Int J. Refrig. Vol. 21, No. 8, pp.

617625, 1998 1998 Elsevier Science and IIR Printed in Great Britain. All rights reserved PII:S0140-7007(98)00030-9 01407007/98/$19.00+00

An experimental study of an ejector cycle refrigeration machine operating on R113

Nehad Al-Khalidy Technical University of Selesia, ul.Konarskiego 22, 44-101 Gliwice, Poland
Received 3 November 1997; revised 27 March 1998; accepted 27 March 1998

This paper discusses an ejector cycle refrigeration machine that can use a wide range of refrigerants including halocarbons. The feature of such a system is the possibility of using a low grade heat source such as solar energy and waste heat to operate the system. A theoretical analysis was carried out to select a suitable refrigerant for the system. The inuence of boiler, condenser and evaporator temperatures on system heat transfer is investigated experimentally under different operating conditions. The experimental machine uses R113 as a working refrigerant. 1998 Elsevier Science Ltd and IIR. All rights reserved.
(Keywords: ejector system; heat pump; solar energy)

rimentale dune machine frigorique Etude expe `e jecteur au R113 a

sente une machine frigorique a `e jection qui peut fonctionner avec beaucoup Cet article pre ` nes, y compris des halocarbures. Linte re t dun tel syste ` me est quon peut le faire de frigorige nergie solaire ou fonctionner avec une source de chaleur de valeur basse telle que de le cupe re e. On a mene une analyse the orique pour choisir un frigorige ` ne adapte . Linuence re ratures du bouilleur, du condenseur et de le vaporateur sur le transfert de chaleur des tempe ` me a e te e tudie expe rimentalement dans diverses conditions. La machine dans le syste rimentale fonctionne au R113. 1998 Elsevier Science Ltd and IIR. All rights reserved. expe s: syste ` me a `e jecteur; pompe a ` chaleur; e nergie solaire) (Mots cle

Introduction Refrigeration and air conditioning machines consume large quantities of energy. Therefore, the conservation of energy in these machines has received a great deal of interest. The conventual air conditioning compression system is one of the more efcient ones among the available systems in terms of the Coefcient of Performance (COP). refrigeration effect Q COP e compressor mechanical power Wc (1)

range of 101000 tons of refrigeration. The advantages of the steam jet refrigeration system were as follows: Operating cost was very low when the waste steam was available. Maintenance cost was also low because of the absence of moving parts (except pumps) in the system.

These system disappeared when the vapour absorption system (cooling and heating) was developed. The major reasons for their disappearance were: They cannot be used for creating temperatures below 4C in the evaporator because the refrigerant (water) would freeze. The very low pressure in the evaporator: a purge unit must be used to remove non-condensable gases from the system1. The overall coefcient of the system COP overall was

This system consumes a large amount of energy. Therefore, many systems that use thermal energy instead of electrical power for their operation have been developed, such as the steam jet refrigeration system and the vapour absorption system. In 1930, steam jet refrigeration systems were used for conditioning large commercial buildings with a capacity

617

618

N. Al-Khalidy

Nomenclature Ac cp COP h k m p Q T concentrator area (m 2) specic heat (J kg 1 K 1) coefcient of performance of the refrigeration cycle enthalpy (J kg 1 K 1) specic heat ratio mass ow rate (kg s 1) pressure (N m 2) heat rate (W) temperature (K) W h work (J) efciency

Subscripts: b boiler c condenser e evaporator i inlet o outlet ss solar system w water

poor. COP overall is dened as COPoverall Qe Qb Wp (2)

The design and experimental investigation of an ejector improvement cycle refrigeration machine have been presented before8,9. The objectives of the studies were: To design and build a refrigerant ejector refrigeration machine using the heat driven from the concentrator collectors4,8. To study the efciency of the ejector8. To determine the efciency of the concentrator collectors9. To investigate the ideal, theoretical and experimental COP of such a system10.

where Q e, Q b and W p denote the heat absorbed at the evaporator, heat added at the boiler, and pump power, respectively. The heat-driven ejector system is a combination of a heat engine and a reverse cycle (refrigerator) device. Therefore, the Overall Coefcient of Performance COP overall is calculated from the following equation COPoverall h COP (3)

where h is the efciency of the heat engine and COP is the coefcient of performance of the refrigeration cycle. Recent interest in the ejector refrigeration system has been stimulated when the idea of using low boiling temperature refrigerants was suggested (see for example Zhadan and Shchetinina2, Huang et al.3, Al-Khalidy4, Sokolov and Hershgal5, Da Wan6 and Nahdi7). This was because the disadvantages of the steam jet refrigeration were minimized.

