereevaporator pres-
sure is increased.
The mass flow rate circulating through the turbine, however,
decreases as the condensereevaporator pressure decreases, as
shown in Fig. 13. As the pressure of condensereevaporator 2 ap-
proaches that of condensereevaporator 1, the allowable mass flow
rate of the working fluid vanishes. Thus, when considering the
circulating mass flow rate, the efficiency increases at higher
Fig. 11. Power output and input from the turbine and the pump in Cycle 2 (W
_
Power output from the turbine (kJ/s), W pump_WF : Power input into working fluid pump
_ _
(kJ/s), W pump_SW : Power input into seawater pump (kJ/s), W net : Net power output from
a cycle (kJ/s)).
I.-H. Choi et al. / Energy 61 (2013) 179e195
Fig. 12. Specific enthalpy difference from Turbine, 2 versus the pressure of condensere
evaporator, 2 in Cycle 7.
condensereevaporator pressures. In consequence, the maximum
power output from the second turbine is limited by the combina-
tion of these two constraints. Fig. 14 presents the power output
versus the second condensereevaporator pressure, indicating that
a condensereevaporator pressure of 2.7 bar produces the
maximum power output.
Fig. 15 presents the power output and input of the second stage
in Cycle 7. Unlike the first stage, the power input into the pump is
comparable to the power output from the turbine, and the opti-
mum second condensereevaporator pressure shifts from 2.7 bar to
2.2 bar.
4.3. Multi-stage cascade Rankine cycle (CRC)
Optimization of a multi-stage CRC is a simple extension of the
optimization of the two-stage CRC, i.e., a sequential optimization of
each stage beginning with the first. For example, to optimize the
three-stage CRC, the first two stages are already optimized, and the
remaining task is to optimize the additional third stage.
_
turbine :
Fig. 13. Mass flow rate of working fluid from Turbine, 2 versus the pressure of
condensereevaporator, 2 in Cycle 7.
 I.-H. Choi et al. / Energy 61 (2013) 179e195 191

Fig. 16. Power output and input of the third stage versus the pressure of condensere
_
evaporator, 3 in Cycle 11 (W turbine;3 : Power output from the turbine in the third stage
Fig. 14. Power output from Turbine, 2 versus the pressure of condensereevaporator, 2 _
(kJ/s), W pump_WF;3 : Power input into working fluid pump in the third stage (kJ/s),
in Cycle 7. W_ : Power input into seawater pump in the third stage (kJ/s), W _ : Net
pump_SW;3 net;3
power output from a cycle in the third stage (kJ/s)).

As an illustrative example, Fig. 16 shows the net power output,

the power output, and the power input of Cycle 11. The maximum
net power output is obtained when the pressure of condensere
evaporator 3 is approximately 3.5 bar.
The net power output of the first stage is 64.66 kJ/kg compared
to 8.06 kJ/kg in the second stage and 1.02 kJ/kg in the third stage.
The second stage generates approximately one eighth of the net
power output of the first stage and approximately eight times the
net power output of the third stage.
4.4. Comparison of performance of cycles
Table 6 presents the performances of the cycles listed in Table 4.
Each cycle is optimized as described in Section 4.3. Under the as-
sumptions used in this study, Cycle 1 has a negative net power
output due to substantial work consumption by the LNG and
seawater pumps. Comparison of Cycle 2 and Cycle 3 indicates that
the two do not exhibit significant differences in the power input
into the pumps but that the power output from the turbine is much
larger in Cycle 3 and in Cycle 2.
Cycle 4 of the combined cycle with C3 results in less net power
output than Cycle 3. Though Cycle 4 produces more power output
than Cycle 3, it also consumes much more work for LNG boosting.
Because Cycle 4 includes Cycle 1, which has a negative net power
output, this observation seems reasonable.
Cycles 5 to 11 of the cascade Rankine cycle exhibit much better
performance than the conventional cycles. Because the turbines of
the added stages produce additional work, the power output from
the turbines generally increases as the number of stages increases.
This trend, however, is not consistent; Cycle 6, the C3/C2 CRC, de-
livers more turbine work than all others except for Cycle 11, which
achieves the highest power output from the turbines. Cycles 5 to 7,
all of which consists of two stages with different working fluids,
produce different levels of power output from the turbines. This is
also the case for Cycles 8 to 11, which are all three-stage CRCs. This
observation demonstrates the importance of the working fluids.
A difference in the power output from the turbine leads directly
to a difference in the net power output. Note that the power
input into the seawater pumps remains relatively constant at

Table 6
Summary of the performance of the cycles.

