Performance of Non-flammable Azeotropic

HFO1234yf/HFC134a mixture for HFC134a Applications

Yohan Lee, Dong-gyu Kang, Dongsoo Jungˋ

Department of Mechanical Engineering, Inha University, Incheon, Korea, 402-751

(ˋ) Corresponding author. Tel.: +82-32-860-7320; Fax: +82-32-868-1716; E-mail address: dsjung@inha.ac.kr

ABSTRACT

In this study, ‘drop-in’ performance of HFC134a, HFO1234yf and HFO1234yf/HFC134a mixture at three

compositions of 5%, 10%, and 15% HFC134a is measured in a heat pump bench tester under summer and

winter conditions. Test results show that the COP, capacity, discharge temperature of HFO1234yf and

HFO1234yf/HFC134a mixture are similar to those of HFC134a. For HFO1234yf/HFC134a mixture,

flammability decreases as more HFC134a is added and at compositions of more than 10% of HFC134a, the

mixture becomes non-flammable. The amount of charge for HFO1234yf and HFO1234yf/HFC134a mixture is

up to 11% lower than that of HFC134a. Since HFO1234yf/HFC134a mixture with 10-11% HFC134a is

non-flammable and azeotropic and has no ODP and GWP of less than 150 meeting the requirement of European

mobile air-conditioner directive, it can be used as a long term environmentally friendly solution for various

HFC134a applications including mobile air-conditioners with minor modifications.

Keywords:

Alternative refrigerant, HFO1234yf, HFC134a, HFO1234yf/ HFC134a mixture, mobile air-conditioner

- Y -G

Nomenclature

CFC Chlorofluorocarbon

COP Coefficient of performance

GTD Gliding temperature difference (°C)

GWP Global warming potential

HFC Hydrofluorocarbon

HCO Hydrofluorolefin

HTF Heat transfer fluid

LFL Lower flammability limit(Vol. % in air)

MAC mobile air-conditioner

ODP Ozone depletion potential

Q Capacity [W]

T Temperature (°C)

Subscript

c condenser

diff difference

dis discharge

e evaporator

- Z -G

1. Introduction

Since 1930s, CFCs have been widely used in refrigeration and air-conditioning equipment. They, however,

were found to be responsible for the destruction of the ozone layer and thus in 1987, the Montreal protocol was

proposed to phase out the ozone depleting substances (UNEP, 1987). At this time, CFCs are entirely phased out

while HCFCs are being phased out in the world thanks to the Montreal protocol.

In order to fill the gap caused by the phase out of CFCs, refrigeration and air-conditioning industry has carried

out extensive research and development activities in an attempt to find alternative refrigerants whose ozone

depletion potential (ODP) is 0. As a result of these efforts, HFC134a has been successfully developed and

adopted in domestic refrigerators and mobile air-conditioners (MACs) for the past two decades. HFC134a is

known to have a similar vapor pressure and performance to those of CFC12.

In these days, global warming has been the one of the most important issues facing mankind and in 1997,

Kyoto protocol was proposed to control greenhouse gases including HFCs (GECR, 1997). Consequently,

HFC134a was identified as one of the controlled greenhouse gases. The 100 year global warming potential

(GWP) of HFC134a is 1430 as compared to that of carbon dioxide (CO2). Therefore, it needs to be replaced by

more environmentally friendly refrigerant in the near future. Thus, EU F-Gases Regulation and MAC directive

bans the use of HFC134a from 2011 in MACs of newly manufactured vehicles for environmental protection

(OJEU, 2006). The same MAC directive specifically prohibits the use of fluorinated greenhouse gases of which

GWP is greater than 150.

Recently, HFO1234yf has been suggested as a possible alternative refrigerant for HFC134a in MACs (Minor

and Spatz, 2008; Zilio et al., 2009) and in beverage coolers (Minor et al., 2010). An azeotropic mixture

composed mainly of HFO1234yf, called DR-11, was suggested for replacing HFC134a in centrifugal chillers

(Kontomaris et al., 2010). HFO1234yf has zero ODP and excellent life cycle climate performance as compared

to HFC134a. The 100 year GWP of HFO1234yf is 4 as compared to that of CO2 (Nielsen et al., 2007) and hence

HFO1234yf meets the current EU regulations.