In this paper, the inuence of boiler, condenser and evaporator temperatures on the performance in a system using R113 as a working refrigerant are evaluated. Concentrator collectors were used to heat a transformer oil which was used to boil R113 in the boiler. Eight stainless steel concentrators were formed as a parabola with a total area of 15 m 2. Each concentrator can be moved manually to follow the movement of the sun (see also Ref. 9).

Figure 1

The ejector jecteur (1. buse primaire; 2. buse secondaire; 3. e vaporateurraccord de le jecteur) Figure 1 Le

Experimental study of an ejector cycle machine

619

Figure 2

Schematic diagram for the refrigerant ejector refrigeration machine ma de la machine frigorique avec e jection de frigorige ` ne Figure 2 Sche

Table 1 Gas dynamic and thermodynamic properties of selected refrigerants te s dynamiques et thermodynamiques de la phase gazeuse de frigorige ` nes se lectionne s Tableau 1 Proprie Gas dynamic properties Refrigerant Water R717 R11 R12 R113 R114 Critical pressure ratio 0.541 0.548 0.576 0.577 0.581 0.587 Specic heat ratio 1.33 1.29 1.13 1.14 1.12 1.09 Molecular weight 18.02 17.03 137.38 120.92 187.39 170.94 Toxic group 6 2 5 6 45 6 Thermodynamic properties Temp (C) 10 10 10 10 10 10 Pressure (bar) Latent heat (kJ kg 1) 0.0122 6.1500 0.6055 4.2330 0.2367 1.2773 2520.0 1471.0 393.92 355.73 366.52 343.65

Principle of ejector cycle refrigeration machine The refrigerant ejector refrigeration machine can be considered as a combination of a heat engine and a refrigerator. The heat engine converts heat Q b into mechanical work. The refrigerator converts this work and produces the refrigerating effect Q e. The ejector refrigeration cycle is similar to the compression air conditioning cycle except for the method of compressing the working uid. An ejector is used instead of a mechanical compressor to compress the refrigerant vapour from the evaporator to the condenser. The ejector consists of two main parts: a primary convergent divergent nozzle and a secondary nozzle (see Figure 1). A schematic diagram showing the experimental equipment that was used is shown in Figure 2. This shows the four major components: ejector (compressor), condenser, evaporator and expansion valve. A ooded shell and tube evaporator was used and the refrigerant level was controlled by a oat. The system operated as follows. Heat from a solar concentrator collector is transferred to the boiler to evaporate R113 refrigerant and so generate the motive vapour (m b). This is then expanded in the primary nozzle creating low pressure. Refrigerant vapour (m e) from a shell-and-tube, threepass, ooded-type evaporator was entrained. The mixture was then compressed by the secondary nozzle into a shell-and-tube, two-pass water-cooled condenser. Some of the condensate from the condenser is returned to the evaporator (fed by a oat valve) and the rest of the refrigerant is transported back to the boiler by the liquid pump.

Selection of a refrigerant for the system A theoretical study was carried out to select a suitable refrigerant for the system. The following refrigerants were considered: R717, R11, R12, R113 and R114. It is worth mentioning that many refrigerants such as R134a and natural substances could be used in the system (see Ref. 10). Thermodynamic, chemical and gas dynamic properties for the refrigerants should be studied to choose a desirable refrigerant for the system. A desirable refrigerant would have the following qualities: (1) The refrigerant should be chemically stable, available and inammable. (2) Physical and thermal properties of the refrigerant play an important role in determining the system performance through decreasing power consumption for each ton of refrigeration. The following parameters should be considered: Latent heat of evaporation: This should be as large as possible in order to minimize the mass ows. This reduces the input power to the system. Specic heat: This should be as small as possible in the liquid phase to increase the degree of subcooling. Also, the specic heat of the vapour should be high so as to reduce the superheating region. The availability of both properties in one refrigerant leads to an increase in the efciency of the heat exchangers11. Compressibility factor (n pv/mRT): It is preferable to use a refrigerant that has a compressibility factor approaching 1. The use of ideal gas assumptions for designing the ejector becomes more realistic when n

