Accepted Manuscript

Comparative thermodynamic analysis of compressed air and liquid air energy storage
systems

Piotr Krawczyk, Łukasz Szabłowski, Sotirios Karellas, Emmanuel Kakaras, Krzysztof

Pobrano z http://repo.pw.edu.pl / Downloaded from Repository of Warsaw University of Technology 2022-11-15

Badyda

PII: S0360-5442(17)31255-0
DOI: 10.1016/j.energy.2017.07.078
Reference: EGY 11266

To appear in: Energy

Please cite this article as: Piotr Krawczyk, Lukasz Szabłowski, Sotirios Karellas, Emmanuel Kakaras,
Krzysztof Badyda, Comparative thermodynamic analysis of compressed air and liquid air energy storage
systems, (2017), doi: 10.1016/j.energy.2017.07.078

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.
 ACCEPTED MANUSCRIPT
Comparative thermodynamic analysis of
compressed air and liquid air energy storage
systems

Piotr Krawczyka, Łukasz Szabłowskib, Sotirios Karellasc, Emmanuel Kakarasd and

Krzysztof Badydae

PT
a,b,e
Institute of Heat Engineering, Warsaw University of Technology, Warsaw, Poland
c,d
National Technical University of Athens, Athens, Greece
a
piotr.krawczyk@itc.pw.edu.pl
b
lukasz.szablowski@itc.pw.edu.pl

RI
c
sotokar@mail.ntua.gr
d
ekak@central.ntua.gr
e
krzysztof.badyda@itc.pw.edu.pl

SC
Abstract:

U
The paper presents a thermodynamic analysis of selected CAES and LAES systems. The LAES cycle is a
combination of an air liquefaction cycle and a gas turbine power generation cycle. CAES and LAES systems
Pobrano z http://repo.pw.edu.pl / Downloaded from Repository of Warsaw University of Technology 2022-11-15

AN
are simulated using Aspen HYSYS software. CAES is modeled in a dynamic mode. A comprehensive
thermodynamic analysis was conducted along with the comparison of storage volumes. The results indicate
that both systems are characterized by high energy storage efficiency, equal to approximately 40% for the
CAES and 55% for the LAES systems. One clear advantage of the LAES over the CAES is the significantly
M

lower volume demanded for energy storage. For the considered LAES system,the liquid air tank volume is
3 3
around 5000 m , while for the CAES the cavern volume is approximately 310000 m . Heat exchangers and
combustion chambers are the main contributors to the total exergy destruction in the analyzed systems.
D

Keywords:
Liquid air energy storage, Compressed air energy storage, energy analysis, storage efficiency.
TE

Abbreviations

a – attraction parameter [J2/(mol2Pa)],

EP

b – van der Waals covolume [J/(mol Pa)],

B – exergy [kJ],
CF – correction factor [kJ/(m3Pa)],
C

Eel c – energy consumed by compressors and pumps [kJ],

Eel g – energy from generator [kJ],
AC

h – specific enthalpy of the working medium [kJ/kg],

h0 – specific enthalpy of the working medium in ambient conditions [kJ/kg],
h1 – specific enthalpy of working medium at the inlet [kJ/kg],
h2 – specific enthalpy of the working medium at the outlet [kJ/kg],
h’2 – specific enthalpy of the working medium at the outlet for the isentropic process [kJ/kg],
I – moment of inertia [kg m2],
k – coefficient representing the inverse of flow resistance (conductivity) [kg/(s Pa0.5)],
M – molecular weight of the working medium [kg/mol];
Mr – the rotating mass [kg];
m& – mass flow [kg/s],
m& shell – flow rate of working medium on the shell side [kg/s],
1
 ACCEPTED MANUSCRIPT
m& tube
– flow rate of working medium on tube side [kg/s],
n – the volume exponent,
n&1 – molar flow rate at the inlet [mol/s],
P – power [kW],
p – pressure [Pa],
p1 – inlet pressure [Pa],
p2 – outlet pressure [Pa].
Q – heat transfer between shell side and tube side [kW],
Qf – chemical energy of fuel [kJ],

PT
Qloss– heat losses [kW],
R – gas constant [kJ/(mol K)],
Rg – the radius of gyration [m],
S – entropy [kJ/K]

RI
s2 – specific entropy of working fluid at outlet [kJ/(kgK)],
s0 - specific entropy of the working fluid in ambient conditions [kJ/(kgK)],
T – absolute temperature [K],

SC
u – specific internal energy of the working medium [kJ/kg],
u2 – specific internal energy of the working medium at the outlet [kJ/kg],
T0 – ambient temperature [K],

U
v – molar volume [m3/mole],
V – volume of holdup in the shell or in the tube [m3]
Pobrano z http://repo.pw.edu.pl / Downloaded from Repository of Warsaw University of Technology 2022-11-15

AN
Greek symbols
ηcycle – round trip efficiency of a cycle [-],
ρ – density [kg/m3],
M

ρ’2 – density of the working medium at the outlet for the isentropic process [kg/m3],
ρ1 – inlet density of the working medium [kg/m3],
ω – the rotation speed [rad/s].
D

1. Introduction
TE

Energy storage plays an increasingly important role in the current energy system due to the very
intensive development of highly fluctuating, intermittent renewable energy sources.
EP

Currently, there are two methods of large scale energy storage: pumped storage and compressed air
energy storage (CAES). The special geographical requirements involved for the application of these
technologies inhibit their development, since there are relatively few places with suitable conditions
C

for their implementation.

