Experimental study on R-134a refrigeration system using

a two-phase ejector as an expansion device

Praitoon Chaiwongsa, Somchai Wongwises *

Fluid Mechanics, Thermal Engineering and Multiphase Flow Research Laboratory (FUTURE), Department of Mechanical Engineering,
King Mongkut’s University of Technology Thonburi, Bangmod, Bangkok 10140, Thailand

Received 8 May 2006; accepted 14 May 2007

Available online 2 June 2007

Abstract

This paper is a continuation of the authors previous work. In the present paper, the performance of the refrigeration cycle using a
two-phase ejector as an expansion device is experimentally investigated. Refrigerant R-134a is used as working fluid. Motive nozzles hav-
ing three different outlet diameters are tested. New experimental data that have never been seen before are presented on the effects of the
external parameters i.e. heat sink and heat source temperatures on the coefficient of performance and various relevant parameters i.e.
primary mass flow rate of the refrigerant, secondary mass flow rate of the refrigerant, recirculation ratio, average evaporator pressure,
compressor ratio, discharge temperature and cooling capacity. The effects of size of the motive nozzle outlet on the system performance
are also discussed.
Ó 2007 Published by Elsevier Ltd.

Keywords: Two-phase ejector; Refrigeration; Liquid-recirculation; Coefficient of Performance; Motive nozzle; Expansion device

1. Introduction To the best of the authors’ knowledge, information on

the application of an ejector as an expansion device in a
In vapor refrigeration cycle, in order to accommodate refrigeration cycle is still limited.
the low temperature evaporation process, a saturated or Kornhauser [1] investigated the thermodynamic perfor-
sub-cooled liquid produced from condensation at high mance of the ejector expansion refrigeration cycle by using
temperature must be expanded through an expansion R-12 as a refrigerant under standard conditions, 15 °C
device to low-pressure. Throttling process or isenthalpic and 30 °C for evaporator and condensor temperatures,
process is the process occurring through an expansion respectively. A theoretical COP improvement of up to
device. This process produces the thermodynamic losses 21% over the standard cycle was found. This result was
and causes a larger amount of the refrigerant to flash into based on ideal cycle and constant mixing pressure in the
a vapor than in the isentropic process. ejector. Harrell et al. [2] used a R-134a two-phase ejector
In order to reduce the thermodynamic loss in the expan- and the test rig to estimate the COP of the refrigeration
sion process, various researchers have attempted to look cycle. It was found that the COP improvement ranged from
for other appropriate devices. Due to the low cost, no mov- 3.9% to 7.6%. Menegay et al. [3] developed a bubbly flow
ing parts and ability to handle two-phase flow without tube to reduce the thermodynamic non-equilibrium in the
damage, an ejector is considered as an attractive expansion motive nozzle with R-12 as the refrigerant. This device
device in the refrigeration system. was installed upstream of the motive nozzle. The COP of
the system using the bubbly flow tube can be improved
up to 3.8% over the conventional cycle under standard con-
*
Corresponding author. Tel.: +662 470 9115; fax: +662 470 9111. ditions. However, they reported that the result was not as
E-mail address: [email protected] (S. Wongwises). good as was expected and study of the ejector expansion

1359-4311/$ - see front matter Ó 2007 Published by Elsevier Ltd.

doi:10.1016/j.applthermaleng.2007.05.005
 Nomenclature

COP coefficient of performance me mixing chamber exit

cp,hw specific heat at constant pressure of the hot ne motive nozzle exit
water (kJ/kg k) Qevap heat transfer rate at evaporator (kW)
Dni inlet diameter of motive nozzle (mm) se suction nozzle exit
Dnt throat diameter of motive nozzle (mm) Tsink heat sink temperature (°C)
Dne outlet diameter of motive nozzle (mm) Tsource heat source temperature (°C)
Lni inlet length of motive nozzle (mm) Thw,in hot water temperature at the inlet of the evapo-
Lnc convergent length of motive nozzle (mm) rator (°C)
Lnd divergent length of motive nozzle (mm) Thw,out hot water temperature at the outlet of the evap-
m_ p primary mass flow rate of refrigerant (kg/s) orator (°C)
m_ s secondary mass flow rate of refrigerant (kg/s) Wcomp electrical power supplied to the compressor
m_ e flow rate of the vaporized refrigerant (kg/s) (kW)
m_ hw mass flow rate of hot water (kg/s)

