Accepted Manuscript

Title: Optimum compressor cylinder volume ratio for two-stage compression

air source heat pump systems

Author: Shuang Jiang, Shugang Wang, Xu Jin, Yao Yu

PII: S0140-7007(16)30032-9
DOI: http://dx.doi.org/doi: 10.1016/j.ijrefrig.2016.03.012
Reference: JIJR 3290

To appear in: International Journal of Refrigeration

Received date: 2-2-2016

Revised date: 16-3-2016
Accepted date: 19-3-2016

Please cite this article as: Shuang Jiang, Shugang Wang, Xu Jin, Yao Yu, Optimum compressor
cylinder volume ratio for two-stage compression air source heat pump systems, International
Journal of Refrigeration (2016), http://dx.doi.org/doi: 10.1016/j.ijrefrig.2016.03.012.

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service
to our customers we are providing this early version of the manuscript. The manuscript will
undergo copyediting, typesetting, and review of the resulting proof before it is published in its
final form. Please note that during the production process errors may be discovered which could
affect the content, and all legal disclaimers that apply to the journal pertain.
 Optimum compressor cylinder volume ratio for two-stage compression air source heat
pump systems
Shuang Jianga, Shugang Wanga, , Xu Jinb and Yao Yuc
a
Faculty of Infrastructure Engineering, Dalian University of Technology, Dalian 116024, China
b
School of Energy and Power Engineering, Northeast Dianli University, Jilin 132012, China
c
Department of Construction Management and Engineering, North Dakota State University, Fargo, ND 58108-6050, USA

Highlights
A method to determine the optimum compressor cylinder volume ratio is developed.
The appropriate design cylinder volume ratio is between 2.0 and 2.1.
The appropriate outdoor design temperature is between -4C and -8C.
A proper design cylinder volume ratio will save investment and operation costs.

Abstract
With the outstanding performance potential and excellent modulation capacity, the Two-Stage Air Source Heat
Pump (TSAHP) may become a substitute for fossil fuels for space heating in cold climates. In TSAHP, the design
cylinder volume ratio is closely related to the evaporating and condensing conditions and will determine the
average heating seasonal coefficient of performance ( C O P ). In this paper, based on weather data and a two-step
optimization approach, a volume-ratio selection method was presented. Eight typical cities in the cold region of
North China were selected for demonstrating the impacts of the optimum cylinder volume ratio at different
outdoor design temperatures on C O P . The results show that, the appropriate outdoor design temperature is
between -4C and -8C, and the corresponding design cylinder volume ratio is between 2.0 and 2.1. The COP
with the volume ratio of 2.0 is 10.3% and 17.6% higher than that with the volume ratio of 3.0.
Keywords: Two-stage system; Volume ratio; Air-to-water heat pump; Optimum; Subcooler

Corresponding author: Tel./fax.: +86-411-84706407; E-mail address: sgwangln@aliyun.com

Page 1 of 19
Nomenclature
Nomenclature Subscripts
a1-a4, c1-c4 coefficients in the efficiency formulas (-) a air
Bo boiling number (-) cy cylinder
COP coefficient of performance (-) d design
h specific enthalpy (kJ kg-1)) dis discharge
f frequency (Hz) el electric
refrigerant mass flow rate (kg s-1) eq euaivalent
Nu Nusselt Number H high-stage compressor
n polytropic exponent (-) h heating
p pressure (kPa) in input
Pr Prandtl number inj injection
Q heat transfer rate (kW) L low-stage compressor / liquid
Re Reynolds number (-) m intermediate / mass
Rm refrigerant mass flow ratio (-) max maximum
Rcy compressor cylinder volume ratio (-) min minimum
T temperature (C) n indoor
V volume (m3) rat ratio
work (kW) suc suction
w specific work (kJ kg-1) th theoretical
x refrigerant quality (-) T total
heat exchanger subcooling parameter (-) vol volume
efficiency (-) w wall
specific volume (m3 kg-1)
viscosity (Pas) Superscript
density (kg m-3) * parameters at base frequency

1. Introduction
Applying heat pumps to space heating for residential buildings in cold regions will reduce the combustion of
gas, oil and other fossil fuels and the emissions of greenhouse gases. An Air Source Heat Pump (ASHP), which
uses the widely available air as heat source, is more easily deployed and applied than other types of heat pumps,
such as geothermal heat pumps. Nevertheless, in the cold areas where space heating is required, the outdoor
temperatures are extremely low, and could be below -20C in the northern parts of China. In this case, the
evaporating temperature of the ASHP decreases and the temperature difference between the heat source and sink
increases, which result in the deteriorated heating capacity, declining coefficient of performance and even
abnormal shut down due to the high discharge temperature.
Different approaches regarding how to improve the thermodynamic performance of high temperature lift heat
pumps were presented, experimentally compared and deeply analyzed by Zehnder (2004). These approaches
included one stage compression with intermediate injection of saturated vapor, two-stage economizer cycle and
booster concept, and liquid subcooling through an auxiliary cycle. Zehnder concluded that, the two-stage
compression cycle represents the next generation of heat pump due to its modulation capacity and the potentially
higher system performance. Tian and Liang (2006) proposed a two-stage compression variable frequency air
source heat pump. The low- and high-stage variable frequency compressors were used, which had the capacities
of 5HP and 3HP, respectively. The experimental results show that the COP is higher than 2.0, the discharge
temperature of the high-stage compressor is below 120C, and the heating capacity is able to meet the heating
load when the condensing and evaporating temperature are 50C and -25C, respectively. The tested system
works in a stable state and the lubricant return is performed well, and is able to meet the heating demand without
auxiliary heat sources in cold regions where the outdoor temperature is not lower than -18 C. Bertsch and Groll
(2008) simulated, designed, constructed and tested a two-stage air-source heat pump for water and air heating,
which is able to operate at ambient temperatures between -30C and 10C with the supply water temperatures of
up to 50C. This system uses a subcooler as the inter-stage structure and could be operated in either single- or
two-stage mode. At the same ambient temperature, two-stage mode operation approximately doubles the heating