In the case of a CAES power plant, a high volume tank is needed to store large amount of
AC

compressed air. The construction of such tanks has negative financial implications and thus only
natural reservoirs appear to be currently economically viable. Instances of such reservoirs include
salt caverns, aquifers, salt mine workings, mines, and limestone and other mineral formations in
hard rock structures.
Meanwhile, pumped-storage plants require the existence of a water reservoir, such as a lake, located
at the foot of a mountain, on top of which an upper reservoir may be constructed.
An alternative to these technologies is liquid air energy storage (LAES) power plants, which can
store large amounts of energy at decreased storage volumes.
CAES technology has been extensively studied. Currently, there are two large plants of this type:
Huntorf in Germany (290 MW) and McIntosh in the USA (110 MW). The first was launched in
1978, the second in 1991.
An overview of the state of the art about CAES technology is presented in [1-6].

2
 ACCEPTED MANUSCRIPT
In [7] a hybrid system using CAES combined with renewable energy sources (e.g. wind turbines)
was investigated and dynamically modeled, and the possibility of implementing it in Poland was
presented [7, 8].
In [9] a simulation and thermodynamic analysis was performed for a compressed air energy storage-
combined cycle (CAES-CC). The overall efficiency of the system was about 10% higher than the
conventional, non-regenerative reference CAES. According to the authors, the heat obtained from
the compressor intercoolers when charging the air reservoir can be used to keep the steam part of
the system at high temperature at stand-by operation.
In [10] a system consisting of a CAES system and a Kalina Cycle to recover waste heat was
presented. The system had an efficiency higher by 4% compared to a standalone, regenerative

PT
CAES system.
An analysis of the impact of selected parameters (pressure ratio, temperature, mass flow rate of air)
on the performance of an adiabatic CAES system was presented in [11]. The calculations were

RI
performed on the tacit assumption of the authors that the system operates at constant pressure.
In [12] a thermodynamic analysis of an adiabatic CAES system with packed bed regenerators was
presented. In this case, the heat of compression was stored in a solid material, with predicted

SC
efficiencies in the range of 70.5 – 71.1%. The maximum packed bed temperature was 713 K (for a
2-stage system).
A completely different approach to the adiabatic CAES system was presented in [13], where a low
temperature CAES concept was developed. The roundtrip efficiency ranged from 52% to 60% and

U
the maximum temperature in the thermal energy storage was 200°C. According to [13], this
solution uses less expensive system elements and the start-up time is less than 5 minutes.
Pobrano z http://repo.pw.edu.pl / Downloaded from Repository of Warsaw University of Technology 2022-11-15

AN
In [14] the off-design operation of a CAES system integrated into a hybrid power plant connected to
a wind power plant was analyzed. The authors simulated the operation of this system on the Italian
Power Exchange market for one year. The cost analysis was performed using the thermo-economic
M

approach.
In [15] the concept of a small CAES combined with photovoltaic panels and operating in order to
cover the needs of a radio base station was introduced. The heat generated during the air
compression was consumed during the expansion stage of the turbine. An additional function of the
D

presented solution was the production of cooling, due to the fact that the temperature of the air
leaving the turbine was equal to 3°C. Despite the small scale of this CAES system, its storage
TE

efficiency was 57%.

The LAES technology is not as well known and tested as CAES. Currently, in the world there is
only one research facility of this kind, with a power capacity of 350 kW and a capacity of 2.5 MWh,
EP

built in 2011 in the UK. However, because of its simplified configuration it has very poor
performance – its efficiency being around 8% [16].
In [17] a concept and thermodynamic analysis of the adiabatic performance of a LAES were
presented. The efficiency of the proposed solution was 49%.
C

In [18, 19] a hybrid system was presented combining CAES and LAES with the capability of
converting compressed air (50 bar) into liquid air at atmospheric pressure. According to the authors,
AC

this solution is cheaper than the LAES and the CAES system (with the use of artificial compressed
air tanks). A preliminary analysis [18] indicated an efficiency of 53%, but subsequent
calculations [19] showed that the efficiency of such energy storage is 42%.
A comparison between LAES systems based on the composition of the liquefied gases (air, nitrogen
and carbon dioxide) and a CAES system was described in [20]. The authors stated that energy
storage systems using liquefied gases have a promising future in large scale energy storage.
An economic analysis of energy storage systems based on compressed air and liquid air for
different mixes of liquid and gaseous air (from 0 to 100%) was performed in [21].
In [22] an energy storage system based on liquid CO2 operating in a closed circuit was presented.
The proposed system had two storage tanks for liquid CO2 at high and low pressure. During the
charging phase, the high pressure tank was being filled and the low pressure reservoir was being

3
 ACCEPTED MANUSCRIPT
emptied, whereas during the discharging phase the opposite process occurred. The efficiency of this
system was estimated equal to 56.64%.
In [23] a number of cryogenic energy storage systems combined with distributed energy sources
were compared. According to the authors, the best solution for air liquefaction was the Claude cycle
(using an expander and a Joule-Thompson valve) and not the Linde-Hampson cycle (whicn only
included a Joule-Thompson valve). On the other hand, according to the authors of the paper, the use
of cold storage can alter this result, since it reduces by several times the work required to liquefy
air. A part of the working agent flowing through the turbine in the Claude cycle is only used to cool
the air before the Joule-Thompson valve.
An interesting energy storage system has been discussed [24]. This solution uses the Rankine cycle,

PT
with air being used as the working medium. The system can operate continuously. The air is
constantly liquefied and then pumped, heated and finally expanded in the turbine. This system is
equipped with a cryogenic tank for liquid air, which stores energy. The efficiency of such a system

RI
was estimated to range from 20 to 50%.
An analysis of the impact of selected operational parameters on the performance of an adiabatic
LAES system can be found in [25].