620

N. Al-Khalidy

Figure 4 T S diagram for the ejector refrigeration machine Figure 3 T S diagram for an ideal ejector cycle `e jection ide al Figure 3 Diagramme TS dun cycle a Figure 4 jection e ` Diagramme TS dune machine frigorique a

is approximately near 110. When using actual gas equations for designing the ejector this parameter may be not important. Molecular weight: This should be high. An experimental observation showed that the mixing ratio (m e/ m b) and coefcient of performance increase with increasing molecular weight12.

(3) The compressibility factor n for R113 is greater under the same conditions. It varies between 0.9 and 0.995.

The coefcient of performance The thermodynamic cycle in a T S diagram for the ideal and theoretical cycles are shown in Figures 3 and 4 respectively. The ejector cycle consists of two subcycles: Power subcycle: Boiler-Ejector-Condenser-Liquid pump; Refrigeration subcycle: Evaporator-Ejector-Condenser-Expansion valve.

Thermodynamic and gas dynamic properties for the refrigerants under study are shown in Table 11315. It was concluded from this study that: (1) Refrigerant R717 has the highest specic refrigerant effect among the refrigerants investigated. However, it is toxic. Also, it becomes explosive and ammable under certain conditions. Therefore, it was considered to be unsuitable for this system. (2) Halocarbons are chemically stable, non-inammable, non-toxic and non-explosive. The disadvantage of R12 is that for a selected operation condition the pressure of the boiler is very high in comparison with the other refrigerants (for example the pressure of R12 reached 17 bar in a solar collector of a solardriven ejector system3). Therefore, R12 was not selected to avoid the heavy constriction of the refrigeration machine. Thus R11, R113 and R114 are more suitable for use in ejector cycle refrigeration machines. To select one of them, mixing ratios and COPs were calculated at T e 10C and T c 43C with boiling temperatures from 80C to 150C. The results showed that COP values for systems using R11 and R113 were similar and greater than that for R114. Refrigerant R113 was chosen for the experimental solar refrigerant ejector refrigeration machine for the following reasons10: (1) R113 has a sonic velocity similar to R11 but R113 is a low pressure refrigerant; therefore, we can avoid the heavy construction for a system using R113. (2) The molecular weight of R113 is greater. Then, the mixing ratio can be increased.

The maximum coefcient of performance (COP ideal) for the refrigerator subcycle equals the reverse Carnot cycle. Thus COPideal Te Tc Te (4)

The overall coefcient of performance of the ideal ejector cycle using the denition provided by eqn (3) is given by COPideal overall Te (Tb Tc ) Tb (Tc Te ) (5)

where T b, T c and T e are boiler temperature, condenser temperature and evaporator temperature, respectively. It seems that eqn (5) provides the basis for designing a real system. Theoretical processes in an ejector refrigeration cycle are shown in Figure 4. In process 34, the motive vapour is assumed to be expanded to the evaporator pressure in a primary convergentdivergent nozzle. In process 45, the secondary vapour in the evaporator (state 5) is mixed with the motive vapour (state 4). State 6 indicates the state of the mixture. In process 67, the mixture is compressed in the mixing section and in the diffuser to the condenser pressure. The theoretical coefcient of performance (COP th) can be

Experimental study of an ejector cycle machine

calculated simply from the equation COPth me (h5 h9 ) mb (h3 h8 ) (6)

621

It seems that for a given ejector the COP th changes with changing mixing ratio for constant evaporator, condenser and boiler temperatures. The mixing ratio depends on the physical properties of the compressed vapour. The actual coefcient of performance depends on many parameters including the geometry of the ejector. Before presenting the equations used to calculate actual COP values, remarks concerning the experimental equipment should be mentioned8: All the measurement devices were calibrated, including the CuCn thermocouple, pressure gauges and rotometers. A pressure safety test was carried out for each part of the system. The air was compressed in each part for a pressure higher than the operated pressure to ensure it was leak free. When the whole machine was

assembled, the system was tested again by dividing the system into high pressure and low pressure regions. A solenoid valve was placed between the boiler and ejector. A check valve was placed between the boiler and liquid pump.

