This document summarizes research on a jet-pump refrigeration cycle as a low-cost option for utilizing low-grade waste heat to provide refrigeration. Key points: 1) A jet-pump refrigerator replaces the compressor in a vapor compression cycle with a boiler and jet-pump. Experimental results show the COP depends on boiler and evaporator temperatures but not condenser temperature. 2) Below a critical condenser pressure/temperature, the COP increases with higher boiler temperatures and lower evaporator temperatures. Above the critical point, the COP drops sharply to zero. 3) Previous research explored jet-pump cycles using water or halocarbon refrigerants powered by waste heat sources like solar collectors. Most focused on

1. Printed in n
THE JET-PUMP CYCLE-A LOW COST REFRIGERATOR
OPTION POWERED BY WASTE HEAT
I. W. EAMES, S. APHORNRATANA and DA-WEN SUN
Department of Mechanical and Process Engineering, University of Sheffield, Mappin Street, Sheffield,
S14DU, U.K.
(Received 4 December 1994)
Abstract-A perennial problem with waste heat is the capital cost of plant required to make its utilisation
justifiable. A good example of this is the use of waste heat to power absorption refrigerators. The capital
cost of absorption refrigerators rises sharply as the temperature of the heat source falls, making waste
heat recovery and use uneconomic. This paper describes and evaluates the potential of the jet-pump cycle
as a low capital cost option for providing refrigeration utilising low grade waste heat. A brief literature
review is provided. An experimental jet-pump refrigerator is described, experimental results are presented
and evaluated and the cost benefits of jet-pump refrigerators compared with vapour compression systems
are discussed.
INTRODUCTION
A perennial problem with waste heat is the capital cost of plant required to make its utilisation
justifiable. A good example of this is the use of waste heat to power absorption refrigerators. The
capital cost of absorption refrigerators rises sharply as the temperature of the heat source falls,
making waste heat recovery and use uneconomic. An example of this is the problem of utilising
the vast quantities of low temperature waste heat generated by municipal incinerators during
summer months. During the winter this heat can be sold to warm offices, homes, public buildings
and so on, through a district heating company. However, during the summer vast quantities of heat
generated by incineration are wasted to the environment because of the prohibitively high cost of
absorption refrigerators, which can be twice that of conventional electrically-powered vapour
compression systems, that could convert some of the waste heat into useful refrigeration for
building cooling. This paper addresses how this wasted heat could be directed in an energy efficient
and environmentally-beneficial way to provide refrigeration. The paper describes and evaluates the
results of an experimental study undertaken to assess the potential application of low capital cost
refrigerators designed on the jet-pump principle powered by waste heat. Water was selected as the
refrigerant for the experimental refrigerator. There has been little published research on steam
jet-pump refrigeration in recent years. Most published work has concentrated on the use of
halocarbon refrigerants. Because of the harmful environmental effects of these compounds and the
growing interest in utilizing waste heat, it was felt to be an opportune time to investigate the
potential of low-temperature, steam-powered, jet-pump refrigeration systems. It is hoped that this
contribution will encourage debate in this important area of research.
THE JET-PUMP REFRIGERATOR CYCLE
Figure 1 shows a schematic view of a jet-pump refrigerator. In principle a jet-pump refrigerator
operates in much the same way as a conventional vapour compression unit, except that the
compressor is replaced by a boiler and a jet-pump. The operation of jet-pump refrigerators is
described in most standard thermodynamics text books. However, in operation the jet-pump
refrigerator cycle is similar to that of the conventional vapour compression cycle, in which the
mechanical compressor in the latter is replaced by a boiler and a jet-pump. The schematic diagram
in Fig. 1 shows the main components of the jet-pump cycle. High pressure and temperature
711

2. JET
PUMP
CONDENSER
I THROTTLE
VALVE
II
I
PUMP
Fig. 1. Schematic representation of a jet-pump cycle-a low cost refrigerator.
