©Strojniški vestnik 47(2001 38,366-373 ©Journal of Mechanical Engineering 471200138,366-373

ISSN 0 0 3 9 - 2 4 8 0 ISSN 0 0 3 9 - 2 4 8 0
UDK 5 3 6 .2 UDC 5 3 6 .2
Izvirni znanstveni članek (1 .01 ) Original scientific paper C1 .OD

THE PERFORMANCE OF THE

REFRIGERANTS R-134a, R-290, R404A, R-407c
AND R-4I0A IN AIR CONDITIONERS AND REFRIGERATORS

Josua P. Meyer
Department o f Mechanical Engineering
Rand Afrikaans University
PO Box 524, Auckland Park, 2006
Johannesburg, South Africa

ABSTRACT
A wide range o f measurements was performed on the agreement, individual countries may have domestic regulations
performance of the refrigerants R134a, R-290 (propane), R- for the earlier phase out o f R-22. For example, Sweden and
404A, R-407c, R-410A, and R-22 in a vapor compression Germany have set the most stringent regulations for the phase
experimental set-up. The results were used to evaluate and out o f HCFCs, i.e. in Germany for R-22 by January 1, 2000
compare the following performance criteria: cooling capacity, (Kruse and Tiedemann 1997).
coefficient o f performance, compressor discharge temperature, The search for a replacement for R-22 is a challenging task
input power, refrigerant mass flow, pressure ratio, and since no one substitute fits all needs. A few o f the medium-
volumetric refrigeration capacity. The measurements were term possible HFC replacements mentioned in literature are:
taken over a wide range o f evaporating temperatures from R-134a, (Sanvorderdenker 1997), R-404A (Weiss and Goguet
-20°C to 20°C at a condensing temperature of 55°C on the 1997), R-407C (Kruse and Tiedemann 1997), and R-410A
same experimental set-up. This wide selection of evaporating (Kruse and Tiedemann 1997; Keller et al. 1997). However, the
temperatures makes it possible to predict the performance of unknown long-term future of the HFCs due to their global
air-conditioning and refrigeration equipment. The results were warming potential has caused the manufacturers of unitary
used to compare the performances o f the different refrigerants equipment in central Europe to employ natural refrigerants,
as possible substitutes for R-22. It was concluded that propane such as R-290, which is propane (Keller, et. al. 1997), in
is a good long-term replacement for R-22. unitary air-conditioning and small heat pumps. The most
important concern regarding the adoption o f propane, is its
INTRODUCTION flammability. While this is an emotive subject which conjures
Environmental concerns repeatedly forced the air up visions of fireballs and dramatic explosions, safety concerns
conditioning and refrigeration industry into the selections of can be addressed (Ritter and Chem 1999). More and more
new and “more appropriate” refrigerants. This was an studies have also shown propane to be an excellent long-term
expensive and complex choice with little actual environmental replacement for R-22 (Meyer 1999; Douglas et al. 1999;
benefits once ozone depletion was addressed. In this regard, Keller et al. 1997; Purkayastha and Bansal 1998).
during the last decade, the air-conditioning and refrigeration The purpose o f this paper is to compare experimentally the
industry has gone through substantial changes due to the performance of the following refrigerants to R-22 as base: R-
Montreal Protocol (Cavallini 1996; Rowland 1997; Dugard 134a, R-290, R-404A, R-407c, and R-410A. Many studies
1997). The Montreal Protocol and Kyota Accord forced the have been conducted previously (Keller et al. 1997; Baskin et
climate control industry to substitute refrigerants for those that al. 1997; Richardson et al. 1996; Wei et al. 1997; Spatz and
were commonly used. The aim of the Protocol is to replace Zheng 1996; Payne et al. 1999), where some of these
traditional refrigerants, which are CFCs and HCFCs. The refrigerants have been compared with each other or to other
import of CFCs was already banned to most countries in 1996 refrigerants. This study differs from these previous studies in
while imports of HCFCs are progressively restricted, with the large number of refrigerants tested on exactly the same
complete phase-out early in the next century. Although experimental set-up, the accuracy of the experiments as well
numerous studies have revealed various possibilities, the as the wide range of evaporating temperatures tested, which
search for more efficient, cheaper, environmentally friendly, are representative of many air-conditioning and refrigeration
and safe refrigerants is most probably a never-ending one. operating conditions.
One of the most widely used HCFCs is R-22, which
represents the largest amount of refrigerant sold (Weiss and EXPERIMENTAL SET-UP
Goguet 1997). R-22 is to be phased out relative to a 1989 A schematic outlay o f the experimental set-up is given in
reference level for developed countries. Production was frozen Figure 1. It was a vapor compression refrigeration and/or heat
at the reference level on January 1, 1996. Production will be pump cycle. The compressor was a hermetically sealed,
limited to 65% of the reference level by January 1, 2004; to reciprocating type with a nominal cooling capacity of 4 kW.
35% by January 1, 2010; to 10% by January 1, 2015; and to The condenser was a water-cooled fluted tube heat exchanger.
0. 5% of the reference level by January 1, 2020. Complete A coriolis mass flowmeter with an error of ±0.1 % was used for
cessation of the production o f HCFCs is called for by January the measurement of the refrigerant mass flow. Sight-glasses
1, 2030 (ASHRAE 1997). In addition to the international were installed before and after the coriolus flowmeter to

