i n t e r n a t i o n a l j o u r n a l o f r e f r i g e r a t i o n 4 6 ( 2 0 1 4 ) 8 6 e9 9

Available online at www.sciencedirect.com

ScienceDirect
w w w . i i fi i r . o r g

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

Comparative analysis of various CO2 configurations

in supermarket refrigeration systems*
Vishaldeep Sharma, Brian Fricke*, Pradeep Bansal
Oak Ridge National Laboratory, PO Box 2008, Oak Ridge, TN 37831-6070, USA

article info

abstract

Article history:

This paper presents an analysis of various CO2 transcritical and cascade/secondary loop

Received 11 April 2013

refrigeration systems that are becoming popular in supermarket applications with the

Received in revised form

objective of optimizing the operating parameters of these systems. In addition, the per-

10 June 2014

formance of selected CO2-based refrigeration systems is compared to the baseline R404A

Accepted 2 July 2014

multiplex direct expansion system using bin analyses in the eight climate zones of the

Available online 9 July 2014

United States. For the refrigeration systems investigated, it was found that the Transcritical Booster System with Bypass Compressor (TBS-BC) had the lowest energy con

Keywords:

sumption for ambient temperatures (Tamb) less than 8

Refrigeration

temperatures the R404A direct expansion system was found to have the lowest energy

Transcritical booster system

consumption. Also, the TBS-BC performs equivalent to or better than the R404A direct

Cascade system

expansion system in the northern two-thirds of the US. For the southern portion of the US,

Secondary loop system

the R404A multiplex DX system performs better than CO2 systems.

Carbon dioxide

C, and for higher ambient

2014 Elsevier Ltd and IIR. All rights reserved.

Thermodynamics

Analyse comparative de diverses configurations au CO2 dans

mes frigorifiques de grandes surfaces
des syste
frige
ration ; Syste

me de suralimentation transcritique ; Syste

me en cascade ; Syste

me a

boucle secondaire ; Dioxyde de
Mots cles : Re
carbone ; Thermodynamique

1.

Introduction

Supermarkets and other large food retail stores commonly

utilize multiplex direct expansion (DX) refrigeration
*

systems in conjunction with synthetic refrigerants such as

R22, R404A and R507. The estimated annual refrigerant leak
rate for these systems ranges from 3% to 35% for in-use
equipment, with the higher annual leak rates (>25%) being

Notice: This manuscript has been authored by UT-Battelle, LLC, under Contract No. DE-AC05-00OR22725 with the U.S. Department of
Energy. The United States Government retains and the publisher, by accepting the article for publication, acknowledges that the United
States Government retains a non-exclusive, paid-up, irrevocable, world-wide license to publish or reproduce the published form of this
manuscript, or allow others to do so, for United States Government purposes.
* Corresponding author. Tel.: 1 865 576 0822; fax: 1 865 574 9332.
E-mail address: frickeba@ornl.gov (B. Fricke).
http://dx.doi.org/10.1016/j.ijrefrig.2014.07.001
0140-7007/ 2014 Elsevier Ltd and IIR. All rights reserved.

i n t e r n a t i o n a l j o u r n a l o f r e f r i g e r a t i o n 4 6 ( 2 0 1 4 ) 8 6 e9 9

Nomenclature
CR
h
m_
MR
P
Q_
T
DT
_
W
WR

circulation ratio
enthalpy (kJ kg1)
mass flow rate (kg s1)
mass flow ratio
pressure (MPa)
rate of heat transfer (W)
temperature (
C)
evaporator superheat
power (W)
work ratio

Greek Symbols

heat exchanger effectiveness

Subscripts
amb
ambient
app
approach
BP
bypass
circ
refrigerant circuit
comp
compressor
cond
condensing
GC
gas cooler
HP
high-pressure
int
intermediate
LP
low-pressure
LT
low-temperature

more characteristic of older equipment and the lower rates

(<15%) being more characteristic of newer equipment (ICF
Consulting, 2005). The high Global Warming Potential
(GWP) of the hydrofluorocarbon (HFC) refrigerants
commonly used in these systems, coupled with the large
refrigerant charge and the high refrigerant leakage rates
leads to significant direct emissions of greenhouse gases
into the atmosphere. Hence, these multiplex refrigeration
systems can directly contribute to the increase in global
warming.
Furthermore, the operation of refrigeration systems contributes to global warming indirectly. For a typical supermarket in the US, which has an average size of 4200 square
meters, the annual energy consumption is roughly 2 million
kWh, and approximately half of that is for refrigeration
(Westphalen et al., 1996; Zhang, 2006). Thus, the indirect
impact on the environment results from the release of
greenhouse gases (mainly CO2) associated with the generation
and transmission of the electrical energy used by the refrigeration system.
The direct environmental impact of the refrigeration system can be reduced by using refrigerants with lower GWP.
Refrigerants such as R32, R134a, R717, R744, R290, R600a and
R1234yf could be potential alternative refrigerants. However,
due to toxicity and/or flammability, some of these refrigerant
options may not be permissible under various municipal
safety codes. Cascade systems and secondary loop systems
(discussed later in Section 2) using CO2 as a refrigerant can be
used to reduce the direct impact on the environment due to
their lower HFC refrigerant charge.

MT
pump
ref
SLHX
total

87

medium-temperature
pump
refrigerating capacity
suction-liquid line heat exchanger
sum around refrigeration cycle

Abbreviations
CFC
chlorofluorocarbon
COP
coefficient of performance
CSC
combined CO2 secondary/cascade system
CSC-G combined glycol secondary/CO2 cascade system
DEC
direct expansion cascade system
DX
direct expansion
GWP
global warming potential
HC
hydrocarbon
HCFC
hydrochlorofluorocarbon
HFC
hydrofluorocarbon
HTC
high temperature circuit
LTC
low temperature circuit
ODP
ozone depletion potential
SC
secondary coolant system
SLHX
suction-liquid line heat exchanger
STBS
standard transcritical booster system
TBS-BC transcritical booster system with bypass
compressor
TBS-UX transcritical booster system with upstream
expansion valve

