Solar Energy VoL 33, No. 2, pp. 155 162, 1984 00384)92X/84 $3.00+ .

00
Printed in the U.S.A. © 1984 Pergamon Press Ltd.

THERMAL PERFORMANCE OF A DIRECT EXPANSION

SOLAR-ASSISTED HEAT PUMP

SUSHIL K. CHATURVEDI~and JAMES Y. SH~N~

Department of Mechanical Engineering and Mechanics, Old Dominion University, Norfolk,
VA 23508-8508, U.S.A.

(Received 29 July 1982; revision received 1 August 1983; accepted 22 September 1983)

Abstract--A direct expansion solar assisted heat pump, in which a bare flat plate collector also acts as
the evaporator for the refrigerant, Freon-12, is designed and operated. The system components, e.g. the
collector and the compressor, are properly matched so as to result in system operating conditions wherein
the collector/evaporator temperature ranges from 0 to 10°C above ambient temperature under favorable
solar conditions. This operating temperature range is particularly favorable to improved heat pump and
solar collector performance. The system thermal performance is determined by measuring refrigerant flow
rate, temperature and pressure at various points in the system. The heat pump COPH and the solar
collector efficiency ranged from 2.0 to 3.0 and from 40 to 70 per cent, respectively, for widely ranging
ambient and operating conditions. Experimental results indicate that the proposed system offers significant
advantage in terms of superior thermal performance when compared with results gotten by replacing the
solar evaporator with a standard outdoor fan-coil unit.

INTRODUCTION pansion solar assisted heat pump that employs a bare

Due to rapidly rising oil and natural gas prices, alter- collector as the heat pump evaporator.
nate energy sources such as solar energy are receiving The direct expansion SAHP offers several advan-
greater attention for heating, cooling and power gen- tages over the conventional SAHP systems. These
eration applications. In addition, the short supply of include superior thermal performance due to direct
natural gas and the uncertainty about the oil supply in expansion and evaporation of the refrigerant in the
recent years has advanced the use of heat pumps for collector and the elimination of components such as a
residential/commercial and industrial heating applica- conventional evaporator and an intermediate heat ex-
tions. Since their introduction in the early fifties, com- changer circuit when compared with a conventional
mercially available heat pumps have undergone SAHP design using water as the collector fluid. The
significant design changes that have made them more use of refrigerant as the collector working fluid elimi-
reliable, and in some cases, even competitive with nates the night-time freeze up problems often associ-
conventional energy sources such as oil. These com- ated with collectors using water as the heat transfer
mercially available air to air heat pumps have instan- fluid. The experience from the refrigeration practice
taneous coefficient of thermal performance (COPH) also indicates that copper (and even aluminum tubes)
that ranges from 1.5 to 3.0 depending on the source can be used with refrigerants Freon-12 and Freon-22
(outdoor) and sink (indoor) temperatures for typical with minimal corrosion problems.
residential units. To present a balanced view of the proposed system,
In order to improve the heat pump COPH and dis- we shall describe here some of the potential problems
place the fossil energy resource, the idea of combining inherent in the proposed design. The use of a long
the heat pump and solar energy in mutually beneficial refrigerant loop in the proposed system can result in
ways has been proposed and developed in several refrigerant leakage and trapping of the oil in the
previous studies[l-4]. In a typical solar assisted heat evaporator. Although these problems were not en-
pump (SAHP), the solar collector and the heat pump countered in the current set-up, they can be elimi-
evaporator are separate units. The working fluid, usu- nated by designing and assembling the refrigeration
ally air or water, is circulated and heated through the loop properly. A more serious problem is due to large
collector and is used as the source for the evaporator fluctuation of load on the collector/evaporator due to
of the heat pump. On the other hand, the so-called large swings in solar insolation and ambient tem-
direct expansion solar assisted heat pumps [5-7] com- perature. As a result, significant mismatch between
bine the collector and evaporator functions into one the evaporative capacity of the collector and the
unit. The refrigerant from the condenser is expanded refrigerant pumping capacity of the compressor can
in the solar collector/evaporator panel where it boils develop at off-design conditions, leading to the dete-
due to direct solar energy input. This study determines rioration of thermal performance of the system. This
the experimental thermal performance of a direct ex- problem can be alleviated partially by using a vari-
able capacity compressor such as the one employing
tAssistant Professor. a 2-speed motor. The use of a fan-coil in parallel with
:~Graduate Student. the solar panel as a backup can also present potential

