Energy Conversion and Management 48 (2007) 2255–2260

www.elsevier.com/locate/enconman

Open air–vapor compression refrigeration system for air

conditioning and hot water cooled by cool water
a,b,*
Shaobo Hou , Huacong Li a, Hefei Zhang a

a
School of Power and Energy, Northwestern Polytechnical University, Xi’an, Shaanxi 710072, PR China
b
College of Engineering, Guangdong Ocean University, East Jiefang Road, No. 40, Xiashan, Zhanjiang, Guangdong 524006, PR China

Received 25 January 2005; received in revised form 22 September 2006; accepted 9 May 2007
Available online 18 June 2007

Abstract

This paper presents an open air–vapor compression refrigeration system for air conditioning and hot water cooled by cool water and
proves its feasibility through performance simulation. Pinch technology is used in analysis of heat exchange in the surface heat exchan-
ger, and the temperature difference at the pinch point is selected as 6 °C. Its refrigeration depends mainly on both air and vapor, more
efficient than a conventional air cycle, and the use of turbo-machinery makes this possible. This system could use the cool in the cool
water, which could not be used to cool air directly. Also, the heat rejected from this system could be used to heat cool water to 33–
40 °C. The sensitivity analysis of COP to gc and gt and the simulated results T4, T7, T8, q1, q2 and Wm of the cycle are given. The sim-
ulations show that the COP of this system depends mainly on T7, gc and gt and varies with T3 or Twet and that this cycle is feasible in
some regions, although the COP is sensitive to the efficiencies of the axial compressor and turbine. The optimum pressure ratio in this
system could be lower, and this results in a fewer number of stages of the axial compressor. Adjusting the rotation speed of the axial
compressor can easily control the pressure ratio, mass flow rate and the refrigerating capacity. The adoption of this cycle will make
the air conditioned room more comfortable and reduce the initial investment cost because of the obtained very low temperature air.
Humid air is a perfect working fluid for central air conditioning and no cost to the user. The system is more efficient because of using
cool water to cool the air before the turbine. In addition, pinch technology is a good method to analyze the wet air heat exchange with
water.
Ó 2007 Elsevier Ltd. All rights reserved.

Keywords: Turbo-machinery; Air cycle; Air conditioning units; Natural working fluid; Refrigeration; Pinch technology

1. Introduction directly, it is not efficient because of the small temperature

The outdoor air temperature can reach 38–40 °C in mid-
dle China in summer, while the temperature of the water
from the water supply system and underground is about
17 °C and that of the refilled underground water could be
lower. These waters are too cool for people to bath directly
at the hotel, and the bath water is usually heated by a boi-
ler. If these waters were used to cool air for air conditioning
*
Corresponding author. Address: School of Power and Energy,
Northwestern Polytechnical University, Xi’an, Shaanxi 710072, PR China.
Tel.: +86 759 2886705.
E-mail addresses: sbhou@163.com, sbhou@tom.com (S. Hou).
difference and large amount of consumed water.
The air compression refrigeration cycle was studied long
ago. Several disadvantages prevented air from being used
as a working fluid in refrigeration. These included low vol-
umetric refrigerating effect, which may result in a large
compressor, and low COP due to low efficiencies of the
compressor and expander. After CFC’s invention in the
1930s, people paid little attention to air compression
refrigeration.
Recently, as a result of the destruction of the ozono-
sphere by chlorofluorocarbons (CFC) and the pressure of
environmental protection, research upon air refrigeration
cycles has received more attention [1–3]. Optimizations of