The indirect environmental impact of the refrigeration

system can be reduced by increasing the energy efficiency or
decreasing the energy consumption of the system. One option
for decreasing energy consumption is to reduce the load on
the refrigeration system. For example, this can be done by
replacing open display cases with doored display cases.
Several studies have shown that doored display cases can
reduce refrigeration system energy consumption by up to 50%
when combined with high efficiency display case components
such as LED lighting, demand defrost, electronically commutated evaporator fan motors and humidity controlled antisweat heaters (Rauss et al., 2008; Fricke and Becker, 2010).
Other energy efficiency measures that can be utilized include
variable speed drives for compressors and condenser fan
motors, as well as floating condensing and suction pressure
controls.
Carbon dioxide has recently received considerable attention as an alternative to the commonly used synthetic refrigerants in supermarket refrigeration systems, in an effort to
develop systems with lower environmental impact (Bansal,
2012; Getu and Bansal, 2008). Although CO2 has a high critical pressure (7.38 MPa) and a low critical temperature
(30.97
C), its high operating pressure leads to a high vapor
density and thus a high volumetric refrigerating capacity. The
volumetric refrigerating capacity of CO2 (22,545 kJ m3 at 0
C)
is 3e10 times larger than CFC, HCFC, HFC and HC refrigerants
(Kim et al., 2004). In addition, carbon dioxide has no Ozone
Depletion Potential (ODP); a GWP of one; and is nontoxic,
nonflammable and inexpensive e all attractive characteristics
when compared to synthetic refrigerants.

88

i n t e r n a t i o n a l j o u r n a l o f r e f r i g e r a t i o n 4 6 ( 2 0 1 4 ) 8 6 e9 9

Table 1 e Schematics and P-h diagrams of various CO2-based refrigeration systems.

Schematic

P-h diagram

Features
 R404A is used as refrigerant
in the system
 System has a rack of
compressors and an SLHX






O reciever um
separador de lquido
vapor

CO2 low-temperature circuit

R404A high-temperature circuit
CO2 receiver
Liquid CO2 is pumped to
either LT or MT evaporators
 R404A circuit has a rack
of compressors

 CO2 low-temperature circuit

 R404A high-temperature circuit
 Both circuits have an SLHX and
a rack of compressors

89

i n t e r n a t i o n a l j o u r n a l o f r e f r i g e r a t i o n 4 6 ( 2 0 1 4 ) 8 6 e9 9

Table 1 e (continued )
Schematic

P-h diagram

Features





CO2 low-temperature circuit

R404A high-temperature circuit
CO2 receiver
LT load is cooled with direct
expansion CO2 while MT load is
cooled with pumped liquid CO2
 Both cascade circuits have a
rack of compressors and an SLHX






CO2 low-temperature circuit

R404A high-temperature circuit
CO2 receiver
LT load is cooled with direct
expansion CO2 while MT load is
cooled with chilled glycol-water
solution
 Both cascade circuits have a
rack of compressors and an SLHX

O glicol no muda de
fase

 CO2 is used as refrigerant

throughout the system
 CO2 receiver
 LT and MT loads are cooled via
direct expansion
 The system has high-pressure
and low-pressure compressors
and two SLHXs
 Refrigerant is expanded after
exiting SLHX1

(continued on next page)

90

i n t e r n a t i o n a l j o u r n a l o f r e f r i g e r a t i o n 4 6 ( 2 0 1 4 ) 8 6 e9 9

Table 1 e (continued )
Schematic

P-h diagram

Features
 CO2 is used as refrigerant
throughout the system
 CO2 receiver
 LT and MT loads are cooled
via direct expansion
 The system has high-pressure
and low-pressure compressors
and two SLHXs
 Refrigerant is expanded before
entering SLHX1

 CO2 is used as refrigerant

throughout the system
 CO2 receiver
 LT and MT loads are cooled
via direct expansion
 The system has high-pressure,
low-pressure and bypass
compressors and two SLHXs
 Bypass refrigerant is
compressed before entering
the Gas Cooler

Carbon dioxide has successfully been used as a refrigerant

in the low-temperature circuit of cascade systems, in secondary loop systems, and in transcritical systems (Bansal, 2012;
Girotto et al., 2004; Hinde and Zha, 2009). However, transcritical CO2 systems tend to be more popular in moderate
climates such as Northern Europe where the refrigeration
system operates a majority of the time in the more efficient
subcritical mode (Denecke et al., 2012; Sawalha and Palm, 2003).
In an effort to increase the efficiency of the transcritical CO2
system and to make it applicable to warmer climates, several
researchers have investigated the energy efficiency of various
configurations of the transcritical refrigeration system (Bell,
2004; Ferrandi and Orlandi, 2012; Ge and Tassou, 2009, 2010;
Mazzola et al., 2012; Sarkar and Agrawal, 2010; Sawalha,
2007; Winter and Murin, 2012). However, these studies
focused on a particular system only and lacked any system
performance comparison with various other possible CO2
system configurations. Also, these studies did not optimize the
design parameters such as the evaporator superheat (DT) and
the suction line heat exchanger (SLHX) effectiveness, in order
to maximize the system coefficient of performance (COP). In
this study, a comprehensive analysis of seven CO2-based
refrigeration system configurations that are currently being
used in the supermarket refrigeration industry around the
world is performed. The paper presents a systematic analysis

of each configuration for the same operating conditions. In

addition, the performance of the more energy-efficient CO2based refrigeration systems is compared with that of a baseline R404A multiplex DX system using bin analyses in sixteen
cities from eight climate zones of the United States.

2.
Description of CO2 refrigeration systems
evaluated
The CO2-based refrigeration systems investigated in this
study include cascade and secondary loop systems as well as
transcritical systems, as shown in Table 1. These systems are
briefly described in the following section, along with the performance comparison of the CO2-based refrigeration systems
with the baseline R404A multiplex DX system (System 1).

2.1.

Secondary loop and cascade systems

A secondary loop system is comprised of two circuits, a primary DX circuit and a secondary pumped loop circuit, coupled
through a secondary fluid heat exchanger as shown in System
2 of Table 1. The primary DX circuit typically utilizes R404A
while the secondary loop uses a pumped liquid such as propylene glycol (single-phase) or CO2 (two-phase).

Table 2 e Refrigeration capacity and required power.