SC
This paper presents a comparative analysis of energy storage systems based on liquefied air (LAES)
and on compressed air (CAES). For this purpose, a CAES and a LAES with generated power
outputs of 290 and 270 MW and storage capacities of 1700 and 1080 MWh, respectively, are

U
considered. Both systems use natural gas as an additional fuel. The aim of the paper is to
numerically investigate the efficiency of these systems and identify the main causes of exergy
Pobrano z http://repo.pw.edu.pl / Downloaded from Repository of Warsaw University of Technology 2022-11-15

AN
destruction.
The capacity of the working medium storage tank constitutes a major design parameter and a key
difference between the two systems, as it is much lower for the LAES compared to the CAES.
M

However, the LAES technology requires additional volume for cold storage. Considering the above,
the necessary storage capacity and the unit energy density for both technologies are estimated.
D

2. Simulation
TE

The calculations shown in this section were done using commercial software [26, 27]. Dynamic
calculations of physical phenomena are described using linear and nonlinear differential equations,
which are presented in detail in [28]. In this study, only governing equations are included.
The mass flow rate of the working medium through valves is described by the following formula:
EP

m& = k ∆p (1)
C

The basic equation defining the characteristics of the valves takes into account the coefficient k and
AC

the pressure drop across the valve as a function of the flow resistance:

m& = f (k , p1 , p 2 ) (2)

The relationship describing the energy balance on the shell side of the heat exchanger is given by
the equation:

d (V ⋅u2 )shell
m& shell ⋅ (h1 − h2 )shell − Oloss + Q = ρ (3)
dt

The relationship describing energy balance on the tube side of the heat exchanger is given by the
equation:

4
 ACCEPTED MANUSCRIPT
d (V ⋅ u2 )tube
m& tube ⋅ (h1 − h2 )tube − Q = ρ (4)
dt

Both the polytropic and isentropic power of a compressor or expander can be calculated using (4).
 n −1 
   
 n   p1   p2   n  
P = n&1 ⋅ M ⋅   ⋅ CF ⋅   ⋅   −1 (5)
 n −1  ρ1   p1  
 

The correction factor is expressed by the following equation:

PT
h' 2 − h1
CF = (6)
 n  p 2 p1 
  − 
 n − 1  ρ ' 2 ρ1 

RI
The power provied to the working medium in the compressor is described as follows:
P1−2 = n&1 ⋅ M ⋅ (h2 − h1 )

SC
(7)

The power taken from the working medium in the expander is expressed by the following equation:
P1−2 = n&1 ⋅ M ⋅ (h1 − h2 ) (8)

U
Pobrano z http://repo.pw.edu.pl / Downloaded from Repository of Warsaw University of Technology 2022-11-15

The rotational inertia is given by the equation:

AN
I = M r ⋅ Rg
2
(9)

The power needed to change the rotation speed is:

M

dω
PI = I ω (10)
dt
D

The round trip efficiency of the cycle is described by the following formula:
TE

Eel g
ηcycle = (11)
Eel c + Q f
EP

The exergy destruction is described using the Gouy-Stodola equation [29, 30]:
δB = T0 ∑ ∆S (12)
C

3. Model description and Results

AC

The Aspen HYSYS software is used for the simulation. The air composition is shown in Table 1. The
air properties are assumed to be in accordance with ISO conditions, i.e. a temperature of 15 °C, and
a pressure of 1.013 bar. Both energy storage systems are charged during the valleys of the load of the
power system and discharged at the peaks. Fuel is also consumed while discharging. The fuel used in both
systems was methane with a LHV of 50035 kJ/kg.

3.1 CAES
A schematic diagram of the modelled CAES system is shown in Fig. 1. The installation consists of
the two-stage compressor with the intercooler, the underground storage with the relatively large
volume (salt cavern) and the two-stage gas turbine with two combustion chambers. The air is taken
from the environment and compressed into the cavern during the charging phase, during which
electricity is consumed. When the system is discharging, after being partially depressurized, the air
is transported to the high-pressure combustion chamber and subsequently to the turbine. The

5
 ACCEPTED MANUSCRIPT
expanded air leaving the HPT is transported to the low-pressure part of the gas turbine. After the
expansion process, the exhaust gases are discharged into the environment.
The polytropic efficiency of the rotating machines (turbines and compressors) is assumed equal to
75%. The volume of the cavern is 310000 m3, while the pressure inside it during normal operation
ranges from 43 to 70 bar. The air temperature after the intercoolers is 50 °C, while the pressure
drop at the intercoolers IC2 and IC2 at nominal operating conditions is 0.1 and 0.5 bar, respectively.
The inlet temperature and pressure of the cooling water are 15°C and 2 bar, while the outlet
temperature and pressure are 75 °C and 1.9 bar. The compressor is divided into 2 parts, each
needing 30 MW of power during the charging phase of the plant.
The temperatures after the first and second combustion chambers are respectively 550 and 825 °C.