0196-8904/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.enconman.2007.05.005
2256 S. Hou et al. / Energy Conversion and Management 48 (2007) 2255–2260
Nomenclature
B wet air pressure, Pa
D humidity ratio of wet air, g/kg(d.a.)
H enthalpy of wet air, kJ/kg(d.a.)
P pressure, Pa
T temperature, K or °C
Twet wet bulb temperature, K or °C
Wc ideal input work of compressor, kJ/kg(d.a.)
Wt ideal output work of turbine, kJ/kg(d.a.)
Wm practical work consumed by system, kJ/kg(d.a.)
gc efficiency of compressor
gt efficiency of turbine
ratio mass flow rate ratio of cooling water to dry air
air cycles are also conducted using finite time thermody-
namics (FTT) or entropy generation minimization
(EGM) [4–7].
Chen et al. investigated the cooling load versus COP
characteristics of a simple [8] and a regenerated [9,10] air
refrigeration cycle with heat transfer loss and/or other
irreversibilities. Luo et al. [11] optimized the cooling load
and the COP of a simple irreversible air refrigeration cycle
by searching for the optimum pressure ratio of the com-
pressor and the optimum distribution of heat conductance
of the hot and cold side heat exchangers for fixed total heat
exchanger inventory.
With the development of the aeronautical industry,
highly efficient axial compressors and turbines have
become a reality. At present, the stagnation isentropic effi-
ciencies of a single stage axial compressor and a turbine can
reach 0.88–0.91 [12]. High speed fans have been used in
ordinary air conditioning systems nowadays.
However, the water vapor in the working fluid was not
considered in all the above researches [1–11] on the air
compression refrigeration cycle, and the used equipments
were a centrifugal compressor and a centripetal turbine,
which have lower efficiencies than axial compressors and
turbines. The amount of water extracted from high pres-
sure wet air can reach 18–30 g/kg(d.a.), and the amount
of latent heat discharged from the vapor condensed is
about 45–75 kJ/kg(d.a.), exceeding the sensible heat from
the air of 30–50 kJ/kg(d.a.).
Hou and Li presented an axial flow air–vapor compres-
sion refrigerating system for air conditioning in 1992 [13] in
which wet air is the working fluid and an axial compressor
and turbine were used, but these have not yet attracted
people’s attention so far.
Hou and Zhang presented an axial flow air–vapor com-
pression refrigerating system for air conditioning cooled by
circulating water in Ref. [14] (2004) in which wet air is the
working fluid, an axial compressor and turbine were used
and the wet air is cooled by circulating water. The paper
proves its feasibility through performance simulation and
also indicates its advantages. These include the possibility
R gas constant, kJ/kg K
n exponent
Subscripts
air dry air
vapor water vapor in moist air
s saturated
last last time
hot hot stream
cold cold stream
w water
to simplify air conditioning systems, to reduce the amount
of initial investment of an air conditioning system and to
make air conditioned rooms more comfortable.
The aim of this paper is to present an open system,
which is an open air–vapor compression refrigeration sys-
tem for air conditioning and hot water cooled by cool
water and its performance from simulation. In this open
air–vapor refrigeration cycle, water from the water supply
system or underground is used. Thus, we could get a lower
wet air temperature before the turbine. In addition, the
cool water is heated for the bath.
2. System
Representation on enthalpy–entropy coordinates and a
circuit diagram of an open air-compression refrigeration
system for air conditioning and hot water cooled by cool
water are shown in Fig. 1.
The outdoor air at 2 is drawn into the atomizing cham-
ber, cooled to saturated air at 3 with some fine water drop-
lets and then compressed by an axial compressor. A flow of
compressed air at 4 with higher temperature, T4, and high
pressure, P4, is obtained. Then, the compressed air at 4 is
cooled to saturated air at 7 with a temperature T7 by cool
water/underground water in a surface heat exchanger after
Fig. 1. Representation on enthalpy–entropy coordinates and circuit
diagram of an open air-compression refrigeration system for air condi-
tioning and hot water cooled by cool water.
 S. Hou et al. / Energy Conversion and Management 48 (2007) 2255–2260
the axial compressor outlet. Some vapor is condensed,
and the latent heat of the vapor is discharged from 4 to
7. Then, the saturated air at 7 is expanded and cooled to
the cool air at 8 in the turbine. The cool air at 8 is then
ducted to the air conditioned rooms. The cool water is