refrigerant vapour is evolved in the boiler to produce the primary (motive) fluid. This enters the
jet-pump nozzle, where it expands to produce a supersonic flow that creates a low pressure region
within the mixing or entrainment chamber. This region of low pressure is fed by a secondary vapour
flow from the evaporator to produce the required refrigeration effect. Down-stream of the mixing
region the combined primary and secondary flows enter a diffuser, where the static pressure rises
to equal that in the condenser. The refrigerant is liquefied in the condenser, from where some is
pumped at pressure to the boiler whilst the remainder is expanded across a throttle valve before
returning to the evaporator.
The Coefficient of Performance (COP) of a jet-pump refrigerator is defined as the ratio between
the cooling capacity (evaporator heat input) and the energy input to the boiler. The power input
to the mechanical pump is normally negligible compared with the heat input to the boiler and,
therefore, is not usually included in thermodynamic performance calculations.
A LITERATURE SURVEY
The first steam jet-pump refrigeration system dates back to 1901 and was designed by LeBlanc
of France and Parsons of England [l]. During the 193Os, steam jet refrigeration units experienced
their first wave of popularity for air-conditioning large buildings. However, these units were later
supplanted by mechanical vapour-compression systems. This was encouraged by compressor
developments, particularly centrifugal machines. Since the 1930s there has been little published
research on steam-driven jet-pump refrigeration.
The most recent paper concerned with steam systems is by Decker [2] who considered the
application of steam jet-pump vacuum cooling systems in paper bleaching operations and in the
pharmaceutical and chemical industries for various manufacturing processes. In addition to
providing chilled water, steam jet-pump systems may also be used for quick chilling of process
fluids. Direct vacuum or ‘ flash’ cooling of food is a further application. In this process water is
evaporated quickly from produce, such as leafy vegetables like lettuce and cabbage, to produce the
cooling effect. Flash cooling may not be practical with mechanical compression systems due to the
large volumes of low pressure water vapour needed to be pumped, since this would require
compressors with high compression ratios and large displacements. Therefore, steam jet-pump refri-
geration systems compete successfully with mechanical compressor systems in these applications.
Bowrey ef al. [3] produced an analysis designed to minimise the energy consumption of a
three-stage steam ejector deodoriser-cooling system for use in the food industry. The analysis,
though simple, demonstrated the need to set operating conditions correctly to reduce energy costs.
Other papers reporting research on steam jet-pumps and worthy of mention are those by Addy
et al. [4], Munday and Bagster [5-l and Fabri and Siestrunck [8].

3. The jet-pump cycle-low cost refrigeration 713
A limitation of the steam jet-pump cycle is that cooling temperatures can only be above the
freezing point of water. This limitation has previously encouraged researchers to investigate the
use of halocarbon fluids as alternative refrigerants. As with water, halocarbon compounds such
as CFCs, HCFCs and HFCs can be used to utilise low grade heat energy to power the jet-pump
cycle, at temperatures ranging upwards from 60°C. This energy is available from a plate solar
collector, waste steam, exhaust from automobiles and flue gases. In some cases the cost of the heat
supply is negligible and, therefore, the operating costs can be significantly lower than for
conventional vapour compression systems.
The earliest reported research on jet-pump refrigeration using a refrigerant other than water was
that carried out by Mizrahi et al. [9], who undertook a theoretical study to determine the
performance of a jet-pump cycle operating on a number of different refrigerants. They assumed
a boiler temperature of 60°C supplied by a solar collector. Of the conventional refrigerants tested,
R22 and R12 proved to provide the best all-round performance.
Misrahi et al.‘ work was extended by Heymann and Resnick [lo] who used Keenan et al.3 [l l]
s
method for analysing jet-pumps. They argued that a design point boiler temperatures of 90°C
would be more appropriate for solar collector heat sources. However, this temperature is still well
within the sphere of ‘ waste’ heat supply.