_ total
W

DX
SC

Q_ ref Q_ LT m_ 3 h4  h3 or Q_ ref Q_ MT m_ 3 h4  h3
Q_ ref Q_ LT m_ 10 h11  h10 or Q_ ref Q_ MT m_ 10 h11  h10

DEC

Q_ ref Q_ LT m_ 9 h10  h9

Combined System 1 (LT SC MT SC)

Q_ ref Q_ LT;SC Q_ MT;SC

Q_ LT;SC m_ 10 h11  h10 LT
m_ 10 h11  h10 
Q_

_ total m_ 5 h6  h5 ;
W
_ pump W
_ comp
_ total W
W
_ comp m_ 5 h6  h5 ;
W
_ pump m_ 9 h10  h9
W
_ total W
_ LP W
_ HP
W
_ HP m_ 5 h6  h5
_ LP m_ 11 h12  h11 ; W
W
_ pump;LT;SC W
_ pump;MT;SC W
_ comp;LT;SC W
_ comp;MT;SC
_ total W
W
_ comp;LT;SC m_ 5 h6  h5 
W
LT
_ comp;MT;SC m_ 5 h6  h5 
W
MT
_ pump;LT;SC m_ 9 h10  h9 
W
LT
_ pump;MT;SC m_ 9 h10  h9 
W
MT
_ total W
_ pump;MT;SC W
_ comp;LT;DEC W
_ comp;MT;SC
W
_ comp;LT;DEC m_ 11 h12  h11 m_ 5 h6  h5 
W
LT
_ comp;MT;SC m_ 5 h6  h5 
W
MT
_ pump;MT;SC m_ 9 h10  h9 
W
MT
_ total W
_ LP W
_ HP W
_ pump
W
_ HP m_ 5 h6  h5
_ LP m_ 16 h17  h16 ; W
W
_ pump m_ 9 h10  h9
W
_ total W
_ LP W
_ HP W
_ pump
W
_ HP m_ 10 h11  h10
_ LP m_ 16 h17  h16 ; W
W
_ pump m_ 18 h19  h18
W
_ LP W
_ HP
_ total W
W
_ LP m_ 12 h13  h12 ; W
_ HP m_ 16 h17  h16
W
_ total W
_ LP W
_ HP
W
_ HP m_ 16 h17  h16
_ LP m_ 12 h13  h12 ; W
W
_ LP W
_ HP W
_ BP
_ Total W
W
_ LP m_ 12 h13  h12 ; W
_ HP m_ 15 h16  h15 and
W
_ BP m_ 4 h5  h4
W

MT;SC

MT

Combined System 2 (LT DEC MT SC)

Q_ ref Q_ LT;DEC Q_ MT;SC

Q_ LT;DEC m_ 9 h10  h9 LT
Q_ MT;SC m_ 10 h11  h10 MT

CSC

Q_ ref Q_ LT Q_ MT
Q_ LT m_ 14 h15  h14 ; Q_ MT m_ 12 h13  h12

CSC-G

Q_ ref Q_ LT Q_ MT
Q_ m_ 14 h15  h14 ; Q_
LT

STBS
TBS-UX
TBS-BC

MT

m_ 19 h20  h19

Q_ ref Q_ LT Q_ MT
Q_ LT m_ 10 h11  h10 ; Q_ MT m_ 7 h8  h7
Q_ ref Q_ LT Q_ MT
Q_ LT m_ 10 h11  h10 ; Q_ MT m_ 7 h8  h7
Q_ ref Q_ LT Q_ MT
Q_ LT m_ 10 h11  h10 ; Q_ MT m_ 7 h8  h7

i n t e r n a t i o n a l j o u r n a l o f r e f r i g e r a t i o n 4 6 ( 2 0 1 4 ) 8 6 e9 9

Q_ ref

System

91

92

i n t e r n a t i o n a l j o u r n a l o f r e f r i g e r a t i o n 4 6 ( 2 0 1 4 ) 8 6 e9 9

Table 3 e Parameters and their ranges.

Parameter

Range

Tamb
Pump Circulation Ratio (CR)
Heat Exchanger Effectiveness (SLHX)
Medium-Temperature Superheat (DTMT)
Low-Temperature Superheat (DTLT)

0e40
C
1.5, 2.5
0, 0.4, 0.7
10, 15 K
10, 15 K

A cascade system is comprised of separate hightemperature and low-temperature circuits, coupled through
a heat exchanger called the cascade condenser. The cascade
condenser functions as an evaporator for the hightemperature circuit and a condenser for the lowtemperature circuit. Generally, the high-temperature circuit
is a single-stage direct expansion system but the lowtemperature circuit can either be a direct expansion system
or a secondary loop system. For a cascade system with highand low-temperature DX circuits, one or both circuits may
have a SLHX as shown in System 3 of Table 1. The refrigerant
in the high temperature circuit is typically an HFC (R404A in
this case), while CO2 is used in the low-temperature circuit.

2.1.1.

CO2 Secondary Coolant (SC) System

In the SC system (System 2, Table 1), a high temperature

R404A circuit provides cooling to a low-temperature pumped
CO2 secondary loop circuit. The pumped CO2 circuit in turn,
provides cooling for the medium- or low-temperature loads.
In the pressure-enthalpy (P-h) diagram, the DX R404A (1-2-3-45-6) and the pumped CO2 (7-8-9-10-11) circuits are represented
by the inner and the outer domes, respectively.
Refrigeration systems with separate SC systems for the
low- and medium-temperature loads are commonly used in
supermarkets. In this paper, such a paired SC system is called
the Combined System 1. This system would be represented by
two separate System 2 cycles as shown in Table 1.

2.1.2.

CO2 direct expansion cascade (DEC) system

In the DEC system (System 3, Table 1), the low-temperature

circuit and the high-temperature circuit are single-stage

Fig. 1 e COP and work ratio vs. ambient temperature with

CR 1.5 for Combined System 1.

direct expansion systems, each having their own SLHXs (if

necessary), coupled through the cascade condenser. In the P-h
diagram, the DX R404A (1-2-3-4-5-6) and the DX CO2 (7-8-9-1011-12) circuits are represented by the inner and the outer
domes, respectively.
A refrigeration system using a DEC system for the lowtemperature (LT) loads and a separate SC system for the
medium-temperature (MT) loads is also studied in this paper.
Such a refrigeration system will be called the Combined System 2. This system would be represented by two separate cycles; System 2 for MT and System 3 for LT, as shown in Table 1.

2.1.3.

Combined CO2 secondary/cascade (CSC) system

In the CSC system (System 4, Table 1), the low-temperature

and medium-temperature loads are managed by a singlestage direct expansion system and a CO2 secondary loop
system, respectively, coupled with the high-temperature circuit through the cascade condenser. In the P-h diagram, the
DX R404A (1-2-3-4-5-6) and the CO2 (7-8-9-10-11-12-13-14-1516-17-18) circuits are represented by the inner and the outer
domes, respectively.