heated in the surface heat exchanger.
Water injection before the axial compressor aims to
decrease both the temperature of the working fluid and
the polytropic exponent in the compression process. Thus,
we can save some compression work. This method has been
used in a jet engine when a fighter plane increases its speed.
However, the difference is that what is injected in a jet
engine is water, alcohol, etc. [12].
The water vapor in the compressed air can easily be
extracted by a surface heat exchanger. With the same tem-
perature, the humidity ratio of the saturated wet air at high
pressure P4 is only about P3/P4 of that at pressure P3. The
method of using compressed air to acquire dry air has been
used in some workshops in southern China.
The system above differs from a conventional air cycle
system. There are many characteristics in this air–vapor
refrigeration circle.
Firstly, an axial compressor and a turbine are used in
the above system. The characteristics of turbo-machines
are large mass flow rate and high efficiency. The other types
of compressor and expander have none of the above
advantages.
Secondly, this refrigeration system intakes precooled
wet air with fine water droplets, and some vapor is con-
densed during the air cooling from 4 to 7. The amount of
water extracted from the high pressure wet air can reach
18–30 g/kg(d.a.), and the amount of latent heat discharged
from the vapor condensed, about 45–75 kJ/kg(d.a.),
exceeds the sensible heat from the air, 30–50 kJ/kg(d.a.).
For this reason, the refrigeration load in this air–vapor
refrigeration system depends on a combination of the sen-
sible heat of air and the latent heat of vapor.
Lastly, the cool from the cooling water was used. Usu-
ally, it cannot be used.
3. Performance simulation
3.1. Wet air
The humidity ratio of wet air, D, is obtained from
P vap
D ¼ 621:98
B  P vap
The enthalpy of wet air, H, is calculated from
H ¼ 1:006t þ 0:001Dð2501 þ 1:805tÞ
The adopted relation for water vapor between satura-
tion pressure and saturation temperature, Ps = f(ts), is
selected from Ref. [15].
To calculate the saturated temperature of the wet air
from the saturated enthalpy, Eqs. (1) and (2) and Ps = f(ts)
are used.
2257
3.2. Axial compressor
During the compression process of the wet air, the fine
water droplets in the air may evaporate. Because the evap-
oration of water takes in heat, we can regard the ideal com-
pression process of the wet air in the compressor as a
polytropic process. Therefore, we can obtain the ideal work
of the compressor per kilogram dry air, Wc, from
"  n1 #
n p4 n
Wc ¼ ðRair þ 0:001DRvapor ÞT 3 1  ð3Þ
n1 p3
in which n is the polytropic exponent for the compression
process.
The practical work consumed by the axial compressor is
Wc/gc in which gc is the thermal efficiency of the
compressor.
3.3. Turbine
The saturated air with a pressure of P7 and a tempera-
ture of T7 before the turbine has been dehumidifed in the
surface heat exchanger by cooling water. At point 7, the
amount of vapor included in the saturated air is very small,
about P3/P7 of the amount included in saturated air at P3.
Thus, the water condensed in the air is in fog. Nevertheless,
expansion of the saturated air in the turbine cannot be
regarded as an adiabatic expansion of an ideal gas. With
the decrease of the wet air pressure in the turbine, the tem-
perature of the wet air decreases, and some heat is dis-
charged during the condensation of some water vapor.
The heat discharged may cause increases in both the tem-
perature of the turbine outlet and the work done in the
expansion.
For this problem, we can imagine that no phase change
exists and that there is some heat added to the wet air dur-
ing the expansion process when we calculate the work done
by the expansion process. According to the above assump-
tion, this problem can be simplified to a problem of the
polytropic expansion of an ideal gas. Consequently, we
can obtain the ideal work done by the expansion, Wt,
through iteration and then obtain the real work generated
by the turbine and the temperature of the turbine outlet.
The following are the steps to calculate Wt and T8.
1. Determine Ps7 from T7, D7 from Ps7 and P7, and H7
ð1Þ
from D7 and T7.
2. Get gas constant R for the saturated air at 7 by using
the formula:
ð2Þ
R ¼ 0:001ð287 þ 0:461D7 Þ: ð4Þ
3. Calculate the initial Wt to iterate according to an adi-
abatic expansion of an ideal gas.
4. Calculate the enthalpy of the saturated air at 8 by using
H 8 = H 7  W t.
5. Determine T8, Ps8 D8 using H8 and P8.
2258 S. Hou et al. / Energy Conversion and Management 48 (2007) 2255–2260