PT
The pressure at the exit of the first turbine (HPT) is assumed to be 22 bar. The Valve V2 is
controlled to maintain a constant flow (416 kg/s) during the discharging phase.
It should be noted that the parameters described above, except for the efficiency of the rotary

RI
machines, are identical to those of the Huntorf power plant.
The storage charging and discharging processes in the CAES technology are transient. This means
that the operating conditions of the rotating machinery are changing depending on the storage

SC
charge level.
The variation of the pressure inside the cavern as well as the injected air mass flow rate during the
charging phase are shown in Fig. 2. The mass flow rate of the air released from the cavern and the

U
V2 valve position during the discharging phase are depicted in Fig. 3. The variation of the pressure
inside the cavern during the discharging phase are shown in Fig. 4.
Pobrano z http://repo.pw.edu.pl / Downloaded from Repository of Warsaw University of Technology 2022-11-15

AN
During the charging the system, there is a non-linear variation of mass flow rate (Fig. 2). The same
behavior would also be observed for the pressure, if it was lower (if the cavern would be filled
starting from atmospheric pressure). The same effect can be observed for the pressure during the
M

discharging phase of the system (Fig. 4).

As mentioned above, while the cavern is being emptied (Fig. 3), the control system maintains the
mass flow rate at a constant level by adjusting the position of the valve V2. The curve showing this
D

parameter is highly nonlinear.

The power consumption of the compressors is 60 MW, while the total energy consumed during the
TE

charging of the system is 1775 MWh.

The power generated in the discharging phase of the CAES cycle is 290 MW, while the energy
generated during the discharging phase is 1723 MWh. The round trip efficiency of the cycle is
EP

approximately 39.8%, taking also into account the chemical energy of the fuel. According to [1] the
round trip efficiency of the Huntorf power plant is about 42%. The difference between those values
may be caused by differences in assumptions of internal efficiencies of rotating machines in the
modelled system and in Huntorf CAES.
C

The energy balance and exergy destruction of the system are shown in Table 2.
AC

3.2 LAES
The LAES cycle contains three principal parts: a charging device, a liquid and various thermal
storage equipment components, and a generation device. Each of these three main parts can be
configured in different ways. The configuration of these elements has a crucial influence on the
efficiency of energy storage.
During the charging phase, excess electricity from the grid is used for air liquefaction. When the
demand for electricity in the grid increases, the air is vaporized and expanded in turbines to
generate electricity.
This paper presents and analyzes a hypothetical LAES system. It is assumed that the charge-to-
discharge time ratio is 2 (charging time – eight hours; discharging time – four hours). The

6
 ACCEPTED MANUSCRIPT
liquefaction plant capacity is assumed at the level of 150 kg/s. The efficiency of the compressors
and turbines is assumed to be 89% and 85%, respectively.
3.2.1. Charging the system
In the analyzed cycle the Hampson-Linde method of air liquefaction is used. In this method, the
Joule-Thomson effect is used to cool the gas. In addition, for a LAES plant it is possible to use an
external source of cooling produced from the air evaporation process during the discharging phase.
This is of fundamental importance for achieving high storage efficiencies. A diagram of the
analyzed LAES system is shown in Fig. 5.
In the proposed system, the air is compressed to 120 bar in a three-stage compressor equipped with

PT
intercoolers. Subsequently, it flows to a multi-stream heat exchanger in which it is condensed and
cooled to temperatures bdlow -180°C. After passing through the heat exchanger, the air is directed
towards the Joule Thomson valve, where it expands to a pressure of 1.5 bar and is thus further
cooled. The cold air is then directed to the separator, in which the separation of the liquid and gas

RI
phases takes place.
The cold air in gaseous form is directed through the multi-stream heat exchanger to the compressor

SC
inlet. Additional streams from cold storage are introduced to the multi-stream heat exchanger (Cold
Box).
The power consumption to drive the compressors in the analyzed liquefaction cycle is 98.8 MW
(corresponding to a specific energy consumption of 0.184 kWh/kg liquid air). Achieving such low

U
specific energy consumption is possible only with the implementation of cooling from the air
Pobrano z http://repo.pw.edu.pl / Downloaded from Repository of Warsaw University of Technology 2022-11-15

evaporation process during the discharging of stored energy, which was described above. The
AN
energy consumption during the charging phase of the system is 790.4 MWh.
3.2.2. Cold Recycle
M

The possibility of storing the cooling produced during the air evaporation process and utilizing it for
the liquefaction process is a key feature which enables achieving high energy storage efficiencies.
Among the solutions regarding cold energy storage for LAES cycles described in the literature,
D

solid mineral material (sand, gravel) beds are mainly considered. In the presented LAES cycle,
another type of cold storage is proposed, based on liquid substances. This concept is based on using
TE

two tanks ‒ a “cold” and a “warm” one. During the cooling production, when the cooling storage is
being charged, i.e. when the energy storage system is being discharged, a fluid stream from the
warm tank flows through the air heater to the cold tank. Similarly, during the cooling storage
EP

discharging, i.e. charging of the energy storage system, a fluid stream from the cold tank goes to the
multi-stream heat exchanger where it is heated and subsequently accumulated inside the warm tank.
A technical obstacle for the implementation of this concept lies in finding substances that remain
liquid at such low temperatures. The only substances that meet the above condition are
C

hydrocarbons.
In the analyzed LAES cycle, a two-step cooling storage is proposed. In the first stage, liquid
AC

propane (refrigerant R290) is used as a cooling medium (Boiling point = -42°C; Freezing point = -
189°C). In the second stage liquid methanol was used (Boiling point = 65°C; Freezing point = -
97.6°C).