Table 4 e Baseline parameters of refrigeration systems.

Baseline parameters

Value

Refrigeration Load
Baseline System

Condensing Temperature

Cascade Systems

Condensing Temperature

Booster Systems

Cascade Condenser Approach Temperature

Pump Power, Glycol
Pump Power, CO2
Intermediate Pressure in Receiver
Condenser/Gas Cooler Outlet Temperature

Condenser/Gas Cooler Pressure (for Tamb  27
C)

Q_ MT 120 kW at 5
C
Q_ LT 65 kW at 30
C
Tcond 21
C for Tamb 
Tcond Tamb 10
C for
Tcond 21
C for Tamb 
Tcond Tamb 10
C for
Tapp 3.3
C
_ pump 0.05 W
_ comp
W
_ comp
_ pump 0.01 W
W

8
C
Tamb > 8
C
8
C
Tamb > 8
C

Pint 3.5 MPa

TGC 10
C for Tamb  0
C
TGC Tamb 10
C for 0
C < Tamb < 18
C
TGC 28 at P 6.9 MPa for 18
C  Tamb < 22
C
TGC 28 at P 7.5 MPa for 22
C  Tamb < 25
C
TGC Tamb 3
C for Tamb  25
C
PSTBS 0:00254*T2amb 0:10003*Tamb 2:9211
PTBS-UX 0:0021*T2amb 0:13516*Tamb 2:2487
PTBS-BC 0:0004*T2amb 0:25237*Tamb 0:362

i n t e r n a t i o n a l j o u r n a l o f r e f r i g e r a t i o n 4 6 ( 2 0 1 4 ) 8 6 e9 9

93

saturation pressure corresponding to the mediumtemperature and low-temperature loads, respectively. After
absorbing heat from the low-temperature loads, the lowpressure refrigerant is further superheated in the SLHX2 and
compressed in the low-pressure compressors. The discharge
from the low-pressure compressors, the mediumtemperature loads and the bypass valve combines before
entering the high-pressure compressors.

2.2.1.

Fig. 2 e COP vs. ambient temperature with DTLT 10 K for

the CSC system (System 4).

2.1.4. Combined glycol secondary/CO2 cascade (CSC-G)

system
In the CSC-G System (System 5, Table 1), the low-temperature
loads are served by a single-stage direct expansion CO2 system
coupled with the high-temperature circuit through the
cascade condenser, while the medium-temperature loads are
managed through a propylene glycol secondary loop (instead
of liquid CO2 as in System 4). In the P-h diagram, the DX R404A
(1-2-3-4-5-6-7-8-9-10-11) and the CO2 (12-13-14-15-16-17) circuits are represented by the inner and the outer domes,
respectively.

2.2.

Transcritical booster (CO2) systems

A transcritical booster system (Systems 6, 7 and 8 in Table 1) is

divided into four pressure levels. At the highest pressure level,
the refrigerant (CO2) from the high pressure compressors enters a gas cooler/condenser and rejects heat to the surroundings. The high pressure CO2 is further cooled in a suction line
heat exchanger (SLHX1) and expanded before being collected
in the receiver. Depending upon the system configuration,
refrigerant is expanded to an intermediate pressure level
either before or after passing through the SLHX1. The saturated refrigerant liquid from the receiver is expanded to the

Standard transcritical booster system (STBS)

In the STBS (System 6, Table 1), the high pressure refrigerant

from the outlet of the gas cooler/condenser passes through
the SLHX1 before expanding to an intermediate pressure level
through the expansion device between locations 2 and 3. In
the P-h diagram, state points 1, 2 and 17 are at the high pressure level, state points 3, 4, 6 and 9 are at the intermediate
pressure level, state points 5, 7, 8, 13, 14, 15 and 16 are at the
medium-temperature level and state points 10, 11 and 12 are
at the low-temperature level.

2.2.2. Transcritical booster system with upstream expansion

valve (TBS-UX)
The TBS-UX (System 7 in Table 1) differs from the STBS in that
the high pressure refrigerant from the gas cooler is expanded
between 2 and 3, prior to passing through SLHX1 and stored in
the receiver. In the P-h diagram, state points 1 and 17 are at the
high pressure level, state points 2, 3, 4, 6 and 9 are at the intermediate pressure level, state points 5, 7, 8, 13, 14, 15 and 16
are at the medium pressure level and state points 10, 11 and 12
are at the low pressure level.

2.2.3. Transcritical booster system with bypass compressor

(TBS-BC)
In the TBS-BC (System 8, Table 1), is similar to the STBS,
however, the refrigerant exiting the receiver bypass at location 4 is compressed by an additional set of compressors
(bypass compressors), which then combines with the
discharge of the high-pressure compressors at location 17,
before finally entering the gas cooler/condenser. In the P-h
diagram, state points 1, 2, 5, 16 and 17 are at the high pressure
level, state points 3, 4, 6 and 9 are at the intermediate pressure
level, state points 7, 8, 13, 14, and 15 are at the medium

Fig. 3 e Performance of the STBS system (System 6) for SLHX 0.4 and DT 10 K. (a) System COP and work ratios, (b) Mass
flow ratios.

94

i n t e r n a t i o n a l j o u r n a l o f r e f r i g e r a t i o n 4 6 ( 2 0 1 4 ) 8 6 e9 9

Fig. 4 e Bypass mass flow ratio vs. ambient temperature

with DT 10 K for the STBS (System 6).
pressure level and state points 10, 11 and 12 are at the low
pressure level.

3.

System analysis

The overall coefficient of performance, COP, of a refrigeration

system is determined by:
COP

Q_ ref
_ total
W

(1)

where Q_ ref is the total refrigerating capacity of the system and

_ total is the total power required by the compressors and
W
pumps (the condenser/gas cooler fan power is excluded in
these analyses). For the eight refrigeration systems shown in
Table 1, the corresponding equations for refrigerating capacity
and total power input are given in Table 2. These equations
were used with the definition of COP (Eq. 1) to determine the
performance of each of the eight refrigeration systems.
Note that for booster and cascade refrigeration systems, it
is difficult to separate cooling energy and work input into
medium-temperature and low-temperature contributions.
Thus an overall COP was used in this study to quantify the
performance of the various refrigeration systems.

4.

Methodology

4.1.