Page 2 of 19
capacity compared to the single-stage mode operation. The discharge temperatures of the compressors in the
two-stage mode maintain below 105C at all times. Based on this technology, Caskey et al. (2012) and their three
industrial partners designed and fabricated a new air-source heat pump system optimized for cold climates, which
was installed in a military barrack for a field demonstration. A simulation model was developed in EES in order
to predict the designed heat pump performance at different ambient conditions. The EES results were
incorporated with a TRNSYS model that was used to estimate the military barrack building load. This coupled
model has the ability to predict the heating energy consumption of the heat pump system according to different
capacities determined by ambient conditions or weather data. The predicted results show that the heat pump
performs well in cold climates and the COPs constantly maintain above 3.5 over the entire heating season.
Additionally, the savings of a cold climate heat pump compared to a natural gas furnace was anticipated to be
over 30% for primary energy consumption and CO2 emissions with a 25% energy cost savings on a yearly basis.
As shown in the most recent report (Hutzel and Groll, 2013), the primary energy and cost savings achieved
during the field demonstration did not match the simulation results; however the results are attractive, and the
commercialization work has been moving on with UNICO, who has licensed two patents owned by Purdue
University. Jin et al. (2008) designed and tested a two-stage vapor compression system with a flash tank as the
inter-stage structure. Both compressors were fixed-speed rotary types, in which R22 was used as the refrigerant.
The experimental results show that the system is capable of providing 50C hot water when the outside
temperature is -20C. The COP is 1.76; the heating capacity of the system is 4.71 kW, and the exhausted
temperature of the compressor is below 100C. Jin (2013) built up an experimental system with a variable-speed
compressor as the low-stage compressor and two 2000-step electronic expansion valves as throttling components.
In this system, the inter-stage structure was known as a subcooler, and a vertical brazed plate heat exchanger was
employed. Besides, a dynamic coupling model between the low- and high-stage compressor was established and
experimentally verified. As shown in the simulation and experimental results, the heating capacity of the system
is very sensitive to the variation of the intermediate pressure, whereas the COP is insensitive to that. The optimal
ratio of the intermediate pressure to the geometric mean of the condensing and evaporating pressure is between
1.1 and 1.3.
When a Two-Stage Air Source Heat Pump (TSAHP) is designed, the selection of the cylinder volume ratio of
the two compressors (Rcy) is one of the most important tasks, which will not only influence the initial cost but also
have a significant impact on the operation cost. The optimum Rcy, however, is closely related to the evaporating
and condensing conditions and will vary with these conditions. For a TSAHP unit, the temperature of the heat
source, i.e. the ambient air temperature, may fluctuate significantly with time. The selection of the outdoor design
temperature, upon which the optimum Rcy is determined, is a topic worth further studying. Heo et al. (2013)
simulated a vapor-injection heat pump with a twin rotary compressor and a flash tank, and obtained the optimum
Rcy at different compressor frequencies and ambient temperatures (0C, 5C and 10C). They pointed out that the
optimum Rcy does not vary much with operating conditions, such as ambient temperature and compressor
frequency. However, the effect of compressor frequency on the efficiency was not considered in the compressor
model in which the loss and efficiency method was used. Besides, the optimum Rcy at temperatures below 0C
was not analyzed. The analysis and optimization of subcritical two-stage vapor injection heat pump systems were
done by Redn et al. (2014). The results show that, for an ideal cycle at the evaporating temperature of -8C and
the condensing temperature of 65C, the optimum displacement ratio of the compressor corresponding to the
maximum COP is 0.5; for an actual system, the optimum value is 0.6, at which the maximum seasonal COP in the
climate zone A defined by European Norm En 14825 is reached. Ko et al. (2013) investigated the performance of
a two-stage vapor injection heat pump equipped with a twin rotary compressor under ARI conditions (ANSI/ARI
Standard 540-2004). Three different cylinder volume ratios, 0.59, 0.68 and 0.82, were tested. The results show
that, the performance of the unit with a cylinder volume ratio of 0.59 is superior to those with other ratios. Baek et
al.(2014) measured the heating performances of the vapor injection CO2 cycles including the flash tank vapor
injection and sub-cooler vapor injection cycles, by adjusting the cylinder volume ratio of the twin rotary
compressor at the ambient temperature of -15C. It was pointed that the optimum cylinder volume ratio is 0.7 in
order to achieve the maximum COP with the designed heating capacity. Li and Yu (2016) investigated the
influences of the thermal conductance allocation ratio on system COP and heating capacity using a lumped
parameter method for a flash tank based air source heat pump system. The analysis results indicate that there exist
optimal volumetric displacement ratios for maximizing the system COP under the given conditions. Abdelaziz

Page 3 of 19
and Shen (2012) presented a step-by-step approach for the design optimization of cold-climate heat pumps.
Optimal designs for two different system configurations (a flash tank cycle and an economizer cycle) were
presented and analyzed. Results indicate that the optimization of system plays a significant role in the system
design process. It was also pointed that a further investigation involving cold-climate heat pump design
optimization is required, in order to assess the technical and economic feasibilities of various systems at different
climate conditions.
Unlike the single-stage system or vapor injection system with a single compressor, the TSAHP system consists
of two independent compressors with respective electric motors. Hence, it is able to apply the variable frequency
technology to adjust the displacements of the two compressors, so that the heating demand is satisfied or/and an
optimum COP is obtained (Jiang et al., 2015; Li and Yu, 2016). As far as we know, in the previous studies, the
selection of compressor cylinders was always performed under specified conditions without a consideration of
the minimum operation cost covering the whole service period of the unit. Moreover, most of the current
researches related to the optimum volume ratio are about vapor injection system with a twin rotary compressor,
and studies regrading a TSAHP system are very limited.
This study presents a volume ratio selection method based on meteorology data and a two-step optimization
approach, in which a simplified heating load calculation model and a general two-stage heat pump model are
utilized. The average COPs over the entire heating season at different design temperatures are compared and
analyzed, whose results are able to be used for the determinations of the optimum design temperature and the
corresponding Rcy. The method proposed in this paper will not only give guidance for the proper selection of
compressors for a TSAHP system, but also provide a reference for equipment selection in heating mode at
different climate conditions.