An interesting application was devised by Chen [12], who optimised the design of an automobile
air-conditioning cooling system, powered by a jet-pump refrigerator, using Elrod’ jet-pump theory
s
[ 131.This refrigerator was powered using waste heat from an engine. He used R113 as the working
fluid. The optimum performance of the cooling cycle at the design point was determined, however,
off-design performance data were not provided.
Hamner [14, 151 investigated the use of Rl 1 for his jet-pump compression heat pump. His
experimental and theoretical investigation concentrated on the overall performance of the
refrigeration cycle and did not consider the performance of the jet-pump within this system. Like
Chen [ 121, Hamner also suggested the use of jet-pump refrigerators powered by waste engine heat
to air-condition vehicles.
Other research papers worthy of mention here include: Faithful1 [16], who constructed a Rll
‘ combined Rankine and vapour compression cycle heat pump’ for teaching purposes; Nahdi et al.
[I 71. who undertook an experimental study of an (Rl 1) jet-pump refrigerator and concluded that
its thermodynamic performance was very sensitive to the design of the jet-pump and the thermal
conditions at the boiler, condenser and evaporator; Tyagi and Murty [18], who investigated the
use of R 11 in jet-pump refrigerators and conducted similar tests using R113.
Huang et al. [ 191also used R113 in their study. They redefined the jet-pump choking theory of
Munday and Bagster [6,7] and used it to analyse the performance of their jet-pump. Huang et al.
[19] found that the COP of the jet-pump cycle can be increased if the liquid refrigerant returning
to the generator is preheated by the ‘ hot’ refrigerant vapour coming from the ejector exhaust, hence
reducing the heat input to the generator. Similar improvements in COP were reported when liquid
refrigerant was pre-cooled on leaving the condenser, before it enters the expansion valve, by using
the cold refrigerant vapour leaving the evaporator.
Chen and Hsu [20] performed a simulation study on an ejector refrigerator using RI 1. In this
case Elrod’ method [ 131 was used to calculate jet-pump performance. Their results showed that
s
halocarbon refrigerants, particularly Rl 1, R113 and R114, are most suitable for jet-pump
applications and with the addition of a regenerator and precooler into the cycle, as suggested by
Huang et al. [ 191and described previously; the COP of an R 11 system could be increased by 17%
when operating at boiling, condensing and evaporating temperatures of 93.3, 43.3 and 10°C
respectively.
Sokolov and Hershgal [21-231 carried out a detailed theoretical and experimental study on
possible improvements to the jet-pump refrigeration cycle powered by low grade heat. For the
jet-pump analysis, they modified Keenan et al’ [l l] method by using real gas (refrigerant) data
s
instead of ideal gas data. A comparison of the refrigerants they tested indicated that R114 was the
most suitable from a thermodynamic standpoint. They compared the performance of a booster-
assisted jet-pump cycle, a hybrid compressor and jet-pump cycle, and a combined
booster-compressor-jet-pump cycle. They also reported experiments on a double jet-pump with
compression-enhanced cycle. Double jet-pumps were used to improve the part-load performance.

4. 714 I. W. EAMES al.
et
One was designed to provide high entrainment ratios when condenser pressures were low, whilst
the other was designed to provide low entrainment ratios when condenser pressures were high. This
need for two or more jet-pumps indicates the sensitivity of the jet-pump cycle to changes in ambient
temperature, particularly at the condenser. An alternative approach is to use a variable geometry
jet-pump.
It should be pointed out that although jet-pump refrigeration systems using halocarbon
compounds as refrigerants have some advantages over steam systems, most halocarbon refrigerants
damage the ozone layer and are ‘ greenhouse’ gases. Also, the production and import of CFCs is
to cease shortly and HCFCs are expected to be subject to end-use controls from 1995, with total
phase-out by no later than 2015. Clearly, it is becoming important that research into the use of
environment-friendly alternative refrigerants for jet-pump refrigeration systems be undertaken,
particularly with regard to their application in the utilisation of low-grade waste heat for
refrigeration.