Baseline parameters

In order to assess the merits and the advantages offered by the

CO2-based refrigeration systems shown in Table 1, a parametric study was performed to determine the effect of various
operating parameters on the energy efficiency of the systems.
Ambient temperature ranging between 0
C and 40
C was
used to determine the performance of the systems in all eight
climate zones of the US. SLHX effectiveness values of either
0 or 0.7 were used to represent a system without a SLHX or the
upper range of a realistic heat exchanger (where 1.0 represents an ideal 100% efficient heat exchanger), respectively.
Evaporator superheat is always required to ensure that no
liquid enters the compressor. Thus, the performance of the
systems was studied using an evaporator superheat of either
10 K or 15 K for both the low- and medium-temperature loads.
A summary of the parameters used in the analysis is given in
Table 3. The thermodynamic properties of the refrigerants
were calculated using the National Institute of Standards and
Technology (NIST) Reference Fluid Thermodynamic and
Transport Properties Database (REFPROP) Version 8.0
(Lemmon et al., 2007).

4.2.

System operating assumptions

For an average supermarket (4200 m2), the typical refrigeration loads and temperatures of the low-temperature and the
medium-temperature loads were assumed to be 65 kW at
30
C and 120 kW at 5
C, respectively. It has been noted
that for typical supermarkets, the MT load is approximately
two to five times more than that of the LT load (Girotto et al.,
2004; Getu and Bansal, 2008). In this study, it was assumed
that the MT load is twice that of the LT load.
For the proper operation of the expansion valves and efficient heat transfer through the condenser in the DX and
cascade system options, the minimum condensing temperature was fixed at Tcond 21
C and the condensing temperature was set to be 10
C greater than the ambient temperature
(Tamb). In addition, no liquid sub-cooling was assumed at the
exit of condenser. The approach temperature, which is the

Fig. 5 e System COP vs. ambient temperature at SLHX 0.4, DT 10 K and CR 1.5. (a) Transcritical booster systems, (b)
Cascade/secondary systems.

95

i n t e r n a t i o n a l j o u r n a l o f r e f r i g e r a t i o n 4 6 ( 2 0 1 4 ) 8 6 e9 9

Table 6 e Impact of increasing SLHX and DT on the

system COPs.
System
Combined System 1
Combined System 2
CSC
CSC-G
STBS
TBS-UX
TBS-BC

SLHX

DT

Increase
Increase
Decrease for Tamb < 15
C
Increase for Tamb > 15
C
Decrease
Negligible*
Negligible*
Negligible*

NA
Negligible*
Negligible*
Negligible*
Negligible*
Negligible*
Negligible*

* Less than 1%.

Fig. 6 e System COP vs. ambient temperature at

SLHX 0.4, DT 10 K and CR 1.5 for R404A DX System
(System 1), CSC (System 4), STBS (System 6), and TBS-BC
(System 8).

difference between the temperature at the outlet of the lowtemperature circuit and the inlet of the high-temperature
circuit of the cascade condenser, was Tapp 3.3
C.
Generally, in transcritical booster systems, the receiver
pressure is approximately 0.5 MPa higher than the saturated
pressure of the medium-temperature loads, and thus, the
intermediate pressure in the receiver was fixed at
Pint 3.5 MPa for this study. In addition, since transcritical CO2
booster systems operate at higher pressures in comparison to
the HFC based DX and cascade system options, the expansion
valves can function properly at lower ambient temperatures.
Thus, the minimum condensing temperature of the transcritical CO2 booster systems can be lower than that of the
HFC-based DX and cascade system options. In this study it
was assumed that the gas cooler/condenser temperature (TGC)
was 10
C greater than the ambient temperature during
subcritical
operation.
In
the
intermediate
stage,
18
C < Tamb < 22
C, TGC and condensing pressure were fixed
to 28
C and 6.9 MPa; 22
C < Tamb < 25
C, TGC and condensing
pressure were fixed to 28
C and 7.5 MPa, respectively. In

supercritical operation, Tamb  27
C, TGC is set to be 3
C

greater than the ambient temperature. Also, in supercritical
operation, an optimum gas cooler pressure exists which
maximizes the COP of the system (Kauf, 1999). These parameters, along with the calculated optimum pressure equations
for each transcritical booster system as functions of ambient
temperature, are given in Table 4. In addition, the following
assumptions were made for the parametric analysis:
 The pressure drop and heat loss/gain in the suction lines
were ignored.
 Only saturated liquid and saturated vapor exit the receiver.
 The isentropic efficiency of the low-pressure, high-pressure and bypass compressors was 0.65.
 The expansion valves were isenthalpic.
 Condenser/gas cooler fan power consumption was
assumed to be roughly equal for all the systems, so fan
power was not included in the COP computations.

5.

Results and discussion

Altogether, seven CO2-based refrigeration system configurations were analyzed and compared, including the Combined
System 1 (MT system 2, LT System 2), Combined System 2 (MT

Table 5 e US climate zones, cities and annual average COP for several CO2-based refrigeration systems.
Climate zone

1
2A
2B
3A
3B
3B
3C
4A
4B
4C
5A
5B
6A
6B
7
8

City

Annual average temperature (
C)

Miami, FL
Houston, TX
Phoenix, AZ
Atlanta, GA
Los Angeles, CA
Las Vegas, NV
San Francisco, CA
Baltimore, MD
Albuquerque, NM
Seattle, WA
Chicago, IL
Boulder, CO
Minneapolis, MN
Helena, MN
Duluth, MN
Fairbanks, AK

24.9
20.7
23.8
17.0
17.3
20.2
14.4
13.3
14.2
11.4
10.0
10.3
8.0
7.2
4.3
2.1

Annual average COP

R404A DX

CSC and combined system 2

TBS-BC

2.60
2.90
2.70
3.14
3.22
2.96
3.51
3.32
3.31
3.60
3.42
3.46
3.47
3.58
3.63
3.70

2.54
2.80
2.62
3.02
3.10
2.86
3.35
3.18
3.17
3.42
3.27
3.30
3.31
3.41
3.46
3.52

2.36
2.68
2.40
3.04
2.75
2.78
3.09
3.44
3.40
3.52
3.78
3.75
3.98
4.08
4.28
4.54

96

i n t e r n a t i o n a l j o u r n a l o f r e f r i g e r a t i o n 4 6 ( 2 0 1 4 ) 8 6 e9 9

Fig. 7 e Comparative advantage of a specific system in various climate zones of the US.

System 2, LT System 3), CSC system (System 4), CSC-G system

(System 5), STBS (System 6), TBS-UX (System 7) and TBS-BC
(System 8). The effects of ambient temperature, evaporator
superheat and SLHX effectiveness on the performance of each
of the seven configurations were investigated. In addition, the
performance of the seven CO2-based refrigeration systems
was compared to the baseline R404A multiplex DX system
(System 1).

