- Email: [email protected]
Simulation of solar cooling system based on variable effect LiBr-water absorption chiller
Accepted Manuscript Simulation of solar cooling system based on variable effect LiBr-water absorption chiller Z.Y. Xu, R.Z. Wang PII:
S0960-1481(17)30574-8
DOI:
10.1016/j.renene.2017.06.069
Reference:
RENE 8936
To appear in:
Renewable Energy
Received Date: 7 April 2016 Revised Date:
27 April 2017
Accepted Date: 18 June 2017
Please cite this article as: Xu ZY, Wang RZ, Simulation of solar cooling system based on variable effect LiBr-water absorption chiller, Renewable Energy (2017), doi: 10.1016/j.renene.2017.06.069. 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.
ACCEPTED MANUSCRIPT
Simulation of solar cooling system based on variable effect LiBr-water absorption chiller Z.Y. Xu, R.Z. Wang*
RI PT
Institute of Refrigeration and Cryogenics, Shanghai Jiao Tong University, Shanghai 200240, China
* Corresponding author. Tel.: +86 21 34206548.
AC C
EP
TE D
M AN U
SC
E-mail address: [email protected] (R.Z. Wang).
ACCEPTED MANUSCRIPT 1
Simulation of solar cooling system based on variable effect
2
LiBr-water absorption chiller
3
Z.Y. Xu, R.Z. Wang*
4
Institute of Refrigeration and Cryogenics, Shanghai Jiao Tong University, Shanghai 200240, China
RI PT
5 Abstract
7
In solar absorption cooling system, the instability of solar power causes mismatch between the
8
solar collector and the absorption chiller. The variable effect absorption cycle was proposed to
9
improve this. In order to investigate its solar driving performance, a Compound Parabolic
10
Collector (CPC) driving variable effect LiBr-water absorption cooling system is simulated. Model
11
of the variable effect LiBr-water absorption chiller is built through artificial neural network (ANN)
12
modeling based on 450 groups of experimental data. Good agreement between the prediction and
13
experimental data is achieved with correlation coefficient of 0.994. The CPC driving absorption
14
cooling system is then built in TRaNsient SYstem Simulation program (TRNSYS) based on the
15
chiller model. The daily performance of this system is calculated and analyzed. The variable effect
16
chiller can work with low driving temperature, which guarantees a long working period. Besides,
17
the variable effect chiller has high COP under high driving temperature, which ensures a
18
competitive overall efficiency. The calculation shows that average chiller COP of 0.88 and solar
19
COP of 0.35 are obtained. The effects of solar collector area, storage tank volume and cut-off
20
driving temperature on the system performance are analyzed. The optimal solar collector area and
21
tank volume are obtained.
22
Keywords
23
Solar cooling, CPC, absorption cooling, simulation, TRNSYS
AC C
EP
TE D
M AN U
SC
6
24 25
Nomenclature
26
COP
coefficient of performance (-)
27
Q
heat transfer amount (kJ)
28
P
power (kW)
1
ACCEPTED MANUSCRIPT 1
η
collector efficiency (-)
2
A
collector area (m2)
3
I
irradiance (W/m2)
4
∆t
time step (s)
Subscripts and superscripts
7
ch1
chilling water inlet
8
ch2
chilling water outlet
9
c1
cooling water inlet
10
c2
cooling water outlet
11
col
collector
12
E
13
G
M AN U evaporation
generation
14 15
SC
6
RI PT
5
1. Introduction
The solar cooling system is an environmental friendly system which consumes renewable
17
energy. Among the potential options for solar cooling, solar powered absorption cooling system is
18
competitive due to the high efficiency and commercialization of absorption chiller. A solar
19
absorption cooling system consists of solar collector, heat storage, absorption chiller, cooling
20
tower, cooling load, pumps and control units. Solar absorption cooling systems with different solar
21
collectors and absorption chillers have been studied [1-4].