HFO1234yf could be used as a ‘near drop-in replacement’ for HFC134a and this implies that automobile

manufacturers would not have to make significant modifications in assembly lines or in vehicle system designs

to accommodate the product. HFO1234yf has the lowest switching cost for automobile manufacturers among

the currently proposed alternatives, although the initial cost of the product is much higher than that of HFC134a.

The product could be handled in repair shops in the similar way as HFC134a, although it would require different,

specialized equipment to perform the service. One of the reasons for that is the mild flammability of

- [ -G

HFO1234yf. Another issue affecting the compatibility between HFO1234yf and HFC134a systems is the

lubricating oil since the current oil is showing signs of damage to some materials.

In April 2012, the European Commission declared that, due to the lack of HFO 1234yf availability from

suppliers, automobile manufacturers are able to use HFC134a to fill new type-approved production vehicles

until the end of 2012. New vehicles still have to be compatible with the MAC directive requiring that the new

refrigerant must have a GWP of less than 150.

Recently, one of the major automobile manufacturers in Europe refused to adopt HFO1234yf due to its

flammability issue in actual situations. In fact, in September 2012, Daimler has provided relevant authorities

with the findings of an investigation which raises questions on the safe usage of HFO1234yf (Autoblog, 2012).

Up to now, the climate-friendly chemical was set to be used worldwide in the automotive industry and was

previously perceived to be safe. This was determined by numerous laboratory and crash tests carried out by

international vehicle manufacturers and independent institutions.

Despite multiple confirmations of non-critical results, Daimler carried out a series of additional tests on

HFO1234yf as part of a new real-life test scenario developed in-house which goes above and beyond the legally

prescribed requirements.

In the new real-life test scenario, HFO1234yf is dynamically dispersed at high pressure near to hot

components of the test vehicle's exhaust system. This corresponds to a serious head-on collision in which the

refrigerant line is severed and the reproducible results demonstrate that HFO1234yf which is otherwise difficult

to ignite under laboratory conditions can indeed prove to be flammable in a hot engine compartment. Similar

tests of the current HFC134a refrigerant did not result in ignition.

Due to the new findings of this study and the high safety demands, HFO1234yf will not be used in its

products. The company therefore wishes to continue to use the proven and safe HFC134a in its vehicles. At this

time, General Motors seems to be the only company that will use HFO1234yf in 2013 for some of their models

in the US.

In this study, an azeotropic mixture of HFO1234yf/HFC134a is proposed for replacing HFC134a in various

applications such as MACs and beverage coolers and centrifugal chillers. By adding 10-11% of HFC134a to

HFO12234yf, the mixture becomes non-flammable with GWP still less than 150. Hence, it can solve

successfully the major issue of HFO1234yf’s flammability while it meets the MAC directive requirement. This

mixture also is cheaper and more compatible than HFO1234yf.

The objectives of this paper are to measure the performance of this mixture in a heat pump bench tester and to

- \ -G

provide the data with comparison against HFC134a and HFO1234yf.

2. Experiments

2.1 Experimental apparatusG

To achieve the goals of this paper, a breadboard type water source heat pump bench tester in which

refrigerant and secondary heat transfer fluid (HTF) flow in a counter current manner was designed and built in

our laboratory (Lee et al., 2012). The bench tester is of the similar structure used by US National Institute of

Standards and Technology (NIST) (Mulroy et al., 1988). For the bench tester, uniform external conditions can

be imposed by controlling the HTF temperatures. Fig. 1 shows the schematic of the experimental bench tester

whose nominal capacity is 1 ton of refrigeration (3.5 kW).

The evaporator and condenser of the heat pump tester were manufactured by connecting 8 pieces of

pre-manufactured double tube commercial pipes (E-stick) in series. Each pipe stick is 740 mm long and inner

and outer diameters are 19.0 mm and 25.4 mm, respectively. Both the evaporator and condenser were designed

to be counter-current and the secondary HTF passed through the inner tube while the refrigerant flowed through

the annulus. Precision chiller and heating bath of 0.1¶C accuracy were employed to control the temperatures of

the water/ethylene glycol mixture and water entering into the evaporator and condenser, respectively.