5.1.

Cascade and secondary loop systems

In general for the cascade and secondary loop systems, an

increase in the ambient temperature leads to an increase in
the work required by the high-temperature circuit with no
impact on the work required by the low-temperature circuit.
This is due to the fact that the pressure ratio of hightemperature circuit compressors changes with change in the
ambient temperature, while the low-temperature circuit
compressors and pumps are decoupled from the ambient
conditions through the cascade condenser. Thus, the system
COP decreases with an increase in the ambient temperature.

5.1.1.

Combined System 1 (MT System 2, LT System 2)

Fig. 1 shows the system COP and the work ratio (WR) vs.
ambient temperature for a given pump circulation ratio of 1.5
for the Combined System 1 (CO2 Secondary Coolant System
for low- and medium-temperature applications, two separate
System 2 cycles as shown in Table 1). The work ratio of the
compressors is defined to be the ratio of the compressor
_ to the total power required by the system,
power, W,
_
_
W Wpump , as follows:
WR

_
W
_
_ pump
WW

(2)

Since the SLHX is in the high-temperature circuit, it has no

impact on the performance of the low-temperature circuit.
However, with an increase in the effectiveness of the SLHX,
the work required by the compressors decreases, as illustrated

by a decrease in the compressor work ratio shown in Fig. 1,

resulting in an increase in the system COP. For example, at an
ambient temperature of 20
C, the system COP increases by
5.8% with an increase in SLHX effectiveness from 0 to 0.7.

5.1.2.

Combined System 2 (MT System 2, LT System 3)

In the Combined System 2 (DEC for low-temperature (System

3), SC for medium-temperature (System 2)), an increase in the
effectiveness of the SLHX in the high-temperature circuit results in a decrease in the mass flow rate which dominates the
increase in enthalpy change occurring across the highpressure compressors. Thus, the power required by the
high-pressure compressors decreases. However, an increase
in the effectiveness of the SLHX in the low-temperature circuit
results in an increase in the enthalpy change across the lowpressure compressors which is greater than the decrease in
the mass flow rate across the compressor. This leads to increase in the power required by the low-pressure compressors. Thus, the impact of SLHX effectiveness on the hightemperature and low-temperature circuits counteracts each
other. The result is an increase in COP, but not as great as that
of Combined System 1. For example, at an ambient temperature of 20
C, the system COP increases by 2% with an increase
in SLHX effectiveness from 0 to 0.7. Finally, the effect of
increasing the LT superheat by 5 K from 10 K to 15 K is insignificant, resulting in a decrease in COP of only 0.02%.

5.1.3.

Combined CO2 secondary/cascade (CSC)

For the Combined CO2 Secondary/Cascade System (CSC, System 4, Table 1), Fig. 2 shows the system COP vs. ambient
temperature for a LT evaporator superheat of 10 K.
With an increase in the effectiveness of SLHX1 and SLHX2,
the mass flow through the system decreases and the enthalpy
at the inlet of compressors increases. In comparison to the
rate of increase in enthalpy change across the compressors,
the rate of decrease in the refrigerant mass flow rate is greater
in the high-temperature circuit than in the low-temperature
circuit. This leads to an increase in the work required by the

i n t e r n a t i o n a l j o u r n a l o f r e f r i g e r a t i o n 4 6 ( 2 0 1 4 ) 8 6 e9 9

low-pressure compressors and a decrease in the work

required by the high-pressure compressors.
As the SLHX effectiveness increases at low ambient temperature, the increase in work required by the low-pressure
compressors dominates the decrease in that of the highpressure compressors, thereby, increasing the total work
required by the system and decreasing the system COP.
However, as the SLHX effectiveness increases at high ambient
temperatures, the total work required by the system decreases and the COP increases. At an ambient temperature of
5
C, an increase in effectiveness of the SLHX from 0 to 0.7
leads to the system COP decreasing by about 1.3%. On the
other hand, at an ambient temperature of 20
C and 40
C, with
change in the effectiveness of the SLHX from 0 to 0.7, the
system COP increases by 0.9% and 9.7%, respectively. The effect of increasing the LT superheat by 5 K, from 10 K to 15 K, is
insignificant, resulting in a decrease in COP of only 0.01%.

5.1.4.
G)

Combined glycol secondary/CO2 cascade system (CSC-

As with the CSC System, low-temperature evaporator superheat has a negligible impact on the COP of the CSC-G System
(System 5, Table 1). However, the work required by the lowtemperature circuit and high-temperature circuit compressors of the CSC-G System increases with an increase in the
effectiveness of all the suction line heat exchangers, leading
to a decrease in the system COP. At an ambient temperature of
20
C, an increase in effectiveness of the SLHX from 0 to 0.7
leads to a 12.7% decrease in the system COP.

5.2.

Transcritical Booster Systems

For the Standard Transcritical Booster System (STBS), Fig. 3

shows the system COP, the normalized work ratios (WRLP,
WRHP) and the normalized mass flow ratios (MRBP), vs.
ambient temperature at constant SLHX effectiveness
(SLHX 0.4). The normalized work ratio is defined as the ratio
of the work required by either the low-pressure or highpressure compressor to the work required by low-pressure
compressor.
WRcomp

w_ comp
w_ LP

(3)

where WRcomp is either WRLP or WRHP, and w_ comp is either w_ LP

or w_ HP , respectively. The normalized mass flow ratio is defined
as the ratio of either the low-temperature, medium-temperature, or bypass mass flow rate to the mass flow rate through
the low-temperature circuit.
MRcirc