2. Model development
2.1. Two-stage compression cycle
Two-stage vapor compression cycle varies depending on different inter-stage configurations. The common
inter-stage configurations include flash tank, flash intercooler, and subcooler. The subcooler cycle is more easily
controlled compared with others and has been implemented in some commercial products. In this paper, a
two-stage compression air source heat pump with a subcooler as the inter-stage configuration has been studied,
which can be operated in either single or two-stage mode. Moreover, the two compressors in the system are
characterized by variable frequency. The system and pressure-enthalpy (ln(p)-h) diagram for this two-stage cycle
are presented in Fig. 1. For this system, the solenoid valve is switched on for the two-stage mode, and off for the
single-stage mode. In the single mode, the 2nd EEV is closed and either of the two compressors can be selected as
the operation compressor. The 4-way valve is opened only in cooling or defrosting operation.

In a previous study, a general model for two-stage heat pump systems was presented and experimentally
validated (Jiang et al., 2015). For the general model, the heating COP for the two-stage cycle mode of the studied
system presented in Fig.1 can be expressed with Equation (1).

h4 h5
C O Ph , (1)
w th ,L w th ,H
Rm
e l,L e l,H

where the compressor theoretical specific work, wth, is the reversible work in a polytropic process defined in
(engel and Boles, 2008), i.e.,
n n 1 / n
w th p s u c s u c p 1 ; (2)
n 1
ra t

Rm is defined as the ratio of mass flow rate of the low-stage compressor to that of the high-stage compressor, i.e.,

Page 4 of 19
 mL f L V c y ,L 1 v o l,L f L 1 v o l,L
Rm = R cy . (3)
mH f H V c y ,H 3 v o l,H f H 3 v o l,H

The volumetric and electric efficiencies at an arbitrary frequency are expressed by using Equation (4) and
Equation (5) (Jiang et al., 2014), respectively, in which the superscript * represents the parameters at a base
frequency,

c 1 a
*

f

2
p ra t
1/ n
vol f f c2 f f a2
* * * *
1 1
, (4)
f

f f 1
2
f c2 f
* *
c1

el el . (5)
f f 1
2
f c4 f
* *
c3

where el is the compressor electric efficiency at a base frequency. For small capacity rotary compressors, el
is determined by using Equation (6) (Jiang et al., 2014),

a* a* p ra t
1/ n

3 4
w h e n p ra t r0
el . (6)
p ra t p ra t
2
w h e n p ra t r0
a4 a5
* * *
a6

The discharge temperature and the heating capacity are calculated by using Equation (7) and Equation (8),
respectively,
n 1

T d is = T s u c p ra t n , (7)

Q h m H h 4 h5 . (8)

The specific enthalpy of the high-stage compressor suction refrigerant is calculated with Equation (9),

hD h5 h 2 h5 h L h D
h3 , (9)
h D h5 h5 hL

where is the inter-stage subcooling parameter, representing a value that is not equal but very close to the heat
exchanger effectiveness, and is defined as follows,

h5 h6
. (10)
h5 h L

Rm can be expressed with by using Equation (11). It is noted that when condensing conditions is kept constant,
the value of Rm calculated with Equation (11) will be only determined by the intermediate pressure and .
h D h5
Rm (11)
h D h5 h5 hL

The determination of the intermediate pressure (pm) is the key for solving the two-stage compression problem.
It is noticed that, for a stable system, Rm calculated with Equation (11) must be equal to that with Equation(3).
Therefore, if the compressor parameters, such as cylinder volume, frequency, etc., are regarded as known
constants, Equation (12) can be established for the calculation of pm. The root of this equation is just the value of
pm, which can be found using iteration methods.
f L V c y ,L 1 v o l,L h D h5
f1 pm 0 (12)
f H V c y ,H 3 v o l,H h D h5 h5 h L

In addition, if the heating capacity is regarded as one of the known constants, and the compressor cylinder
volume or operation frequency are the parameters to be solved, Equation (13) may be used to solve the
intermediate pressure pm.

Page 5 of 19
 f L V c y ,L 1 v o l,L h D h5
f2 pm 0 (13)
Q h / h 4 h5 h D h5 h5 h L

As the equation of pm is established, i.e., Equation (12) or Equation (13), and there exist many root-finding
algorithms (Press H. et al., 1992), the determination of f1(pm) or f2(pm) with the given pm plays a significant role in
obtaining the correct solution of pm. A flow chart is shown in Fig. 2, indicating the steps to calculate f1(pm) or f2(pm)
with the given pm, which is the basis of the program to solve the two-stage system and the following optimum
problem. The flow chart of the two-stage compression heat pump system model is shown in Fig. 3, in which the
root-finding procedure may be implemented with any of the one-dimensional root-finding routines presented by
Press H. et al. (1992) without using the derivative of the function.

2.2. Heat exchangers

The zone-based lumped-parameter model is applied to the three heat exchangers, including air-to-refrigerant
evaporator, refrigerant-to-water condenser and refrigerant-to-refrigerant subcooler. Both the condenser and
subcooler are Vertical Brazed Plate Heat Exchangers (VPHE), and the finned-tube heat exchanger is employed as
the evaporator.
For two-phase evaporating process in a subcooler, heat transfer coefficient correlations obtained from Han et al.
(2003) is employed, i.e.,

N u C R e e q B o e q P rL
m 0 .3 1 3
, (14)

where, the equivalent Reynolds number Reeq, the equivalent boiling number Boeq, and m can be found in the
study of Han et al. (2003). A linear correction is performed for the coefficient C according to experimental data
(Jiang et al., 2014), duo to the small equivalent Reynolds number in the subcooler as a result of the minimized
injection refrigeration mass flow rate.
For the finned-tube heat exchanger, the heat transfer coefficient correlations for the air side (Equation (15)) and
the two-phase refrigerant side (Equation (16)), obtained from Kang et al.(1994) and Gungor and Winterton (1986)
respectively, are employed,
0 .0 8 8 7 0 .1 5 9 0
s N s2
N u 0 .9 8 2 R e
0 .4 2 4
, (15)
d3 d3

T P E L S nb , (16)

where the two-phase heat transfer coefficient, TP, is the weighed sum of the liquid heat transfer coefficient L
and the poling boiling heat transfer coefficient nb; E and S are enhancement factor and suppression factor,
respectively; L is given by the classic Dittus-Boelter equation (Bergman et al., 2011) for liquid only flow in the
duct, i.e.,