JET-PUMP REFRIGERATOR EXPERIMENTS
The aim of the experiments was to determine the performance of a laboratory-scale jet-pump
refrigerator operating with water as the refrigerant and using a low-temperature heat source of a
super heater I w
--
I7
,3 to vacuum pump
cooling water
Fig. 2. Schematic view of experimental jet-pump refrigerator.

5. The jet-pump cycle--low cost refrigeration 715
Fig. 3. Experimental jet-pump refrigerator.
degree commonly available as waste heat from an industrial process. The experimental refrigerator
is shown schematically in Fig. 2 and a photograph of the unit is included in Fig. 3.
The boiler design was based on the thermosiphon principle with baffle plates located at its upper
end to prevent liquid droplets being carried over with the saturated vapour. The maximum heating
capacity of the boiler was 7 kW, provided by two 3.5 kW electric heaters. A 500W electrically-
powered superheater was positioned in the steam line between the boiler and the jet-pump, to dry
the vapour prior to entering the primary nozzle. In practice the addition of this super-heater was
found to be unnecessary as its effect on the performance of the refrigerant was insignificant. The
evaporator design was based on the flash-evaporation principle. A single 3.5 kW electric heater
located within the evaporator vessel was used to provide a cooling load. The output of all electric
heaters was controlled using variable transformers. The condenser was a shell and coil type cooled
by water taken from the laboratory’ cooling tower.
s
The test jet-pump was designed using a one-dimensional compressible flow method reported by
Keenan and Neumann [l I]. A sectional drawing showing the dimensions of the test’ jet-pump is
s
shown in Fig. 4. The position of the nozzle was fixed at 26 mm in from the bell mouth entry to
the mixing chamber.

6. 716 I. W. EAMFSet al.
Primary nozzle throat diameter 2mm
Primary nozzle exit diameter gmm
4Omm 1OOmm 40mm 2lOmm
I
nozzle exit positive (NXP)
Fig. 4.
REFRIGERATOR PERFORMANCE CHARACTERISTICS
Experiments on the jet-pump refrigerator were carried out over a range of boiler, evaporator and
condenser temperatures. The electric power input to the boiler and the evaporator were measured
and the coefficient of performance (COP) of the cycle was calculated using the following equation:
Cop = Electrical power consumption at evaporator
Electrical power consumption at boiler ’
The results of these experiments are shown in Figs 5 and 6 for various operating temperatures.
Referring to these results it can be seen that the COP is dependent on boiler and evaporator
temperatures only and independent of condenser temperature. However, at a certain value of
condenser temperature the COP was found to fall sharply to zero. If the condenser temperature
was further increased above this critical value, the COP remained at zero. The condenser pressure
at which the COP just began to fall was named by Huang et al. [19] as “the critical condenser
pressure”. The performance of the experimental refrigerator at the critical condenser operating
temperature is shown in Figs 7 and 8.
T cond (“‘I
25 28 31 34 37
I I I I I
nozzle exit Position 26.15mm
evaporator temperature 10.0 ‘C
A Tboiler = 120°C. Pbbtler = 1.98 bar
o Tbiler = 12S°C, Pboilar = 2.32 bar
130°c. Pboiler = 2.70 bar
3S”c. Pboiler = 3.13 bar
??Tboilcr = 140°C. Pboiler = 3.61 bar
30 40 50 60
P cond (mbar)
Fig. 5. Experimental results showing the variation in COP with condenser pressure over a range of boiler
pressures and with an evaporator temperature of 10°C.