Cycle no. Notation W_ _ _ _ hth (%) hex (%)

turbine W pump W pump_SW W net
(kJ/s) (kJ/s) (kJ/s) (kJ/s)

1 DEC 7.7 19.5 15.7 27.5 e e

2 C2 ORC 66.1 3.5 18.9 43.7 5.1 56.7
3 C3 ORC 84.2 1.5 18.0 64.7 7.6 57.7
4 C3 CC 91.9 21.0 22.2 48.7 5.7 62.1
5 C2/C3 CRC 83.0 4.4 20.6 58.0 6.8 63.3
6 C3/C2 CRC 125.4 4.6 20.0 100.8 11.9 65.1
7 C3/C3 CRC 97.7 3.5 21.7 72.5 8.5 64.7
Fig. 15. Power output and input of the second stage versus the pressure of condensere 8 C2/C2/C2 CRC 96.7 1.8 20.9 74.0 8.7 64.7
_
evaporator, 2 in Cycle 7 (W turbine;2 : Power output from the turbine in the second stage 9 C2/C2/C3 CRC 113.7 11.2 23.2 79.3 9.3 64.8
_
(kJ/s), W pump_WF;2 : Power input into working fluid pump in the second stage (kJ/s), 10 C2/C3/C3 CRC 120.0 10.0 22.7 87.3 10.3 64.9
_
W pump_SW;2 : Power input into seawater pump in the second stage (kJ/s), W _
net;2 : Net 11 C3/C3/C3 CRC 133.6 4.3 23.3 106.0 12.5 65.2
power output from a cycle in the second stage (kJ/s)).
192 I.-H. Choi et al. / Energy 61 (2013) 179e195

approximately 20 kJ/kg. Adding stages increases both the turbine
power output and the pump power input, but the exact changes are
significantly affected by the working fluids.
It should be noted that Cycle 11 exhibits the best thermal and
exergy efficiency. This observation is natural because the thermal
efficiency is obtained by normalizing the net power output by the
energy input. Table 7 shows the pressure, temperature, mass flow
rate, specific enthalpy, specific entropy and exergy transfer rate of
each steams of Cycle 11 (C3/C3/C3 CRC) illustrated in Fig. 5. Exergy
transfer rate of all streams for Cycle 11 is calculated by Eq. (18) and
exergy loss rate of equipment in the entire process is obtained from
Eqs. (19) and (20). Fig. 17 shows the exergy loss and exergy effi-
ciency of Cycle 4, Cycle 7 and Cycle 11. The exergy efficiency of Cycle
11 is 65.23% compared to 57.72% in Cycle 4 and 64.7% in Cycle 7. This
phenomenon is affected by the reduction of exergy loss of the
turbine, vaporizer, pump and heater. Most of the exergy loss occurs
in the condensereevaporator because the temperature differences
between hot and cold stream is quite large in the condensere
evaporator. The exergy loss in the condensereevaporator is 27.27%
in Cycle 11, compared to 25.67% in Cycle 4 and 27.12% in Cycle7. It
can be observed that exergy loss of condensereevaporator slightly
increases as the number of stages increases.

Table 7
Pressure, temperature, mass flow rate, specific enthalpy, specific entropy and exergy transfer rate of each streams of C3/C3/C3 CRC (Cycle 11): Ambient pressure 1 bar and
temperature 15
C.