The bench tester was equipped with an open type variable speed compressor connected to an electric motor

with an inverter. Throughout the tests, however, no change was made in the compressor speed since ‘drop-in’

tests were performed. A fine metering needle valve was used as an expansion device to control the refrigerant

mass flow rate. Even though a suction line heat exchanger (SLHX) was installed initially to examine the effect

of SLHX, it has not been used during this study.

To properly compare the performance of various refrigerants, a fair test condition should be employed. For

this purpose, all tests were conducted with the external HTF temperatures fixed. In this study, tests were

performed under two sets of different evaporator/condenser saturation temperatures for HFC134a: 7¶C/45¶C,

-7¶C/41¶C, respectively as done by Lee et al. (2012). The first condition is for normal air-conditioning during

summer. On the other hand, the second condition is for normal heat pumping during winter in Korea. For a

given condition, first of all, tests were carried out for HFC134a with the adjusted external HTF temperatures to

provide the required saturation temperatures in the evaporator and condenser. And then, subsequent tests were

performed under the same external conditions for HFO234yf and HFO1234yf/HFC134a mixture at three

- ] -G

compositions of 5%, 10% and 15% HFC134a. For a given external condition, actual saturation temperatures of

the various refrigerants in the evaporator and condenser varied a little due to the difference in heat transfer

characteristics of these fluids.

The subcooling and superheat at the exits of the condenser and evaporator were fixed to be 5¶C with ·1¶C

variation. Under this condition, heat transfer characteristics of each refrigerant are reflected in the measured

energy efficiency.

Table 1 lists all refrigerants tested and their gliding temperature differences (GTDs) and GWPs and lower

flammability limits (LFLs). Since the mixture is azeotropic (or very close to azeotropic), GTD, the temperature

difference between the beginning and ending temperatures during evaporation, is close zero as listed in Table 1.

All thermodynamic properties were calculated by REFPROP routine developed by US NIST (Lemmon et al.,

2010). For all fluids tested, a PAG oil for HFC134a was used. Since Lee et al. (2012) contains all the details of

the test apparatus, measurements, and experimental procedure, they will not be presented again here. An

interested reader is referred to Lee et al. (2012) for the details.

3. Results and discussion

Table 2 lists the ‘drop-in’ tests results for HFC134a, HFO1234yf, and HFO1234yf/HFC134a mixture under

summer and winter conditions. For these refrigerants, tests were repeated at least three times to confirm the data.

All data in Table 2 have good repeatability showing less than 1% scatter.

As seen in Table 2, the COPs of HFO1234yf and HFO1234yf/HFC134a mixture are slightly lower than that

of HFC134a under summer condition. On the other hand, they are very similar to that of HFC134a under winter

condition. Measured capacities (Q in Table 2) of HFO1234yf and HFO1234yf/HFC134a mixture are again

slightly lower than that of HFC134a under summer condition. On the other hand, they are a little bit higher than

that of HFC134a under winter condition. Discharge temperatures (Tdis in Table 2) of HFO1234yf and

HFO1234yf/HFC134a mixture are always lower than those of HFC134a by up to 8.7¶C. Based upon the results,

it can be concluded that the performance of HFO1234yf and HFO1234yf/HFC134a mixture is similar to that of

HFC134a. This trend would be even truer if the system is optimized for new fluids with proper lubricating oil.

As listed in Table 1, GTDs of HFO1234yf/HFC134a mixture are close to zero. This means that this

practically azeotropic mixture would act like any pure refrigerant showing no composition shift even though

there is a leakage in the system. As well known, for non-azeotropic mixtures, fractionation due to a leakage is

- ^ -G

one of the major issues.