7. The jet-pump cycle-low cost refrigeration 717
T cond (“‘
)
25 28 31 34 37
0.5
nozzleexit posttion 26.15mm
evaporator temperature 5.0 OC
A *boiler = 120°C. Pboiler= 1.98 bar
= 125°C. Pboiler= 2.32 bar
? ?*boiler
s 130°C. Pboilcr = 2.70 bar
O %oiler
’ *boiler = 135°C. Pboilcr= 3.13 bar
??*boiler
= 14OT. Pboilcr= 3.61 bar
0.0
30 40 50 60
P mod cmbar)
Fig. 6. Experimental results showing the variation in COP with condenser pressure over a range of boiler
temperatures and with an vaporator temperature of 5°C.
The reason for the sudden cut-off in COP, shown in Figs 5 and 6, is not yet fully understood.
However, a possible explanation comes from the behaviour of supersonic flows through conver-
gent-divergent nozzles. The pressure ratio across a nozzle (Pboiler/Pevaporator)
above which supersonic
flow at its outlet can be expected to occur for steam is approximately 1.86. Throughout these
experiments this pressure ratio was always between 160 and 300, producing an estimated nozzle
outlet Mach number of between 3.5 and 4.5. It is certain, therefore, that during all experiments
the primary nozzle always operated in a choked condition. With a fixed nozzle throat area the
primary flow was, therefore, only a function of boiler temperature (assuming the steam to be
saturated at entry to the nozzle). In other words, for a fixed boiler temperature the primary flow
was constant and independent of both evaporator and condenser temperatures. For the same
reasons the diffuser was always choked and, therefore, the structure of the flow up-stream of the
diffuser throat was independent of the condenser temperature up to its critical value, as indicated
by the step characteristic of the results shown in Figs 5 and 6. This meant that, as long as
evaporator and boiler temperatures remained constant, the primary flow from the nozzle would
entrain the same quantity of secondary flow refrigerant from the evaporator, regardless of the
1000
900
1ooc
800 7.YC
3
PC
g 700
Q
f 600
1
& 500
nozzle(l): 2mm throat dia.
400
i 8mm dia. outlet
300 1 I I 1 I I I
2s 27 29 31 (OF) 35 37
T(,,,)
Fig. 7. Experimental results showing the variation in cooling capacity with condenser temperature over
a range of boiler and evaporator temperatures.

8. 718 I. W. EAMESef al.
0.40
0.35
= 10°C
0.30
= 7.5v
0.25 = 3°C
%
” 0.20 I
0.15
0.10
0.05
I
0.00 1
25
I
27
I
29
I
31 c:
I I
35
I
31
T(co.d)
Fig. 8. Experimental results showing the variation in COP with condenser temperature over a range of
boiler and evaporator temperatures.
conditions in the condenser, so long as its temperature did not exceed the critical value. Therefore,
at condenser temperatures less than critical the refrigeration capacity and the COP of the cycle were
dependent only on boiler and evaporator temperatures.
When operating with condenser temperatures (pressures) less than critical, it was thought that
a (normal) shock wave was produced in the divergent section of the diffuser, thus allowing the static
pressure of the flow to rise at the diffuser outlet to equal the conditions in the condenser. As the
condenser pressure increased the shock wave would have moved up-stream towards the throat and
at the critical pressure it would enter the throat. It was noticeable during experimentation that there
was a sharp rise in steam temperature at a section of the diffuser throat when operating the
refrigerator at a critical condenser condition. This is thought to have resulted from the sudden
compression effect of the shock wave. Any further increase in condenser pressure, above its critical
value, caused a sudden rise in evaporator temperature. This could only be explained by ‘ hot’ steam
from the primary nozzle flowing directly to the evaporator from the mixing chamber. At this point
the refrigeration capacity and COP fell sharply to zero. Refrigeration could then only be
re-established by increasing the boiler temperature, however, as shown in Figs 5 and 6, the COP
was reduced.