155
156 S.K. CHATURVEDIand J. Y. SHEN

control problems resulting from switching of the evated evaporating temperatures. His experimental
refrigerant from the solar panel to the fan-coil and results indicate that high COPs values, in the range
vice-versa. of 3-7, are possible by maintaining the evaporator
It is well known that solar energy collectors operate temperature 11-37°C for a 46°C condensing tem-
more efficiently at lower collector temperatures due perature. A very high evaporating temperature (be-
to reduced radiation and convection losses. Heat yond 25-30°C) may not be desirable because of
pumps, on the other hand, operate more efficiently at reduced marginal improvement in COPH values at
higher evaporating temperature. In order to improve elevated evaporating temperatures. At these high
the thermal performance of both components, the temperatures, the parasitic power becomes compara-
collector temperature must be lowered and the evap- ble to the compressor power and the COPH value
orator temperature must be raised simultaneously. does not improve as fast. Also, the high evaporating
The proposed direct expansion solar assisted heat temperatures impose a penalty on the collector ther-
pump achieves the dual objective by integrating the mal performance[7].
collector and evaporator functions into one com- Chaturvedi, Chiang and Roberts[7] have the-
ponent, henceforth referred as the collector/ oretically analyzed a solar assisted heat pump in
evaporator panel. which solar collector also acts as the evaporator for
The direct circulation and boiling of cool refrig- the heat pump. To analyze the 2-phase flow in the
erant in the collector/evaporator tubes results in collector tubes, an equalibrium homogeneous model
quenching of the hot collector surface which is cooled is adopted and the analysis extends the single phase
and maintained at temperatures slightly above ambi- collector theory to 2-phase flow regime. The local
ent (0-10°C) under favorable solar conditions. The collector fluid temperature, pressure and quality are
heat loss from the collector/evaporator is reduced as determined by formulating differential equations for
it is proportional to the collector and ambient tem- mass, momentum and energy balance. The resulting
perature difference, and this results in significant set of coupled first order differential equations for
improvement in the collector efficiency. This im- pressure (equivalently temperature in 2-phase region)
provement occurs despite an increase in collector loss and quality are solved numerically by the fourth-
conductance (UL) due to a bare collector surface, order Runge-Kutta method. A parametric evaluation
which is more than compensated by the decreasing of thermal performance is carried out to identify
collector and ambient temperature difference. Con- parameters that affect the system performance
versely, the evaporator (synonymous with collector in significantly. Theoretical results indicate that col-
this study) temperature is 0-10°C above ambient lector efficiency ranging from 45 to 80 per cent and
temperature, and this usually means significant im- COPn ranging from 3 to 6 per cent are feasible with
provement in the thermal COPH. In contrast, the a combination of an inexpensive bare collector and a
conventional fan-coil evaporator typically operates commercially available hermetic compressor.
with evaporating temperature 5-15°C below ambient In the present study, the steady state thermal
temperature, depending on the specific design. This performance of a direct expansion solar assisted heat
translates into very low COPH values at low ambient pump is investigated experimentally. The system is
temperatures. operated over widely ranging ambient and operating
The concept of direct expansion solar assisted heat conditions, and the collector efficiency and heat
pump is not a new one and was first proposed in an pump COPH are determined by measuring the refrig-
experimental study by Sporn and Ambrose[5]. De- erant mass flow rate, pressure and temperature at
spite an apparent mismatch between the compressor's different points in the circuit. The experimental re-
mass pumping capacity and the collector's evapo- suits are then compared with the theoretical predic-
rative capacity, their experiments indicated that the tion from a simplified version of the method
use of refrigerant collector as the evaporator for the presented in Ref.[7].
heat pump has merits from the point of view of
improved thermal performance of the collector and EXPERIMENTAL SET-UP
the heat pump. Franklin, Saaski and Yamagiwa[6] The proposed system is shown schematically in
have developed a direct expansion solar assisted heat Fig. 1. Although the heat pump can be operated in
pump at Sigma Research Inc., under the Department both heating and cooling modes, only the heating
of Energy sponsorship. A bare collector, without any mode is investigated due to its significant potential
glazing or back insulation, is used as evaporator for for year around water heating applications. A 1/2 hp
the heat pump. The collector is so mounted that it belt driven "Tecumseh" open compressor is em-
collects energy both directly from the top surface and ployed for compressing the refrigerant at low pres-
indirectly from the rear surface after reflection from sure and temperature (state 1) to state 2, which
a highly reflective supporting surface. They have represents superheated vapor at condensing pressure.
found that system performance is significantly im- The corresponding thermodynamic states of the re-
proved when compared with conventional solar heat- frigerant, Freon-12 for this investigation, are shown
ing systems under similar ambient conditions. on a T - s diagram in Fig. 2.
Kush [8] has summarized the thermal performance of The working fluid enters a "Standard Refrig-
residential size water to water heat pumps that el- eration" shell and tube type condensor and leaves it
 Thermal performance of a direct expansion solar-assisted heat pump 157