refrigeration cycle should be extended. Domanski [4] con- 2. Experimental apparatus and procedure
cluded that the ejector efficiency was very sensitive to the
theoretical COP of the ejector expansion refrigeration Fig. 1 shows the schematic diagram of the experimental
cycle. Nakagawa et al. [5] showed that the longer divergent apparatus. A commercial R-134a is used as the working
part provided a longer period of time for the two-phase fluid. The refrigerant loop consists of the vapor compres-
flow to achieve equilibrium. He concluded that the longer sion cycle components: compressor, condenser, expansion
the length of the divergent part of the motive nozzle, the valve and evaporator, and other accessory parts – the oil
higher the motive nozzle efficiency could be obtained. separator, liquid receiver, filter/drier, sight glass and the
The papers mentioned above focused on operating with accumulator. The operating conditions of the apparatus
a dry-expansion evaporator by installing an expansion are similar to those of a typical air-conditioning applica-
valve downstream of the liquid-vapor separator. However, tion. The principal modifications from the standard refrig-
the purpose of the ejector is to replace the throttling valve. eration system are the addition of a two-phase ejector and
With this consequence, any throttling device in the system a liquid–vapor separator.
should be avoided. Refrigerant is discharged by a two-cylinder single stage
Up to now, there have been only two papers, published reciprocating unit, driven by an electric motor. The speed
by Disawas and Wongwises [6] and Wongwises and of the motor is varied by an inverter to regulate the refrig-
Disawas [7], dealing with this issue. In their experimental erant flowing through the motive nozzle. Compact plate
apparatus, the evaporator was flooded with refrigerant heat exchangers are used for the condenser and evapora-
and became a liquid-recirculation system, in which, in addi- tors. The evaporator referred to in this paper is the main
tion to serving as an expansion device, the ejector also evaporator as shown in Fig. 1. A filter/drier, placed down-
acted as a refrigerant pump for the low-pressure side of stream of the receiver, is provided to keep the circulating
the system. refrigerant free from harmful substances: moisture and for-
Although, some information is currently available on eign particles that might remain in the system. An oil sep-
the refrigeration cycle using a two-phase ejector as an arator is used to keep the oil content in the refrigerant to a
expansion device, the detailed investigation is still lacking. minimum.
Especially, it can be noted that no attention has been paid The motive and the suction mass flow rates are mea-
to the effects of geometric parameters on the performance sured by volumetric flow meters located downstream of
of the refrigeration system. This paper is the third in a ser- the sight glass and of the liquid-vapor separator, respec-
ies and is a continuation of the authors’ previous work. tively. All flow meters are specially calibrated for R-134a
The main concern of the present study is to experimentally from the manufacturer. The total capacity of all refrigerant
investigate the effect of the external parameter i.e. heat sink flow meters is 0.3 to 3.3 LPM. The manufacturer’s listed
temperature and heat source temperature, the effect of rel- accuracy is 0.1% of the full scale. The temperatures are
evant geometric parameters of the ejector on the perfor- measured by T-type thermocouples having accuracy of
mance of the refrigeration cycle using a two-phase ejector 0.1 °C. All the temperature-measuring devices are well cal-
as an expansion device. Moreover, the effect of the outlet ibrated in a controlled temperature bath using standard
diameter of the motive nozzle on the performance of the precision mercury glass thermometers. Bourdon gauges,
refrigeration system, which has never before appeared in calibrated against the dead weight test, are used to measure
the open literature, is presented. the pressures. All static pressure taps are mounted flush in
 Oil
Condenser
Separator

2′
Stirrer
RTD By-Pass
Valve
2
3 1 Inverter
CDU Flow
Electrical Meter
Heater Sight
Glass
Condensing Cold-Water Cold-Water
Unit Tank Pump Compressor Motor
.
mp Filter-Drier
Receiver
Sight
Glass
Primary . OE/CC
Flow mp
Meter OC/CE
. .
Ejector m p+ ms
OE/CC 3′ 4
Liquid-
Vapor Accumulator
.
OE/CC
Separator
ms
OC/CE
6 .
Expansion ms
Main Valve
Evaporator Secondary
Flow
Meter