m_ circ
m_ LT

97

bypass flow rate and consequently, the bypass flow ratio

(MRBP) increases, as shown in Fig. 3(b).
The bypass mass flow ratio is constant for
18
C  Tamb < 22
C and 22
C  Tamb  25
C due to fixed
condensing pressure of 6.9 MPa and 7.5 MPa, respectively,
during the transition from subcritical to transcritical operation. The change in bypass mass flow ratio has a direct impact
on the work required by the high-pressure compressors. This
can be seen in Fig. 3(a) where the work ratio of the highpressure compressor increases, and consequently, the system COP decreases.
An increase in evaporator superheat results in a decrease
in the total refrigerant mass flow rate. In general, increasing
the evaporator superheat leads to an increase in the refrigerant enthalpy throughout the system between the lowpressure compressor and the gas cooler inlet. However, the
increase in enthalpy associated with superheat is nearly
counterbalanced by the decrease in the total refrigerant flow
rate. As a result, the rate of change of COP with an increase in
superheat is insignificant (<0.5%). Similarly, in the Transcritical Booster System with Upstream Expansion Valve (TBSUX) and the Transcritical Booster System with Bypass
Compressor (TBS-BC), the evaporator superheat has insignificant impact on the system COP.
Fig. 4 shows the bypass mass flow ratio (MRBP) vs. ambient
temperature for a given LT and MT evaporator superheat of
10 K for the STBS (System 6). With an increase in the effectiveness of SLHX1 and SLHX2, the quality of refrigerant at the
inlet of both the receiver and the low-temperature evaporator
decreases, thereby decreasing the bypass mass flow ratio
(MRBP) and the low-temperature mass flow ratio (MRLT).
In addition, for the STBS (System 6), an increase in the
SLHX effectiveness increases the enthalpy of the refrigerant at
the inlet of the low- and high-pressure compressors. In
subcritical operation, the rate of increase in the enthalpy
difference across the compressors and the rate of decrease in
the mass flow through the low-temperature evaporator and
the bypass, leads to changes in the relative work input to the
low-pressure and high-pressure compressors. This results in a
decrease in the system COP during subcritical operation but
an increase during transcritical operation. For example, as the
SLHX effectiveness increases from 0 to 0.7, the system COP
decreases by 1.4% at an ambient temperature of 10
C; however, the COP increases by 1.9% at an ambient temperature of
35
C. Similarly, in the TBS-UX (System 7) and the TBS-BC
(System 8), as the SLHX effectiveness increases from 0 to 0.7,
the system COP decreases by 1.2% and 1.3%, respectively, at an
ambient temperature of 10
C, whereas the system COP increases by 0.2% and 1.9%, respectively, at an ambient temperature of 35
C.

(4)

where MRcirc is either MRLT, MRMT or MRBP, and mcirc is either

mLT, mMT or mBP, respectively.
Since the loads in the low-temperature circuit and the
medium-temperature circuit are fixed in this analysis, the
refrigerant flow rate through the LT and MT circuits are constant, and ambient temperature has no impact on these flow
rates. However, an increase in ambient temperature leads to
an increase in vapor entering the receiver. As a result, the

5.3.
Comparison of CO2-based refrigeration systems
with the baseline system
Fig. 5 shows the system COP vs. ambient temperature for all
the Transcritical Booster Systems and the Combined Cascade/
Secondary Loop Systems for a constant SLHX effectiveness
(SLHX 0.4) and evaporator superheat (DT 10 K). It can be
seen that the Transcritical Booster System with Bypass
Compressor (TBS-BC) and both the Combined System 2 and

98

i n t e r n a t i o n a l j o u r n a l o f r e f r i g e r a t i o n 4 6 ( 2 0 1 4 ) 8 6 e9 9

the Combined CO2 Secondary/Cascade (CSC) System are the

most efficient systems among the transcritical booster and
the cascade/secondary loop systems, respectively.
Fig. 6 shows the system COP vs. ambient temperature for a
baseline system consisting of separate R404A multiplex DX
circuits for the MT and LT loads, the TBS-BC (System 8), the
STBS (System 6), the Combined System 2, and the Combined
CO2 Secondary/Cascade (CSC) (System 4), for a constant SLHX
effectiveness (SLHX 0.4) and evaporator superheat
(DT 10 K). At ambient temperatures between 0
C and 40
C,
the COP of the R404A multiplex DX system is higher than the
COP of the Combined System 2 and the Combined CO2 Secondary/Cascade (CSC) System by up to 6.8%, with an average
difference in COP of approximately 5.8%.
At ambient temperature less than 8
C, the COP of the TBSBC (System 8) is higher than the COP of R404A multiplex DX
system (System 1), but at higher temperatures, the COP of the
R404A multiplex DX system is higher than the COP of TBS-BC
(System 8). For example, at an ambient temperature of 5
C,
the system COP of TBS-BC (System 8) is 14.3% higher than that
of R404A multiplex DX system (System 1). Whereas, at an
ambient temperature of 35
C, the system COP of R404A
multiplex DX system (System 1) is 15% higher than TBS-BC
(System 8). The STBS behaves similar to the TBS-BC, however, over the temperature range of 0
Ce40
C, the average COP
of the TBS-BC is 3.3% greater than that of the STBS. Also, the
average COP's over the range 0
Ce40
C for the R404A multiplex DX system (System 1) and the TBS-BC (System 8) are
nearly equal.

5.4.

Climate zones

Bin analyses were performed to determine the annual average

COP of the TBS-BC (System 8), the Combined System 2, Combined CO2 Secondary/Cascade (CSC) System and the R404A
multiplex DX system (System 1) in sixteen cities selected from
the eight climate zones of the United States (ICC, 2009). The 16
cities selected for the bin analyses are shown in Table 5, along
with their annual average temperature and annual average
COP of the four systems. The hourly weather data from these
sixteen cities were used to determine the annual average COP
for the four refrigeration systems. It can be seen that the TBSBC (System 8) is the most efficient system in Zones 5, 6, 7 and
8. In these zones, the annual average COP of the TBS-BC
(System 8) is higher than that of the R404A DX system (System 1) by up to 13%. In Zones 1, 2 and 3, the R404A DX system
performs the best. The annual average COP of the R404A DX
system (System 1) is up to 8.3% higher than that of the TBS-BC.
However, in zone 4, the R404A DX system (System 1) and the
TBS-BC (System 8) perform similarly, within 4% of each other.
Also, in all climate zones, the average performance of the
Combined System 2 and the CSC system are 4.2% lower than
that of the R404A DX system.

6.