Pyranometer
Ambient Tamp. /r ~ 7 I~-=¢
Pres. Transducer

J
s=
J 5V~r¢" n eatI

Datalogger

Dry-filter Liq.~

A , , , , , - - , IO°°dense'l Y--I l=-%t-,,,<,te,

CompressorUnit
Fig. 1. Experimental set-up

refrigerant at the entrance of the collector (point 4).

The refrigerant pressure at the collector inlet is also
measured by a "Serta System" pressure transducer
with a capacitance type sensor. The milivolt output
from the thermocouple and the pressure transducer is
ku
tz continuously monitored and recorded at 10-min in-
I tervals.
Or"
hi The solar collector consists of a single panel of
0_
~r 3.39 m 2 area and is formed from a serpentine tube
1' and sheet made out of copper. The specifications and
characteristics of solar collector and all other com-
ponents are listed in Table 1. The collector does not
employ any glazing and rests on an aluminum struc-
tural support with mineral wool back insulation. A
conventional fan-coil evaporator with an expansion
ENTROPY( s ) valve is employed in parallel with the solar panel as
Fig. 2. Heat pump cycle on T - s diagram. the back-up system. In the absence of solar energy,
the refrigerant is routed through the fan-coil to
achieve evaporation. The conventional heat pump
as subcooled liquid at state 3. The condensing of using the fan-coil, and the solar assisted heat pump
refrigerant is achieved by circulating cold water from using the collector/evaporator panel are operated
the municipal water supply (state 6). The warm water under identical ambient and operating conditions to
from the condenser (state 7) can be used for hot water evaluate the improvement in thermal performance
applications. The subcooled refrigerant from the due to the use of solar energy as the source for the
condenser is pumped to a solar collector through a heat pump.
9.6 mm (3/8 in) dia. copper tube. The refrigerant at state 4 boils off in the serpentine
A "Flow Technology" O M N I F L O flowmeter (8) is collector tube due to incident solar energy and the
provided near the entrance to the collector for mea- resultant superheated vapor leaves the collector at
suring the liquid refrigerant flow rate continuously. state point 5. The pressure and the temperature and
The milivolt output from this turbine flowmeter is the collector exit are measured by another pressure
continuously monitored on a twenty channel data- transducer and thermocouple pair. The refrigerant in
logger that samples and records the millivolt output superheated state is then returned to the compressor
at 10-min intervals. The refrigerant is then throttled through a filter-drier and a suction line receiver which
to the collector/evaporator pressure by an expansion is incorporated to trap any liquid returning to the
valve. compressor.
An " O M E G A - M C J " chromel-constantan thermo- The solar flux in the plane of the collector is
couple measures the temperature of the saturated measured directly by mounting an Eppeley pyr-
158 S. K. CHATURVEDland J. Y. SrmN

Table 1.