L = 0 .0 2 3 R e L P r L
0 .8 0 .3
kL / d . (17)

The expression for nb was proposed by Cooper (Cooper, 1984),

0 .5 5
lo g 1 0
0 .5
nb = 5 5 p r
0 .1 2 0 .6 7
pr M q , (18)

where pr, M, q denote the reduced pressure, molecular weight, and heat flux, respectively.
For the two-phase refrigerant condensing process in VPHE, the heat transfer coefficient correlation obtained
from Yan et al. (1999) is employed, i.e.,
0 .4 0 .3
N u = 4 .1 1 8 R e e q P r L . (19)

For the single phase refrigerant and water flow in VPHE, the heat transfer coefficient correlation obtained
from Hsieh and Lin (2003) is employed, i.e.,

Page 6 of 19
 w
0 .1 4
N u 0 .2 0 9 2 R e
0 .7 8 1 3
Pr . (20)

2.3. Simplified heating loads

When heat pump systems are used in heating model during winter seasons and the indoor heating source is
assumed to be uniformly distributed, the space heating load is approximately linear to the temperature difference
between the indoor and outdoor environments, due to the fact that, the temperature difference between the inside
and outside is significant and the heat is transferred in one direction from the inside to outside (Wang et al., 2008;
Zehnder, 2004). Therefore, at the moment of i, the heating load of a building is,

Q i K F t n t a ,i , (21)

where, tn, ta,i, K, and F denote indoor air temperature, outdoor air temperature at time i, comprehensive heat
transfer coefficient, and the area of the building envelope, respectively. K, related to the building envelope, can be
obtained from the given index of building heat loss and the corresponding outdoor air temperature (both can be
found in local design standards for building energy efficiency).

The average heating seasonal coefficient of performance ( C O P ) of the heat pump system can be determined
by dividing the total heat supplied by the total power consumed, and is expressed as,

Qi Qi t n
t a ,i
COP , (22)
wi Qi t n t a ,i
C O Pi

C O Pi

where wi is the system power consumption at time i, including both of the power consumption of the heat pump
unit and the auxiliary electric heating power input. It can be inferred from Equation (22) that, the simplified
heating load makes C O P independent from K and F but only constrained by tn, ta,i and COPi. Furthermore,
based on the heat pump model presented in the Section 2.1, the outdoor air temperature is the only external factor
that affects COPi when the condensing condition is assumed to be constant; therefore, C O P only depends on tn
and ta,i.
2.4. Model validation
The simulation results of the two-stage compression cycle model are compared with experimental data, and the
experimental devices were described in the previous study (Jiang et al., 2015). Although the finned-tube coil heat
exchanger was not involved in the previous study, the results are applicable in the current study if the evaporating
temperature is set at the same level. Besides, the purpose of the finned-tube coil heat exchanger model employed
in the current study is to reflect the impact of the outdoor air temperature on the evaporating temperature, and
therefore the performances indices of the units equipped with compressors of different Rcy are quantitatively
comparable.
The experiments were performed with the evaporating and condensing temperatures of -20C and 40C,
respectively, as well as varied frequencies of the low-stage compressor and the openings of the 2nd EEV.
According to ASHRAE (1996), an uncertainty analysis was carried out for the experimental data. The heating
capacity and the COP were evaluated based on the measurements of temperatures, volume flow rates and the
power inputs. The uncertainties of the heating capacity and the COP were estimated by using the single-sample
analysis with the uncertainties propagate principle per ASHRAE (1996). The uncertainty analysis results are
reported in Table 1. The relative uncertainties are acceptable according to the research of Ertesvg (2011).

Fig. 4 displays the comparison results between simulation and experiment, the abscissas of which show the
openings of the second electronic expansion valve (the injection expansion valve) during the experiments, and
Table 2 shows the deviations of the simulation results, which together indicate a reasonable agreement between
the simulation results and the experimental data. Therefore, the current system model described in this paper is
acceptable and can be used in the following studies.

Page 7 of 19
2.5. Problem description
2.5.1 Assumptions
Due to the facts that the emphasis of this study is placed on the optimum cylinder volume ratio of the
compressors in a two-stage compression cycle under different outdoor design temperatures and there exist many
studies about optimum designs for ASHP heat exchangers (Domanski et al., 2004, 2005; Wu et al., 2008), the
parameters of the heat exchanger selected in this study are based on a common approach, in which under design
conditions, the temperature difference between the evaporating and outdoor air temperatures is set to be 8C and
5C is used representing the temperature difference between the condensing and outlet water temperature.
Detailed information of the heat exchangers can be found in Table 3. Besides, it is assumed that, under off-design
conditions, the refrigerant flow rates of each loop and the air flow rates of the finned-tube evaporators are
uniform. The water temperature difference between the inlet and outlet of the condenser is kept constant as 5C,
and the outlet water temperature is set to be 45C. Other assumptions are listed as follows,
(1) The refrigerant employed is R410A, and the thermodynamic property of the refrigerant is obtained from
REFPROP (Lemmon et al., 2013);
(2) The injection refrigerant is a saturated gas at the intermediate pressure (point D);
(3) The two variable-frequency compressors are adopted; the coefficients for the calculations of the volumetric
and electric efficiencies are obtained from the experimental data (Jiang et al., 2015); the compressor frequency
range is between 15 Hz and 120 Hz, and its operation frequency is reduced so that the discharge temperature is
always kept below 125C;
(4) Only the high-stage compressor is operating in a single-stage mode;
(5) The subcooling parameter of the subcooler is set to be 0.80, and the two compressors are operating at their
base frequencies under design conditions;
(6) The degree of superheat at the low-stage compressor suction line is 5C, and the degree of subcooling at the
condenser outlet is 2C;
(7) A continuous heating system is established, in which, when the heating capacity of the unit is insufficient,
auxiliary electric heaters are used, and when the heating capacity is excess, the unit will be operated at its
minimum capacity;
(8) The power consumptions of the fan and pump are not considered.