01 -8 gnn^gbjiRoiigcgn I
s tra n 3 S 6
 I
J. P. Meyer: The performance of the refrigerants R - 1 34a, R - 2 9 0 , R 40 4A , R - 4 0 7 c

ensure that only liquid flows through it. A filter drier followed change of the refrigerant determined with thermodynamic
and a hand-controlled expansion valve. A water-heated fluted properties obtained from REFPROP (1993). This was
tube evaporator was used and a suction accumulator at the compared, for each experiment, to the heat loss from the
compressor inlet on the low-pressure side to complete the waterside, which was determined from the water temperature
refrigerant loop. change and water mass flow. The energy balance for all
Two main water loops were used, one flowing through the experiments was within ±3%.
condenser and one flowing through the evaporator. Chi the The refrigerants were evaluated and compared by making
condensing side the water loop was connected to a 1 000 liter use of the following parameters: cooling capacity, compressor
insulated storage tank connected to a 20 kW chiller. The water input power, cooling coefficient o f performance (COP),
flow rate through the test section could be controlled with a compressor input power, compressor discharge temperature,
hand-controlled valve. The flow rate through the condenser pressure ratio over the compressor and the volumetric
was measured with a coriolis mass flowmeter with an error of refrigeration capacity. The volumetric refrigeration capacity is
±0.02%. A similar flow loop was used on the evaporating side, the evaporator capacity divided by the swept volume o f the
also with a coriolis mass flowmeter and an insulated 1000 liter compressor.
storage tank, but connected to a 24 kW resistance heater. The
water temperatures in both loops could be thermostatically RESULTS AND DISCUSSION
controlled at a constant temperature with an error of ±1 °C. The results of the measurements are given in Figures 2 to
Temperatures were measured with RTDs calibrated to 8. From Figure 2 it follows that the cooling capacity of R-410A
measure temperature differences with errors less than is the highest. It is followed by R-22 and R-404A whose
±0.03°C. Temperatures were measured at all the refrigerant cooling capacities are almost identical, but approximately 6%
and water inlets and outlets of both the condenser and lower than that of R410A. Then R-407c follows with R-290,
evaporator as shown in Figure 1. At each o f these locations, which have similar cooling capacities except at high
four PTIOOs were located at the top, sides and bottom of the evaporating temperatures where R-407c has a slightly better
tube to take care of any circumferential temperature variation. capacity than R-290. R-134a has the lowest cooling capacity,
The average temperature of the four values was used as the which is on average 56% lower than the cooling capacity of
temperature measurement. Pressures were measured as R410A.
indicated in Figure 1 with 160 mm dial pressure gauges. On Measurements on the compressor input power (Figure 3)
the high-pressure side the gauges were calibrated to an error of show that R-134a needs the lowest energy input followed by
±5 kPa and to an error o f ±2 kPa on the low-pressure side. A R-290, R407c, R-22, R404A and R-410A. The compressor
kilowatt hour meter was used to measure the electric power inputs required by R-407c, R-22 and R 404A are very similar.
input to the compressor to an accuracy of ±0.5% of the The energy input of R134a is on average 27% lower than
reading. that o f R-290 and 61% lower than R-410A.
The cooling coefficients of performance (COP) are given in