Components Specifications and characteristics

A. Collector:
1. Configuration Serpentine tubing and plate is tube-in-fin 100 per cent
capillary soldered.
2. Absorber panel Plate material: copper; thickness: 0.0003048 m
surface treatment: black paint coating (absorptivity
=90 per cent, emmissivity = 0.9).
3. Tube material and spacing Copper 1/2" nominal type M on 0.1143 m centers.
4. Insulation Edge: 0.0254 m Thmax polyisocyanurate.
Back: 0.0254 m Thmax foil faced 2 sides.
(k = 0.2249 W/mK).
5. Case material 6063T5 extruded aluminum frame, back 0.0008128 m
aluminum sheet, mill on brown anodized.
B. Heat pump 1/2 HP; bore: 0.0381 m; stroke: 0.0365 m;
cylinder: 2; displacement: 0.0000832 m 3.
C. Thermocouple connector Model No. MCJ-E, Chromel Constant,
accuracy: +½°C, 15-30°C; +I°C, 10~5°C.

anometer in the plane of the collector. The milivolt rator as heat of vaporization and for superheating;
output from the pyranometer is continuously mon- 1--.2 represents compression; 2 ~ 3 is condensation
itored and recorded at 10-min interval. The power process in which heat is rejected from the refrigerant
input to the electrical motor-compressor set is mea- for water heating applications, and the process 3--*4
sured and recorded by connecting a wattmeter to the represents throttling in the expansion valve. The
motor terminals. The low and the high pressure collector/evaporator temperature and the mass flow
across the compressor are measured by Bourdon type rate through the collector are principal unknowns of
pressure gages. To protect the compressor/motor set the problem and are determined by specifying col-
from overload, low pressure and high pressure cut off lector, ambient, compressor and working fluid param-
switches are connected to the compressor. The tem- eters.
perature of water entering and leaving the condenser Neglecting the small pressure (also temperature)
is measured by a set of thermocouples and monitored drop through the collector (method 2), the steady state
at 10-min intervals. The measurement of water tem- energy balance across the isothermal collector be-
perature rise across the condenser and the condenser comes
water flow rate enables one to calculate the energy
rejected by the refrigerant in the condenser. th(h, -- h4) = F'Ac[S(r~) - U L ( T f - Ta)]. (1)

THEORETICAL PREDICTION OF THERMAL Where F ' is the collector efficiency factor; Ac is the
PERFORMANCE collector area (m2); S is the solar flux available in the
Thermal performance of the solar refrigerant col- plane of the collector (W/m2); (z~) is transmittance-
lector can be predicted by two methods. In method I, absorptance product. Tf and Ta are collector fluid and
the pressure drop in the collector/evaporator tubes is ambient temperatures respectively, th is the refrigerant
considered by formulating the governing equations mass flow rate (kg/sec); h is the enthalpy (kJ/kg) and
for 2-phase flow in the tubes with the assumption of UL is the collector loss coefficient.
equilibrium homogeneous flow. After some manipu- The collector efficiency factor is given by the stan-
lations, one gets coupled first order ordinary dard expression in Hottel-Whilliar-Bliss model for
differential equations for pressure (or temperature) flat plate collector.
and quality in the saturated region. These equations
are solved by an iterative method described in detail in 1/UL
F"
Ref. [7]. In method 2, a simplified version of method 1, [ 1 1+1] .(2)
the pressure drop through the collector is neglected w U~tD+ ( w - o ) r l + ~
and an energy balance on it leads to a set of algebraic
governing equations. Results from both methods are Where various terms have their usual meaning.
in excellent agreement for cases in which the pressure The expression for tube internal heat transfer
drop through the collector tubes is small (<20 kPa). coefficient, h~, for refrigerant-12 is taken from Ref. [9]
For the collector compressor combination chosen in which applies to 2-phase flow of refrigerant in hori-
this study, the typical pressure drops are of the order of zontal tubes.
10 kPa ( - 1 . 4 psi) and thus have negligible effect on
the thermal performance of the system. 0.0082 K f / R e 2 hfg'~°4°
The heat pump cycle is shown on a T - s diagram in
Di (3)
Fig. 2. Evaporation of the refrigerant takes place
along process 4--* 1 in which the net energy absorbed Where Kf is the liquid thermal conductivity; J is a
(incident energy minus the losses) is utilized by evapo- dimensional constant with value 778; Ax is the change
 Thermal performance of a direct expansion solar-assisted heat pump 159