EP
AC C
22
TE D
16
In the past decades, solar absorption cooling system composed of low temperature solar
23
collector and single effect LiBr-water absorption chiller was popular in research. This system
24
adopts flat plate collector or evacuated tube collector as heat source and supplies hot water below
25
100oC to the absorption chiller. Assilzadeh et al. [5] studied an evacuated tube solar collector
26
driving single effect LiBr-water absorption cooling system in Malaysia. The system performance
27
was calculated in TRaNsient SYstem Simulation program (TRNSYS). The optimum design for
28
3.5 kW cooling output needed 35m2 collector, and worked about 5 hours a day. Syed et al. [6]
29
studied a 49.9m2 flat plate collector driving 35kW single effect LiBr-water absorption cooling
2
ACCEPTED MANUSCRIPT system in Madrid. Maximum instantaneous COP and daily average COP of 0.60 and 0.42 were
2
experimentally obtained. Lizarte et al. [7] studied a vacuum flat plate collector driving 4.5 kW
3
air-cooled single effect LiBr-water absorption cooling system in Madrid. Mean COP and solar
4
COP of 0.53 and 0.06 were experimentally obtained. These systems had short working time and
5
required large area of solar collector. Solar collector with higher working temperature can be
6
adopted to improve this. Mazloumi et al. [8] studied a Parabolic Trough Collector(PTC) driving
7
17.5kW single effect LiBr-water absorption cooling system by simulation. The chiller worked for
8
12 hours per day with COP from 0.67 to 0.76. The working time was longer but the system
9
efficiency was limited by the single effect absorption chiller.
SC
RI PT
1
In order to increase the system efficiency, absorption chillers with higher efficiencies including
11
double effect absorption chiller and Generator Absorber eXchange (GAX) absorption chiller can
12
be used. Bermejo et al. [9] experimentally studied a Linear Fresnel Reflector (LFR) driving double
13
effect LiBr-water absorption cooling system in Spain. The chiller was driven by pressurized hot
14
water powered by LFR with a backup natural gas burner. The chiller started to work at 12 a.m.
15
Daily average COP of the chiller was 1.1~1.25 and the solar cooling ratio was 0.44. Qu et al. [10]
16
experimentally studied a linear parabolic trough collector driving double effect LiBr-water
17
absorption cooling system. The chiller started to work after 12 a.m. The product of the chiller COP
18
and the collector efficiency was about 0.33~0.44. Velázquez et al. [11] studied a LFR direct
19
driving ammonia-water GAX absorption cooling system through simulation. The air-cooled GAX
20
cycle COP and solar COP reached 0.85 and 0.53 respectively. These systems required high heat
21
source temperature that they started to work late in the daytime.
TE D
EP
AC C
22
M AN U
10
The single effect and double effect LiBr-water absorption chillers obtain COPs about 0.7 and
23
1.2 under driving temperature around 90oC and 150oC respectively [12]. The chillers are not
24
flexible under variable driving temperature: their COPs vary little with the driving temperature
25
and their required driving temperatures are constrained in a small range. However, the solar heat
26
power has variable temperature. This makes the coupling between solar collector and absorption
27
chiller unstable. When the heat source has low temperature, it is unable to activate the absorption
28
chiller. When the heat source temperature is high, there will be a waste of energy grade. In order to
29
enhance the compatibility between solar collector and absorption chiller, various absorption cycles
30
have been developed [13] including the variable effect absorption refrigeration cycle previously 3
ACCEPTED MANUSCRIPT 1
proposed by the author [14]. The COP of variable effect cycle increases with driving temperature.
2
Later, a variable effect LiBr-water absorption chiller was built and tested based on this cycle [15]. As the variable effect absorption chiller is designed for high efficient solar cooling, the
4
characteristics and performance of solar powered variable effect absorption cooling system should
5
be investigated. In this paper, this solar cooling system is studied through simulation. Considering
6
the TRNSYS is a popular modeling software for solar absorption cooling system [5, 16-18], the
7
solar powered variable effect absorption cooling system is built in TRNSYS. An ANN model of
8
the variable effect absorption chiller is built from the experimental data and used in the TRNSYS
9
model. The performance of the solar driving variable effect absorption cooling system is studied
SC
10
RI PT
3
through simulation.