n logðP 7 =P 8 Þ q2
6. Obtain ¼ logðT : ð5Þ COP ¼ ð11Þ
n1 7 =T 8 Þ Wm
n
7. W t;new ¼ n1 RðT 7  T 8 Þ: ð6Þ
W t þW t;new
8. Wt ¼ 2
. 5. Results
9. H 8 = H 7  W t.
10. T8last = T8. There are many factors that may influence the COP of
11. Determine T8, Ps8, D8 using H8 and P8. this refrigerating system for air conditioning and hot water
12. If abs (T8last  T8) < 0.1 go to step 13. otherwise, go to cooled by cool water. These include the pressure ratio of
step 5. the axial compressor, P4/P3, the efficiencies of the axial
13. Obtain the enthalpy using the formula, H8 = H7  compressor and turbine, the wet bulb temperature of the
W t Æ g t. atmosphere Twet and the cool water temperature.
14. Determine T8, Ps8, D8 using H8 and P8. During simulation, the pressure ratio of the axial com-
pressor was varied from 1.6 to 2.5, the wet bulb tempera-
3.4. Surface heat exchanger ture of the outdoor air from 20 to 30 °C and the cooling
water temperature from 15 to 25 °C.
Pinch technology is used in the analysis of heat exchange There is 300 Pa pressure loss before the axial compres-
in the surface heat exchanger, and the temperature differ- sor, 300 Pa between the axial compressor and turbine and
ence at the pinch point is 6 °C. Pinch technology is a graph- 600 Pa after the turbine.
ical method of identifying technically and economically Some encouraging results are illustrated in Figs. 2–6.
interesting energy efficiency measures. The minimum cool- The results are calculated from a lower pressure ratio of
ing and heating demands in the system can thereby be the axial compressor than that in Ref. [14].
determined, together with the net heat for each tempera- The sensitivity of COP to the efficiencies of the axial
ture level. The concepts and methodology of pinch technol- compressor and turbine is illustrated in Fig. 2. The lines
ogy are well explained in the works of Linhoff et al. [16], in Fig. 2 are the COP lines of an open air compression
Eastop and Croft [17], Linnhoff [18] and Mubarak Ebra- refrigeration system for air conditioning and hot water
him [19]. cooled by cool water when T7 = 296 K and the efficiencies
The cold stream is cooling water, and the enthalpy is of the axial compressor and turbine are 90%, 88%, 86% and
determined by Hcold = cpwt. 80%, respectively. From Fig. 2, the efficiencies of the axial
The hot stream is hot wet air with a pressure of P4, and compressor and the turbine influence the COP heavily. The
the enthalpy is determined by COP is higher because of a lower T7 than we can get than
in Ref. [14]. Also, the optimum pressure ratio of the axial
H hot ¼ 1:006t þ 0:001Ds ð2501 þ 1:805tÞ ð7Þ compressor declines. That is, the number of stages could
in which be decreased to two.
Although the COP is sensitive to the efficiencies of the
P s ðtÞ axial compressor and turbine, these cycles are feasible.
Ds ¼ 621:98 ð8Þ
P 4  P s ðtÞ Firstly, this new turbo-machinery works near the design
The optimum mass flow rate ratio of cool water to dry point, and the efficiencies of axial compressors and turbines
air can be obtained from the hot and cold curves according
to pinch technology.
5.5
nt=nc=0.9
4. Performance 5.0
nt=nc=0.88
4.5 nt=nc=0.86
The refrigerating capacity per kilogram dry air, q2, can
be determined by the enthalpy difference between the inlet 4.0 nt=nc=0.80
of the compressor and the outlet of the turbine by using the
COP