in quality from the inlet of the collector tube to the Where A T is the degree of superheating and is an
exit; h:g is the latent heat of vaporization and ReD, is the experimentally controllable parameter. Data from
Reynold's number based on internal tube diameter D i A S H R A E handbook[10] is used to fit a third order
and liquid refrigerant viscosity. The symbol L repre- polynomial to vv.
sents the length of the collector tube.
For a constant compressor speed operation, the vv = a0 + alP + a2P 2 + aaP 3. (9)
mass pumped and circulated by the compressor is
given by the following expression Equations (1-9) represent nine equations govern-
ing nine variables Ty, rh, qv, v~, vv, p, 1"1, h~, UL,
q, VD respectively. Eliminating all variables except T¢, the
rh - (4)
system of 9 equations can be reduced to the following
functional form,
The symbol VD is the displacement volume of the
compressor (m3/sec), v I is the specific volume at the G[T:; fluid parameters a0, al, a2, a3; VD, Pcorld, A0, AI,
inlet of the compressor and r/~ is the volumetric WS, S, T o, Ao AT, W, D] = 0 . (10)
efficiency. Although qv is dependent on several param-
eters, it is usually correlated with pressure ratio across The solution (Tf) of this functional eqn (10) can be
the compressor. The volumetric efficiency is deter- obtained iteratively by Newton-Raphson technique
mined experimentally by measuring the mass flow rate for given set of compressor parameters, e.g. A0, A~,
and the pressure ratio across the compressor. The least VD; ambient parameters To, WS, S; collector param-
square fit through the volumetric efficiency data is eters, Ao W, D and operational parameters such as
given by the following expression the condensing pressure Pcond and the degree of
superheat AT. Once the evaporator/collector tem-
P perature (T:) is determined the system performance
qv = A° + B° ~])con--" (5) parameters such as the collector efficiency and the
heat pump COPH can be obtained by determining
Where P is the pressure at the inlet to the compressor, enthalpies at various points in the system.
Pcond is the condensing pressure and A0, B0 are con- The collector efficiency qco:: is defined as,
stants. For saturated states, P can be expressed as a
function of saturated collector/evaporator tem- useful energy collected
?/coil --
perature T:.'~ energy incident in the plane of the collector

P = f(Ty). (6) ~h[hl - h4]

(11)
AcS
The specific volume at states 1 and 1'? can be related
by generalized equation of state The heat pump C O P H is defined as

vl T1 Zlpv Thermal energy rejected by the condenser

COPH =
vv Tv Zv Pl " Electrical work imput to the compressor

Where Z~ and Zv are compressibility factors at states = n~dhw0 - h~i]