The qualitative impact of an increase in the SLHX effectiveness (SLHX) and evaporator superheat (DT) on the performance of the transcritical booster systems and the combined
cascade/secondary loop systems is summarized in Table 6.
Generally, the evaporator superheat has negligible impact on
the performance of the CO2-based refrigeration systems. An
increase in SLHX effectiveness has negligible impact on the
performance of the transcritical booster systems. However, an
increase in SLHX effectiveness leads to an increase in the
performance of the Combined Systems 1 and 2 and a decrease
in the performance of the CSC-G (System 5). Also, an increase
in the SLHX effectiveness in the CSC (System 4) leads to an
increase in the system performance for Tamb < 15
C but a
decrease in performance for Tamb > 15
C.
The TBS-BC and both the Combined System 2 and the
Combined CO2 Secondary/Cascade System (CSC) are the most
efficient systems among the transcritical booster and cascade/
secondary loop systems, respectively. Fig. 7 shows the refrigeration systems which perform most efficiently for each
climate zone in the US. In Zones 5, 6 and 7, shown in the green
shaded regions on the map, the TBS-BC is the most efficient
system. In Zones 1, 2, and 3, shown in the red regions, the
R404A DX system (System 1) is most efficient. Finally, in Zone
4, shown in yellow, the R404A DX system (System 1) and the
TBS-BC (System 8) perform similarly.
Implementation of transcritical booster systems or
cascade/secondary loop systems using optimized operating
conditions will lead to reduced direct greenhouse gas emissions while achieving comparable energy consumption as
compared to current HFC-based multiplex DX systems.

Conclusions

The comparative analysis of the CO2-based refrigeration systems with the baseline R404A system has revealed the
following conclusions:

references

Bansal, P.K., 2012. A review e status of CO2 as a low temperature

refrigerant: fundamentals and R&D opportunities. Appl.
Therm. Eng. 41, 18e29.
Bell, I., 2004. Performance increase of carbon dioxide refrigeration
cycle with the addition of parallel compression
economization. In: Proceedings of the 6th IIR Gustav
Lorentzen Conference on Natural Working Fluids, Glasgow,
UK, 29 August e 1 September 2004.
Denecke, J., Hafner, A., Eikevik, T., Ladam, Y., 2012. Heat recovery
solutions for R744 booster commercial refrigeration systems.
In: Proceedings of the 10th IIR Gustav Lorentzen Conference
on Natural Working Fluids, Delft, The Netherlands, 25e27 June
2012.
Ferrandi, C., Orlandi, M., 2012. Theoretical analysis of cold storage
device effects on the performance and regulation of a CO2
supermarket refrigeration plant. In: Proceedings of the 10th IIR
Gustav Lorentzen Conference on Natural Working Fluids,
Delft, The Netherlands, 25e27 June 2012.
Fricke, B.A., Becker, B.R., 2010. Energy use of doored and open
vertical refrigerated display cases. In: Proceedings of the 13th
International Refrigeration and Air Conditioning Conference.
Purdue University, West Lafayette, IN, 12e15 July 2010.
Ge, Y.T., Tassou, S.A., 2009. Control optimization of CO2 cycles for
medium-temperature retail food refrigeration systems. Int. J.
Refrigeration 32, 1376e1388.
Ge, Y.T., Tassou, S.A., 2010. Performance evaluation and control
optimization of a CO2 booster refrigeration system in
supermarket. In: Proceedings of the 1st IIR International

i n t e r n a t i o n a l j o u r n a l o f r e f r i g e r a t i o n 4 6 ( 2 0 1 4 ) 8 6 e9 9

Conference on Sustainability and the Cold Chain, Cambridge,

UK, 29e31 March 2010.
Getu, H.M., Bansal, P.K., 2008. Thermodynamic analysis of an
R744-R717 cascade refrigeration system. Int. J. Refrigeration
31, 45e54.
Girotto, S., Minetto, S., Neksa, P., 2004. Commercial refrigeration
system using CO2 as the refrigerant. Int. J. Refrigeration 27,
717e723.
Hinde, D., Zha, S., 2009. Natural refrigerant applications in North
American supermarkets. In: Proceedings of the IIAR Industrial
Refrigeration Conference and Exhibition, Dallas, TX, 22e25
March 2009.
ICC, 2009. International Energy Conservation Code. International
Code Council, Washington, D.C.
ICF Consulting, 2005. Revised Draft Analysis of U.S. Commercial
Supermarket Refrigeration Systems.
Kauf, F., 1999. Determination of the optimum high pressure for
transcritical CO2 refrigeration cycles. Int. J. Therm. Sci. 38,
325e330.
Kim, M.H., Pettersen, J., Bullard, C.W., 2004. Fundamental process
and system design issues in CO2 vapor compression systems.
Prog. Energy Combust. Sci. 30, 119e174.
Lemmon, E., McLinden, M., Huber, M., 2007. NIST Standard
Reference Database 23: Reference Fluid Thermodynamic and
Transport Properties-REFPROP, Version 8.0, Standard
Reference Data Program. National Institute of Standards and
Technology, Gaithersburg, MD.
Mazzola, D., Toffolo, A., Orlandi, M., 2012. Supermarket
application. CO2 system with groundwater sink. Model
simulation. In: Proceedings of the 10th IIR Gustav Lorentzen

99

Conference on Natural Working Fluids, Delft, The

Netherlands, 25e27 June 2012.
Rauss, D., Mitchell, S., Faramarzi, R., 2008. Cool retrofit solutions
in refrigerated display cases. In: Proceedings of the 2008 ACEEE
Summer Study on Energy Efficiency in Buildings, Pacific Grove,
CA, 17e22 August 2008.
Sarkar, J., Agrawal, N., 2010. Performance optimization of
transcritical CO2 cycle with parallel compression
economization. Int. J. Therm. Sci. 49, 838e843.
Sawalha, S., 2007. Theoretical evaluation of trans-critical CO2
systems in supermarket refrigeration. Part 1: modeling,
simulation and optimization of two system solutions. Int. J.
Refrigeration 31, 516e524.
Sawalha, S., Palm, B., 2003. Energy consumption evaluation of
indirect systems with CO2 as secondary refrigerant in
supermarket refrigeration. In: Proceedings of the 21st IIR
International Congress of Refrigeration, Washington, D.C.,
17e22 August 2003.
Westphalen, D., Zogg, R.A., Varone, A.F., Foran, M.A., 1996. Energy
savings Potential for Commercial Refrigeration Equipment.
Building Equipment Division, Office of Building Technologies,
U.S. Department of Energy, Washington, D.C.
Winter, J., Murin, S., 2012. Energy saving and increasing reliability
at CO2 transcritical boosters. A case study. In: Proceedings of
the 10th IIR Gustav Lorentzen Conference on Natural Working
Fluids, Delft, The Netherlands, 25e27 June 2012.
Zhang, M., 2006. Energy analysis of various supermarket
refrigeration systems. In: Proceedings of the 11th
International Refrigeration and Air Conditioning Conference.
Purdue University, West Lafayette, IN, 17e20 July 2006.