2.5.2 Description of the optimum problem with heating demand restriction

The first step is to determine the unit parameters under design conditions. Firstly, the design heating demand is
calculated with Equation (21). For convenience, K has a value of 1 W m-2 and F is equal to 100 m2. As described
in the previous section, the values of these two parameters will not have an effect on C O P . Secondly, the
optimum cylinder volumes of the two compressors and the areas of the three heat exchangers (the subcooler,
evaporator and condenser) are calculated by using the design heating demand and the specified subcooling
parameter. The first step of the optimum problem involves a two-dimensional optimization method, which is
described in Equation (23).

M a x im iz e C O Pd
s u b je c t to Qh Qd
fL fL
*
, (23)
fH fH
*

0 .8

where Qd is the design heating load, and Q d K F tn td .

Once the compressor cylinder volumes and the heat exchanger areas are determined, another process is started

Page 8 of 19
for the determination of the appropriate outdoor air temperature, i.e., the balance point, at which the operation
should be switched from two-stage to single-stage mode. At the balance point, the performances of the two-stage
and single-stage cycles are equal. When the outdoor air temperature is higher than the balance point, the
single-stage mode is superior to two-stage mode and thus should be selected, and conversely, the two-stage mode
should be selected.
Lastly, the program reads the hourly weather data and determines the operation mode according to the balance
point. Hourly heating load is calculated with Equation (21) and then the compressor frequencies, the power
consumption of the heat pump unit and the auxiliary electric power consumption are determined. The optimum
problem for the two-stage operation mode is described in Equation (24). The description of the optimum problem
for the single-stage mode is similar to that for the two-stage mode, except that the low-stage compressor
frequency constraint is not included the single-stage mode.

M a x im iz e C O Pi

s u b je c t to Qh Qi
(24)
f L ,m in f L f L ,m a x

f H ,m im f H f H ,m a x

The flow chart of the calculation steps for the optimum problem is shown in Fig. 5. The extremum problems
are solved with Brents method (Press H. et al., 1992), and the root-finding problems are solved with Van
Wijngaarden-Dekker-Brent Method (Press H. et al., 1992).

3. Results and discussions

Eight representative cities located in different latitudes in cold regions of North China (Wang et al., 2008),
including Beijing, Jinan, Shijiazhuang, Taiyuan, Tianjin, Lanzhou, Xian and Zhengzhou, are selected in this
study. The hourly weather data are obtained from Chinese Standard Weather Data (CSWD) through Energyplus
website (Energyplus, n.d.). These weather data were developed for use in estimating building heating and cooling
loads, simulating energy use, and determining renewable energy utilizations (China Meteorological Bureau,
Climate Information Center, Climate Data Office and Tsinghua University, 2005). Table 4 shows the
geographical information of the studied cities, and Fig. 6 shows the weather data bins derived from CSWD with
statistics, in which the bins are in 2 increments and the bin at ti represents a temperature range of ti -1 < ti ti
+1. It can be found that both latitude and elevation will influence citys weather data. For example, Beijing,
Lanzhou and Taiyuan have more cold hours than other cities as shown in Fig. 6, due to the fact that the latitude of
Beijing is highest among these cities, and the elevations of Lanzhou and Taiyuan are higher than others. Xian and
Lanzhou are warmer than other cities because of their relatively lower latitudes.

Fig. 7 shows the results of C O P of the eight cities and the corresponding optimum Rcy at different design
temperatures. As shown in these results, first of all, the optimum Rcy decreases with an increase in the outdoor
design air temperature. For example, when the design temperature is -30C, the optimum Rcy is 3.32, and when
that is 0C, the optimum Rcy is 1.86. This can be explained with Equation (25) that is derived from Equation (3).
In Equation (25), fH and fL can be canceled out each other, because they are equal to the base frequency at the
design point. Besides, in the optimum cycle, the intermediate pressure is close to the geometric average pressure,
so the value of vol,H / vol,L is approximate to unity and will not change much. Hence, the optimum Rcy mainly
depends on Rm and 3 / 1. With an increase in the outdoor design temperature, the speed of growth in 1 is much
higher than that in 3 and Rm, so the 1 dominates the optimum Rcy. Therefore, the optimum Rcy decreases with an
increase in the outdoor design temperature.

Page 9 of 19
 f H 3 v o l,H
R cy = R m (25)
f L 1 v o l,L

Furthermore, for the eight cities, the COP s increase in the beginning and then drop with an increase in the
outdoor design temperature. There exists an optimum design temperature maximizing the COP . Take the city of
Beijing for example, the C O P for this city is 2.29 when the design temperature is -30C; the C O P has a
maximum value of 2.65 when the design temperature is -8C; and when the design temperature increases up to
0C, the C O P drops to 2.42. There are two reasons for this variation. Firstly, when the heat pump operation
condition deviates much from its design condition, its performance gets deteriorated. The deviated conditions
include too high or too low operation frequency of the compressor, too big or too small pressure ratio, etc.
Secondly, in the hourly weather data during the whole heating period, the bin hours for the extreme outdoor
temperatures are small, compared to the moderate temperatures. As a result, the optimum design temperature at
which the COP is maximum appears at a moderate temperature level.

Finally, the optimum design temperatures maximizing the C O P are various according to the eight cities.
However, there are no enormous differences. The optimum design temperatures are -8C, -5C, -8C, -6C, -8C,
-7C, -4C and -4C for the cities, Beijing, Jinan, Lanzhou, Shijiazhuang, Taiyuan, Tianjin, Xian and Zhengzhou
respectively, and the corresponding optimum Rcy are 2.12, 2.02, 2.12, 2.05, 2.12, 2.09, 1.99 and 1.99, respectively.
Obviously, the differences come from the unique weather data of each city, and the relatively warmer cities have
smaller optimum Rcy, and vice versa.