1 and 1'. A detailed calculation of Z~/Z1, for most cases 14,el~tn¢ (12)
of interest reveals that although Z~ and Z1, values are
typically in the range of 0.80-0.90; their ratio can be Where rh wis the condenser water flow rate and hw, and
approximated by unitY with the resultant error in/)1//)l' hw0 are enthalpies of water at inlet and exit of the
being less than 2 per cent. The small pressure drop condenser.
between point 1 and 1' of single phase superheated
region is also neglected, thus resulting in condition DISCUSSION OF RESULTS
Pv = Pt. With the above approximations, the specific The experiments were performed with the objective
volume relationship reduces to of obtaining the basic understanding of operational
aspects of direct expansion SAHP system and to
v, T, validate the theoretical model governing its steady
(7)
/)1" TI' state thermal performance. Consistent with above
objectives, and to minimize the number of variables,
and the experiments were performed under clear sky
conditions, so that a quasi steady state could be
T, = T , , + AT. (8) realized. For most runs, the system was started well
ahead of solar noon, and the quasi steady state of the
operation was usually achieved around noon-time.
?Under the assumption of small pressure (temperature) The data was obtained for various ambient and
drop through collector, T: and Tv are same. operational conditions with ambient temperature
160 S.K. CHATURVEDIand J. Y. S~mN

ranging from - 4 to 22°C, condensing temperature slightly due to a small (negligible) pressure drop in
from 40 to 50°C, and solar radiation in the plane of the saturated region, and then rises to superheated
the collector from 700 to 1000 W/m 2. state at the exit of the collector. Since the 2-phase
Figure 3 shows the collector/evaporator tem- region occupies roughly 95 per cent of the collector
perature (Ti) as a function of time of the day. The area for most experimental runs, the collector is
incident solar radiation (S), ambient temperature roughly at uniform temperature (Ti). From Fig. 3, it
(To) and Ts - To, are also shown on the same figure. is also apparent that the collector fluid temperature
It is evident that around solar noon, the (Ti) can be maintained at temperatures 0-10°C above
collector/evaporator temperature is roughly constant the ambient temperature under clear sky conditions
with time. Since the compressor operates between the encountered during this and most other similar runs.
constant evaporator temperature and a constant Figure 4 compares the theoretically predicted
(controlled) condensing temperature, most system collector/evaporator inlet pressure with the experi-
parameters show steady values. After entering the mentally observed inlet pressure for several experi-
collector, the refrigerant temperature first decreases mental runs. The straight line at 45°C angle repre-

-1000 m"
® ® ® ® ®
22 ® ® ® -900 •
[]
E) I~ [3 mm~ B m 800
X
3
20. [] ® LL
[] "700 rr
,,¢
18.
._J
16. .600 o

_ 14. O O
(J
o
--
W 12.
x
D 10.
I,- X
0 X
n,," X X
uJ 8
13_ X X
'" 6 Q solar f l u x (S) X
t.--
E) c o l l e c t o r temperature(Tfi)
O ambient temperature(To)
x Tf~ - Ta IZ~TI
O.
Nov.12 1980
l O'~oli:OOli:2o li~o 1~oo1~.2o1~o ~'~o 1':2o ;~o :;:oo ~o
LOCAL TIME(EST)
Fig. 3. Variation of T~, T= and S with local time.

sents a perfect match between theoretical and

u 560- experimental results. Theoretical results are obtained
by simulating experimental runs, using the simplified
o 530- / theoretical model discussed earlier. Figure 5 shows
the comparison of experimentally measured and the-
500- oretically predicted refrigerant mass flow rates
L,U through the collector. Figure 6 shows the comparison
cr
a. 470- of experimentally and theoretically predicted col-
._c lector efficiency. It is also evident from Fig. 6 that
a_ 440-
high collector efficiencies (40-70 per cent) are feasible
with bare collector operating under ambient condi-
410. A
tions encountered in winter months. Despite some
380. / scatter in Figs. 4-6, the general trend is that the
theoretical predictions are roughly in line with experi-
mentally observed values for pressure level, mass flow
3~o ~o ~io ~o s6o s~o s~o rate and collector efficiency. The simplified method is
Pin ( MEASURED ), kPo adequate for predicting the thermal performance for
Fig. 4. Comparison of predicted and measured collector those cases where the pressure drop through the
inlet pressure. collector is small.
 Thermal performance of a direct expansion solar-assisted heat pump 161

