Applied Thermal Engineering 73 (2014) 41e50

Contents lists available at ScienceDirect

Applied Thermal Engineering

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

Experimental evaluation of a R134a/CO2 cascade refrigeration plant

Carlos Sanz-Kock a, Rodrigo Llopis a, *, Daniel Sa

nchez a, Ramo

n Cabello a,
b
Enrique Torrella
a
n, Spain
Jaume I University, Dep. of Mechanical Engineering and Construction, Campus de Riu Sec s/n, E-12071 Castello
b
Polytechnic University of Valencia, Department of Applied Thermodynamics, Camino de Vera, 14, E-46022 Valencia, Spain

h i g h l i g h t s

The experimental evaluation of a R134a/CO2 cascade refrigeration plant is presented.

Temperature difference in cascade heat exchanger varied from 3.3 to 5.3  C.

Variation of intermediate level produced maximum COP variations of 6%.

Cooling capacity ranged from 4.5 kW (40 and 40  C) to 7.5 kW (30 and 30  C).

COP ranged from 1.05 (40 and 40  C) to 1.65 (30 and 30  C).

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

Article history: We present the experimental evaluation of a R134a/CO2 cascade refrigeration plant designed for low
Received 26 May 2014 evaporation temperature in commercial refrigeration applications. The test bench incorporates two
Accepted 16 July 2014 single-stage vapour compression cycles driven by semi hermetic compressors coupled thermally through
Available online 23 July 2014
two brazed plate cascade heat exchangers working in parallel and controlled by electronic expansion
valves. The experimental evaluation (45 steady-states) covers evaporating temperatures from 40
Keywords:
to 30  C and condensing from 30 to 50  C. In each steady-state, we conducted a sweep of the
Cascade
condensing temperature of the low temperature cycle with speed variation of the high temperature
Two-stage
Low GWP
compressor. Here, the energy performance of the plant is analysed, focussing on the compressors' per-
R134a formance, temperature difference in the cascade heat exchanger, cooling capacity, COP and compressors
Carbon dioxide discharge temperatures.
Energy efficiency © 2014 Elsevier Ltd. All rights reserved.

1. Introduction overcomes 40 tonnes of equivalent CO2; from 2022 in centralized

commercial refrigeration systems with capacity of more than
The high warming impact associated to refrigeration systems, 40 kW, refrigerants with GWP  150 will not be permitted except
due to the direct leakage of refrigerants and to the indirect emis- for the primary refrigerant of cascade systems, where refrigerants
sions of CO2 by electricity consumption, moves scientific commu- with GWP up to 1500 will be allowed. Both agreements represent
nity and institutions to more environmental friendly solutions. the future disappearance of the most used refrigerants in central-
Among refrigeration groups, centralized commercial refrigeration ized commercial refrigeration at low temperature in Europe [4], the
highlights because of its high annual leakage rate, commonly R404A and the R507A with GWP of 3700 and 3800, respectively [5].
higher than 10% of the total charge of the system [1] and because of Commercial refrigeration section, specially supermarkets, needs to
its high energy consumption [2]. Both negative aspects motivated adapt their refrigeration systems and fluids to the new F-Gas
European Commission to approve the revision of the F-Gas regu- regulation. The adaptation will be based on low GWP refrigerants,
lation [3], focussing the efforts on reducing the emissions to the and generally it would need to replace the common single-stage
atmosphere of greenhouse refrigerants. The agreements that most refrigeration systems by new refrigeration configurations adapted
affect centralized commercial refrigeration in Europe are: from to the new fluids. The system that now attracts more attention is
2020 on refilling of systems with refrigerants of GWP>2500 will cascade refrigeration using CO2 as LT refrigerant, although different
not be allowed if the total equivalent charge of the system refrigerant options are considered for the HT cycle.
In literature, the most analysed cascade refrigeration system is
* Corresponding author. Tel.: þ34 964 72 8136; fax: þ34 964 72 8106. the combination NH3/CO2. For this pair, Lee et al. [6] evaluated
E-mail address: rllopis@uji.es (R. Llopis).

http://dx.doi.org/10.1016/j.applthermaleng.2014.07.041
1359-4311/© 2014 Elsevier Ltd. All rights reserved.
42 C. Sanz-Kock et al. / Applied Thermal Engineering 73 (2014) 41e50

Nomenclature D increment
r density, kg m3
COP coefficient of performance
cp specific isobaric heat, kJ kg1 K1 Subscripts
GWP global warming potential (100 years integration) C1 cascade heat exchanger number 1
h specific enthalpy, kJ kg1 C2 cascade heat exchanger number 2
HT high temperature cycle casc cascade
LT low temperature cycle dis discharge
m_ mass flow rate, kg s1 env environment
N compressor's speed, rpm exp expansion
P pressure, bar gc gas-cooler
PC compressor power consumption, kW H high-temperature cycle
Pgc gas-cooler power consumption, kW i inlet
Q_ heat transfer rate, kW K condensing level
SF secondary fluid L low-temperature cycle
t compression ratio o output
T temperature,  C O evaporating level
V_ volumetric flow rate, m3 h1 ref refrigerant
ws isentropic specific compression work, kJ kg1 res receiver
xv vapour title s isentropic
sf secondary fluid in LT evaporator (water-tyfoxit)
Greek symbols suc suction
hG compressor global efficiency w secondary fluid in HT condenser (water)