5 OE/CC Sight Sub-Evaporator

Glass

Flow
Meter
Stirrer
RT

By-Pass
Valve OE/CC : Open for Ejector Mode, Closed for
Conventional Mode
OC/CE : Open for Conventional Mode, Closed
Electrical for Ejector Mode
Heater
Hot-Water Hot-Water P : Pressure Gauge (High/Low)
Pump Tank T : Thermocouple (T-Type)

3 2′
Condenser

Liquid-Vapor
Separator . 2
3′ ne se me 4 mp
. .
Ejector m p + ms 1

Compressor
.
ms
6 5
Evaporator

Fig. 1. Schematic diagram of experimental apparatus.

the tube wall. Please note that the sub-evaporator shown in loop. The water in the tank is heated with a 4.5 kW electric
Fig. 1 was not used in the present study. It was prepared heater and supplied through the evaporator by the circulat-
for the experiment with high cooling load. The main evap- ing pump. The condenser rejects heat to the water coming
orator has enough capacity for the cooling load used in the from a cold water tank. The water is cooled by a separated
present study. refrigeration system using R-22 as refrigerant. The sepa-
In order to collect data at various conditions, the heat rated refrigeration system consists of a condensing unit
load to the evaporator is supplied by using the hot water with a capacity of approximately 2.6 tons of refrigeration,
 Motive Nozzle Mixing Chamber Diffuser

Mixed
Primary Refrigerant
Refrigerant

Suction Chamber
Secondary
Refrigerant

Nozzle Rod
Motive Nozzle Mixing Chamber and Diffuser
Locking Nut

φ22.0
φ10.0
φ30.0

O-Rings Suction Chamber

30.0 110.0 90.0
Fig. 2. Ejector assembly.

a capillary tube, filter/drier, sight glass and a helical copper gent length (Lnd) of all three motive nozzles are 6, 32, 6,
tube coil immersed in water being cooled in an insulated 0.9, 20 mm, respectively. Usually, a comparison of the
tank. The test runs are done at the cooling load ranging cycle performance can be made by two approaches. The
between 1.8 and 3.0 kW. first one is based on internal parameters i.e. evaporating
The schematic diagram of the two-phase ejector as and condensing temperatures. This method requires the
shown in Fig. 2, is designed in three main parts: the motive different modes to be compared at the same evaporating
nozzle, the suction chamber, and the mixing chamber with and condensing temperatures. The second one is based
diffuser. The motive nozzle throat area is designed accord- on external parameters such as the inlet temperature and
ing to the Henry and Fauske model [8]. This model is used the flow rate of the heat transfer fluid (HTF) (Hoegberg
because it considers the metastable effect of the expansion et al. [11], Giuliani et al. [12]). This method allows each
of saturated liquid into the liquid–vapor mixture region. mode of operation to operate under the same external con-
The remaining cross-sectional areas of the ejector are ditions. In this paper, the comparison is based on the sec-
designed according to the homogeneous equilibrium model ond method. This method is selected because it is more
(HEM) (Kornhauser [1], Sherif et al. [9]). HEM is based on likely in real life. Water is used as the heat transfer fluid.
the assumption that vapor and liquid are in thermal and Hot water acts as a heat source of the evaporator and heat
mechanical equilibrium. Furthermore, the mixing process source temperature means temperature of hot water. While
is assumed to occur at constant pressure. The other dimen- cold water acts as a heat sink of the condenser and heat
sions, including the lengths of each section and the conver- sink temperature means temperature of cold water.
gent and divergent angles, are based on recommendations According to the experimental conditions, the test runs
from the ASHRAE Handbook [10] and from Nakagawa are done at heat source temperatures ranging between
and Takeuchi [5]. Brass is used as material for the ejector. 8 °C and 16 °C while the volume flow rate of hot water is
The three main parts of the ejector are connected by fine kept constant at 12 LPM. Each value of heat source tem-
screws. Three o-rings, as shown in Fig. 2, are used to pre- perature is tested at varying heat sink temperatures of
vent refrigerant leakage. The detail drawing of the tested 26.5, 29.5, 32.5, 35.5 and 38.5 °C. The volume flow rate
motive nozzle is shown in Fig. 3. of cold water is fixed at 14 LPM. These volumetric flow
In the present study, three motive nozzles having differ- rates of cold water and hot water are employed to prevent
ent outlet diameter (Dne) of 2.0, 2.5 and 3.0 mm are inves- water from freezing on the surface of the equipment. The
tigated. The inlet diameter (Dni), inlet length (Lni), compressor speed is maintained at 450 rpm by controlling
convergent length (Lnc), throat diameter (Dnt) and diver- the inverter frequency. It should be noted that this speed
 Dnt
Dni