LNG streams

A B C D E F

Pressure (bar) 1 60 60 60 60 60
Temperature (
C) 162 159 47 21 8 10
Mass flow rate (kg/s) 1 1 1 1 1 1
Specific enthalpy (kJ/kg) 5173 5157 4602 4494 4452 4402
Specific entropy (kJ/kg K) 4.25 4.29 7.44 7.89 8.06 8.24
Exergy transfer rate (kJ/s) 6397.0 6392.5 6744.7 6766.3 6773.3 6775.1

CRC 1-stage streams

1 2 3 4 5 6

Pressure (bar) 1 6.3 6.3 6.3 6.3 1

Temperature (
C) 42 40 21 8 10 42
Mass flow rate (kg/s) 1.306 1.306 1.306 1.306 1.306 1.306
Specific enthalpy (kJ/kg) 2889 2888 2841 2808 2400 2464
Specific entropy (kJ/kg K) 1.427 1.429 1.623 1.75 3.197 3.267
Exergy transfer rate (kJ/s) 4310 4309 4321 4325 4337 4447

C3 CRC 2-stage streams

10 11 12 13 14

Pressure (bar) 2.2 6.3 6.3 6.3 2.2

Temperature (
C) 17 16.9 7.5 9.7 10
Mass flow rate (kg/s) 0.425 0.425 0.425 0.425 0.425
Specific enthalpy (kJ/kg) 2831 2830 2808 2400 2431
Specific entropy (kJ/kg K) 1.663 1.664 1.75 3.195 3.225
Exergy transfer rate (kJ/s) 1406.7 1406.4 1407.6 1411.1 1427.9

C3 CRC 3-stage streams

20 21 22 23

Pressure (bar) 3.5 6.3 6.3 3.5

Temperature (
C) 3 2.9 9.7 3
Mass flow rate (kg/s) 0.246 0.246 0.246 0.246
Specific enthalpy (kJ/kg) 2797 2796 2400 2415
Specific entropy (kJ/kg K) 1.79 1.793 3.19 3.2
Exergy transfer rate (kJ/s) 814.9 814.8 816.4 820.8

Seawater streams (1)

S1 S2 S3 S10 S11 S12

Pressure (bar) 1 3 3 1 3 3
Temperature (
C) 15 15 13 15 15 13
Mass flow rate (kg/s) 61.3 61.3 61.3 61.3 61.3 61.3
Specific enthalpy (kJ/kg) 15,930 15,930.7 15,939 15,930 15,930.7 15,939
Specific entropy (kJ/kg K) 2.834 2.833 2.804 2.834 2.833 2.804
Exergy transfer rate (kJ/s) 1,026,541.6 1,026,566.8 1,026,563.6 1,026,541.6 1,026,567 1,026,564

Seawater streams (2)

S20 S21 S22 S30 S31 S32

Pressure (bar) 1 3 3 1 3 3
Temperature (
C) 15 15 13 15 15 13
Mass flow rate (kg/s) 11.24 11.24 11.24 11.24 11.24 11.24
Specific enthalpy (kJ/kg) 15,930 15,930.7 15,939 15,930 15,930.7 15,939
Specific entropy (kJ/kg K) 2.834 2.833 2.804 2.834 2.833 2.804
Exergy transfer rate (kJ/s) 188,227.2 188,231.8 188,231.2 188,227.2 188,232 188,231
 I.-H. Choi et al. / Energy 61 (2013) 179e195 193

Fig. 17. Exergy loss and exergy efficiency of Cycle 4, Cycle 7 and Cycle 11.

4.5. LCC evaluation of cycles

Life Cycle Cost (LCC) includes all the costs involved in initial Table 8
CAPEX methodology.
capital investment and operation of the system. LCC assessment is
applicable in the concept design stage to choosing the best optimal Items Percentage of delivered
design among alternatives. The conventional LCC generally consists equipment cost (%)

of Capital Expenditure (CAPEX) and Operational Expenditure Purchased equipment 100

(OPEX). Purchased equipment installation 47
Instrumentation and controls 36
Table 8 shows the CAPEX methodology using Percentage of
Piping and Electrical systems 79
Delivered Equipment Cost. Percentage of Delivered Equipment Cost Building 18
is the method which estimates the other items from the equipment Service facilities 70
cost. After determining the equipment cost, the other items Engineering and supervision 33
included in the CAPEX is estimated. The equipment costs of each of Legal expenses 4
Contractor’s fee 22
power generation cycles are determined by a commercial code Contingency 44
ASPEN Process Economic Analysis in Table 9.
194 I.-H. Choi et al. / Energy 61 (2013) 179e195

Table 9 Table 10
Equipment cost for each of power generation cycles. OPEX methodology.