016 - 15-
G
o
lO-
015-
~_~ 5-
014-
i i i ~ i I i i
~_)

~ 013- 014-
n,-
Q.
013-
- - 012-
012-
011- 011-
.E
010-
010- ~'
I i I i i i i I

I I I I I I
10 .011 .012 .013 .014 .015 016 07-
mf (MEASURED). kg/s 06- ~
05. ~
Fig. 5. Comparison of predicted and measured refrigerant
mass flow rate. ~ o,..~ "~

0.8 212 2+, 2+6 2+8 260 2~2 26, 266 288
TQ (AMBIENT). k

0.7 Fig. 7. Variation of experimentally measured values of

W Tj~-Ta, rhs and 0r as a function of ambient temperature.
_o 0.6
E3
ILl higher temperatures. With the solar energy and other
a. 0.5" controllable parameters remaining constant, the col-
lector temperature elevation over ambient tem-
0.4"
perature decreases due to an increased mass flow rate.
0.3" The decreasing value of Ts - Ta implies reduced
energy losses from the collector and consequently
0.2 higher collector efficiency as seen from Fig. 7.
Figure 8 shows the variation of COPH as a function
0.1 of ambient temperature for the solar assisted and
fan-coil heat exchanger modes under similar oper-
0'1 ;.2 cJ.3 &,. d5 ;.6 6.7 &8 ating and ambient conditions. The COPH calculation
for the fan mode also includes the parasitic power for
Q c o l l (MEASURED)
the evaporator fan. The data for the SAHP mode
Fig. 6. Comparison of predicted and measured collector
shows that the COPH increases with ambient tem-
efficiency.
perature. Although the data points for the fan oper-
ated mode are limited, they clearly show the trend
Figure 7 shows the variation of experimentally that COPH in the solar mode is roughly 30-50 per
measured values of Ti - Ta, rn and ~/c, as a function
of ambient temperature. For the data points shown
in Fig. 7, all experimentally controllable parameters SOLAR MODE
were maintained in quite a narrow operational range. ra FAN NODE
From example, the condensing temperature ranged 5

between 47 and 49°C; the solar energy flux in the

4
plane of the collector between 850 and 950 W/m2; the
"I-
degree of superheating A T between 3 and 8°C. The Q_
3
O
solid lines represent the least square fit through the L_)

data points. The small scatter in the data points is due 2

partly to a slight variability in above parameters and
also to uncontrollable wind conditions. The experi- 1
mental results show the same trend as predicted in a
previous theoretical study[7]. As the ambient tem-
perature increases, the mass flow rate through the 2~o 2~s 260 265 2~o
collector increases. With the compressor displace- TO (AMBIENT) (*K)
ment volume remaining fixed, more mass is pumped Fig. 8. Variation o f COP H as a Function of ambient tem-
through the circuit because refrigerant is denser at perature for the solar mode and the fan mode.
162 S. K. CHATURVEDIand J. Y. SH~N

gible effect on the thermal performance. In view of

1/.- small pressure drops encountered in most experi-
g mental runs, the pressure (temperature) drop is alto-
-- 12. gether neglected in the theoretical analysis, thus
O
iii
resulting in isothermal approximation (simplified
A A
~- 10- method 2).
N
111