Dne
Lni Lnc Lnd

Fig. 3. Motive nozzle.

is appropriate for this experimental set up. This speed is chamber, refrigerant vapor originating from the evapora-
employed after problems were encountered at several dif- tor at point 6, is entrained with the high velocity refrigerant
ferent speeds used in previous experiments. It has been jet and compressed through mixing chamber into the dif-
found that if compressor speeds of over 450 rpm are used, fuser at point 4. Vapor refrigerant from the liquid–vapor
the pressure in the liquid–vapor separator decreases. This separator is sucked by compressor while sub-cooled liquid
enables the liquid in the separator to become increasingly passes on to the inlet of the evaporator at point 5 and
vaporized and finally result in the increase of the vapor absorbs heat from hot water supplied to the evaporator.
temperature. This high temperature vapor probably causes The outlet of the evaporator appears a point 6. The cycle
the compressor to become damaged. Also if compressor continues on and on.
speeds of lower than 450 rpm are used, the amount of Motive nozzles used in the present study have different
liquid in the separator gradually increases. Finally, the sizes of outlet diameter. This different outlet diameter
liquid refrigerant floods the outlet of the separator and directly affects the shape of the divergent angle of the
flows to the compressor. This results in compressor failure. motive nozzle. The following results are gained from exper-
The water temperatures are kept constant at the required imental conditions in which the refrigerant at the entries of
values in both the hot and the cold water tanks. They are the motive nozzle and the evaporator are sub-cooled with
then circulated through the condenser and evaporator at temperatures of 0.1–3.5 °C and 0.1–1.1 °C, respectively,
constant temperatures and constant volume flow rates by while the refrigerant at the inlet of the compressor is super-
the circulating pumps. heated with the temperature of 2.0–4.5 °C. The experiment
yields similar variations of most variables affecting the per-
3. Results and discussion formance of the cycle, with significant differences in certain
conditions. However, the variations of variables tested with
Fig. 4 shows a pressure-enthalpy diagram of the this range of outlet diameter are not favorably obvious.
two-phase ejector refrigeration cycle for Dne = 2.5 mm, at Fig. 5 shows the variation of the primary mass flow rate
Tsource = 8 °C and Tsink = 26.50 °C. Fig. 4 is only an exam- with heat sink temperature from using the various sizes of
ple, however the experimental results from other experi- motive nozzle for the different heat source temperatures of
mental conditions give the same tendency. The cycle is 8, 12 and 16 °C. The graph indicates that an increase or
separated into two loops; primary refrigerant loop and sec- decrease of the heat source temperature does not affect
ondary refrigerant loop. In the primary refrigerant loop, the primary mass flow rate, even though it has direct effect
vapor from the liquid–vapor separator is drawn in the com- on the increase or decrease of temperature and pressure in
pressor cylinder during its suction stroke and is compressed the evaporator. This varying pressure has very little effect
to pressure p2 and temperature T2 during the compression on the pressure difference between the inlet and the dis-
stroke and delivered out from the compressor at condition charge of the motive nozzle. Therefore, the variation of
2 passes on to condenser at 2 0 in which cooling water is the heat source temperature has no significant effect on
supplied to remove heat from the vapor. Vapor is therefore the primary mass flow rate.
first cooled to the saturated temperature at pressure p2 and Concerning the variation of the primary mass flow rate
further removal of heat. Condensation at the high temper- with heat sink temperature at heat source temperatures of
ature produces a sub-cooled liquid refrigerant at point 3. 8, 12 and 16 °C, as shown in Fig. 5, it is found that the pri-
The high pressure liquid is now further expanded through mary mass flow rate tends to increase when the heat sink
motive nozzle at point 3 0 . The primary refrigerant will be temperature increases. This is because the temperature
mixed with the secondary refrigerant at mixing chamber. and pressure of the condenser increase with increasing heat
The mixture will be compressed through diffuser and flow sink temperature. This higher condenser pressure results in
to point 4 and to the liquid–vapor separator. In the second- the increase of the pressure at the inlet of the motive nozzle.
ary refrigerant loop, since high pressure primary refrigerant As a result, the pressure difference between the inlet and the
is supplied to the nozzle inlet and is expanded in the mixing outlet of the motive nozzle also increases which, in turn,
 10.0 10.0

Motive Nozzle Dne = 2.5mm.