theoretically the optimal condensing temperature of the cascade 24% respect the single-stage configuration and the refrigerant
heat exchanger and the COP for evaporating levels between 45 charge is reduced, however, the investment cost of the cascade is
and 55  C. Dopazo et al. [7], also theoretically, analysed the influ- 18.5% higher. However, they do not present the energy evaluation of
ence of the cycle parameters on its efficiency and evaluated the the cascade system under different operating conditions.
optimal condensing temperature, too. Both works presented poly- As can be observed, most part of scientific work deals with
nomials to evaluate the optimum condensing level. And finally, theoretical performance evaluation of cascade systems and with
Messineo [8] evaluated theoretically the performance of this system establishing their optimum LT condensing conditions. Few experi-
regards a direct two-stage R404A system, stating that the cascade mental works have been reported up to now and most of them deal
system is an interesting alternative in commercial refrigeration for with NH3/CO2 systems for industrial applications, which are not
energy, security and environmental reasons. For other combinations recommended for commercial systems because of security reasons.
of refrigerants, Getu and Bansal [9] analysed theoretically cascades of UNEP [14] promotes the R134a/CO2 cascade refrigeration systems
CO2 with ammonia, propane, propylene, ethanol and R404A, as a high-security low-GWP solution for centralized commercial
concluding that the best pair from an energy point of view was refrigeration, especially for supermarkets, and some brands have
ethanol/CO2 followed by NH3/CO2. And, Xiao and Liu [10] studied selected this system as a future solution to overcome the new F-Gas
theoretically the application of R32 as HT refrigerant. regulation. However, no experimental work exists about this
Regarding experimental evaluation of cascade systems, Bingm- configuration. Therefore, this communication pretends to cover this
ing et al. [11] presented the experimental results of a NH3/CO2 lack of research and presents the experimental evaluation of R134a/
cascade driven by two screw compressors for a condensing tem- CO2 cascade refrigeration prototype driven by semi hermetic
perature of 40  C and evaporating temperatures between 50 compressors over a wide range of working conditions.
and 30  C. They evaluated experimentally the influence of the
temperature difference in the cascade heat exchanger, the 2. Experimental plant
condensing temperature in the LT cycle and the degree of superheat
in this heat exchanger. Also, they presented an experimental The experimental plant, which can be appreciated in Fig. 1,
comparison with single-stage and two-stage NH3 systems, corresponds to a R134a/CO2 cascade refrigeration system designed
concluding that the cascade system is very competitive in low to operate at the low evaporating temperature level of commercial
temperature applications, specially below 40  C. With recipro- refrigeration (40 to 30  C). The plant is driven by two single-
cating compressors, Dopazo and Ferna
ndez-Seara [12], evaluated stage reciprocating compressors, the heat exchangers are of
the performance of a NH3/CO2 cascade for evaporating tempera- brazed plate type and incorporates electronic expansion valves. We
tures between 50 and 35  C and for a condensing temperature of present the schematic diagram of the plant, the position of the
30  C. They studied experimentally the optimum condensing measurement devices and the designation in Fig. 2. Next, we pre-
temperature of the LT cycle and compared the performance of the sent the details of the plant, of the thermal support system and of
plant with a direct two-stage NH3 system, concluding that the the measurement instrumentation.
cascade better performs for evaporating temperatures
below 35  C. Finally, da Silva [13] presents a comparison of a 2.1. Refrigeration cycle
R404A/CO2 cascade with semi hermetic compressors with direct
expansion single-stage systems of R404A and R22 in a supermarket The refrigerating cycle is detailed in Fig. 2. It is an indirect two-
application designed for operation at 30  C. The study concludes stage or cascade system composed of two single-stage cycles
that the cascade can reduce energy consumption between 13 and coupled thermally with two cascade heat exchangers working in
 C. Sanz-Kock et al. / Applied Thermal Engineering 73 (2014) 41e50 43

Fig. 1. View of the cascade refrigeration prototype.

parallel, which act as evaporator of the HT cycle and condenser of an NTC and a pressure gauge. The heat load to the evaporator is
the LT cycle. We use CO2 in the LT cycle and R134a in the HT cycle. provided with a loop working with a tyfoxit-water mixture (84% by
volume) which allows to operate up to 45  C. This loop allows
regulating the inlet temperature of the SF to the evaporator (sf,i)
2.1.1. LT cycle
and varying the SF flow rate (Vsf).
A CO2 variable speed semi hermetic compressor for subcritical
applications, with a displacement of 3.48 m3/h at 1450 rpm and a
nominal power of 1.5 kW, drives the LT cycle. The lubricant oil is 2.1.2. High temperature cycle (HT cycle)
POE C55E. The compressor absorbs the vapour at the suction point A R134a variable speed semi hermetic compressor, with a
(suc,L) and compresses it to the LT high pressure (dis,L), then we displacement of 32.66 m3/h at 1450 rpm and nominal power of
separate the lubricant oil. Following, the refrigerant gets into a gas- 3.7 kW, drives the single-stage HT cycle. The lubricant oil is POE
cooler (gc,i,L) where the CO2 is desuperheated with an air cooler SL32. The compressor absorbs the superheated refrigerant coming
heat exchanger before entering to the cascade heat exchangers from the cascade heat exchangers (suc,H) and delivers it com-
(gc,o,L), since generally the discharge temperature is higher than pressed to the HT high pressure (dis,H). Then, lubricant oil is
the environment temperature. This cross flow heat exchanger, separated and it feeds the condenser (k,i,H), a brazed plate heat
driven at its nominal speed with a fan of 75 W of power con- exchanger with a heat transfer area of 2.39 m2. At the exit of the
sumption, has a heat transfer area of 0.6 m2 in the refrigerant side condenser the HT refrigerant mass flow rate (Mref,H) is measured
and of 3.36 m2 in the air side. Next, CO2 flow is divided and con- with a coriolis mass flow meter. Then the refrigerant gets into the
densated it in two plate heat exchangers (C1,i,L and C2,i,L) with a receiver. Next, the refrigerant feeds two electronic expansion valves
total heat transfer area of 3.52 m2 CO2 leaving the condensers (C1,exp,i,H and C2,exp,i,H) that regulate the evaporation process in
(C1,o,L and C2,o,L) is joined and its mass flow rate is measured with the two cascade heat exchangers. The valves regulate indepen-
a Coriolis mass flow meter (Mref,L). Then, it enters to the receiver dently the degree of superheat at the exit of these heat exchangers
and feeds the expansion valve of the cycle (exp,i,L), an electronic (C1,O,o,H and C2,O,o,H) with NTC sensors and pressure gauges.
expansion valve that controls the evaporating process in a brazed Heat rejection in the condenser is performed with a loop working
plate evaporator with a heat transfer area of 2.39 m2. The valve with water, which allows controlling the inlet temperature (w,i)
regulates the degree of superheat at the evaporator exit (O,o,L) with and varying the flow rate (Vw). More information about the