12
M AN U
11 2. Solar driving variable effect absorption cooling system
The system discussed in this paper is shown in Fig.1. It includes the solar collector, storage
14
tank, auxiliary heater, absorption chiller, cooling tower, cooling load and the pumps. Absorption
15
chiller integrated in the system is the variable effect LiBr-water absorption chiller. As the driving
16
temperature of the chiller is higher than 100oC, the medium temperature solar collector is essential
17
to supply heat source with sufficient temperature.
TE D
13
18 19
EP
Fig. 1 - The solar driving variable effect absorption cooling system
20
22
2.1. Variable effect LiBr-water absorption chiller
AC C
21
The chiller studied in this paper is built based on the variable effect absorption cycle.
23
According to the thermodynamic calculation, the variable effect LiBr-water absorption cycle
24
obtained COP from 0.8 to 1.06 under generation temperature from 95oC to 135oC, evaporation
25
temperature of 5oC, absorption temperature of 35oC and condensation temperature of 40oC
26
[14].The chiller COP was calculated from Eq.(1). COP =
(1)
27
The chiller prototype with rated power of 50kW was built based on this variable effect cycle.
28
The variable effect chiller prototype is also shown in Fig.1. Experimental results showed that the 4
ACCEPTED MANUSCRIPT 1
variable effect water-LiBr absorption chiller obtained COP from 0.69 to 1.08 under generation
2
temperature from 95.0oC to 120.0oC [15]. The COP increased as it is predicted by the calculation.
3
Table 1 shows some of the experimental results.
4 5
Table 1- Experimental data of the variable effect absorption chiller [15]
7
RI PT
6 2.2. Solar collector
Among different medium temperature solar collectors, the CPC is able to offer heat source of
9
enough temperature for the variable effect absorption chiller with simple stationary configuration.
10
A type of commercial CPC is used in this paper. It is an improved version of the collector used in
11
a previous study of solar driving cooling system [19]. In this stationary non-imaging CPC, copper
12
tube inside glass evacuated-tube is used for heat transfer. According to the tested data, the
13
collector achieves efficiencies of 50% and 48% under temperature of 130 oC and 150 oC
14
respectively, when solar irradiance is 900W/m2 and ambient temperature is 35 oC. The efficiency
15
of the solar collector can be calculated from the efficiency data from the manufacturer.
M AN U
SC
8
TE D
16 17
3. ANN model of the variable effect absorption chiller
18
3.1. Choice of modeling method
In order to precisely predict the performance of the variable effect absorption chiller, a proper
20
model is necessary. The commonly used modeling methods include the thermodynamic model
21
considering the heat transfer [12], the adapted characteristic equation model, the multivariate
22
polynomial regression model and the ANN model [20]. As parameters of the variable effect chiller
23
change a lot under different temperature conditions and cooling loads [14], the heat transfer
24
coefficients cannot be set as constant for simulation, and the thermodynamic model is not effective.
25
Among other modeling methods, the ANN model has better accuracy [20].
26
3.2. Preparation of experimental data
AC C
EP
19
27
In order to build the ANN model, the chiller was tested under different conditions, and 450
28
groups of data under steady state working condition were selected. Each group of data contained
29
the heat source temperature, generation temperatures of high pressure generator and low pressure 5
ACCEPTED MANUSCRIPT generator, inlet and outlet cooling water temperatures, condensation temperature, evaporation
2
temperature, cooling power and COP. The experimental parameters here were obtained under
3
constant flow rates. The heat source temperature varied between 85.34oC and 140.96 oC. The
4
cooling water inlet temperature varied between 24.74 oC and 34.25 oC. The cooling water outlet
5
temperature varied between 28.40 oC and 39.22 oC. The chilled water inlet temperature varied
6
between 9.05 oC and 20.83 oC. The chilled water outlet temperature varied between 4.43 oC and
7
16.25 oC. The cooling power varied between 30kW and 60kW, and the COP varied between 0.6
8
and 1.1.