Power Equipment Type Equipment Total Items Method of the estimation

generation cost ($) equipment
Operation labor Using operating labor graph
cycles cost ($)
Utilities Using commercial simulation model
DEC Vaporizer DHE FLOAT HEAD 10,300 357,300 Maintenance and repair 7% of OPEX
Heater DHE FLOAT HEAD 8500 Operating supplies 15% of maintenance and repair
Turbine DTURTURBOEXP 78,300 Local Taxes 3% of CAPEX
Pump_LNG DCP CENTRIF 237,600 Insurance 1% of CAPEX
Pump_SW,1 DCP CENTRIF 13,100 Plant overhead costs 60% of the operating labor and maintenance
Pump_SW,2 DCP CENTRIF 7700

ORC Vaporizer DHE FLOAT HEAD 17,600 382,200

Heater DHE FLOAT HEAD 34,000
Fig. 18 indicates the LCC, CAPEX, OPEX and Equipment cost of
Turbine DTURTURBOEXP 105,700
Condensere DHE PLAT FRAM 35,400 cycles. It should be noted that Equipment cost and CAPEX are
evaporator increased from Cycle 1 to Cycle 11 because CAPEX is affected by the
Pump_LNG DCP CENTRIF 154,700 equipment cost of power generation cycles. The life cycle of all
Pump_WF DCP CENTRIF 14,500
power generation cycles is assumed to be 20 years. The LCC results
Pump_SW1 DCP CENTRIF 13,100
Pump_SW2 DCP CENTRIF 7200
indicate that the key factor is OPEX. OPEX varies with the opera-
tional labor and the electricity cost. In this study, electricity
ORC combine Vaporizer DHE FLOAT HEAD 15,800 510,600
generated from power generation cycles is assumed to sale the
with DEC Heater DHE FLOAT HEAD 8500
Turbine,1 DTURTURBOEXP 102,900 markets and the cost of generated electricity considered for the
Turbine,2 DTURTURBOEXP 78,000 estimation of OPEX. OPEX of power cycles more generated elec-
Condensere DHE PLAT FRAM 32,300 tricity are decreased. Consequently, LCC of power cycles is also
evaporator varied with the generated electricity. In the thermodynamic as-
Pump_LNG DCP CENTRIF 237,600
Pump_WF DCP CENTRIF 14,400
pects, Cycle 11 has the best net power output, but LCC of Cycle 11 is
Pump_SW,1 DCP CENTRIF 12,900 the fourth-largest value of power cycles.
Pump_SW,2 DCP CENTRIF 8200