EXPERIMENTS Figure 4. The heating COPs are not given, as they would
The system was pressurized first with nitrogen up to a theoretically be equal to the cooling COPs plus one. It was
pressure o f 3 MPa and placed in a hot-water bath for 48 hours found from the measurements that for all cases the heating
at 70°C and checked very well for leaks. The test area was also COPs were equal to the cooling COPs plus a factor of between
kept well ventilated. Tests were initially started with R-22 to 0.98 to 1.02. From Figure 4 it follows that the COP of R-134a
set up the base reference under identical operating conditions is the highest although the differences in cooling COP between
before the other refrigerants considered were loaded. Every R-134a, R-290 and R-22 are on average less than 2%. It is
time after one o f the other refrigerants was tested, the tests followed by R 404A and R410A, which have cooling COPs of
were repeated with R-22 to ensure that the base reference was approximately 12% lower than that of R134a. The cooling
still constant. The refrigerant charge was loaded until 5°C of COP o f R-407c is the lowest, on average 17% lower than that
subcooling occurred at ARI conditions (7.2°C evaporating and o f R-134a.
54.4°C condensing). A polyolester ISO 68 oil was used for In Figure 5 the refrigerant mass flows are given as function
compressor lubrication. of different evaporating temperatures while the condensing
Tests were conducted at different evaporating temperatures temperature is being kept constant at 55°C. The mass flows of
from -20°C to 20°C in steps of 5°C while maintaining the R 404A and R-410A are very similar and are the highest when
condensing temperature at 55CC. When 20°C was reached the compared to the other refrigerants. R-22, R-407c, R-134a and
tests were repeated in the reverse direction. These cycles were R-290 follow them. The refrigerant mass flow of R-290 is on
completed five times after which the average of ten values was average 65% lower than that o f R-404A and R-410A.
taken as the measured values. The temperatures were obtained The compressor discharge temperatures are given in Figure
by manual adjustments o f the water flow rates and different 6. R-22, R-410A and R 407c have high discharge temperatures
water temperatures. Care was taken to ensure that no pinch while the discharge temperatures of R-134a, R-290 and R-
points occurred in the condenser and evaporator. The 404A are much lower.
superheat of the refrigerant was kept constant for all tests at From Figure 7 it follows that the compressor pressure ratio
10°C. It was possible to keep the evaporating and condensing for R-290 is the lowest. R 404A , R-410A and R-22 follow it
temperatures constant at ±0.5°C during tests. closely. On average the pressure ratio of R-22 is only 9%
The 400 series refrigerants had temperature glides higher than the pressure ratio of R-290. The pressure ratios of
(typically 7.2°C for R 407c, 0.8'C for R-404A and 0.1 °C for R-134a and R407c are much higher than those o f the other
R-410A) during condensation and evaporation. In these cases refrigerants. The pressure ratio of R-407c is 37% higher than
that o f R-290. High-pressure ratios are in general detrimental
the condensing/evaporating temperatures were calculated by
averaging the bubble point and dew point temperatures. The to compressor life as they cause high stresses on the
heat transfer to the evaporator and condenser was calculated compressor bearings and crankcase. In general the volumetric
with the product of the refrigerant mass flow and enthalpy