9
3.3. ANN model
SC
RI PT
1
The commonly used ANN models include feed-forward neural network, feedback neural
11
network, etc. The Back Propagation (BP) network is a simple and effective feed-forward neural
12
network. Considering that the chiller modeling in this paper is simple and doesn’t need feedback,
13
the BP network is chosen.
M AN U
10
In the chiller model, the input parameters include the heat source inlet temperature, cooling
15
water inlet temperature, chilled water inlet temperature and the required cooling output. With the
16
constant water flow rates, these four parameters are enough to decide the cooling water outlet
17
temperature and chilled water outlet temperature. The cooling power and COP can be further
18
calculated from the temperatures and flow rates. In this case, the network needs 4 inputs and 2
19
outputs. According to the calculation, one hidden layer with 6 units is enough for the precise
20
prediction. Fig.2 shows the construction of the network including input layer (P), hidden layer (A),
21
output layer (O) and the transfer functions. Considering that the calculation from inlet
22
temperatures and required cooling output to the outlet temperatures is nonlinear, the hyperbolic
23
tangent sigmoid function (tansig) is used both for the hidden layer calculation and output layer
24
calculation. The mean square root error is minimized during the iteration.
AC C
EP
TE D
14
25 26
Fig. 2 - Construction of the ANN model
27 28
The ANN model is built in the MATLAB neural network tool box. Normalization of the
29
experimental data is carried out in the range of (-1, 1). In the model building, part of the
6
ACCEPTED MANUSCRIPT 1
experimental data are used for model training, and the rest data are used for model validation.
2
Table 2 shows the summary of this ANN model. A correlation coefficient (R2) of 0.994 is obtained
3
which proves the effectiveness of the ANN model.
4 5 6
4. Simulation of the solar cooling system in TRNSYS
SC
7
RI PT
Table 2- ANN model of the variable effect absorption chiller
The system performance is studied in the software TRNSYS. New modules (named as “type”
9
in TRNSYS) are built both for the variable effect absorption chiller and the commercial CPC. In
10
the chiller module, the flow rates of cooling water and chilled water are set as 20m3/h and 10m3/h
11
respectively, which are the same with experimental values [15]. Considering the chiller was tested
12
under limited temperature range, the ANN model can be invalid when its inputs deviate too much
13
from the tested range. In the simulation, this kind of invalid inputs are avoided to ensure
14
reasonable prediction. Module of the CPC is built from the manufacturer supplying data. The
15
variable effect absorption chiller module and CPC module are named as “type 151” and “type 161”
16
respectively.
19 20
TE D
EP
18
Fig. 3 - CPC driving variable effect LiBr-water absorption cooling system in TRNSYS
AC C
17
M AN U
8
Fig.3 shows the system in TRNSYS. The blue blocks, grey blocks and white blocks represent
21
the self-built modules, component modules and controller modules respectively. The component
22
modules are connected according to Fig.1. The controllers control the operation of solar collector
23
pump, auxiliary heater, hot water pump, cooling water pump and chilled water pump. The
24
controlling logics are as follows: the solar collector pump operates when the collector working
25
temperature is higher than the tank temperature; the auxiliary heater operates when its inlet
26
temperature is lower than the setting temperature; the chiller works when the hot water
27
temperature and room temperature are both higher than their setting values. The weather data
7
ACCEPTED MANUSCRIPT 1
supplied by the software are used for simulation. The TMY2 file named “US-FL-Miami-12839” is
2
selected. For building simulation, the single zone model of Type 12c is used. The air conditioning
3
temperature of the building is set as 25oC. Several efficiencies are used to evaluate the system performance. The average chiller COP is
5
defined as the ratio of total cooling output to total heat input in Eq. (2). The average collector
6
efficiency is defined as the ratio of total heat output to total irradiance input when the collector
7
works in Eq. (3). The solar COP is defined as the ratio of total cooling output to total irradiance
8
input in Eq. (4).