3.5
following formula
3.0
q2 ¼ H 3  H 8 ð9Þ
2.5
The work consumed by the refrigeration cycle is calcu-
lated by 2.0

Wc 1.5
Wm ¼  W t  gt ð10Þ 20 22 24 26 28
gc
o
T3 / Twet, C
The COP of this refrigeration system is calculated by the
following formula. (The work consumed by the cool water Fig. 2. The sensitivity of COP to efficiencies of the axial compressor and
system is not included in Wm.) turbine when T7 = 23 °C.
 S. Hou et al. / Energy Conversion and Management 48 (2007) 2255–2260 2259

60

50 T4 4.5
T7
40
T8
30 4.0
T, oC

COP
20

3.5
10

0 28
3.0 27
26
-10 25
20 24

C
t ,o
20 22 24 26 28 21 23
o 22 22
T3 / Twet, C

we
23

T
T ,o 21
24

3 /
7 C 20
Fig. 3. The simulated temperatures of an open air-compression refriger-

T
25
ation system for air conditioning and hot water cooled by cool water when
gc = 0.90 and gt = 0.90 and T7 = 23 °C. Fig. 5. The COP of an open air-compression refrigeration system for air
conditioning and hot water cooled by cool water when gc = 0.90 and
gt = 0.90.

100

80

90
60
q , kJ/kg(d.a)
1
q , kJ/kg(d.a) 80
2
q2 , kJ/(d.a)

40 wm, kJ/kg(d.a)
70
20

60
0
20 22 24 26 28 28
50 27
T3 / Twet,oC 26
25

C
t ,o
20 24
21 23
Fig. 4. The simulated q1, q2, Wm of an open air-compression refrigeration