Fig. 2. Schematic diagram of the experimental plant.

44 C. Sanz-Kock et al. / Applied Thermal Engineering 73 (2014) 41e50


secondary loops can be found in the work of Torrella et al. [15] and TK;L ¼ f P ¼ PC;i;L ; xv ¼ 1; CO2 (2)
Llopis et al. [16,17]


TO;H ¼ f P ¼ PC1;o;i;H ; xv ¼ 0; R134a (3)
2.2. Measuring system


TK;H ¼ f P ¼ Pdis;H ; xv ¼ 1; R134a (4)
The experimental plant is fully instrumented to analyse the
energy performance of the cycle. The allocation of sensors is pre- Then, the temperature difference in the cascade heat exchanger
sented in Fig. 2. It is equipped with 24 T-type immersion thermo- is computed with Eq. (5).
couples for measuring refrigerant and secondary fluids
temperature and a T-type air thermocouple for the environment DTcasc ¼ TK;L  TO;H (5)
one. Pressure is measured with 7 piezoelectric gauges for the LT
Eqs. (6) and (7) represent the pressure ratios of the CO2 and
cycle and 4 for the HT cycle, refrigerant mass flow rates with two R134a compressors, respectively.
Coriolis mass flow meters (Mref,L and Mref,H) and secondary volu-
metric flow rates (Vsf and Vw) with electromagnetic flow meters. 
tL ¼ Pdis;L P (6)
Power consumption is obtained with two digital watt meters (Pc,L suc;L

and Pc,H) and the compressors' speed (NL and NH) with the signal

from the inverter drives, calibrated using accelerometers and a tH ¼ Pdis;H P (7)
suc;H
frequency analyzer system. The calibration range and accuracies of
the measurement devices are detailed in Table 1. Regarding compressors' performance, the global efficiencies of
All the information obtained from the sensors is gathered by a the LT and HT compressors are calculated with Eqs. (8) and (9),
cRIO data acquisition system (24 bits of resolution) and handled respectively, where the specific isentropic compression works are
online with an own developed application based on LabView [18]. obtained with relations Eqs. (10) and (11). The compression work is
evaluated with the specific enthalpy at compressor inlet and the
subsequent isentropic specific enthalpy at discharge.
3. Data reduction
ws;L
hG;L ¼ m_ ref;L $ (8)
The analysis of the experimental plant is based on the infor- Pc;L
mation provided by the detailed measurement system. With the
measurements, thermodynamic properties of the refrigerants in ws;H
the cycles are evaluated using Refprop 9.1 database [19]. hG;H ¼ m_ ref;H $ (9)
Pc;H
Phase change temperatures in the heat exchangers are calcu-
lated using measured inlet pressure values and considering satu-
ws;L ¼ hs;L  hsuc;L (10)
rated state. LT evaporating temperature is calculated with Eq. (1)
using pressure at the inlet of the evaporator, and LT condensing
temperature with Eq. (2) with pressure at the inlet of the cascade ws;H ¼ hs;H  hsuc;H (11)
condensers. For the HT, the evaporating temperature is evaluated
About the energy parameters of the plant, the heat transfer rates
with Eq. (3) using pressure at the inlet of cascade condenser 1, and
in the LT cycle are evaluated as follows: The cooling capacity, which
the condensing level with Eq. (4) using the discharge pressure.
corresponds to the cooling capacity provided by the plant, with Eq.

 (12), considering the expansion process as isenthalpic. Heat rejec-
TO;L ¼ f P ¼ PO;i;L ; xv ¼ 0; CO2 (1)
tion at the gas-cooler with Eq. (13). Condensation heat transfer in
the cascade heat exchanger with Eq. (14), where we average the
enthalpy difference of both condensers. And the individual refrig-
Table 1 erating COP of the LT cycle with Eq. (15).
Accuracies and calibration range of the transducers.