Park and Jung (2010) and Park et al. (2011) measured the nucleate boiling and external condensation heat

transfer coefficients of HFC134a and HFO1234yf and reported that they are very similar for various surfaces

under typical evaporation and condensation temperature ranges. Since HFO1234yf/HFC134a mixture is

azeotropic, heat transfer degradation associated with non-azeotropic mixtures would not be expected, which is

another excellent feature for efficient heat exchanger design.

Table 1 also lists the lower flammability limits (LFLs) of HFO1234yf and HFO1234yf/HFC134a mixture.

LFLs are measured using a flammability testing apparatus manufactured according to ASTM E681-04 standard

(ASTM, 2004) with a 12 liter spherical glass flask as prescribed in detail in ASHRAE standard(ASHRAE,

2010). LFLs were determined by visual observation of the flame propagation recorded by the camcorder. As the

spark occurred at the ignition source, the flame occurred near the ignition source. Refrigerants were determined

to be flammable when the flame went outside the 90¶ fan marked by tapes on the flask (ASHRAE, 2010).

To check the reliability of the apparatus, many tests using a few well known flammable refrigerants were

performed. As listed in Table 1, LFL of HFO1234yf at 23¶C is 6.5% and LFLs of HFO1234yf/HFC134a

mixture increase as more non-flammable HFC134a is added to HFO1234yf. One thing to be noted is that

HFO1234yf/HFC134a mixture becomes non-flammable with more than 10% HFC134a. The flame was

confined inside the 90¶ fan marked by tapes on the flask.

Finally, the amount of charge for HFO1234yf and HFO1234yf/HFC134a mixture is up to 11% lower than that

of HFC134a. This is due to a decrease in density with HFO1234yf (Lemmon et al., 2010). This is also a good

feature from the viewpoint of environmental pollution caused by refrigerant leaks. This will help alleviate

further the direct emission of refrigerant which is responsible for the global warming.

4. CONCLUSIONS

In this study, ‘drop-in’ performance of new refrigerant HFO1234yf and HFO1234yf/HFC134a mixture at

three compositions 5%, 10%, and 15% HFC134a is measured in a heat pump bench tester under summer and

winter conditions. Based upon the test results, following conclusion can be drawn.

1. The COP, capacity, discharge temperature of HFO1234yf and HFO1234yf/HFC134a mixture are

similar to those of HFC134a.

2. For HFO1234yf/HFC134a mixture, flammability decreases with more HFC134a and at compositions

- _ -G

above 10% of HFC134a, the mixture becomes non-flammable.

3. The amount of charge for HFO1234yf and HFO1234yf/HFC134a mixture is up to 11% lower than

that of HFC134a due to a decrease in density with HFO1234yf.

4. Since HFO1234yf/HFC134a mixture with 10-11% HFC134a is both non-flammable and azeotropic

and has no ODP and low GWP of less than 150, it can be used as a long term environmentally friendly

solution for various HFC134a applications with minor modifications.

Acknowledgement

This work was financially supported by the National R&D project of the “Development of Energy utilization

technology with Deep Ocean Water” supported by the Korean Ministry of Land, Transport and Maritime

Affairs(2012). Inha University also supported this work.

References

Autoblog, 2012. Daimler sounds alarm on new ac refrigerant maybe flammable.

http://www.autoblog.com/2012/09/28

ASHRAE, 2010. Designation and safety classification of refrigerants. ASHRAE Standard, ANSI/ASHRAE

Standard 34-2010.

ASTM, 2004. Standard test method for concentration limits of flammability of chemicals. ASTM E681-04,

Philadelphia: American Society for Testing and Materials.

Global Environmental Change Report, 1997. A brief analysis of the Kyoto protocol. Vol. IX, p. 24.

Kontomaris, K., Leck, T., Hughes, J., 2010. A non-flammable reduced GWP, HFC-134a replacement in

centrifugal chillers: DR-11. In: International Refrigeration and Air Conditioning Conference at Purdue, West

Lafayette, IN, USA, Paper No. 2142.

Lee, H., Kim, H., Kang, D., Jung, D., 2112. Thermodynamic performance of R32/R152a mixture for water

source heat pumps. Energy 40, 100-106.