TSource = 8oC
TSink = 26.5oC
Pressure (MPa)

1.0 2 1.0
3
3' 2'

5 4
6 1

Ideal
Actual
ne me se
0.1 0.1
150 200 250 300 350 400 450 500
Enthalpy (kJ/kg)

Fig. 4. Pressure–enthalpy diagram of the two-phase ejector refrigeration cycle in the present study.

.04

Subcooling at the Inlet : Motive Nozzle :

- Motive Nozzle : 0.1-3.5 oC Dne = 3.0 mm.
: 0.1-1.1 oC
Primary Mass Flow Rate (kg/s)

- Evaporator Dne = 2.5 mm.

.03
Dne = 2.0 mm.
TSource = 8 oC
TSource =12 oC
.02 TSource =16 oC

.01

0.00
24 26 28 30 32 34 36 38 40
Heat Sink Temperature (οC)
Fig. 5. Comparison of primary mass flow rates of refrigerant at heat source temperatures of 8, 12 and 16 °C.

increases the primary mass flow rate. When comparing the Fig. 6 shows the comparison of the secondary mass flow
primary mass flow rate gained from the three motive noz- rate gained from using various sizes of motive nozzle at the
zles, it is found that all motive nozzles give similar primary heat source temperatures of 8 and 16 °C. The graph shows
mass flow rate. that the change of the heat source temperature has a small
 .04

Subcooling at the Inlet :

- Motive Nozzle : 0.1-3.5 oC

Secondary Mass Flow Rate (kg/s)

- Evaporator : 0.1-1.1 oC

.03

.02
Motive Nozzle :
Dne = 3.0 mm.
TSource = 8 oC
Dne = 2.5 mm.
TSource =16 oC
Dne = 2.0 mm.

.01
24 26 28 30 32 34 36 38 40
Heat Sink Temperature (οC)
Fig. 6. Comparison of secondary mass flow rates of refrigerant at heat source temperatures of 8 and 16 °C.

effect on the secondary mass flow rate. In addition, the The comparison of the recirculation ratio when using
results suggested that the secondary mass flow rate depends motive nozzles at heat source temperatures of 8 and
mainly on the primary mass flow rate and the pressure at 16 °C as shown in Fig. 7, indicates that motive nozzle hav-
the outlet of the motive nozzle. ing Dne = 3.0 mm produces the higher recirculation ratio
The comparison of the secondary mass flow rates at than the others. This is because the recirculation ratio is
Tsource = 8 °C, reveals that motive nozzle having the ratio of the secondary mass flow rate and vaporized
Dne = 3.0 mm produced a slightly higher secondary mass mass flow rate in the evaporator obtained from the calcu-
flow rate than the other nozzles, as more clearly evident lation. The calculation result indicates that the vaporized
at high heat sink temperatures. Concerning the changes mass flow rate is lowest when using motive having
in the secondary mass flow rate and heat sink temperature Dne = 3.0 mm. Considering the change in the recirculation
when using 3 different sizes of motive nozzle, it is found ratio and heat sink temperature gained from the three dif-
that the secondary mass flow rate increases when the heat ferent sizes of motive nozzle, it is found that the recircula-
sink temperature increases. This is because the higher heat tion ratio tends to increase when the heat sink temperature
sink temperature makes the pressure difference between the increases. This is due to the ejector behavior. That is
inlet and the outlet of motive nozzle higher, resulting in a because the secondary mass flow rate increases with
higher velocity of the primary mass flow rate that causes increasing heat sink temperature, while the vaporized mass
the pressure to drop at the outlet of the motive nozzle. This flow rate drops. The higher heat sink temperature causes
pressure drop at the outlet of the motive nozzle increases the heat transfer at the evaporator to drop.
the pressure difference between the outlet of the motive Fig. 8 shows the variation in the average evaporator
nozzle and the outlet of the evaporator. Hence, the second- pressure as the heat sink temperature changes, from using
ary mass flow rate increases accordingly. the various sizes of motive nozzle at heat source tempera-
The recirculation ratio is defined as the ratio between tures of 8, 12 and 16 °C. The graph indicates that the
the refrigerant mass flow rate delivered to the evaporator higher heat source temperature gives higher average evap-
ðm_ s Þ and flow rate of the vaporized refrigerant ðm_ e Þ. orator pressure. This is because the change in the heat
Fig. 7 shows the variation of the recirculation ratio with source temperature directly affects the changes of tempera-
heat sink temperature from using the various sizes of ture and pressure in the evaporator.
motive nozzle at heat source temperatures of 8 and The changes in the average evaporator pressure and heat
16 °C. The result shows that the recirculation ratio sink temperature as shown in Fig. 8, indicate that when the
decreases with an increase in heat source temperature. This heat sink temperature increases, the average evaporator
is because the heat source temperature increases with pressure will also increase slightly – even though the
increasing rate of heat transfer at the evaporator, which increase or decrease of the heat sink temperature does
results in the increasing vaporized mass flow rate, while not have a direct effect on the change of pressure in the
the secondary mass flow rate changes slightly. Ultimately, evaporator. The increase of the average evaporator pres-
the recirculation ratio also increases. sure may result from other causes, such as, the heat transfer
 5.0