Sensors Measured Measurement Calibration Calibrated Q_ O;L ¼ m_ ref;L $ hO;o;L  hO;i;L (12)
variable device range accuracy

62 to 125  C ±0.5  C


Q_ gc;L ¼ m_ ref;L $ hgc;i;L  hgc;o;L
25 Temperature T-type
thermocouple
(13)
3 Low pressure Piezoelectric gauge 0 to 60 bar ±0.18 bar
LT cycle "

#
hC1;k;i;L  hC1;k;o;L þ hC2;k;i;L  hC2;k;o;L
4 High pressure Piezoelectric gauge 0 to 100 bar ±0.3 bar
Q_ K;L ¼ m_ ref;L $
LT cycle 2
2 Low pressure Piezoelectric gauge 0 to 10 bar ±0.03 bar
HT cycle (14)
2 High pressure Piezoelectric gauge 0 to 25 bar ±0.075 bar
HT cycle
1 LT mass flow Coriolis 8.5 kg min1 ±0.15% Q_ O;L
COPL ¼ (15)
rate of lecture PC;L þ Pgc;L
1 HT mass flow Coriolis 52.8 kg min1 ±0.15%
rate of lecture Regarding the HT cycle, cooling capacity in the cascade heat
2 SF volumetric Magnetic flow 0 to 4 m3 h1 ±0.33%
exchanger is evaluated with Eq. (16), where the specific cooling
flow rates meter of lecture
2 Compressor Digital wattmeter 0 to 6 kW ±0.5% capacity is the average of the enthalpy difference in both cascade
power of lecture heat exchangers. In this case, the expansion processes are consid-
consumption ered isenthalpic too. The heat rejection in the HT condenser is
2 Compressor inverter drive 0 to 1800 rpm ±1.3% calculated with Eq. (17). And the individual refrigerating COP of the
speed signal of lecture
HT cycle with Eq. (18).
 C. Sanz-Kock et al. / Applied Thermal Engineering 73 (2014) 41e50 45

Table 2
Test summary of the cascade refrigeration plant evaluation.

TO,L ( C) TK,L ( C) DTSH,L ( C) NL (rpm) TO,H ( C) TK,H ( C) DTSH,H ( C) Tenv ( C) NH (rpm) Steady-states

Test 1 29.98 ± 0.05 4.98 to 0.21 9.01 ± 0.26 1450 9.17 to 3.43 50.07 ± 0.07 8.74 ± 0.82 27.54 ± 1.57 906.8 to 1309.6 5
Test 2 35.00 ± 0.04 6.66 to 0.35 9.03 ± 0.23 1450 11.29 to 3.50 50.01 ± 0.07 10.04 ± 0.50 26.60 ± 0.92 704.7 to 1208.7 5
Test 3 39.96 ± 0.02 11.52 to 4.83 9.04 ± 0.22 1450 17.10 to 9.52 49.96 ± 0.07 9.46 ± 1.85 24.31 ± 1.27 806.4 to 1409.4 5
Test 4 30.01 ± 0.03 6.93 to 1.59 9.44 ± 0.44 1450 10.68 to 5.00 40.06 ± 0.04 9.51 ± 0.65 21.80 ± 1.3 806.4 to 1208.7 5
Test 5 35.03 ± 0.06 11.00 to 6.34 9.28 ± 0.27 1450 15.29 to 10.28 40.05 ± 0.09 9.43 ± 0.46 21.39 ± 1.40 906.8 to 1309.6 5
Test 6 40.03 ± 0.03 12.80 to 8.02 9.32 ± 0.07 1450 17.96 to 12.49 40.07 ± 0.05 9.65 ± 0.44 20.68 ± 0.69 806.4 to 1208.7 5
Test 7 30.05 ± 0.12 9.23 to 3.82 9.31 ± 1.08 1450 12.71 to 7.25 30.17 ± 0.14 10.49 ± 1.19 22.75 ± 0.71 806.4 to 1208.7 5
Test 8 34.94 ± 0.05 12.17 to 7.23 9.48 ± 0.40 1450 16.03 to 10.57 29.97 ± 0.07 9.67 ± 1.09 20.98 ± 1.39 806.4 to 1208.7 5
Test 9 40.04 ± 0.05 15.36 to 10.27 9.39 ± 0.07 1450 20.15 to 14.43 30.07 ± 0.06 9.43 ± 0.90 19.99 ± 1.13 806.4 to 1208.7 5



 test, the compression rates were kept constant while varying the
hC1;O;o;H  hC1;O;i;H þ hC2;O;o;H  hC2;O;i;H
Q_ O;H ¼ m_ ref;H $ compressor's speed.
2
(16) 4.2. Data validation


Q_ K;H ¼ m_ ref;H $ hk;i;H  hk;o;H (17) Data validation was done comparing the heat transfer rates in
the main heat exchangers of the plant. In Fig. 3, we represent in red
dots (in web version) the heat transfer rate of water, Eq. (21), versus
Q_ O;H the heat rejection of R134a in the HT condenser, Eq. (17); in green
COPH ¼ (18)
PC;H diamonds the heat rejection of CO2, Eq. (12), versus the cooling
capacity of R134a in the cascade heat exchangers Eq. (16); and in
Finally, refrigerating COP of the whole plant is evaluated with
blue squares (in web version) the heat transfer rate of the sec-
Eq. (19), which considers the cooling capacity of the plant, Eq. (12),
ondary fluid, Eq. (20), versus the cooling capacity of CO2, Eq. (12), in
the electrical power consumption of both compressors and of the
the LT evaporator.
fan of the gas-cooler (75 W).
Regarding the heat balance at the HT condenser 99.6% of values
present a deviation below ±10%, at the cascade heat exchanger
Q_ O;L
COP ¼ (19) 97.0% of data are inside ±10%, and data in the LT evaporator 92.4%
Pc;L þ Pc;H þ Pgc deviates less than ±10%.
The heat transfer rates of the secondary fluids are also calcu-
lated: in the LT evaporator with Eq. (20) and in the HT condenser 5. Experimental results
with Eq. (21). Tyfoxit properties provided by the manufacturer [20]
are used and water properties are evaluated with Refprop 9.1. These A detailed energy balance of the cascade plant for a given
heat transfer rates are used for data validation. operating condition is given in Section 5.1, the compressors' per-
  formance is discussed in Section 5.2 and the global performance of
Q_ sf ¼ V_ sf $rsf $cp;sf $ Tsf;i  Tsf;o (20) the plant over a wide range of operating conditions is analysed in
Section 5.3.