3.2.3. Power Recovery

During the discharging of the energy storage system, the pressure of the liquid air is firstly
increased by a pump. In the next step, the air is evaporated and superheated. The heat required for
this purpose is supplied by refrigerant R290, which is cooled from -60°C (the temperature in the
warm tank) to -185°C (cold tank temperature). The air ‒ in gaseous form ‒ then flows into the
second superheater where its temperature is increased to about 20°C. At the same time the methanol
temperature is lowered from 25°C (warm tank temperature) to - 60°C (cold tank temperature). In
the next step, the air is heated in the regenerative heat exchanger and flows into the combustion

7
 ACCEPTED MANUSCRIPT
chamber. The gas exiting the turbine is cooled in the regenerative heat exchanger. Table 3 contains
the basic thermodynamic data in the selected points of the LAES cycle.
In contrast to the CAES system, in the LAES system the compressor and turbine parameters are
constant during the charging and the discharging processes. Moreover, the aforementioned
parameters are independent of the storage charge level.
The power generated during the discharging phase of the LAES cycle is 271.5 MW, while the
energy generated during the energy recovery is 1086 MWh. The round trip efficiency of the cycle is
around 55.2%, taking into account the use of the chemical energy of the fuel.
Achieving such high efficiency is possible only due to use of the cooling produced during the air
evaporation process in liquefaction, since it allows attaining an air temperature before the J-T valve

PT
at a range of -180°C. In the simple liquefaction system (without cooling storage), the maximum
value of this temperature is approximately -135°C. The lower the temperature of the air before the
J-T valve, the lower the gas-to-liquid ratio in the air after the throttling valve. Consequently, a

RI
smaller fraction of the working medium is being recycled.
In the considered system, approximately only 10% of the air flow passing the compressors,
recirculates back at the first point of the liquefaction installation. On the other hand, when the air

SC
temperature before the J-T valve equals to -135°C and the pressure equals to 120 bar, 65% to 70%
of the air stream has to be recycled. In this case, in order to keep constant the performance of the
liquefaction system, the air flow rate to the compressors is increased, and thus the energy

U
expenditure of the process is higher.
Table 4 shows the exergy destruction in selected parts of the analyzed LAES cycle. Exergy
Pobrano z http://repo.pw.edu.pl / Downloaded from Repository of Warsaw University of Technology 2022-11-15

AN
destruction occurs mainly in the heat exchangers and air throttling.

3.3 Comparison of calculation results

M

The comparison of the major parameters for both systems is presented in the Table 5.
The data presented indicates that both systems can generate electricity with relatively similar power.
The CAES system is characterized by a larger capacity, which is 58% higher. However, the fuel
D

consumption for the single storage cycle in the CAES is more than two times higher compared to
that of the LAES system. The aforementioned feature is reflected in the energy efficiency of both
TE

energy storage systems.

Beyond the efficiency, the unit energy density achievable in the storage was compared for both
technologies. The major data for both systems are presented in the Table 6.
EP

In this analysis, the volume of technological devices is neglected. The only ones taken into account
are: the volume of the working medium tank and the volume of the cooling storage tanks (in the
LAES case). Based on the obtained results, the LAES system is characterized by approximately six
times greater unit energy density in comparison with the CAES system, which is an undoubtedly
C

important advantage of the former. The analysis also covered the comparison of the exergy
destruction of these two technologies. The results are presented in Table 7 and Fig. 6.
AC

For both systems, the largest exergy destruction occurred in the combustion chambers. According to
the results, the LAES system unit exergy destruction is lower by about approximately 40%
compared to that of the CAES system.

4. Conclusions
The presented technologies can be applied for energy storage in large scale systems. In this paper, a
comparison of the relatively well examined technology based on compressed air with the currently
developing technology based on liquefied air was presented.
In both described cases, the heat generated during the air compression is not stored. The CAES
system has 2 air coolers (IC1 and IC2), while the LAES has 3. This heat could be used for district
heating, which would increase the overall efficiency of the entire system (the energy could be

8
 ACCEPTED MANUSCRIPT
consumed at the same time by compressors of the CAES or LAES systems). Both systems have a
combustion chamber (CAES – 2 and LAES – 1). The round trip efficiency of the CAES power plant
was 39.77%, whereas the efficiency of the LAES was 55.2%.
The LAES system was additionally equipped with cooling storage, which lead to a significant
increase of the efficiency of the liquefaction process. However, the higher round trip efficiency of
the modeled LAES compared to the CAES derived from the application of regeneration (pre-
heating the air before the combustion chamber with heat from the turbine exhaust gases).
Without the regeneration, both systems (CAES and LAES) would have comparable round trip
efficiency (LAES – about 40%).
The results presented in the article prove that the LAES technology may have greater benefits than

PT
the CAES solution. The major benefit is the larger density of the energy stored. As mentioned
earlier, the LAES technology is not constrained by location, unlike the CAES. The storage
capacities for both systems can be similar. The LAES higher efficiency also leads to lower exergy

RI
destruction.
Considering the above, in the opinion of the authors, the LAES technology is potentially the
prospective alternative for the currently used CAES technology.