Fig. 8 illustrates the COP comparison between the cylinder volume ratios of 2.0 and 3.0. The latter represents
the cylinder volume ratio used for a regular freezer. As shown in Fig. 8, the C O P s with Rcy of 2.0 are all
increased by between 10.3% and 17.6%, compared to those with Rcy of 3.0 for the eight cities.
Fig. 9 and Fig. 10 illustrate the total cylinder volumes and the areas of the heat exchangers at different outdoor
design temperatures, respectively. The results are obtained in the first step of the optimum problem described in
Equation (23). As shown in the two figures, both of the total cylinder volumes and total areas of the heat
exchangers decrease with an increase of the outdoor design temperature, due to the decrease of the heating
demand as the result of an increase in the outdoor design temperature, as expressed in Equation (21). The total
cylinder volumes at the outdoor design temperatures of -5C and -26C are 13.08 cm3 (the corresponding Rcy is
close to 2.0) and 45.19 cm3 (the corresponding Rcy is close to 3.0), respectively. The former is 71.1% smaller than
the latter. The total area of the heat exchangers at the outdoor design temperatures of -5C and -26C are 2.3 m2
and 3.0 m2, respectively, and the former is 23.3% smaller than the latter. In general, the cost of a TSAHP unit
depends on the total cylinder volumes of the compressors installed, so it will be much more expensive for this
studied heating project where a TSAHP unit with a total cylinder volume of 45.19 cm3 is chosen. To show this
further, the performances of the units designed at different outdoor design temperatures (i.e., with different
optimum cylinder volume ratios) are evaluated at the outdoor dry-bulb temperature of -12C according to GB/T
25127.2 (2010). For comparison, the frequencies of the compressors are set to equal to their base frequencies. Fig.
11 illustrates the heating capacity and heating COP of the unit designed at different outdoor temperatures. As
shown in this figure, with an increase of the outdoor design temperature, the heating COP of the unit operating at
the outdoor dry-bulb temperature of -12C increases, whereas the heating capacity drops. The variations of the
COP are mainly caused by the variations of the evaporating temperature. When an air source heat pump unit with
a constant compressor frequency operates below its outdoor design temperature, the difference between the
evaporating temperature and the outdoor air temperature is smaller than the design value, and vice versa. As a
result, the unit designed at a higher outdoor temperature will have a higher evaporating temperature than those
designed at relatively lower outdoor temperatures, when the design difference between the evaporating
temperature and the outdoor air temperature is identical. Since the evaporating temperature is a dominant factor,
the heating COP of the unit designed at a higher temperature will be larger than those designed at relatively lower

10

Page 10 of 19
outdoor temperatures when the units are evaluated at the same outdoor dry-bulb temperature. The variation of the
heating capacity are due to the variation of the total cylinder volumes with outdoor design temperatures, as shown
in Fig. 9. The heating capacities at the outdoor design temperatures of -5C and -26C are 1.98 kW (the
corresponding Rcy is close to 2.0) and 6.04 kW (the corresponding Rcy is close to 3.0), respectively. The former is
only 32.8% of the latter.
In summary, it can be concluded that, the appropriate compressor cylinder volume ratio is between 2.0 and 2.1,
and the corresponding outdoor design temperature is between -4C and -8C. Moreover, a proper selection of the
outdoor design temperature will not only contribute to operation cost savings but also bring about a reduction of
both the total compressor cylinder volume and the areas of the heat exchangers.

4. Conclusions
An optimum compressor cylinder volume ratio selection method for two-stage compression air source heat
pump was developed in this work, based on hourly weather data, a simplified heating load model, and the TSAHP
model. A two-step optimization approach was employed, in which the heating C O P s for eight typical cities in
the cold regions of North China were compared each other at different outdoor design temperatures. The main
conclusions are the following,
Under the modeling conditions described in this paper, the appropriate compressor cylinder volume ratio
is between 2.0 and 2.1, and the corresponding outdoor design temperature is between -4C and -8C, for
TSAHP design in the cold regions of North China;
For the eight typical cities selected in the simulation, C O P with Rcy of 2.0 is 10.3% and 17.6% higher
than that with Rcy of 3.0. The total cylinder volumes with Rcy of 2.0 are 71.1% smaller than those with Rcy
of 3.0, and the total heat exchanger areas are 23.3% smaller.
In summary, it is important to select a proper design temperature when to design a TSAHP system. The method
and conclusions presented in this study may give guidance, not only for the appropriate selection of compressors
in a TSAHP system, but also for TASHP heating equipment selection at different climate conditions. Although
this study does not include the effect of economy, it is worth studying it in the future for the sake of achieving an
outstanding balance between performance and economy.

Acknowledgments
This research was supported by the Research Fund for the Doctoral Program of Higher Education of China
(No.20120041110006). The authors are grateful for the given support.

References
Abdelaziz, O., Shen, B., 2012. Cold Climates Heat Pump Design Optimization. ASHARE Trans. 9, 3442.
ASHRAE, 1996. ASHRAE GUIDELINE, Engineering Analysis of Experimental Data. Ashrae Guidel. 2-1986.
Baek, C., Heo, J., Jung, J., Cho, H., Kim, Y., 2014. Effects of the cylinder volume ratio of a twin rotary
compressor on the heating performance of a vapor injection CO2 cycle. Appl. Therm. Eng. 67, 8996.
doi:10.1016/j.applthermaleng.2014.03.007
Bergman, T.L., Incropera, F.P., Lavine, A.S., 2011. Fundamentals of heat and mass transfer. John Wiley & Sons.
Bertsch, S.S., Groll, E. a., 2008. Two-stage air-source heat pump for residential heating and cooling applications
in northern U.S. climates. Int. J. Refrig. 31, 12821292. doi:10.1016/j.ijrefrig.2008.01.006
Caskey, S.L., Kultgen, D., Tobias, M., Groll, E.A., Hutzel, W., Bertsch, S.S., 2012. Field test simulation of an
air-source heat pump with two-stage compression and economizing for cold climates, in: International
Refrigeration and Air Conditioning Conference. Purdue.