Q_ w ¼ V_ w $rw $cp;w $ Tw;o  Tw;i (21) 5.1. Energy balance of the cascade plant

For clear understanding of the operation of the cascade refrig-

4. Experimental procedure and data validation eration plant, the temperature-entropy diagram of both

4.1. Experimental procedure

The test campaign used to evaluate the performance of the

cascade refrigeration system covered LT evaporating temperatures
from 40 to 30  C and HT condensing temperatures from 30 to
50  C, those values regulated and maintained by the secondary
fluid loop systems. For each combination of temperatures (9 tests),
the operation of the cascade was registered at five LT condensing
temperatures regulating the HT compressor speed. The LT
compressor speed was maintained at its nominal value, as
explained before. Tests were done fixing a degree of superheat in
the valves of the R134a cascade condensers and of the CO2 evap-
orator at 10  C. Also, the fan of the gas-cooler was always kept on,
that consuming a constant value of 75 W.
In total, 45 steady-states of the plant were measured, each lasting
at least 20 min, with 5 s sampling rate, with maximum oscillation of
the phase-change temperatures of 2%, as detailed in Table 2.
Additionally, two additional tests were carried out to analyse the
performance of the compressors under speed variation. In these Fig. 3. Validation of heat transfer rates in the cascade plant.
46 C. Sanz-Kock et al. / Applied Thermal Engineering 73 (2014) 41e50

refrigeration cycles is presented in Fig. 4, where the allocation of all

measurement points in the plant is reflected (Fig. 2). Also, in Fig. 5,
the energy flows through the cascade plant are detailed. Both fig-
ures represent the operation of the plant an LT evaporating tem-
perature of 29.98  C, an HT condensing temperature of 40.08  C
for compressors' speeds of 1450 rpm in the LT compressor and
806.4 rpm in the HT compressor. For this condition, the COP of the
cascade plant is of 1.42.
LT cycle absorbs heat in the evaporator (Q_ O;L ) and through the
suction line (Q_ suc;L ), this last accounting for less than 1% of the
cooling capacity. It consumes electricity in the CO2 compressor
(Pc,L) and in the fan of the gas-cooler (Pgc,L), this last accounting for
3.6% of the total consumption of the LT cycle. The cycle rejects
energy in the cascade condenser (Q_ K;L ), through the discharge line
(Q_ dis;L ) and in the gas-cooler (Q_ gc;L ), this last meaning 15.3% of the
total heat rejection. It is worth highlighting the objective of the gas-
cooler. If environment temperature allows it, the gas-cooler rejects
heat to the environment, thus avoiding pumping it to the
condensing temperature of the HT cycle. This way the COP of the
plant improves. The individual refrigerating COP of the LT cycle for
this condition is of 3.10.
HT cycle absorbs heat in the cascade condenser (Q_ O;H ) and in the
suction line (Q_ suc;H ), this last accounts for 1.2% of the heat taken by
the cycle. About electricity consumption, it absorbs energy in the
R134a compressor (PC,H). Cycle rejects heat in the HT condenser
(Q_ K;H ) and through the discharge line (Q_ dis;H ), this last means 6.9%
of the total heat rejection. The individual refrigerating COP of the
HT is of 2.84.

5.2. Compressors' performance

The way of modifying the intermediate conditions of the

cascade cycle (TK,L or TO,H) for a given operating condition (TO,L and
Fig. 5. Energy flow through the cascade plant at TO,L ¼ 29.98  C, TK,H ¼ 40.08  C,
TK,H) is through the variation of the compressors speed, either of NL ¼ 1450 rpm, NH ¼ 806.4 rpm.
the LT compressor or the HT compressor. If for a constant NH, NL is
increased, the LT cycle moves more refrigerant and the heat rejec-
tion at the cascade condenser increases, thus increasing TK,L and In Fig. 6, the evolution of the global efficiency is presented, Eqs.
TO,H, and obviously temperature difference in the cascade heat (8) and (9) (continuous line, left axis) and the discharge tempera-
exchanger also increases. The same applies for variation of NH. ture (dashed line, right axis) of each compressor under speed
Accordingly, to verify the best way to regulate the intermediate variation for fixed compression ratios. Although both compressors
conditions, both compressors were subjected to a speed variation are ready for variable speed operation, a big degradation of the
test under fixed compression ratios. Two individual tests were global efficiency of the CO2 compressor at low speeds was observed
performed, one for each compressor, where their energy perfor- (5.2% of reduction each 100 rpm in average), which consequence
mance was evaluated. We summarize the test in Table 3. was an increase of the discharge temperature (from 80 to 122  C).
This last effect is associated to the refrigeration of the electrical
motor of the compressor, as analysed for another CO2 compressor

nchez et al. [21]. About R134a compressor, efficiency variation
by Sa
with the speed is observed, but more soft (0.9% of reduction each
100 rpm). Accordingly, the experimental evaluation of the cascade
plant was carried out fixing NL at the nominal design condition of
the LT compressor (1450 rpm) and NH was varied to modify the
intermediate conditions.
In Fig. 7, the global efficiencies of both compressors versus the
compression ratio are presented, and in Eqs. (22) and (23) the
adjusted polynomials for the global efficiency of the CO2 (at
1450 rpm) and R134a compressors respectively, of all experimental
data presented in Table 2. We observe the performance of CO2
compressor is more dependent to the compression ratio than that
of R134a one.

hG;L ¼ 0:7245  0:0852$tL (22)

Fig. 4. Temperature-entropy of the cascade plant at TO,L ¼ 29.98  C, TK,H ¼ 40.08  C,

NL ¼ 1450 rpm, NH ¼ 806.4 rpm. hG;H ¼ 0:5871  0:01021$tH þ 6:067$105 $NH (23)
 C. Sanz-Kock et al. / Applied Thermal Engineering 73 (2014) 41e50 47

Table 3
Test summary of compressors' evaluation.