SC
Acknowledgments

U
The first two authors were supported by the European Union in the framework of the European
Pobrano z http://repo.pw.edu.pl / Downloaded from Repository of Warsaw University of Technology 2022-11-15

AN
Social Fund through the “Didactic Development Program of the Faculty of Power and Aeronautical
Engineering of the Warsaw University of Technology”.

References
M

[1] Luo, X., J. Wang, M. Dooner, J. Clarke, and C. Krupke: 2014, Overview of current
development in compressed air energy storage technology. Energy Procedia 62, 603–611.
D

[2] Lund, H., G. Salgi, B. Elmegaard, and A. N. Andersen: 2009, Optimal operation strategies of
compressed air energy storage (CAES) on electricity spot markets with fluctuating prices.
TE

Applied thermal engineering 29(5), 799–806.

[3] Mason, J. E. and C. L. Archer: 2012, Baseload electricity from wind via compressed air energy
storage (CAES). Renewable and Sustainable Energy Reviews 16(2), 1099–1109.
[4] Satkin, M., Y. Noorollahi, M. Abbaspour, and H. Yousefi: 2014, Multi criteria site selection
EP

model for wind-compressed air energy storage power plants in Iran. Renewable and Sustainable
Energy Reviews 32, 579 – 590.
[5] Yucekaya, A.: 2013, The operational economics of compressed air energy storage systems
C

under uncertainty. Renewable and Sustainable Energy Reviews 22, 298–305.

AC

[6] Milewski J, Badyda K, Szablowski L. Compressed air energy storage systems, Journal of
Power Technologies 2016;96(4):245-260.
[7] Badyda K., Milewski J., Compressed air storage systems as a peak looping power station in
Polish conditions, in: Proceeding of International Conference on Power Engineering, ICOPE
2009, 2009.
[8] Badyda, K. and J. Milewski: 2012, Thermodynamic analysis of compressed air energy storage
working conditions. Archiwum Energetyki 42(1), 53–68.
[9] Liu, W., L. Liu, L. Zhou, J. Huang, Y. Zhang, G. Xu, and Y. Yang: 2014, Analysis and
Optimization of a Compressed Air Energy Storage—Combined Cycle System. Entropy 16(6),
3103–3120.

9
 ACCEPTED MANUSCRIPT
[10] Zhao, P., J. Wang, and Y. Dai: 2015, Thermodynamic analysis of an integrated energy
system based on compressed air energy storage (CAES) system and Kalina cycle. Energy
Conversion and Management 98, 161–172.
[11] Jubeh, N. M. and Y. S. Najjar: 2012, Green solution for power generation by adoption of
adiabatic CAES system. Applied Thermal Engineering 44, 85–89.
[12] Barbour, E., D. Mignard, Y. Ding, and Y. Li: 2015, Adiabatic Compressed Air Energy
Storage with packed bed thermal energy storage. Applied Energy 155, 804–815.
[13] Wolf, D. and M. Budt: 2014, LTA-CAES–a low-temperature approach to adiabatic
compressed air energy storage. Applied Energy 125, 158–164.

PT
[14] De Bosio, F. and V. Verda: 2015, Thermoeconomic analysis of a Compressed Air Energy
Storage (CAES) system integrated with a wind power plant in the framework of the IPEX
Market. Applied Energy 152, 173–182.

RI
[15] Jannelli, E., M. Minutillo, A. L. Lavadera, and G. Falcucci: 2014, A small-scale CAES
(compressed air energy storage) system for stand-alone renewable energy power plant for a
radio base station: A sizing-design methodology. Energy 78, 313–322.

SC
[16] Morgan, R., S. Nelmes, E. Gibson, and G. Brett: 2015, Liquid air energy storage–Analysis
and first results from a pilot scale demonstration plant. Applied Energy 137, 845–853.
[17] Xue, X., S. Wang, X. Zhang, C. Cui, L. Chen, Y. Zhou, and J. Wang: 2015, Thermodynamic

U
Analysis of a Novel Liquid Air Energy Storage System. Physics Procedia 67, 733–738.
[18] Kantharaj, B., S. Garvey, and A. Pimm: 2014, Thermodynamic analysis of a hybrid energy
Pobrano z http://repo.pw.edu.pl / Downloaded from Repository of Warsaw University of Technology 2022-11-15

AN
storage system based on compressed air and liquid air. Sustainable Energy Technologies and
Assessments.
[19] Kantharaj, B., S. Garvey, and A. Pimm: 2015, Compressed air energy storage with liquid air
capacity extension. Applied Energy 157, 152–164.
M

[20] Wang, S., X. Xue, X. Zhang, J. Guo, Y. Zhou, and J. Wang: 2015b, The Application of
Cryogens in Liquid Fluid Energy Storage Systems. Physics Procedia 67, 728–732.
D

[21] Pimm, A. J., S. D. Garvey, and B. Kantharaj: 2015, Economic analysis of a hybrid energy
storage system based on liquid air and compressed air. Journal of Energy Storage 4, 24–35.
TE

[22] Wang, M., P. Zhao, Y. Wu, and Y. Dai: 2015a, Performance analysis of a novel energy
storage system based on liquid carbon dioxide. Applied Thermal Engineering 91, 812–823.
[23] Abdo, R. F., H. T. Pedro, R. N. Koury, L. Machado, C. F. Coimbra, and M. P. Porto: 2015,
EP