Table 2 Inlet and outlet oil temperatures from collectors for oil ow rates of 0.017 kg s 1 ratures dhuile a ` lentre e et a ` la sortie de Tableau 2 Tempe bits de 0,017 kg s 1 collecteurs, pour des de Time Solar ration intensity (kW/m 2) 0.762 0.834 0.874 0.872 0.872 0.836 0.786 Inlet oil temperature (C) 35.5 39.0 47.5 54.2 63.0 71.0 71.0 Outlet oil temperature (C) 47.0 70.5 88.9 118.0 109.5 91.0 89.0

9.0 10.0 11.0 12.0 13.0 14.0 15.0

Figure 5

The relation between T b and mixing ratio at constant T e and T c lange pour des T e et T c constantes Figure 5 Relation entre T b et le ratio de me

Figure 6 The relation between T b and COP at constant T e and T c ` des tempe ratures T e et T c constantes Figure 6 Relation entre T b et le COP a

622

N. Al-Khalidy
energy required to pump the oil around the solar system and that required by the water pump were very low and, therefore, were negligible for the COP calculation presented here. The results of one experiments were as follows: Q b 1.3 kW, Q e 0.21 kW and Q c 1.59 kW. Thus, the required power for all pumps was 0.07 kW. The main objective of this work is to study the inuence of boiler, condenser and evaporator temperatures on the system performance. Three groups of experiments are presented as follows: (1) The boiling temperature was varied between 60C and 100C at a constant condenser pressure of 0.85 bar (T c 42.3C) and constant evaporator pressures of 0.236 bar (T e 10C) and 0.186 bar (T e 5.5C), respectively. (2) The condenser temperature is changed from 40C to 50C at a constant boiler pressure of 3.2 bar (T b 87.1C) and constant evaporator pressures of 0.236 bar (T e 10C) and 0.186 bar (T e 5.5C), respectively. (3) The evaporator temperature is changed from 5C to 18C at a constant boiler pressure of 3.2 bar (T b 87.1C) and constant condenser ressures of 0.85 bar (T c 42.3C) and 1.05 bar (T c 48C), respectively.

The system was evacuated and charged with the refrigerant R113. The system was insulated with berglass. Devices were used to control those parameters that have a major inuence on system performance. For example; a pressure control device to choose the demand boiler pressure is used. It was connected electrically with the solenoid valve. We can also control the following: refrigerant owrate, oil owrate to the solar system, water owrate to the evaporator and condenser and inlet water temperatures. Sight glasses were used to determine the level of refrigerant in both the boiler and evaporator. The actual COP and solar system efciency were determined from eqn (7) and eqn (8) mw cpw (Twi Two ) mb (hb hfc ) Wp (7)

COP

hss

mb (hb hfc ) I Ac

(8)

where m w, T wi, T wo, h fc, W p, h ss, I and A c denote the water ow rate inside the evaporator tubes, inlet water temperature, outlet water temperature, the enthalpy of the condensate vapour, the input power to the liquid pump, the efciency of the solar system, solar radiation intensity and the area of the concentrator collector. The

Figure 7

The inuence of T b on the heat transfer in the system ` me Figure 7 Inuence de T b sur le transfert de chaleur dans le syste

Experimental study of an ejector cycle machine

623

Figure 8

The inuence of T c on the heat transfer in the system ` me Figure 8 Inuence de T c sur le transfert de chaleur dans le syste