Psuc (bar) Pdis (bar) tL () Tsuc ( C) N (rpm) mref (kg/s) PC (kW) hG () Tdis ( C) Steady-states

LT Comp test 11.78 ± 0.11 27.44 ± 0.45 2.33 ± 0.06 12.46 ± 0.51 1100 to 1600 0.013 to 0.024 1.66 to 1.83 0.29 to 0.57 78.0 to 122.6 6
HT Comp test 1.24 ± 0.01 10.18 ± 0.04 8.23 ± 0.09 4.71 ± 0.48 1200 to 1600 0.031 to 0.045 2.71 to 3.63 0.70 to 0.77 88.0 to 89.6 5

Fig. 8. Phase-change temperatures vs. NH (at TO,L ¼ 30  C, Tk,H ¼ 40  C,

Fig. 6. Global efficiency of the compressors and discharge temperature vs. compres-
sor's speed (fixed compression ratios). NL ¼ 1450 rpm).

for the operation of the cascade plant at constant TO,L ¼ 30  C for
5.3. Energy performance of the cascade refrigeration plant three TK,H values in Fig. 9 and at constant TK,H ¼ 40  C for three TO,L
in Fig. 10. From the results, we observe the variations of TK,L versus
Here, the measured performance of the cascade plant over a NH are similar in both variation tests, with an average slope
wide range of operating conditions is presented, as detailed in of 1.3  C for each 100 rpm increment. Regarding the modification
Table 2. In the test, the degrees of superheat in the heat exchangers of TO,L (Fig. 10), TK,L increases 2.5  C for an increment of 5  C of the
were kept constant. The plant was evaluated at fixed operating TO,L. Nonetheless, when we modify TK,H (Fig. 9), the increments of
conditions (TO,L and Tk,H), for constant NL ¼ 1450 rpm, while varying TK,L are not uniform, it increases 2  C when TK,H rises from 30 to
the intermediate level through modification of NH, as presented in 40  C and 3  C when TK,H increases from 40 to 50  C. Eq. (24) rep-
Fig. 8. In it, NH was increased while maintaining constant TO,L and resents the dependence of the LT condensing temperature in ( C)
TK,H. When NH increases, the HT cycle provides more cooling ca- with NH, TK,H and TO,L adjusted from the experimental data.
pacity, the result being a decrease of the intermediate temperature
level (for both TK,L and TO,H). This modification of the intermediate TK;L ¼ 14:3617207  0:0111471$NH þ 0:2922665$TK;H
level modifies the individuals COP (if NH increases, COPL increases þ 0:6279067$TO;L (24)
and COPH decreases) and the temperature difference in the cascade.

5.3.1. LT condensing temperature versus HT compressor speed 5.3.2. Temperature difference in the cascade heat exchanger
As mentioned, the way of modifying the intermediate condi- Modification of NH causes variations of the compression ratios,
tions is through the variation of NH. We present experimental refrigerant mass flow rates, and energy parameters in each single-
measurements of the condensing temperature of the LT cycle (TK,L) stage cycle. When NH is increased, the cooling capacity of the HT
cycle rises, and it modifies the temperature difference in the

Fig. 7. Compressors' efficiencies vs. compression ratio. Fig. 9. LT condensing temperature vs. HT compressor speed (TO,L ¼ 30  C).
48 C. Sanz-Kock et al. / Applied Thermal Engineering 73 (2014) 41e50

Fig. 10. LT condensing temperature vs. HT compressor speed (TK,H ¼ 40  C). Fig. 11. Cascade temperature difference vs. HT compressor speed (TO,L ¼ 30  C).