Subcooling at the Inlet :

- Motive Nozzle : 0.1-3.5 oC
4.0
- Evaporator : 0.1-1.1 oC

Recirculation Ratio
3.0

2.0

Motive Nozzle :
1.0 Dne = 3.0 mm. TSource = 8 oC
Dne = 2.5 mm. TSource =16 oC
Dne = 2.0 mm.
0.0
24 26 28 30 32 34 36 38 40
Heat Sink Temperature (οC)
Fig. 7. Comparison of recirculation ratio at heat source temperatures of 8 and 16 °C.

0.6
Superheating at the Compressor Inlet : 2.0-4.5 oC
Average Evaporator Pressure (MPa)

0.5

0.4

0.3

Motive Nozzle :
Dne = 3.0 mm. TSource = 8 oC
0.2 TSource =12 oC
Dne = 2.5 mm.
Dne = 2.0 mm. TSource =16 oC

0.1
24 26 28 30 32 34 36 38 40
Heat Sink Temperature (οC)

Fig. 8. Comparison of average evaporator pressure at heat source temperatures of 8, 12 and 16 °C.

from surrounding into the system. When comparing the ture. This is because the heat sink temperature has a direct
average evaporator pressures gained from the three effect on the temperature and pressure in the condenser.
different sizes of motive nozzle, it is found that all motive When the condenser pressure increases, the pressure at
nozzles produce almost the same average evaporator the condenser outlet also increases. This results in the
pressure. increase of the compressor pressure ratio. The compressor
Fig. 9 shows the relation between the compressor pres- pressure ratio for specific heat source and heat sink temper-
sure ratio and the heat source temperature from using the atures obtained from all motive nozzles is similar.
various sizes of motive nozzle at the heat sink temperatures Fig. 10 shows the discharge temperature plotted against
of 26.5, 32.5 and 38.5 °C. It is clear that the compressor heat source temperature for the various sizes of motive
pressure ratio increases with increasing heat sink tempera- nozzle at heat sink temperatures of 26.5, 32.5 and
 5.0
Superheating at the Compressor Inlet : 2.0-4.5 oC

4.0

Compressor Pressure Ratio 3.0

2.0
Motive Nozzle :
Dne = 3.0 mm. TSink = 26.5 oC
1.0 Dne = 2.5 mm. TSink = 32.5 oC
Dne = 2.0 mm. TSink = 38.5 oC

0.0
8.0 10.0 12.0 14.0 16.0
ο
Heat Source Temperature ( C)
Fig. 9. Comparison of compressor pressure ratio at heat sink temperatures of 26.5, 32.5 and 38.5 °C.