Results and discussion The aim of this work was to study experimentally the performance of an ejector cycle machine. To achieve this, an experimental ejector refrigeration machine powered by energy derived from solar concentrator collectors was manufactured. The performance of the collectors was encouraging. Results indicated a maximum operating efciency of 20%9. The economic feasibility of a solar-driven ejector cycle depends on the availability of the free solar heat source3. Table 2 gives an indication of the availability of free solar source throughout the daylight hours in the region where the experiments were done. The average daily solar radiation data for a selected day in May (15 May) were adopted from the Solar Energy Research Center, Baghdad. Table 2 shows also the inlet and outlet oil temperatures from the collectors for selected oil owrates. One sees that the circulating oil tends to increase the inlet oil temperature to the absorber. That is one reason why the efciency of the collectors was low. The energy output and the energy requirement were investigated experimentally at different operating conditions where R113 was used. Figures 5 and 6 show the inuence of T b on mixing ratio and COP at constant evaporator and condenser pressures. It was found that both mixing ratio and COP increase with increasing T b. This is caused by increasing the Mach number at the outlet of the primary convergentdivergent nozzle.

Then, the mixing pressure decreases for a given evaporator pressure. Increasing the difference between the evaporator and mixing pressure leads to an increase in the amount of entrained vapour from the evaporator m e for a given m b. Therefore, the mixing ratio increases with increasing T b. We can see that the ideal mixing ratio curve is similar to the actual curve but there is a large difference in their values. The ideal mixing ratio equation can be derived from the laws of thermodynamics Ideal mixing ratio hb hc hc he (9)

In Figure 6, both COP ideal and actal COP increase with increasing T b. This is because COP is proportional with the mixing ratio. It can be seen that COP ideal is approximately sixfold greater than the actual COP. This is because of the assumptions used in calculating COP ideal, which depends only on T b, T e and T c while the actual COP depends on T b, T e, T c, physical properties of the refrigerant, ejector geometry, etc. Figure 7 shows a set of performance curves of the solar refrigerant ejector refrigeration machine. This gure shows the variation of the refrigerating capacity Q e, heat added to the boiler Q e and heat rejected from the condenser Q e, against varying boiler temperatures and a set of evaporator pressures. The following conclusions can be summarized: The refrigerating capacity increases with increasing

624

N. Al-Khalidy

Figure 9 The inuence of T e on the heat transfer in the system ` me Figure 9 Inuence de T e sur le transfert de chaleur dans le syste

T b at constant P e and P c. This results due to increasing m e, then Q e can be calculated from the following equation Qe me (heo hen ) (10)

where h eo and h ei denote the enthalpy of the refrigerant entering and leaving the evaporator, respectively. When T b increases from 80C to 100C the increasing ratio in Q e was 91% at P e 0.236 bar and P c 0.85 bar while, the increasing ratio was only 79% when P e is reduced to 0.186 bar. The heat rejected from the condenser can be calculated from the following equation Qc Qb Qe Wp (11)

were selected to examine system performance in a very hot climate. Performance of the system was very poor when T c 46C. In fact, operation of the system was unstable when T c 50C. Also, it can be seen that the inuence of T c on system performance becomes greater at low evaporator pressures. Q e increased by 20% when the evaporator pressure was allowed to rise from 0.186 to 0.234 bar. It is clear that this system can be used for heating because of the high heating load in the condenser (eqn (11)). The coefcient of performance of the heating system COP h can be calculated from the following equation COPh COP 1 (12)

where W p is the power of the pumps in the system. Thus, Q e increases with T b. It should be mentioned that an electrical heater was used when the demand P b was higher than 4 bar (T b 95C). The heat added by the electrical heater was lower than 20% of the total heat added to the boiler. Figure 8 shows the performance curves of the proposed system against varying condenser temperatures at a set of evaporator pressures. The capacity of the condenser is greatly affected by cooling water temperature and ow. It can be seen that Q e reduces with increasing T c. This gure was plotted for condenser temperatures ranged from 40C to 50C. Such conditions