we
22 22

T
system for air conditioning and hot water cooled by cool water when 23

3 /
T ,o 21
24
gc = 0.90 and gt = 0.90 and T7 = 23 °C. 7 C
25 20 T

Fig. 6. The q2 of an open air-compression refrigeration system for air

are very high at the design point, about 0.89–0.91. Sec-
ondly, the intake air is clean and without dust, and there-
fore, the efficiencies of the axial compressor and turbine
will not drop greatly while working. Thirdly, there is no
great complexity of the combustion chamber and high tem-
perature turbine in the turbo-machinery. Consequently, it
is much easier to accomplish than many people imagine.
Lastly, the efficiencies of the axial compressor and turbine
have room for improvement with additional design
measures.
The simulations of an open air–vapor compression
refrigeration system for air conditioning and hot water
cooled by cool water when gc = 0.90 and gt = 0.90 and
T7 = 296 K are illustrated in Figs. 3 and 4. Fig. 3 gives
the relations of the temperature after the compressor, T4,
the temperature before the turbine, T7, and the tempera-
ture after the turbine, T8 to the temperature before the
conditioning and hot water cooled by cool water when gc = 0.90 and
gt = 0.90.
compressor, T3 (Twet). Fig. 4 gives the relations of the
refrigerating capacity per kilogram dry air, q2, the dis-
charging heat per kilogram dry air, q1, and the work con-
sumed by the refrigeration system, Wm, to the
temperature before the compressor, T3 (Twet). The relation
of COP to the temperature before the compressor, T3 (Twet)
can be located in Fig. 2.
The simulations of an open air–vapor compression
refrigeration system for air conditioning and hot water
cooled by cool water when gc = 0.90and gt = 0.90 and
T7 = 20–25 °C are illustrated in Figs. 5 and 6. Fig. 5 gives
the variation of COP with T3 (Twet) and T7. Fig. 6 gives the
variation of refrigerating capacity per kilogram dry air, q2,
with T3(Twet) and T7.
2260 S. Hou et al. / Energy Conversion and Management 48 (2007) 2255–2260
70
4 rests mainly on gc and gt. The temperature of cool water
60
4'
o
4': T3=20 C
50 o
4: T3=27 C
T, oC
40
Outlet
30
7 8. Pinch technology is a good method to analyze the heat
Cool Water
20
Inlet
10
40 60 80 100 120 140
H, kJ/kg
Fig. 7. The hot and cold curves of dehumidification process in surface
heat exchanger when ratio = 1.
Fig. 7 gives the hot and cold curves of the cooling and
dehumidification process in the surface heat exchanger
when the mass flow rate ratio of cool water to dry air
is one. In Fig. 7, the temperature difference at the pinch
point is 6 °C, and there are two hot lines, which is when
T3 = 20 °C and T3 = 27 °C, respectively. From Fig. 7, we
can get hot water with a temperature of about 33–40 °C
and the same mass flow rate as the dry air mass flow
rate.
6. Conclusions
This study shows the feasibility of an open air–vapor
compression refrigeration system for air conditioning and
hot water cooled by cool water. The calculation results
show:
1. The open air–vapor compression refrigeration system
for air conditioning and hot water cooled by cool water
given in this paper could use the cool in cool water,
which could not be used to cool air directly. Also, the
heat rejected by this system could be used to heat water.
2. Humid air is a perfect working fluid for refrigeration in
central air conditioning.
3. To compare with air–vapor compression refrigeration
system for air conditioning cooled by circulating water
[14], this system is suitable to be employed in the place
where we could get cold water easily.
4. According to the axial compressor performance charac-
teristics among rotation speed, mass flow rate and pres-
sure ratio, we can easily control its mass flow rate and
pressure ratio and the refrigerating capacity by adjusting
the rotation speed of the axial compressor.
5. The COP of this refrigeration air conditioning circle var-
ies with the wet bulb temperature of the atmosphere.
The higher the wet bulb temperature of the atmosphere,
the higher is the COP of this refrigerating air condition-
ing cycle.
6. The COP of this refrigeration air conditioning system
will also affect it heavily. Although the COP is sensitive
to gc and gt, an open air–vapor compression refrigera-
tion cycle for air conditioning and hot water cooled by
cool water is still feasible.
7. The optimum pressure ratio of the axial compressor
could be lower in this system. This results in a fewer
number of stages of the axial compressor.
exchange of wet air cooling.
References
16 0
[1] Hamlin S, Hunt R, Tassou SA. Enhancing the performance of
evaporative spray cooling in air cycle refrigeration and air condition
technology. Appl Therm Eng 1998;18(10):1139–48.
[2] Mc Cormick JA. Progress on the development of miniature
turbomachines for low capacity reverse Brayton cryocooler. Proceed-
ings of nineth international cryocooler conference. 1996:1;255–67.
[3] Saygin H, Sisman A. Brayton refrigeration cycles working under
quantum degeneracy conditions. Appl Energy 2001;69(2):77–85.
[4] Sieniutycz S, Salamon P. Advances in thermodynamics. Finite-time
thermodynamics and thermoeconomics, vol. 4. New York: Taylor &
Francis; 1990.
[5] Bejan A. Entropy generation minimization. Boca Raton, FL: CRC
Press; 1996.
[6] Berry RS, Kazakov VA, Sieniutycz S, Szwast Z, Tsirlin AM.
Thermodynamic optimization of finite-time processes. Chiches-
ter: Wiley; 1999.
[7] Chen L, Wu C, Sun F. Finite-time thermodynamics optimization or
entropy generation minimization of energy systems. J Non-Equilib
Thermodyn 1999;24(3):327–59.
[8] Chen L, Wu C, Sun F. Cooling-load versus COP characteristics for
an irreversible air refrigeration cycle. Energy Convers Manage
1998;39(1/2):117–25.
[9] Chen L, Wu C, Sun F. Finite-time thermodynamic performance of an
isentropic closed regenerated Brayton refrigeration cycle. Int J Energy
Environ Econ 1997;4(4):261–74.
[10] Chen L, Zhou S, Sun F, et al. Performance of heat-transfer
irreversible regenerated Brayton refrigerators. Phys D: Appl Phys
2001;34(5):830–7.
[11] Luo J, Chen L, Sun F, Wu C. Optimum allocation of heat-exchanger
inventory of irreversible air refrigeration cycles. Phys Scripta
2002;65(5):410–4.
[12] Northwestern Polytechnic University etc, Principle of gas turbine
aero-engine, National Defense Industry Publisher, Beijing, 1981.
[13] Hou SB, Li HC 1992, Chinese patent, 92109202.4.
[14] Hou SB, Zhang HF. An axial-flow air-vapor compression refriger-
ating system for air conditioning cooled by circulating water.
ASHRAE Transactions 2004;110(Pt 2):125–9;
Hou SB, Zhang HF. Annual Meeting – Technical and Symposium
Papers. American Society of Heating, Refrigerating and Air-Condi-
tioning Engineers; 2004. p. 78–82.
[15] ASHRAE Handbook – fundamentals volume, 2001, ASHRAE 1791
Tulie Circle, Atlanta, GA 30329, United States of America.
[16] Linnhoff B et al. User guide on process integration for efficient use of
energy. UK: AIChE; 1982.
[17] Eastop TD, Croft DR. Energy efficiency for engineers and technol-
ogists. UK: Longman Scientific and Technical/John Wiley & Sons;
1990.
[18] Linnhoff B. Pinch analysis – a state-of-the-art overview. Trans Icheme
Part A 1993;71(September):503–22.
[19] Mubarak Ebrahim A. Pinch technology: an efficient tool for chemical
plant energy and capital cost saving. Appl Energy 2000;65(1–4):45–9.