Performance evaluation of various cryogenic energy storage systems. Energy 90, 1024–1032.
[24] Ameel, B., C. T’Joen, K. De Kerpel, P. De Jaeger, H. Huisseune, M. Van Belleghem, and
M. De Paepe: 2013, Thermodynamic analysis of energy storage with a liquid air Rankine cycle.
Applied Thermal Engineering 52(1), 130–140.
C

[25] Krawczyk P, Szablowski L, Badyda K, Karellas S, Kakaras E. Impact of selected parameters

AC

on performance of the adiabatic liquid air energy storage system, Journal of Power
Technologies 2016;96(4):238-244.
[26] HYSYS 3.2 Operations Guide. AspenTech. 2003.
[27] HYSYS 3.2 Dynamic Modeling. Hyprotech. 2003.
[28] Szabłowski, Ł. and J. Milewski: 2011, Dynamic analysis of compressed air energy storage
in the car. Journal of Power Technologies 91(1), 23–36.
[29] Szargut, J. and R. Petela: 1965, Egzergia. Warszawa: Wydawnictwa Naukowo-Techniczne.
[30] Szargut, J.: 1996, Exergy analysis of thermal processes [Analiza egzergetyczna procesów
cieplnych]. Bulletin of Institute of Heat Engineering (currently Jourmal of Power Technologies)
84.

10
 ACCEPTED MANUSCRIPT

Table 1. Mass composition of air

Component Unit Value
Oxygen %wt 23.052
Nitrogen %wt 74.99
Carbon dioxide %wt 0.046
Argon %wt 1.276
Water %wt 0.636

PT
Table 2. Results for CAES model
Parameter Unit Value
Efficiency % 39.77

RI
Energy
LPC MWh/cycle 887.5
HPC MWh/cycle 887.5

SC
HPT MWh/cycle 283.7
LPT MWh/cycle 1439.6
CCH 1 MWh/cycle 1400.5
CCH 2 MWh/cycle 1157.8

U
Exergy destruction
LPC MWh/cycle 151.5
Pobrano z http://repo.pw.edu.pl / Downloaded from Repository of Warsaw University of Technology 2022-11-15

AN
HPC MWh/cycle 140.0
IC 1 MWh/cycle 206.5
IC 2 MWh/cycle 247.7
V1 MWh/cycle 0.28
M

Cavern MWh/cycle 7.16

V2 MWh/cycle 94.4
CCH 1 MWh/cycle 802.3
CCH 2 MWh/cycle 401.5
D

V3 MWh/cycle 0.56
V4 MWh/cycle 2.59
TE

HPT MWh/cycle 35.24

LPT MWh/cycle 162.6

Table 3. Thermodynamic data in the selected nodes of analyzed LEAS system

EP

Node Pressure, bar Temperature, °C Mass flow, kg/s

1 1 15 150
2 1 3.2 166.1
3 5 186.6 166.1
C

4 4.9 20 166.1
5 25 219 166.1
AC

6 24.7 30 166.1
7 120 227.8 166.1
8 119 25 166.1
9 118 -183.9 166.1
10 1.5 -190.7 166.1
11 1.5 -190.7 16.1
12 1.5 -190.7 150
13 1.4 -129.3 16.1
14 1.5 -185 230.5
15 1.4 -60 230.5
16 1.5 -60 49
17 1.4 25 49
18 1.5 -190.7 300
19 22 -189.4 300
1
 20
ACCEPTED
21.5
MANUSCRIPT
-65.6 300
21 21.2 21.1 300
22 21 550 300
23 21 1300 306
24 1.03 582 306
25 1 84 306
26 21 15 6
27 2 25 98
28 1.5 -60 98
29 2 -60 461
30 1.5 -185 461

PT
Table 4. Exergy destruction in selected parts of the analyzed LAES cycle

RI
LAES cycle Unit Exergy destruction
Efficiency % 55.2
Charging system - liquefaction

SC
Mixer MWh/cycle 7.8
Intercooler 1 MWh/cycle 36.3
LPC MWh/cycle 22.8
Intercooler 2 MWh/cycle 48.2

U
MPC MWh/cycle 23.2
Intercooler 3 MWh/cycle 54.5
Pobrano z http://repo.pw.edu.pl / Downloaded from Repository of Warsaw University of Technology 2022-11-15

AN
HPC MWh/cycle 22.4
Cold Box MWh/cycle 31.3
J-T valve MWh/cycle 64.5
Liquid air separator MWh/cycle 0.1
M

Power Recovery
Pump MWh/cycle 5.2
HE1 MWh/cycle 117.9
HE2 MWh/cycle 3.9
D

Regenerative HE MWh/cycle 36.7

Combustion chamber MWh/cycle 336.6
TE

Turbine MWh/cycle 46.1

Table 5. The comparison of the major parameters for the analyzed storage systems
Parameter CAES cycle LAES cycle
EP