2-Stage CRC Vaporizer,1 DHE FLOAT HEAD 12,000 471,430 5. Conclusions

Vaporizer,2 DHE FLOAT HEAD 9700
Heater DHE FLOAT HEAD 12,200
Turbine,1 DTURTURBOEXP 106,100
Cascade Rankine Cycles (CRCs) consisting of multiple organic
Turbine,2 DTURTURBOEXP 80,100 Rankine cycles were proposed and sequentially optimized. The first
Condensere DHE PLAT FRAM 51,700 stage was optimized based on the pressures of the condensere
evaporator,1 evaporator and vaporizer, and optimizing the subsequent stages
Condensere DHE PLAT FRAM 1230
required process simulations considering the equipment efficiency
evaporator,2
Pump_LNG DCP CENTRIF 154,700 and the thermodynamic properties of the working fluid. The new
Pump_WF,1 DCP CENTRIF 14,500 cycles exhibited better performances than the conventional direct
Pump_WF,2 DCP CENTRIF 4200 expansion cycle, the organic Rankine cycle, and the combined cycle
Pump_SW,1 DCP CENTRIF 12,900
under the given conditions. The effects of the working fluids were
Pump_SW,2 DCP CENTRIF 6700
Pump_SW,3 DCP CENTRIF 5400
significant enough to overcome the improvements achieved by
increasing the number of stages.
3-stage CRC Vaporizer,1 DHE FLOAT HEAD 21,600 563,930
Several topics are recommended for further investigation. The
Vaporizer,2 DHE FLOAT HEAD 13,500
Vaporizer,3 DHE FLOAT HEAD 9800 first is a sensitivity study on the conditions used in the simulations.
Heater DHE FLOAT HEAD 12,100 For example, increasing the seawater temperature will increase the
Turbine,1 DTURTURBOEXP 105,700 power output from the turbines by elevating the vaporizer pressure
Turbine,2 DTURTURBOEXP 80,300 and will decrease the power input into the pumps by decreasing
Turbine,3 DTURTURBOEXP 76,800
Condensere DHE PLAT FRAM 35,400
the mass flow rate of the heating medium. Another parameter that
evaporator,1 may exert a significant influence is the export gas pressure. The
Condensere DHE PLAT FRAM 790 power input into the LNG pump will decrease as the export pres-
evaporator,2 sure increases. Comparing Cycles 8 to 11 with each other demon-
Condensere DHE PLAT FRAM 740
strates that the power input depends strongly on the working fluid.
evaporator,3
Pump_LNG DCP CENTRIF 154,700 Combining this CRC with a relatively hot heating medium may
Pump_WF,1 DCP CENTRIF 14,500 produce a significant amount of power. One possibility is to
Pump_WF,2 DCP CENTRIF 4200 combine the cycle with cogeneration of power using gas turbines
Pump_WF,3 DCP CENTRIF 3800 whose fuel gas is supplied from the CRC. In this case, the range of
Pump_SW,1 DCP CENTRIF 13,100
Pump_SW,2 DCP CENTRIF 6700
feasible operating pressures and temperatures of the power gen-
Pump_SW,3 DCP CENTRIF 5500 eration cycles is widened, providing more freedom in choosing the
Pump_SW,4 DCP CENTRIF 4700 working fluids and the system’s configuration.
Another point worthy of investigation is the optimization of real
processes. The optimizations in this study considered the energetic
Table 10 indicates OPEX methodology using the CAPEX except aspects (the net power output, the thermal efficiency, and the
operating labor and utilities. Operating labor is estimated by exergy efficiency) and economic aspects. An optimal process in the
operating labor requirement graph for chemical process industries real world should consider additional factors such as the space
[27]. It is assumed that the salary for an operator is approximately limitations, equipment reliability, and maintenance. For example,
50,000 USD/yr. The electricity consumption is estimated on the Cycle 11 is better than Cycle 6 in terms of the net power output,
basis of process simulation. thermal efficiency and exergy efficiency but LCC of Cycle 11 is
 I.-H. Choi et al. / Energy 61 (2013) 179e195
Fig. 18. LCC, CAPEX, OPEX and equipment cost of power generation cycles.
higher than Cycle 6 due to more complicated to configure and
operate. Consequently, it is not certain whether Cycle 11 is better
than Cycle 6 when additional considerations are included.
Acknowledgements
This work was supported by the Global Leading Technology
Program of the Office of Strategic R&D Planning (OSP) funded by the
Korea Evaluation Institute of Industrial Technology, Ministry of
Trade, Industry and Energy, Republic of Korea