This behaviour of the refrigerator was interesting because, as shown in Figs 7 and 8, the greater
the boiler temperature the lower the COP and refrigeration capacity. This was clearly contrary to
what would normally be expected. A reason for this result might be that the flow from the primary
(fixed geometry) nozzle was progressively more under-expanded as boiler temperature and pressure
were increased. This was confirmed by calculation. Therefore, as the flow became more under-
expanded with increasing boiler temperature, a network of oblique expansion waves would have
projected progressively further into the mixing chamber from the nozzle outlet. It is thought that
this expansion region tended to resist mixing and entrainment between the primary and secondary
flow streams because the pressure within the (still expanding) primary stream would, by definition,
be greater than that within the secondary flow stream. In effect the secondary flow would have been
pushed aside by the primary stream. If this hypothesis is correct then mixing and entrainment
would be delayed until the pressure in the primary flow stream had fallen to that within the
secondary and consequently there would be less area and time for refrigerant to be entrained into
the primary stream before it entered the diffuser throat, thus resulting in the characteristic curves
shown in Figs 7 and 8.
It was observed that the critical condenser temperature increased with boiler temperature. There-
fore, in order to operate the refrigerator at high condenser (ambient) temperatures it would seem
necessary to have a high temperature boiler heat source. At first this appeared to limit the usefulness
of low temperature waste heat for refrigeration purposes to conditions when the condenser ambient
temperature is low, for example less than 25°C. However, it is believed, and experiments have
shown, that if the geometry of the nozzle and diffuser were to be made variable then the critical
condenser temperature can be increased whilst maintaining a constant boiler temperature.

9. The jet-pump cycle-low cost refrigeration 719
Table I. Estimated variation in COP values based on
mass evaporated and electrical energy input
T,bOLlW, T,CO”d)
(“(3 (“(3 cop,c,c4 cop,rnaw
120 28.3 0.37 0.54
125 29.8 0.33 0.49
130 31.9 0.29 0.44
135 34.0 0.24 0.36
140 26.3 0.19 0.29
EXPERIMENTAL ERRORS
The COP characteristics of the experimental refrigerator shown in Figs 5-8 were based on electric
power input to the generator and evaporator. These values, therefore, describe the net performance
of the refrigerator because they include unwanted heat losses and gains. Estimates of COP based
on mass flows have shown that the unwanted heat losses at the boiler were in the order of 25%,
whilst the heat gains at the evaporator were 20%. On average the combination of these two effects
produced a 30% underestimate of COP values shown in Figs 5-8. Taking this underestimate into
account, the best COP value recorded was a respectable 0.544 with a boiler temperature of 120°C
evaporator temperature of 10°C and condenser temperature of 28°C. A summary of typical
comparisons is shown in Table 1.
GENERAL DISCUSSION
The experimental results described in the previous section have shown that it is technically
feasible to design a refrigerator based on the jet-pump principle and that this will provide a
reasonably efficient performance when powered by low grade waste heat. It is recognised that the
COP values given for the jet-pump refrigerator are not as high as those commonly reported for
absorption systems (typically from 0.6 to 0.9). However, the capital cost of a refrigerator designed
to the jet-pump cycle principle will be significantly less than that of a similar capacity absorption
refrigerator, making the jet-pump cycle particulary attractive when the cost of heat energy is low.
In order that refrigerators based on the jet-pump principle become economically feasible, it is
necessary that their life-cycle costs at least equal those of conventional vapour compression units.
Of importance here is the comparison between the two systems in terms of capital, operating and
maintenance costs throughout the life cycle of the plant.
Capital costs
In order to compare the capital costs of the two systems it is only necessary to consider the cost
of a compressor with that of the boiler and ejector. For example, a 200 kW jet-pump refrigerator
would require a 600 kW boiler. If a plate heat exchanger arrangement were used the cost of this
will be approximately El800 and the cost of a suitable ejector will be f 1500 making a total cost
of f3300. A recent quotation from a leading refrigerator manufacturer gave the purchase price of
f5000 for an electrically-powered compressor that would supply 200 kW of cooling. Assuming the
cost of the remaining system components in each case to be similar, then it could be expected that
the cost of a chiller designed to the jet-pump principle will be slightly less than that of the
conventional vapour compression system. Also, because water can be used both as the refrigerant
and the primary driving fluid, low cost materials can be used to construct the jet-pump refrigerator
because of the low pressures involved.