11

Page 11 of 19
engel, Y.A., Boles, M.A., 2008. Thermodynamics: An Engineering Approach, McGraw-Hill series in
mechanical and aerospace engineering. McGraw-Hill Higher Education.
China Meteorological Bureau, Climate Information Center, Climate Data Office and Tsinghua University, D. of
B.S. and T., 2005. China Standard Weather Data for Analyzing Building Thermal Conditions. Beijing:
China Building Industry Publishing House.
Cooper, M.G., 1984. Saturation nucleate pool boilinga simple correlation, in: Inst. Chem. Eng. Symp. Ser. pp.
785793.
Domanski, P. a., Yashar, D., Kim, M., 2005. Performance of a finned-tube evaporator optimized for different
refrigerants and its effect on system efficiency. Int. J. Refrig. 28, 820827.
doi:10.1016/j.ijrefrig.2005.02.003
Domanski, P., Yashar, D., Kaufman, K., Michalski, R., 2004. An Optimized Design of Finned-Tube Evaporators
Using the Learnable Evolution Model. HVAC&R Res. 10, 201211.
doi:10.1080/10789669.2004.10391099
Energyplus, n.d. http://apps1.eere.energy.gov/buildings/energyplus/weatherdata_about.cfm [WWW Document].
URL http://apps1.eere.energy.gov/buildings/energyplus/weatherdata_about.cfm
Ertesvg, I.S., 2011. Uncertainties in heat-pump coefficient of performance (COP) and exergy efficiency based on
standardized testing. Energy Build. 43, 19371946. doi:10.1016/j.enbuild.2011.03.039
GB/T 25127.2, 2010. Low ambient temperature air source heat pump (water chilling) packages - Part 2: Heat
pump (water chilling) packages for household and similar application (in Chinese).
Gungor, K.E., Winterton, R.H.S., 1986. A general correlation for flow boiling in tubes and annuli. Int. J. Heat
Mass Transf. 29, 351358. doi:10.1016/0017-9310(86)90205-X
Han, D., Lee, K., Kim, Y., 2003. Experiments on the characteristics of evaporation of R410A in brazed plate heat
exchangers with different geometric configurations. Appl. Therm. Eng. 23, 12091225.
doi:10.1016/S1359-4311(03)00061-9
Heo, J., Yun, R., Kim, Y., 2013. Simulations on the performance of a vapor-injection heat pump for different
cylinder volume ratios of a twin rotary compressor. Int. J. Refrig. 36, 730744.
doi:10.1016/j.ijrefrig.2012.10.015
Hsieh, Y.Y., Lin, T.F., 2003. Evaporation Heat Transfer and Pressure Drop of Refrigerant R-410A Flow in a
Vertical Plate Heat Exchanger. J. Heat Transfer 125, 852. doi:10.1115/1.1518498
Hutzel, W.J., Groll, E.A., 2013. Air Source Cold Climate Heat Pump ESTCP Project EW-201136.
Jiang, S., Wang, S., Jin, X., Zhang, T., 2014. Simulation on a Two-Stage Compression Heat Pump with Focus on
Optimum Control, Proceedings of the 8th International Symposium on Heating, Ventilation and Air
Conditioning, Lecture Notes in Electrical Engineering. Springer Berlin Heidelberg, Berlin, Heidelberg.
doi:10.1007/978-3-642-39581-9
Jiang, S., Wang, S., Jin, X., Zhang, T., 2015. A general model for two-stage vapor compression heat pump
systems. Int. J. Refrig. 51, 88102. doi:10.1016/j.ijrefrig.2014.12.005
Jin, X., 2013. Interstage Matching and Coupling Characteristics in a Two-Stage Compression Heat Pump System
with Variable Capacity. Dalian University of Technology (in Chinese).
Jin, X., Wang, S., Huo, M., 2008. Experimental Investigation of a Novel Air Source Heat Pump for Cold Climate,
in: The First International Conference on Building Energy and Environment. Dalian, China.
Kang, H., Li, W., Li, H., Xin, R., Tao, W., 1994. Experimental study on heat transfer and pressure drop for plane
fin-and-tube heat exchanger. J. Xian Jiaotong Univ. 28, 9198.
Ko, Y., Park, S., Jin, S., Kim, B., Hwan, J., 2013. The selection of volume ratio of two-stage rotary compressor
and its effects on air-to-water heat pump with flash tank cycle. Appl. Energy 104, 187196.
doi:10.1016/j.apenergy.2012.11.021
Lemmon, E., Huber, M., McLinden, M., 2013. NIST standard reference database 23: reference fluid
thermodynamic and transport properties (REFPROP), version 9.1, National Institute of Standards and
Technology, Standard reference data program, Gaithersburg,.
Li, Y., Yu, J., 2016. Theoretical analysis on optimal configurations of heat exchanger and compressor in a
two-stage compression air source heat pump system. Appl. Therm. Eng. 96, 682689.

12

Page 12 of 19
 doi:10.1016/j.applthermaleng.2015.11.132
Press H., W., Teukolsky A., S., Vetterling T., W., Flannery P., B., 1992. Numerical Recipes in C: The Art of
Scientific Computing (2nd Edition).
Redn, A., Navarro-Peris, E., Pitarch, M., Gonzlvez-Macia, J., Corbern, J.M., 2014. Analysis and optimization
of subcritical two-stage vapor injection heat pump systems. Appl. Energy 124, 231240.
doi:10.1016/j.apenergy.2014.02.066
Tian, C., Liang, N., Shi, W., Li, X., 2006. Development and experimental investigation on two-stage
compression variable frequency air source heat pump, in: International Refrigeration and Air Conditioning
Conference. Purdue.
Wang, J., Xie, Q., Shi, W., Zhang, L., 2008. Research on IPLV (H) for Low Temperature Air Source Heat Pump
(Water Chilling) Packages. J. Refrig. 29, 914 (in Chinese).
Wu, Z., Ding, G., Wang, K., Fukaya, M., 2008. Application of a genetic algorithm to optimize the refrigerant
circuit of fin-and-tube heat exchangers for maximum heat transfer or shortest tube. Int. J. Therm. Sci. 47,
985997. doi:10.1016/j.ijthermalsci.2007.08.005
Yan, Y., Lio, H., Lin, T., 1999. Condensation heat transfer and pressure drop of refrigerant R-134a in a plate heat
exchanger. Int. J. Heat Mass Transf. 42, 9931006. doi:10.1016/S0017-9310(98)00217-8
Zehnder, M., 2004. Efficent air-water heat pumps for high temperature lift residential heating, including oil
migration aspects. EPFL. doi:10.5075/epfl-thesis-2998