COP =
∑ Δ ∑ Δ
η =
SC
∑ Δ ∑ Δ
∑ Δ ∑ Δ
9 10
5. Results and discussions
M AN U
η =
RI PT
4
(2)
(3) (4)
The system is simulated with a time step of 0.1 hour. The performance in a sunny day in July
12
is studied. In order to study the solar driving performance of the system, the auxiliary heater is not
13
activated during the simulation. The daily performance is calculated based on the following
14
conditions: the rated power of variable effect absorption chiller is 50kW; the collector area is
15
200m2; the storage tank has volume of 3m3; the chiller starts to work when the heat source
16
temperature exceeds 100 oC. The daily performance is presented, and the effects of some key
17
parameters on the system performance are studied.
18
5.1. Daily performance
EP
AC C
19
TE D
11
Fig. 4 shows the solar irradiance and ambient temperature. The irradiance is higher than zero
20
from 6:12 to 19:00. It reaches the maximum of 989.4W/m2 at 13:12. The ambient temperature
21
varies between 25.6 oC and 30.6oC.
22 23
Fig.4 – Solar irradiance and ambient temperature
24 25
Fig. 5. Daily performance of the solar collector 8
ACCEPTED MANUSCRIPT 1 Fig.5 shows the solar collector temperatures and efficiency. The following points are
3
essential to be mentioned. (1) The collector efficiency varies between 0.3 and 0.51 from 9:00 to
4
16:00. (2) From 7:48, the collector starts to deliver thermal power to the tank, when the collector
5
temperature reaches 95.1oC and exceeds the tank temperature. The tank temperature stays stable
6
with a high temperature before 7:48 because the heat loss coefficient of the tank is small. (3) The
7
collector temperature reaches its maximum value of 148.3oC at 15:30, which is not at the time of
8
maximum irradiance. This is caused by the consumption by chiller.
RI PT
2
Fig. 6 shows the temperature and COP of the variable effect absorption chiller. The chiller
10
works between 8:54 and 19:24 when the hot water temperature is higher than its cut-off driving
11
temperature. During this period, the inlet and outlet heat source temperatures vary in the range of
12
100oC~146.6 oC and 94.5 oC~143.3 oC respectively. The inlet and outlet cooling temperatures vary
13
in the range of 27.2 oC ~29.8 oC and 31.0 oC~ 34.7oC respectively. The inlet and outlet chilled
14
water temperatures vary in the range of 13.4oC~14.6oC and 9.9oC~10.3oC respectively. The COP
15
varies between 0.78 and 1.1. The variations of heating source temperature and COP are similar.
M AN U
SC
9
17 18
TE D
16
Fig. 6. Daily performance of the absorption chiller
From 8:54 to 9:18, the heat input from solar collector is less than the consumption of chiller,
20
the tank temperature cannot keep higher than the cut-off driving temperature of absorption chiller,
21
and fluctuations happens. From 9:18 to 11:00, the COP rises and falls due to the change of solar
22
irradiance. Later, the COP grows with the heat source temperature and achieves the highest value
23
from 15:18 to 15:54. The COP and heat source temperature increase slowly from 8:54 to 15:18 but
24
fall quickly from 15:54 to 19:24. The reasons are as follows. (1) From 8:54 to 15:18, the chiller
25
COP is low which means higher heat consumption rate. The large heat consumption rate slows
26
down the increase of tank temperature. (2) When the heat input from solar collector becomes
27
smaller than the heat consumption of absorption chiller at 15:18, the heat source temperature and
28
chiller COP start to decrease. (3) From 15:54 to 19:24, solar irradiance keeps decreasing. On the
29
other hand, the chiller consumes more thermal energy due to lower COP. In this case, the heat
30
source temperature and COP decrease quickly.