Lemmon, E. W., Huber, M. L., McLinden, M. O., 2010. NIST Reference fluid thermodynamics and transport

properties, REFPROP version 9.0.

Minor, B., Spatz, M., 2008. HFO-1234yf low GWP refrigerant update. In: International Refrigeration and Air

Conditioning Conference at Purdue, West Lafayette, IN, USA, Paper No. 2349.

Minor, B., Montoya, C., 2010. HFO-1234yf performance in a beverage cooler. In: International Refrigeration

- ` -G

and Air Conditioning Conference at Purdue, West Lafayette, IN, USA, Paper No. 2422.

Mulroy, W., Kauffeld, M., McLinden, M.O. , Didion, D.A., 1988. Experimental evaluation of two refrigerant

mixtures in a bread-board air conditioner. In: Proceedings of the 2nd DOE/ORNL Heat Pump Conference,

Washington, DC, USA, 55-61.

Nielsen, O.J., Javadi, M.S., Sulbak, A., Hurley, M.D., Wallington, T.J., Singh, R., 2007. Atmospheric

chemistry of CF3CF=CH2; Kinetics and mechanisms of gas-phase reactions with Cl atoms, OH radicals, and O3.

Chem. Phys. Letters 439, 18-22.

Official Journal of the European Union, 2006. Directive 2006/40/EC of the European Parliament and of the

Council. 14.6.2006.

Park, K., Jung, D., 2010. Nucleate boiling heat transfer coefficients of R1234yf on plain and low fin surfaces.

Int. J. Refrigeration 33, 553-557.

Park, K., Kang, D., Jung, D., 2011. Condensation heat transfer coefficients of R1234yf on plain, low fin, and

Turbo-C tubes. Int. J. Refrigeration 34, 317-321.

United Nations Environmental Programme, 1987. Montreal protocol on substances that deplete the ozone

layer. Final act, New York; United Nations.

Zilio, C., Brown, S.S., Cavallini, A., 2009. Simulation of R-1234yf performance in a typical automotive

system. In: Proceedings of the 3rd Conference on Thermophysical Properties and Transfer Processes of

Refrigeration, Boulder, CO, USA, Paper No. 128.

- XW -G

Table 1 GTD, GWP, and LFL of HFC134a, HFO1234yf, and HFO1234yf/HFC134a mixture

Refrigerant GTD

(¶C) GWP

LFL

(Vol % in air)

HFC134a 1430

HFO1234yf 4 6.8

95%HFO1234yf/5%HFC134a 0.2 75.3 7.2

90%HFO1234yf/10%HFC134a 0.3 146.6 7.5

85%HFO1234yf/15%HFC134a 0.3 217.9 7.8

- XX -G

Table 2 Performance of HFC134a, HFO1234yf, and HFO1234yf/HFC134a mixture

Refrigerant COP

Q

(W)

Tdis

(¶C)

COPdiff

(%)

Qdiff

(%)

Tdis.diff

(¶C)

Summer

(Te: 7¶C,

Tc: 45¶C)

HFC134a 2.01 3506 76.5

HFO1234yf 1.95 3368 69.8 -2.7 -4.0 -8.7

95%HFO1234yf/5%HFC134a 1.95 3444 72.3 -3.0 -1.8 -5.5

90%HFO1234yf/10%HFC134a 1.94 3458 72.2 -3.5 -1.4 -5.6

85%HFO1234yf/15%HFC134a 1.93 3476 72.4 -3.9 -0.9 -5.4

Winter

(Te: -7¶C,

Tc: 41¶C)

HFC134a 2.64 3244 76.3

HFO1234yf 2.62 3240 69.9 -0.8 -0.1 -8.4

95%HFO1234yf/5%HFC134a 2.64 3362 71.6 -0.0 3.6 -6.2

90%HFO1234yf/10%HFC134a 2.63 3355 71.4 -0.4 3.4 -6.4

85%HFO1234yf/15%HFC134a 2.63 3352 71.3 -0.3 3.3 -6.6

G

G

- XY -G

G

G

Fig.1 – Schematic of the heat pump bench tester

Fig. 1 Schematic of a water source heat pump bench tester.