fjŠT?TS(5)J)[i^Ž[KD|
 efficiency also decreases with an increase in pressure ratio Zeotropic Refrigerant Mixtures in a Lorenz-Meutzner
(Stoecker and Jones 1982). Refrigerator/Freezer, ASHRAE Transactions, 103(2): 59-67.
The volumetric refrigeration capacities are a measure of Douglas, J.D., J.E. Braun, E.A. Groll, and D.R. Tree.
the size of the compressor required for particular operating 1999. A Cost-based Method for Comparing Alternative
conditions and are given in Figure 8. The higher the Refrigerants Applied to R-22 Systems. International Journal
volumetric capacity o f the refrigerant, the smaller the size of o f Refrigeration 22(2): 107-125.
the compressor will be. The volumetric refrigeration capacity Dugard, P.H. 1997. The Montreal Protocol - an Exercise in
of R-410A is the highest by far. It is followed by R404A and International Environmental Diplomacy. ASHRAE Journal
R-22, which are almost the same, and on average 29% lower September: 38-41.
than the volumetric refrigeration capacity of R-410A. R-407c Keller, F.J., H. Liang, and M. Farzad. 1997. Assessment of
and R-290 follow with R-I34a that has the lowest volumetric Propane as a Refrigerant in Residential Air-Conditioning and
refrigeration capacity. The volumetric refrigeration capacity of Heat Pump Applications. In Proc. o f ASHRAE/NIST
R-I34a is on average 56% lower than that o f R-410A. Refrigerants Conference, Gaithersburg, Maryland: 57-65.
Although the compatibility of the refrigerants with oils and Kruse, H.H. and T. Tiedermann. 1997. Experience with
materials has not been investigated in detail it was found to be HC Refrigerants and Projections for Future Applications. In
compatible with all components of the system. Adverse effects Proc. of ASHRAE/NIST Refrigerants Conference,
of small traces of corrosive substances (e g. moisture or sulfur) Gaithersburg, Maryland: 44-56.
were not noted on either the compressor or the heat Meyer, J.P. 1999. Evaluation of Energy Efficient and
exchangers. Environmentally Acceptable Pure and Zeotropic Refrigerants
The results o f the experiments given in Figures 2 to 8 are in Air-conditioning and Refrigeration. Energy and the
summarized in Table I , together with other relevant criteria Environment, Voi. 15. Ed. Bejan, A., Vadasz, P. and Kröger,
such as: effect on the environment, cost, toxicity (ASHRAE D.G., Dordrecht: Kluwer: 239-246.
1997), flammability and compatibility with compressor oil and Payne, V.W. II, P.A. Domanski, and J. Muller 1999. A
refrigeration materials. The costs of the different refrigerants Study o f a Water-to-water Heat Pump using Hydrocarbon and
relative to R-22 are given in Figure 9. Hydrofluorocarbon Zeotropic Mixtures, NIST, Gaithersburg,
It can be concluded from Table 1 as well as from Figures 2 MD., USA.
to 8 that not one o f the refrigerants outperformed all the other Purkayastha B. and P.K. Bansal. 1998. An Experimental
refrigerants on all the criteria considered. In the short term R- Study on HC290 and a Commercial Liquefied Petroleum Gas
134a and the R-400 series refrigerants are excellent (LPG) Mix as Suitable Replacements for HCF22.
replacements (especially R-404A and R-410A) and in many International Journal o f Refrigeration 21(1): 3-17.
instances are better than R-22. Propane (R-290) performed REFPROP, 1996. Standard Reference Database 23, NIST
marginally better than the other refrigerants considered. Thermodynamic properties o f refrigerant mixtures database.
Taking into consideration that it has an excellent ozone Version 5, Gaithersburg, MD.
depletion potential (ODP) and global warming potential Richardson, R.N. and T.J. Ritter. 1996. The Performance
(GWP) it is a good long-term alternative for R-22. However, it of Hydrocarbon Alternatives to R12 and R22. A1RAH Journal,
is important that its flammability potential is addressed (Ritter August: 21-28.
and Chem 1999). Ritter, T.J. and C. Chem. 1999. Flammability -
Hydrocarbon refrigerants. Refrigeration and Airconditioning
CONCLUSIONS 14(3): 39-50.
All the refrigerants considered in this paper (R-134a, R- Rowland, F.S. 1997. Stratospheric Ozone Depletion:
290, R404A, R-407c, R410A) can be used as alternatives for Identifying the Problem and the Solutions. ASHRAE Journal
R-22. In the short term R134a is an excellent replacement. September: 29-31.
However, the global wanning potential of R134a is high. Sanvordenker, K.S. 1997. Status of CFC and HCFC
Other alternatives are R-404A, R-410A and R407c (in the Alternatives. In Proc. o f ASHRAE/NIST Refrigerants
order given) with R-404A and R-410A performing the best. Conference, Gaithersburg, Maryland: 111-116.
Propane in general performed marginally better than all the Spatz, M.W. and J. Zheng. 1996 R-22 Alternative
refrigerants considered. It is a good long-term replacement for Refrigerants: Performance in Unitary Equipment, ASHRAE
R-22 as it is very environmentally friendly. The flammability
Technical Data Bulletin 9(4): 48-54.
of propane should, however, be addressed. Stoecker, W.F. and J.W. Jones. 1982. Refrigeration & Air-
conditioning, McGraw-Hill, Singapore, 2nd ed.
REFERENCES Wei, C, S. Lin, and C. Wang. 1997. System Performance of
ASHRAE. 1997. ASHRAE Handbook - Fundamentals. a Split-Type Unit Having R-22 and R^l07c as Working Fluids.
Atlanta: ASHRAE. ASHRAE Transactions 103(2): 797-802.
Cavallini, A. 1996. Working Fluids for Mechanical Weiss, P. and J. Goguet. 1997. Current and Projected Use
Refrigeration. International Journal o f Refrigeration 19(8): of Refrigerants in Europe. In Proc. o f ASHRAE/NIST
485-496. Refrigerants Conference, Gaithersburg, Maryland: 38-43.
Baskin, E.B., E.S. Bayoglu, and F.R. Delafield. 1997.
Evaluation of Ozone-Friendly Hydrofluoropropane-Based