References
[1] Timmerhaus KE, Reed RP. Fifty years of progress: fifty-years’ development of
cryogenic liquefaction processes. New York: Springer; 2007.
[2] Liu H, You L. Characteristics and applications of the cold heat exergy of liq-
uefied natural gas. Energy Convers Manage 1999;40:1515e25.
[3] Liu BT, Chein KH, Wang CC. Effect of working fluids on Organic Rankine Cycle
for waste heat recovery. Energy 2004;29(8):1207e17.
[4] Wei D, Lu X, Lu Z, Gu J. Performance analysis and optimization of organic
Rankine cycle (ORC) for waste heat recovery. Energy Convers Manage
2007;48:1113e9.
[5] Desai NB, Bandyopadhyay S. Process integration of organic Rankine cycle.
Energy 2009;34:1674e86.
[6] Tsatsaronis G, Morosuk T. Advanced exergetic analysis of a novel system for
generating electricity and vaporizing liquefied natural gas. Energy 2010;35:
820e9.
[7] Hisazumi Y, Yamasaki Y, Sugiyama S. Proposal for a high efficiency LNG
power-generation system utilizing waste heat from the combined cycle. Appl
Energy 1998;60:169e82.
[8] Kim TS, Ro ST. Power augmentation of combined cycle power plants using
cold energy of liquefied natural gas. Energy 2000;25:841e56.
[9] Miyazaki T, Kang YT, Akisawa A, Kashiwagi T. A combined power cycle using
refuse incineration and LNG cold energy. Energy 2000;25:639e55.
[10] Lu T, Wang KS. Analysis and optimization of a cascading power cycle with
liquefied natural gas (LNG) cold energy recovery. Appl Thermal Eng 2009;29:
1478e84.
[11] Staicovici MD. Further research zero CO2 emission power production.
‘COOLENERG’ Process Energy 2002;27:831e44.
195
[12] Deng S, Jin H, Cai R, Lin R. Novel cogeneration power system with liquefied
natural gas (LNG) cryogenic exergy utilization. Energy 2004;29:497e512.
[13] Zhang N, Lior N. A novel near-zero CO2 emission thermal cycle with LNG
cryogenic exergy utilization. Energy 2006;31:1666e79.
[14] Zhang F. LNG cold energy recovery and power generation. In: Power and
energy engineering conference. Wuhan, China 2009. p. 1e3.
[15] Vélez F, Segovia J, Chejne F, Antolín G, Quijano A, Martín C. Low temperature
heat source for power generation: exhaustive analysis of a carbon dioxide
transcritical power cycle. Energy 2011;36:5497e507.
[16] Qiang W, Yanzhong L, Jiang W. Analysis of power cycle based on cold energy
of liquefied natural gas and low-grade heat source. Appl Thermal Eng
2004;24:539e48.
[17] Shi X, Che D. A combined power cycle utilizing low-temperature waste heat
and LNG cold energy. Energy Convers Manage 2009;50:567e75.
[18] Roy JP, Mishra MK, Misra A. Parametric optimization and performance anal-
ysis of a waste heat recovery system using Organic Rankine Cycle. Energy
2010;35:5049e62.
[19] Chen H, Goswami DY, Rahman MM, Stefanakos EK. A supercritical Rankine
cycle using zeotropic mixture working fluids for the conversion of low-grade
heat into power. Energy 2011;36:549e55.
[20] Wang D, Ling X, Peng H, Liu L, Tao L. Efficiency and optimal performance
evaluation of organic Rankine cycle for low grade waste heat power gener-
ation. Energy 2013;xxx:1e10.
[21] Baik YJ, Kim MS, Chang KC, Lee YS, Yoon HG. Power enhancement potential of
a mixture transcritical cycle for a low-temperature geothermal power gen-
eration. Energy 2012;47:70e6.
[22] Wang J, Yan Z, Wang M, Dai Y. Thermodynamic analysis and optimization of
an ammoniaewater power system with LNG (liquefied natural gas) as its heat
sink. Energy 2013;50:513e22.
[23] Liu Y, Guo K. A novel cryogenic power cycle for LNG cold energy recovery.
Energy 2011;36:2828e33.
[24] Szargut J, Szczygiel I. Utilization of the cryogenic exergy of liquid natural gas
for the production of electricity. Energy 2009;34:827e37.
[25] Yamada N, Minami T, Mohamad MNA. Fundamental experiment of pumpless
Rankine-type cycle for low-temperature heat recovery. Energy 2011;36:
1010e7.
[26] Yuanwei L, Hongchang Y, Chongfang M. Analysis and optimization of the
power cycle based on the cold energy of liquefied natural gas. In: Interna-
tional conference on measuring technology and mechatronics automation.
Shanghai, China 2011. p. 455e8.
[27] Peters MS, Timmerhaus KD, West RE. Plant design and economics for chemical
engineers. 5th ed. New York: McGraw-Hill; 2003.