Operating costs
It is clear from the experiments described in this paper that the COP values of the jet-pump
refrigerator fall short of those of electrically-powered refrigerators. However, electricity to drive
a compressor is normally more expensive to the user in terms of kWh than waste heat. In the UK
the current general commercial tariff for electricity is approximately f72.2 per MWh. In order to
make economic use of waste heat for refrigeration purposes, it is necessary for the cost of the heat
to be less than f9.6 per MWh. This figure assumes that a conventional vapour compression
refrigerator has a COP of 3.0 and the jet-pump system has a COP of 0.4

10. 720 I. W. EAMES al.
et
Maintenance costs and reliability
Reliability and low maintenance costs incurred are important factors in the selection of
refrigeration units. Conventional absorption systems are more reliable and have lower maintenance
costs than vapour compression systems because they have fewer highly-stressed moving parts. The
jet-pump refrigerator too has few moving parts and, therefore, such machines should present no
increase in maintenance cost or reductions in reliability. In all probability maintenance costs will
be much less for the jet-pump refrigerator.
Energy savings
The use of heat-powered refrigerators offers the potential for savings in electrical energy
associated with refrigeration by utilising waste heat energy normally rejected to atmosphere. This,
in turn, will save burning fossil fuels and reduce COz emissions. There are also additional
environmental advantages if water is used as the refrigerant for higher temperature applications.
CONCLUSIONS
The potential of the jet-pump refrigeration cycle to utilise low-grade waste heat as its power
source is now recognised. A theoretical study undertaken by Bevilacqua provided further
confirmation that the jet-pump refrigeration cycle is suitable for the utilisation of waste heat at
temperatures above 60°C. The use of water as a refrigerant has a number of obvious environmental
advantages over halocarbon compounds. Also, the relatively low system pressures and low
corrosion potential of water offer the possibility of using low-cost materials in the construction of
a refrigerator.
An experimental refrigerator designed on the jet-pump cycle was described. Tests were carried
out with boiler temperatures between 120 and 140°C and evaporator temperatures between 5 and
10°C. Results indicated that COP values in excess of 0.5 are possible from this type of machine.
The results have also shown that refrigerators based on the same principle are economically
feasible, particularly if powered by waste heat. It is also concluded that the capital cost of a
jet-pump refrigerator could be less than that of a conventional vapour-compression unit for the
same cooling capacity. Also, because the jet-pump refrigerator has no moving parts, it is potentially
more reliable than conventional systems.
REFERENCES
ASHRAE, Steam-jet refrigeration equipment. 1979 Equipment Handbook, Chap. 13, pp. 13.1-13.6. ASHRAE, Atlanta,
GA, U.S.A. (1979).
L. 0. Decker, Consider the cold facts about steam-jet vacuum cooling. Chem. Engng. Prog. 89 (l), 7477 (1993).
R. G. Bowrey, V. B. Dang and G. D. Sergeant, An energy model to minimise energy consumption in a low-temperature
operation, steam ejector-cooling system. J. Inst. Energy 45, 45-48 (1986).
A. L. Addy, J. C. Dutton and C. D. Mikkelsen, Supersonic ejector-diffuser theory and experiments. Report No.
UILU-ENG-82-4001, Department of Mechanical and Industrial Engineering, University of Illonois at Urbana-
Champaign, Urbana, Illonois, U.S.A. (1981).