80
Superheating at the Compressor Inlet : 2.0-4.5 oC
70
Discharge Temperature (oC)

60

50

40
Motive Nozzle :
TSink = 26.5 oC
Dne = 3.0 mm.
TSink = 32.5 oC
30 Dne = 2.5 mm.
Dne = 2.0 mm. TSink = 38.5 oC

20
8.0 10.0 12.0 14.0 16.0
ο
Heat Source Temperature ( C)
Fig. 10. Comparison of discharge temperature at heat sink temperatures of 26.5, 32.5 and 38.5 °C.

38.5 °C. The plot shows that the increase of heat sink tem- from all motive nozzles for specific heat source and heat
perature makes the discharge temperature rise accordingly. sink temperatures are similar.
This is because when the heat sink temperature increases, The cooling capacity can be calculated according to the
the compressor pressure ratio will also increase and, conse- following equation:
quently, the compressor temperature ratio defined as the Qevap ¼ m_ hw cp;hw ðT hw;in  T hw;out Þ
ratio between the temperature at the compressor outlet
and the temperature at the compressor inlet, will increase where m_ hw is the mass flow rate of hot water (kg/s), cp, hw is
accordingly. the specific heat at constant pressure of the hot water (kJ/
The result from this figure indicates that the discharge kg k), Thw, in is the hot water temperature at the inlet of the
temperature tends to decrease when the heat source tem- evaporator (°C), Thw, out is the hot water temperature at the
perature increases. The discharge temperatures gained outlet of the evaporator (°C).
 4.0
Superheating at the Compressor Inlet : 2.0-4.5 oC

3.0
Cooling Capacity (kW)

2.0

Motive Nozzle :
1.0 Dne = 3.0 mm. TSink = 26.5 oC
Dne = 2.5 mm. TSink = 38.5 oC
Dne = 2.0 mm.

0.0
8.0 10.0 12.0 14.0 16.0
Heat Source Temperature (οC)
Fig. 11. Comparison of cooling capacity at heat sink temperatures of 26.5 and 38.5 °C.

Fig. 11 shows comparison of the cooling capacity The changes in the cooling capacity and heat source
obtained from all motive nozzles at heat sink temperatures temperature as shown in Fig. 11, indicate that the cooling
of 26.5 and 38.5 °C. It can be clearly seen that the cooling capacity increases as the heat source temperature increases.
capacity decreases with increasing heat sink temperature. The change of the heat source temperature is similar to the
As a consequence of ejector operation, the higher heat sink change in the cooling load of the system. When the
temperature causes the temperature difference between the temperature of the hot water from the heat source
cold water from the heat sink and the refrigerant from the increases, the heat absorption of the refrigerant in the
condenser to decrease. As a result, the heat transfer rate of evaporator also increases. The amount of heat absorbed
the condenser decreases accordingly. Considering the by the refrigerant is actually the average heat transfer rate
energy conservation, it is found that when the heat transfer of the evaporator, which equals the cooling capacity of the
rate of the condenser decreases, the heat transfer rate of the system. The comparison of the cooling capacity gained
evaporator will also decrease accordingly. from using the 3 different outlet diameter of motive nozzle

7.0
Superheating at the Compressor Inlet : 2.0-4.5 oC
6.0
Coefficient of Performance (COP)

5.0

4.0

3.0

2.0 Motive Nozzle :

Dne = 3.0 mm. TSink = 26.5 oC
Dne = 2.5 mm. TSink = 32.5 oC
1.0
Dne = 2.0 mm. TSink = 38.5 oC