theoretically by other authors [8,22] and verified experimentally

cascade heat exchanger (DTcasc), Eq. (5). We present the experi- for a NH3/CO2 cascade [12], however in our plant we could not
mental evolutions of DTcasc for constant operation of the cascade at reach an optimum value apparently. The last experimental point we
TO,L ¼ 30  C in Fig. 11 and for constant operation at TK,H ¼ 40  C in could measure for each test condition (the one at the maximum TK,L
Fig. 12. We observe the increase of NH negatively affects DTcasc. The in each line) was the limit of the CO2 compressor previous to the
increase of TK,H affects DTcasc (Fig. 11), but its variation is small, activation of the thermal protection of the compressor. Nonethe-
approximately 0.7  C increase of DTcasc when TK,H rises from 30 to less, we observe that the COP dependence on TK,L is not much sig-
50  C. However, variation of TO,L (Fig. 12) has greater influence on nificant, since over all the range of variation of TK,L the maximum
DTcasc, we measured an increment of 0.5  C when TO,L drops COP variation we measured was of 6%. Regarding COP dependence
from 30 to 35  C and of 0.8  C when it drops from 35 to 40  C. on TK,H (Fig. 15), we observe the cascade presents an average
Eq. (25) represents the dependence of the DTcasc in ( C) with NH, reduction of 18% for each 10  C increment of this temperature. On
TK,H and TO,L adjusted from the experimental data. the other side, COP dependence on TO,L, we observe a COP reduction
of 12% in average for each reduction of 5  C of this temperature
DTcasc ¼ 1:5858839 þ 0:0025688$NH þ 0:0348272$TK;H (Fig. 16).
 0:0515079$TO;L
(25) 5.3.5. Discharge temperature of the CO2 compressor
Finally, the measurements of the discharge temperature of the
CO2 compressor for constant TO,L ¼ 30  C are presented in Fig. 17
5.3.3. Cooling capacity
and for constant TK,H ¼ 40  C in Fig. 18. About the influence of TK,L,
The experimental cooling capacity provided by the plant, Eq.
we observe the discharge temperature rises linearly with TK,L at an
(12), at constant TO,L ¼ 30  C is presented in Fig. 13 and at constant
average ratio of increase of Tdis,L of 3  C for 1  C increment of TK,L.
TK,H ¼ 40  C in Fig. 14. For this representation we select the
About the external conditions, TK,H does not affect to the discharge
condensing temperature of the LT cycle for better understanding, as
temperature (Fig. 17) but TO,L strongly affects this value (Fig. 18). We
selected previously by other authors [12]. First and regarding TK,L,
measured in average increments of 20  C in Tdis,L when TO,L
we observe the capacity of the cascade plant reduces linearly when
increased 5  C.
the inter-stage level increases. This effect is caused by the reduction
As mentioned in the description of the plant, when the
of the refrigerant mass flow rate through the LT cycle. Although the
discharge temperature of the CO2 compressor is higher than the
COP of the plant can improve when TK,L increases (Section 5.3.4), its
environment temperature (in all tests in our experimental
increase always implies an increment of the low-stage compression
ratio (tL), which consequence is a reduction of the refrigerant mass
flow rate through the LT cycle. About the dependence of the ca-
pacity on the external conditions, we observe the capacity of the
cascade is much more dependent on the evaporating level (TO,L,
Fig. 14) than on the condensing one (TK,H, Fig. 13). The reason is that
when we modify TK,H the states of the refrigerant in the LT cycle
remain practically constant, whereas when we vary TO,L the mod-
ifications directly affect the states of the refrigerant at the LT
compressor suction. Obviously, we can affirm that the capacity of
the cascade is not much dependent on the environment
temperature.

5.3.4. COP
The evaluated COP of the cascade plant, Eq. (19), for operation at
TO,L ¼ 30  C is presented in Fig. 15 and for TK,H ¼ 40  C in Fig. 16.
About the COP, we observe it is dependent on the condensing
temperature of the LT cycle, it improves when TK,L increases. It
seems that a maximum COP exists for a given TK,L, as analysed Fig. 12. Cascade temperature difference vs. HT compressor speed (Tk,H ¼ 40  C).
 C. Sanz-Kock et al. / Applied Thermal Engineering 73 (2014) 41e50 49

Fig. 13. Cooling capacity vs. LT condensing temperature (TO,L ¼ 30  C). Fig. 16. COP vs. LT condensing temperature (Tk,H ¼ 40  C).

level of commercial refrigeration. The plant, fully instrumented,

allows measuring its experimental energy performance. Here, we
analysed the plant over a wide range of operating conditions,
evaporating levels from 40 to 30  C, condensing levels from 30
to 50  C with variation of the condensing temperature of the low-
temperature cycle.
It was observed the performance of the compressors was
different when subjected to variations of speed. While the R134a
compressor presented a small improvement when the speed
increased, the CO2 compressor presented low efficiencies at speeds
below the nominal point. However, at nominal conditions the
performance of the CO2 compressor was similar to that of R134a
one. We observed a quick decrease of efficiency of CO2 compressor
at high compression ratios.
In the plant we modified the intermediate level, condensing
temperature of the low-temperature cycle, with speed variation of
the high-temperature compressor. It was observed a negative linear
Fig. 14. Cooling capacity vs. LT condensing temperature (Tk,H ¼ 40  C). dependence of this temperature with the compressor speed. Also,
when this speed increased, temperature difference in the cascade
evaluation), it is recommended to reject heat in the LT cycle pre- heat exchanger increased too, and this increment was more sig-
vious entering to the cascade heat exchanger for improving COP nificant at low evaporating temperatures. The measured tempera-
(Figs. 2 and 5). In our measurements, the gas-cooler of the plant ture differences in the cascade heat exchanger in the tests ranged
extracted between 0.8 and 1.4 kW. from 3.3 to 5.3  C.
About the energy performance, it was observed the cooling
6. Conclusions capacity is negatively linear dependent with the condensing tem-
perature of the low-temperature cycle, and it is not much affected
This work presents the experimental evaluation of a R134a/CO2 with variations of the high condensing temperature, but its changes
cascade refrigeration plant driven by two semi-hermetic com- are very significant when subjected to modifications of the low
pressors designed to operate at the low evaporating temperature evaporating temperature. The measured cooling capacities ranged

Fig. 15. COP vs. LT condensing temperature (TO,L ¼ 30  C). Fig. 17. LT discharge temperature vs. LT condensing temperature (TO,L ¼ 30  C).
50 C. Sanz-Kock et al. / Applied Thermal Engineering 73 (2014) 41e50