5. J. T. Munday and D. F. Bagster, Design and performance of a steam-jet refrigeration system. ThermofIuids Conf of
National Committee on Thermodynamics and Fluid Mechanics of Institution of Engineers of Australia, Melbourne,
December 1974, pp. 57-61. National Conference Publication, Australia (1974).
6. J. T. Munday and D. F. Bagster, The choking phenomena in ejector with particular reference to steam-jet refrigeration,
ThermoJluirls Conf of National Committee on Thermodynamics and Fluid Mechanics of Institution of Engineers of
Australia, Hobart, December 1976, pp. 8488, National Conference Publication, Australia (1976).
*-__ . .---
7. J. I. Munday and D. r‘ Bagster, A new ejector theory applied to steam-jet refrigeration. Ind. Engng. Chem. Proc. Res.
.
Deu. 16, 442&!49 (1977).
8. J. Fabri and R. Siestrunck, Supersonic air ejectors. In Aduances in Applied Mechanics, Vol. V., pp. l-34, edited by von
Mises and T. von Karman. Academic Press, New York, U.S.A. (1958).
9. J. Mizrahi, M. Solomiansky, T. Zisner and W. Resnick, Ejector refrigeration from low temperature energy sources.
BUN Res. Council of Israel, 6C, l-8 (1957).
10. M. Heymann and W. Resnick, Optimum ejector design for ejector operated refrigeration cycles. Israel J. Technol. 2,
242-247 c 1964).
11. J. H. Keenan, E. P. Neumann and F. Lustwerk, An investigation of ejector design by analysis and experiment. ASME
J. Appt. Me& September, 299-309 (1950).
12. L.-T. Chen, A heat-driven mobile refrigeration cycle analysis. Energy Comers. 18 (1) 25-29.
13. H. G. Elrod Jr., The theory of ejectors. J. Appl. Mech. Trans. ASME 67, AllO-A114.

11. The jet-pump cycle-low cost refrigeration 721
14. R. H. Hamner, An investigation of an ejector-compression refrigeration cycle and its applications to heating, cooling
and energy conservation, Ph.D. Thesis, University of Alabama, Birmingham.
15. R. H. Hamner, An alternate source of cooling: the ejector-compression heat pump. ASHRAE J. 22, 62-66.
16. D. C. Faithfull, A combined Rankine and vapour compression cycle heat pump for teaching purposes. In Directly Fired
Heat Pump-For use in Domestic and Commercial Premises, Proc. of hr. Conf., 19-21 Sept 1984, University of Bristol,
U.K., Paper No. 3.1, pp. (3.1) 1-7, edited by P. W. Fitt and R. T. Moses (1984).
17. E. Nahdi, J. C. Champoussin, G. Hostache and J. Cheron, Optimal geometric parameters of a cooling ejector-
compressor. hr. J. R&g. 16, (1) 67-72 (1993).
18. K. P. Tyagi and K. N. Murty, Ejector-compression systems for cooling: utihsing low grade waste heat. Heal Recovery
Systems & CHP 5 (6), 545-550 (1985).
19. B. J. Huang, C. B. Juang and F. L. Hu, Ejector performance characteristics and design analysis of jet refrigeration
system. J. Engng. Gas Turbines and Power, Trans. ASME 187, 792-802 (1985).
20. F. C. Chen and C.-T. Hsu, Performance of ejector heat pumps. Energy Res. 11, 289-300.
21. M. Sokolov and D. Hershgal, Enhanced ejector refrigeration cycles powered by low grade heat. Part 1. Systems
characterisation. Inr. J. Refrig. 13, 351-356 (1990).
22. M. Sokolov and D. Hershgal, Enhanced ejector refrigeration cycles powered by low grade heat. Part 2. Design
procedures. Znr. J. Refrig. 13, 357-363 (1990).
23. M. Sokolov and D. Hershgal, Enhanced ejector refrigeration cycles powered by low grade heat. Part 3. Experimental
results. ZnZ.J. Refrig. 14, 24-31 (1991).