Energy consumption during charging 1775 MWh/cycle 790.4 MWh/cycle

the system
Power generated during discharging 290 MW 271.5 MW
C

Energy generated during energy 1722 MWh/cycle 1086 MWh/cycle

recovery
AC

Supplied chemical energy of fuel 2557 MWh/cycle 1200 MWh/cycle

Round trip efficiency of the cycle 39.8% 55.2%

Table 6. The comparison of the major storage volume required for the analyzed systems
Parameter CAES cycle LAES cycle
Volume of the cavern 310 000 m3 -
Volume of the liquid air tank - 5000 m3
Low temperature cold storage – 2 - 2 x 12000 m3
tanks (warm and cold)
High temperature cold storage – 2 - 2x 2000 m3
tanks (warm and cold)
Total volume of tanks 310 000 m3 33 000 m3
3
Unit volume density of the energy 180 m /MWh 30 m3/MWh
stored

2
 ACCEPTED MANUSCRIPT
Table 7. The comparison of exergy destruction for the analyzed systems
Parameter CAES cycle LAES cycle
Total exergy destruction, 2250 MWh/cycle 860 MWh/cycle
The energy generated during energy 1722 MWh/cycle 1086 MWh/cycle
recovery
Unit exergy destruction 1.3 MWh/MWh 0.8 MWh/MWh

PT
RI
U SC
Pobrano z http://repo.pw.edu.pl / Downloaded from Repository of Warsaw University of Technology 2022-11-15

AN
M
D
TE
C EP
AC

3
 ACCEPTED MANUSCRIPT

Fuel 1 Fuel 2
0 or 4.7 kg/s 0 or 3.9 kg/s
15°C, 50 bar 15°C, 50 bar

Water 1 in Water 2 in
14
122.6 kg/s 11
105.5 kg/s
15°C 15°C
V3 V4
3
7
Air in
max. 91 kg/s IC1 50°C IC2 12 15
15°C, 1 bar 9 10 550°C, 43 bar 13 825°C, 22 bar
1

PT
5
4 CCH1 CCH2
6
HPC 8
LPC
V1 V2 16
2 HPT LPT Exhaust gases

RI
Water 1 out Water 2 out 338°C, 1 bar
75°C 75°C
50°C 0 or 416 kg/s
50°C, 43-70 bar

SC
Cavern
310 000 m3

Fig. 1. Schematic of the modeled CAES power plant

U
Pobrano z http://repo.pw.edu.pl / Downloaded from Repository of Warsaw University of Technology 2022-11-15

AN
Mass flow Pressure
95 80

70
M

90
60
Mass flow, kg/s

85 50
Pressure, bar
D

40
80 30
TE

20
75
10
EP

70 0
0 3 6 9 12 15 18 21 24 27

Time, h
C

Fig. 2. Changes of pressure inside the cavern and injected air mass flow during charging
AC
 ACCEPTED MANUSCRIPT
Mass flow Valve opening
450 120

400
100
350

300 80

Valve opening, %
Mass flow, kg/s

250
60
200

PT
150 40

100
20

RI
50

0 0
0 1 2 3 4 5

SC
Time, h
Fig. 3. Mass flow of the air released from the cavern and V2 valve openness during discharging

U
75
Pobrano z http://repo.pw.edu.pl / Downloaded from Repository of Warsaw University of Technology 2022-11-15

AN
70

65
M
Pressure, bar

60

55
D

50
TE

45

40
EP

0 1 2 3 4 5 6
Time, h

Fig. 4. Changes of pressure inside the cavern during discharging

C
AC
 ACCEPTED MANUSCRIPT
16.1 kg/s, -129.3°C

Water in Water in Water in 29 461 kg/s, -60°C

13 25
26
166.1 kg/s
Make up air 1 2 3 4 5 6 7 25°C, 119 bar 22 23
150 kg/s IC2
IC1 IC3
Mixer Warm tank 1 Turbine
8
Combustion chamber

15 -60°C Warm tank 2

17
LPC MPC 24
HPC 21
25°C
Water out Water out Water out 306 kg/s, 582°C
27
Cold box 98 kg/s

9 -60°C 16

11 14 Cold tank 2 Superheater

230.5 kg/s 28
-60°C
10

PT
20
Cold tank 1
12
Evaporator
-
150 kg/s 30
185°C
-190.7°C, 1.5 bar

18
19 300 kg/s

RI
300 kg/s -189.4°C, 22 bar

Liquid Air Tank

5000 m3

Fig. 5. Schematic of the modeled LAES system

SC
Exergy destruction in CAES system Exergy destruction in LAES system
HPT LPT LPC Liquid air Pump
V 4 2%

U
7% 7% separator 1% HE2
0% Regenerative
HPC 0%
V3 HE1 0% HE
6% J-T valve
0% 14% 4%
Pobrano z http://repo.pw.edu.pl / Downloaded from Repository of Warsaw University of Technology 2022-11-15

8%
AN
Cold Box
IC 1 4%
CCH 2 9%
18% HPC
3%
M

Intercooler 3
6%
IC 2
11%
MPC
V1 3%
D

0%

Cavern Intercooler 2
0%
TE

6% LPC
V2 3%
4% Combustion
CCH 1 Intercooler 1
4% Mixer chamber
36% 39%
1% Turbine
5%
EP

Fig. 6. Comparison of exergy destruction in CAES and LAES systems

C
AC
 ACCEPTED MANUSCRIPT
Research Highlights
• Research on compressed air and liquid air energy storage systems;
• Analysis of cold recycle for the LAES cycle;
• Comparison of two energy storage systems;

PT
RI
U SC
Pobrano z http://repo.pw.edu.pl / Downloaded from Repository of Warsaw University of Technology 2022-11-15

AN
M
D
TE
C EP
AC