[2] J. Arias, Energy Usage in Supermarkets e Modelling and Field Measurements,

Department of Energy Technology, Royal Institute of Technology, Sweden,
2005.
[3] European Commission, Regulation (EU) No 517/2014 of the European Parlia-
ment and of the Council of 16 April 2014 on Fluorinated Greenhouse Gases
and Repealing Regulation (EC) No 842/2006, 2014.
[4] UNEP, Report of the Refrigeration, Air Conditioning and Heat Pumps Technical
Options Committee (RTOC), 2010. Assessment, in, 2011.
[5] J.M. Calm, G.C. Hourahan, Physical, safety, and environmental data for current
and alternative refrigerants, in: I.I.o. Refrigeration (Ed.), 23rd IIR International
Congress of Refrigeration, IIR, Prage, 2011.
[6] T.S. Lee, C.H. Liu, T.W. Chen, Thermodynamic analysis of optimal condensing
temperature of cascade-condenser in CO2/NH3 cascade refrigeration systems,
Int. J. Refrig. 29 (2006) 1100e1108.
[7] A. Dopazo, J. Fern
andez-Seara, J. Sieres, F.J. Uhía, Theoretical analysis of a
CO2eNH3 cascade refrigeration system for cooling applications at low tem-
peratures, Appl. Therm. Eng. 29 (2009) 1577e1583.
[8] A. Messineo, R744-R717 cascade refrigeration system: performance evalua-
tion compared with a HFC two-stage system, Energy Procedia 14 (2012)
56e65.
[9] H.M. Getu, P.K. Bansal, Thermodynamic analysis of an R744-R717 cascade
refrigeration system, Int. J. Refrig. 31 (2008) 45e54.
Fig. 18. LT discharge temperature vs. LT condensing temperature (TK,H ¼ 40  C).
[10] J. Xiao, Y.F. Liu, Thermodynamic analysis of a R32/CO2 cascade refrigeration
cycle, Adv. Mater. Res. 732e733 (2013) 527e530.
from 7.5 kW at an evaporating temperature of 30  C at a [11] W. Bingming, W. Huagen, L. Jianfeng, X. Ziwen, Experimental investigation on
condensing of 30  C to 4.5 kW at 40  C and 40  C. About refrig- the performance of NH3/CO2 cascade refrigeration system with twin-screw
erating COP, and inside the test range, it was observed a depen- compressor, Int. J. Refrig. 32 (2009) 1358e1365.
[12] J.A. Dopazo, J. Ferna
ndez-Seara, Experimental evaluation of a cascade refrig-
dence of this parameter with the condensing temperature of the eration system prototype with CO2 and NH3 for freezing process applications,
low temperature cycle, increasing COP when rising the condensing Int. J. Refrig. 34 (2011) 257e267.
temperature. An optimum condensing temperature of the low [13] A. da Silva, E.P. Bandarra Filho, A.H.P. Antunes, Comparison of a R744 cascade
refrigeration system with R404A and R22 conventional systems for super-
temperature cycle could not be measured, if it existed, it was placed markets, Appl. Therm. Eng. 41 (2012) 30e35.
out of the application range of the LT cycle. Nonetheless, variations [14] UNEP, Low-gwp Alternatives in Commercial Refrigeration: Propane, CO2 and
of COP for a given external condition with this temperature were HFO Case Studies, 2014. www.unep.org/ccac/portals/50162/docs/Low-GWP_
Alternatives_in_Commercial_Refrigeration-Case_Studies-Final.pdf (29/06/
below 6%. The measured COP ranged from 1.05 at 40 and 40  C to
2014).
1.65 at 30 and 30  C. We measured reductions of COP of 18% when [15] E. Torrella, R. Llopis, R. Cabello, Experimental evaluation of the inter-stage
the high condensing temperature increased 10  C and reductions of conditions of a two-stage refrigeration cycle using a compound compressor,
12% for each 5  C reduction of the low evaporating temperature. Int. J. Refrig. 32 (2009) 307e315.

nchez, Performance evaluation of R404A
[16] R. Llopis, E. Torrella, R. Cabello, D. Sa
Finally, it was observed in the experimental measurements and R507A refrigerant mixtures in an experimental double-stage vapour
discharge temperatures of the CO2 compressor higher than the compression plant, Appl. Energy 87 (2010) 1546e1553.

nchez, HCFC-22 replacement with drop-
[17] R. Llopis, E. Torrella, R. Cabello, D. Sa
environment temperature, which brings the possibility of using a
in and retrofit HFC refrigerants in a two-stage refrigeration plant for low
gas-cooler to reject heat in the low temperature cycle before temperature, Int. J. Refrig. 35 (2012) 810e816.
entering the cascade condenser, which improves the COP of the [18] R. Cabello, R. Llopis, D. S
anchez, E. Torrella, REFLAB: an interactive tool for
plant. supporting practical learning in the educational field of refrigeration, Int. J.
Eng. Educ. 27 (2011) 909e918.
[19] E.W. Lemmon, M.L. Huber, M.O. McLinden, REFPROP, NIST Standard Reference
Acknowledgements Database 23, v.9.1, National Institute of Standards, Gaithersburg, MD, U.S.A,
2013.
[20] TYFO, High-performance Low Viscous Secondary Refrigerant for Applications
The authors gratefully acknowledge Jaume I University of Spain, Down to -55 C, 2013. www.tyfo.de/uploads/TI/IT-TYFOXIT-F15-F50_es_2013.
who financed the present study through the research project pdf (29/06/2014).
P1$B2013-10. [21] D. Sa
nchez, E. Torrella, R. Cabello, R. Llopis, Influence of the superheat asso-
ciated to a semihermetic compressor of a transcritical CO2 refrigeration plant,
Appl. Therm. Eng. 30 (2010) 302e309.
References [22] G. Di Nicola, F. Polonara, R. Stryjek, A. Arteconi, Performance of cascade cycles
working with blends of CO2 þ natural refrigerants, Int. J. Refrig. 34 (2011)
[1] D. Cowan, J. Gartshore, I. Chaer, C. Francis, G. Maidment, Real zero e reducing 1436e1445.
refrigerant emissions & leakage e feedback from the IOR project, Proc. Inst.
Refrig. 2009e10 (2010) 1e7.