Fig. 1 - The system diagram and ln(p)-h diagram for two-stage cycle

13

Page 13 of 19
Fig. 2 Flow chart to calculate f1(pm) or f2(pm) with the given pm

Fig. 3 Flow chart of the two-stage system model

14

Page 14 of 19
 5.0 2.7

1800
4.5 2.6

1600
4.0 2.5

heating capactiy (kW)

1400 3.5
pm (kPa)

COP (-)
2.4

1200 3.0
2.3

2.5
1000 2.2

2.0
800 2.1
1.5
100 150 200 250 300 350 100 150 200 250 300 350 100 150 200 250 300 350
nd nd nd
2 EEV openings 2 EEV openings 2 EEV openings
30 Hz, calculated 30 Hz, tested 40 Hz, calculated 40 Hz, tested 50 Hz, calculated 50 Hz, tested
60 Hz, calculated 60 Hz, tested 70 Hz, calculated 70 Hz, tested

Fig. 4 - Comparison between the simulation results and experimental data

Fig. 5 - Calculation process for the optimum problem

15

Page 15 of 19
 550
Beijing
500
Jinan
450 Lanzhou
Shijiazhuang
400
Taiyuan

Weather data bin (h)

350 Tianjin
Xian
300
Zhengzhou
250

200

150

100

50

0
-20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14
Temperature bin (C)

Fig. 6 Weather data bins of the studied cities

3.1 3.8
Beijing Jinan Rcy
3.0 Lanzhou Shijiazhuang 3.6
2.9 Taiyuan Tianjin 3.4
Xian Zhengzhou
2.8 3.2

Optimum Rcy (-)

2.7 3.0
COP (-)

2.6 2.8
2.5 2.6
2.4 2.4
2.3 2.2
2.2 2.0
2.1 1.8
2.0 1.6
-30 -25 -20 -15 -10 -5 0 5
Outdoor design temperature (C)

Fig. 7 - COP and optimum Rcy at different outdoor design temperatures

3.0
Rcy = 3.0 Rcy = 2.0
2.84
2.81
2.8 2.78
2.73
2.63 2.65
2.60 2.59
2.6
2.45
2.40 2.42
2.4 2.36 2.39 2.38
2.35 2.33
COP

2.2

2.0

1.8

1.6
g an ou ng an nji
n n u
Be
ijin Jin nzh hua iyu Tia Xia zho
La ijia
z Ta eng
Sh Zh
City

Fig. 8 - COP comparison between optimum Rcy of 2.0 and 3.0

16

Page 16 of 19
 3.4
60
Vcy,L+Vcy,H Rcy
3.2

Total cylinder volumes (cm3)

50 3.0

2.8

Optimum Rcy (-)

40
2.6

30 2.4

2.2
20 2.0

1.8
10
1.6

0 1.4
-30 -25 -20 -15 -10 -5 0 5
Outdoor design temperature (C)

Fig. 9 - Total cylinder volumes at different outdoor design temperatures

3.2 Area of condenser

Area of evaporator (ref. side)
2.8 Area of subcooler

2.4

2.0
Area (m2)

1.6

1.2

0.8

0.4

0.0
-30 -25 -20 -15 -10 -5 0
Outdoor design temperature (C)

Fig. 10 - Heat exchanger area at different outdoor design temperatures

2.40 8
COP @-12C
Heating capacity @-12C
2.35 7

6
Heating capacity (kW)

2.30

5
2.25
COP (-)

4
2.20

3
2.15
2
2.10
1
2.05
-30 -25 -20 -15 -10 -5 0 5
Outdoor design temperature (C)

Fig. 11 Performance comparison at outdoor dry-bulb temperature of -12C

Table 1 Uncertainty analysis results

Uncertainties Frequency of the Low-stage compressor

17

Page 17 of 19
(Confidence level: 95%) 30 Hz 40 Hz 50 Hz 60 Hz 70 Hz
Maximum pm 4.7 3.6 2.8 2.4 2.4
(%) Heating capacity 5.4 9.4 6.7 5.9 7.0
Heating COP 5.5 9.5 6.7 6.0 7.1
Average pm 4.1 3.3 2.5 2.2 2.2
(%) Heating capacity 4.9 6.8 5.4 5.7 6.8
Heating COP 5.0 6.9 5.4 5.7 6.9

Table 2 Deviations of the simulation results

Relative deviation Frequency of the Low-stage compressor

(Cal.-Test)/Test100
30 Hz 40 Hz 50 Hz 60 Hz 70 Hz
Maximum pm 3.25 2.64 -8.57 -2.80 -3.32
(%) Heating capacity 6.67 2.06 -2.32 -3.01 -5.16
Heating COP 7.32 2.03 2.23 -4.08 2.29
Average pm 1.07 -0.82 -4.26 -1.62 -3.02
(%) Heating capacity 5.46 0.79 -0.50 -0.81 -4.12
Heating COP 5.44 0.83 0.46 -1.67 2.10

Table 3 Heat exchanger parameters in TSAHP

Type Parameter Value

Subcooler (VPHE) Plate width (mm) 106.0
Enlargement factor 1.1836
Angle of the corrugation (rad) 0.9599
Corrugation amplitude (mm) 2.0
Number of plates 10
Chevron angle (degree) 45.0
Condenser (VPHE) Plate width (mm) 124.0
Enlargement factor 1.1836
Angle of the corrugation (rad) 0.9599
Corrugation amplitude (mm) 2.0
Number of plates 50
Chevron angle (degree) 45.0
Evaporator (Finned tube coil) Number of tubes 4
Number of rows 1
Tube pitch (mm) 25.0
Row pitch (mm) 21.65
Tube O.D. 10.15
Tube I.D. 9.33
Fin pitch (mm) 2.0
Fin thickness (mm) 0.2

Table 4 Locations and elevations of the cities studied

Location
Cities Elevation
Latitude Longitude
Beijing N 3947' E 11628' 31.3
Jinan N 3635' E 11703' 170.3
Lanzhou N 3602' E 10352' 1517.2
Shijiazhuang N 3801' E 11425' 81
Taiyuan N 3746' E 11233' 778.3
Tianjin N 3904' E 11704' 2.5
Xian N 3417' E 10855' 397.5
Zhengzhou N 3443' E 11339' 110.4

18

Page 18 of 19
19

Page 19 of 19