0.0
8.0 10.0 12.0 14.0 16.0
Heat Source Temperature (οC)
Fig. 12. Comparison of coefficient of performance at heat sink temperatures of 26.5, 32.5 and 38.5 °C.
reveals that there is no significant effect on the cooling refrigerant charge, high refrigerant flow, piping insulation
capacity. and installation cost.
The system COP is defined as the ratio between the cool-
ing capacity and the electrical power supplied to the com- Acknowledgements
pressor and can be written as
Qevap The authors would like to express their appreciation to
COP ¼ the Thailand Research Fund (TRF) for providing financial
W comp
support in this study.
where Wcomp is the electrical power supplied to the com-
pressor (kW). This electrical power is directly obtained
from the buit-in function of the inverter. References
Fig. 12 shows the variation of the system’s COP versus
[1] A.A. Kornhauser, The use of an ejector as a refrigerant expander, in:
heat source temperature from using the various sizes of Proceedings of the 1990 USNC/IIR-Purdue Refrigeration Confer-
motive nozzle at heat sink temperatures of 26.5, 32.5 and ence, 1990; 10–19.
38.5 °C. It can be clearly seen that the COP decreases as [2] G.S. Harrell, A.A. Kornhauser, Performance tests of a two-phase
the heat sink temperature increases. This is because when ejector, in: Proceedings of the 30th Intersociety Energy Conversion
Engineering Conference, Orlando, FL, 1995; 49–53.
the heat sink temperature increases, the cooling capacity
[3] P. Menegay A.A. Kornhauser, Improvements to the ejector expan-
of the system decreases, while the compressor pressure sion refrigeration cycle, in: Proceedings of the 31th Intersociety
ratio increases, resulting in an increase of electrical power Energy Conversion Engineering Conference, Washington DC, 1996;
input of the compressor. Ultimately, this results in the 702-6.
decrease of the COP of the system. Comparison of the sys- [4] P.A. Domanski, Theoretical evaluation of the vapor compression
cycle with a liquid-line/suction-line heat exchanger, economizer, and
tem’s COP, at heat sink temperature of 26.5, 32.5 and
ejector. Nistir-5606, National Institute of Standards and Technology,
38.5 °C, all motive nozzles tend to give similar COP. March, 1995.
[5] M. Nakagawa, H. Takeuchi, Performance of two-phase ejector in
4. Conclusions refrigeration cycle, in: Proceedings of the Third International Con-
ference on Multiphase Flow. Lyon. France, Jun. 8–12, 1998; 1-8.
[6] S. Disawas, S. Wongwises, Experimental investigation on the
Variables affecting performance and varying directly
performance of the refrigeration cycle using a two-phase ejector as
with the heat sink temperature include the primary mass an expansion device, Int. J. Refrigeration 27 (6) (2004) 587–594.
flow rate, secondary mass flow rate, recirculation ratio, [7] S. Wongwises, S. Disawas, Performance of the two-phase ejector
compressor pressure ratio, and discharge temperature. On expansion refrigeration cycle, Int. J. Heat Mass Transf. 48 (2005)
the other hand, the cooling capacity varies inversely with 4282–4286.
[8] R.E. Henry, H.K. Fauske, The two-phase critical flow of one-
the heat sink temperature while the average evaporator
component mixtures in nozzles, orifices, and short tubes. ASME
pressure varies only a little and tends to vary directly. Con- Trans. J. Heat Transf. May, 1971; 179–87.
cerning the effect of the heat source temperature on the var- [9] S.A. Sherif, W.E. Lear, J.M. Steadham, P.L. Hunt, J.B. Holladay,
iation of variables affecting performance, the results are Analysis and modeling of a two-phase jet pump of a thermal
found to be contrary with the effect of the heat sink temper- management system for aerospace applications, Int. J. Mech. Sci. 42
(2000) 185–198.
ature, with the exception of the average evaporator pres-
[10] ASHRAE, ASHRAE Handbook – Guide and Data Book. American
sure which varies directly. On the other hand, the Society of Heating, Refrigerating and Air Conditioning Engineering.
primary mass flow rate and the secondary mass flow rate 1969; Chapter 13, 151–58.
tend to be slightly increased as the heat source temperature [11] M. Hoegberg, L. Vamling, T. Berntsson, Calculation methods for
increases. The use of motive nozzles having different outlet comparing the performance of pure and mixed working fluids in heat
pump applications, Int. J. Refrigeration 16 (6) (1993) 403–413.
diameters in the range of 2.0–3.0 mm yields insignificant
[12] G. Giuliani, N.J. Hewitt, F. Marchesi Donati, F. Polonara, Compo-
effects on the system performance. However, although the sition shift in liquid-recirculation refrigeration systems: an experi-
ejector cycle has higher performance over the standard mental investigation for the pure fluid R134a and the mixture R32/
cycle, some disadvantage should be considered e.g. high 134a, Int. J. Refrigeration 22 (6) (1999) 486–498.