Figure 9 shows the performance variation against varying evaporator temperature at a set of condenser pressures and a given boiler pressure P b 3.2 bar. The refrigerating capacity increases with increasing T e. Increasing T e leads to increase in m e on the one hand and to a reduction in the latent heat of evaporation on the other. The increasing ratio in m e is higher and therefore, Q e increases with increasing P e for xed boiler and condenser pressures. The following can be concluded from Figure 9 The reduction ratio in Q c was 25.4% when P c increased from 0.85 bar to 1.05 bar. Q c increases with increasing Q b and Q e (eqn (11)). The maximum Q e was obtained at T e 18C for a

Experimental study of an ejector cycle machine

given P c 0.85 bar. At this point the COP reaches 0.42. This proves that COP is greatly increased with increasing T e. The COP may become similar to the solar absorption system when the system is only used for air conditioning purposes. This makes the solar ejector system more attractive because the system can be cost effective and the absorption system is usually custom made with special components. References

625

Concluding remarks Based on the theoretical analysis and experimental results, the following remarks can be concluded: The use of high molecular weight refrigerants in an ejector cycle refrigeration machine reduces most of the disadvantages of the steam jet refrigeration machine. An economical system can be built when it is operated by waste heat. This leads to a decrease in the operating cost in comparison with the conventual systems. Because of the absence of moving parts (except the pumps), the system maintenance cost is very much less. The system can operate efciently at condenser temperatures lower than 45C and at boiler temperatures higher than 70C.

One of the main advantages of the ejector refrigeration machine is the possibility of using a wide range of refrigerants with the system. Therefore, further studies are important to evaluate the performance of the system using a refrigerant such as R134a or natural substances. It worth mentioning that the efciency of the ejector was good when butane or hexane was used16. Also, we recommend using a multistage ejector system or to change some of the assumptions used for designing the ejector. This may increase the coefcient of performance. Acknowledgements This work was supported by the Mechanical Engineering Department, University of Technology, Baghdad, Iraq.

1. ASHRAE, ASHRAE Handbook Equipment, American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc., Atlanta, GA (1979) pp. 13.1-13.6 2. Zhadan S., Shchetinina N., Selection of cycle design parameters for solar ejector freon refrigeration machine Geliotekhnika (1980) 16(1), 4447 3. Huang B., Jiang C., Hu F., Ejector performance characteristics and design analysis of jet refrigeration system J. Eng. Gas. Turbo. Power, ASME Transection (1985) 107, 792 802 4. Al-Khalidy, N., Experimental investigation of cooling ejector system using concentrators collectors, M.Sc. Thesis, Uni. of Tech., Baghdad, Iraq (1989) 5. Sokolov M., Hershgal D., Optimal cooling and feasibility of a solar-powered year-around ejector air conditioner Solar Energy (1993) 50(6), 507516 6. Da Wan S., Eama I., Performance characteristics of HCPC-123 ejector refrigeration cycle Int. J. of Energy Research (1996) 20, 871885 7. Nahdi, E., Champoussin, J., Hostache, G., Cheron, J., Optimal geometric parameters of a cooling ejector-compressor Int. J. Refrigeration (1993) 67-72 8. Al-Khalidy N., Zayonia A., Design and experimental investigation of an ejector in an air conditioning and refrigeration system ASHRAE Transaction (1995) 101(2), 383 391 9. Al-Khalidy N., Experimental investigation of solar concentrator collectors in a refrigerant ejector refrigeration machine Int. J. of Energy Research (1997) 21, 11231131 10. Al-Khalidy N., Performance of solar refrigerant ejector refrigeration machine ASHRAE Transaction (1997) 103(1), 5565 11. Dossat, R., Principles of Refrigeration, John Wiley, New York (1961) 12. Holton, W., Columbs, Ohio, Effect of molecular weight of entrained uid on the performance of steam jet ejectors ASME Transaction (1951) 905-910 13. Arrora, G., Refrigeration and Air Conditioning, McGrawHill Company, New Delhi (1984) 14. Jain, V., Refrigeration and Air Conditioning, 1st edn., S. Chand and Company Ltd, New Delhi (1986) 15. Software Systems Corporation, Thermodynamics properties, John Wiley and Sons Educational Software, John Wiley and Sons, USA (1988) 16. Khoury F., Heyman M., Resnk W., Performance characteristics of self entrainment ejector I and ES Process Design and Development (1967) 6(3), 331340