IgrifWäLDIMa^lXfll
 1 C o m p re ssor
2 W o te r-c o o le d co nd en se r
3 C o rio lis flo w m e te r
4 F ilte r d rie r
5 H o n d - c o n tro lle d e xpansion volve
6 W a te r-h e a te d e v a p o ra to r
7 S u ctio n a c c u m u la to r
8 C o ld - w a te r s to ra g e tan k
9 A ir - c o o le d ch ille r
10 H o t- w a te r s to ra g e ta n k
11 E le c tric a l h e a tin g e le m e n t
— » R e frig e ra n t flow
— — W ater flow
— tx j— H a n d -c o n tro lle d valve
----- < PT100 te m p e ra tu re m e a s u re m e n t
— [£ ] P re ssure m e te r
— ^3“ S ig h t-g la s s
- -^ )— P um p

Figure 1: Schematic representation o f the experimental set-up.

Figure 2: Cooling capacity at a condensing temperature o f 55°C.

 Figure 3: Compressor input power at a condensing temperature o f 55°C.

Figure 4: Cooling coefficient o f petformance at a condensing temperature o f 55 °C.

□ 1 -B
IVSsfFlMDIK s tra n 3 7 0
Figure S: Refrigerant mass flow at a condensing temperature o f 55°C.

Figure 6: Compressor discharge temperature at a condensing temperature of55°C.

Figure 7: Compressor pressure ratio at a condensing temperature o f 55°C.

Evaporating temperature [ deg. C ]

Figure S: Volumetric refrigeration capacity at a condensing temperature o f 55X1.

 Refrigerant

Figure 9: Relative cost per kilogram o f refrigerants compared to R-22.

Criteria Refrigerant
1 = poor R -2 2 R -2 9 0 R -1 3 4 a R -4 0 4 A R -407C R -4 1 0 A
5 = excellent
Cooling capacity 3 2 1 3 2 5
Input power 3 4 5 2 3 1
Cooling COP 4 5 5 3 2 3
Refrigerant mass flow 4 1 2 5 3 5
Discharge temperature 3 5 5 5 3 3
Pressure ratio 5 5 2 5 1 5
Vol. refrig. capacity 4 3 1 4 3 5
ODP 1 5 5 5 5 5
GWP 2 5 3 1 2 2
Cost 3 5 2 2 1 1
Non-toxicity 5 5 5 5 5 5
Non-flammability 5 1 5 5 5 5
Compatibility 5 5 5 5 5 5
TOTAL 47 51 46 50 40 50

Table 1: Evaluation and Comparison o f Refrigerant Properties

Q1 -S