n CONCLUSIONS
a. 6" A direct expansion SAHP system using bare col-
lector is investigated for hot water heating applica-
1.- tions. The experimental results indicate that high col-
lector efficiencies ranging grom 40 to 70 per cent can be
realized with bare collector in winter-like conditions.
The heat pump COPH in the solar mode ranged from
2.0 to 3.0, while in the fan mode, the values were
~. ~ ~ 1'0 1'2 1'1.
around 1.5, thus indicating a significant improvement.
Ap ( MEASURED ),kPa Still higher COPH values are possible if larger capacity
Fig. 9. Comparison of predicted and measured pressure (more efficient) hermetic compressors are used instead
drop across the collector. of the small open compressor used in the present
study. The theoretical model for calculating pressure
drop in the evaporator underestimates by 25 to
cent higher than the fan (conventional) mode for the 35 per cent. For low pressure drop in the
fancoil chosen in this study. A proper comparison of collector/evaporator panel, the isothermal approxi-
the two modes must be based on the economics of the mation can be applied (method 2) and results obtained
respective modes. Such a comparison, although be-
from this simplified method are in good agreement
yond the scope of this study, is essential in establish- with experimental results.
ing the superiority of one mode over the other. The
COP. values are themselves quite low because a small
open compressor is used in this study. The motor heat REFERENCES
is lost to the atmosphere instead of being put in the 1. R. C. Bosio and N. V. Suryanarayana, Solar assisted
refrigerant, as in a hermetic compressor. Since small heat pump system: a parametric study for space heating
compressors are also less efficient, the resultant COPH of a characteristic house in Madison, Wisconsin. A S M E
values for the solar mode ranged from 2.0 to 3.0. It Paper 75-WA/Sol-8 (1975).
2. J. W. Andrews, Development of a cost effective solar
is conjectured that the use of large size hermetically
assisted heat pump. Proc. 1978 Ann. Meet. Am. Section
sealed compressors will result in COPH in the range Int. Solar Energy Soc. Inc., Vol. 2.1, pp. 281-287.
from 3.0 to 6.0, much in line with the theoretical 3. T. L. Freeman, J. W. Mitchell and T. E. Audit, Per-
predictions made in a previous study[7]. formance of combined solar-heat pump systems. Solar
Figure 9 shows the comparison of the predicted Energy 22, 125-135 (1979).
4. R. T. Duncan and G. J. Van Zuiden, Economic viability
and the measured pressure drop across the collector of the solar-assisted industrial heat pump system. Solar
for widely ranging ambient and operating conditions. Indust. Process Heat Conf. Proc. Solar Energy Research
The theoretical pressure drop is obtained from Institute/TP-333-429,San Francisco, California, (1979).
method 1 by detailed consideration of equilibrium 5. P. Sporn and E. R. Ambrose, The Heat Pump and Solar
Energy Proc. Worm Symp. Appl. Solar Energy, Phoenix,
homogeneous model for 2-phase flow in the
Arizona, (1955).
collector/evaporator tubes[7]. The detailed the- 6. J. L. Franklin, E. W. Saaski and A. Yamgiwa, A high
oretical model, as seen from Fig. 9, underestimates efficiency,direct expansion solar plane. Proc. 1977 Flat-
the measured pressure drop by amounts ranging from Plate Solar Collector Conf, pp. 187-195, Orlando, Flor-
25 to 35 per cent. Although the deviations are large, ida, (1977).
7. S. K. Chaturvedi, Y. F. Chiang and A. S. Roberts, Jr.,
they are by no means uncommon in 2-phase flow Analysis of two-phase flow solar collectors with applica-
literature[l 1]. More experimental data is required tions to heat pumps. A S M E Paper 80-WA/Sol-32
before one can draw any definite conclusion. If the (1980).
theoretical underestimation trend persists in the en- 8. E. A. Kush, Performance of heat pumps at elevated
evaporating temperatures with application to solar in-
larged experimental data base, then perhaps the
put. J. Solar Energy Engng 102, 203-210 (1980).
frictional coefficient can be modified by a constant 9. M. Altman, R. H. Norris and F. W. Staub, Local and
factor to produce a better fit between the theory and average heat transfer and pressure drop for refrigerants.
experiments. Alternately, more sophisticated models A S M E Transa. 189 (1960).
for 2-phase flow may be adopted to predict pressure 10. American Society of Heating, Refrigeration and Air
Conditioning Engineers.Fundamentals Handbook (1965,
drops; but this seems unnecessary, for the present 1966).
experimental set up where the pressure drops for 11. G. B. Wallis, One-Dimensional Two-Phase Flow.
most runs are very small (<11 kPa) and have negli- McGraw-Hill, New York.