AC C
EP
19
9
ACCEPTED MANUSCRIPT 1 2
Fig. 7. Daily energy flows
3 Fig. 7 shows the energy flows in the whole day. The received solar power, collector gained
5
thermal power, heat consumption power, cooling output power and cooling load are presented.
6
The received solar power and collector gained thermal power have similar variations. The heat
7
consumption power varies between 38.2kW and 68.4kW. The cooling load varies between 39.7kW
8
and 49.8kW. Both the heat consumption power and cooling load are high at the beginning of
9
operation due to the high initial room temperature. The average chiller COP and solar COP are
SC
10
RI PT
4
0.88 and 0.35 respectively.
In this section, the daily variations of system parameters are presented. In the following parts,
12
the overall system performance will be evaluated. The analyzed parameters include the solar
13
collector area, storage tank volume and cut-off driving temperature.
14
5.2 Effect of solar collector area
17 18 19
efficiency and solar COP.
TE D
16
Fig. 8 shows the impact of solar collector area on the average chiller COP, average collector
Fig. 8 - Effect of solar collector area on system efficiencies
EP
15
M AN U
11
When the solar collector area increases from 150m2 to 275m2, the overall collector efficiency
21
falls from 49.1% to 43.7% and the chiller average COP increases from 0.82 to 1.00. The reason is
22
that large collector area increases the average tank temperatures. This further decreases the
23
average collector efficiency and increases the average chiller COP. The increase of solar COP
24
from 0.24 to 0.37 indicates that the large collector area is beneficial to the solar COP.
AC C
20
25
Fig. 9 shows the impact of solar collector area on the chiller working time. The working
26
time increases from 6.8 hours to 10.8 hours when the area increases from 150m2 to 275m2. At
27
the same time, the total cooling output increases from 309.1 kWh to 461.3kWh. Both the solar
28
COP and total cooling output increase with larger solar collector area. However, they increase
10
ACCEPTED MANUSCRIPT 1
slowly when the collector area exceeds 200m2.
2 3
Fig. 9 - Effects of solar collector area on working time and cooling output
4 Considering that the collector area is the key parameter affecting the economic performance
6
[16], economic performance under different collector areas is calculated. There are many criteria
7
for economic analysis including the annualized operation cost, lift cycle cost, payback period and
8
net present value [16, 21, 22]. Here, the payback period is used which refers to the time needed for
9
the cumulative fuel savings to equal the initial investment[16]. The investment of absorption
10
chiller is about 22000$ according to our experiment. The investment for CPC is 165$/m2 [21]. The
11
investment for a 100kW cooling tower and 3m3 storage tank are 6760$ and 2700$ respectively[16].
12
The fuel prices of 0.08$/kWh, 0.1$/kWh and 0.12$/kWh are used for calculation [16]. The annual
13
simulation is carried out. When the chiller is driven by fuel, it obtains the highest COP about 1.1.
14
In this case, the saved fuel consumption is calculated from annual cooling output divided by COP
15
of 1.1. Fig. 10 shows the payback period under different solar collector areas and fuel prices. The
16
payback period decreases a little when the area increases from 150m2 to 200m2. The reason is that
17
larger area of solar collector increases the working time and average system efficiency, and more
18
fuel consumption is saved. The payback period becomes longer when the area is larger than 225m2.
19
This means that the solar collector area of 225m2 is enough for the system. Further increase of
20
solar collector area mainly increases the initial investment. Areas of 200m2 and 225m2 are the
21
optimal areas for solar collector.
23 24
SC
M AN U
TE D
EP
AC C
22
RI PT
5
Fig. 10. Effect of solar collector area on payback period
5.3 Effect of storage tank volume
25 26
Fig. 11. Effect of tank volume on system efficiencies
27 28
Fig. 11 shows the impact of storage tank volume on the system efficiencies. When the tank
29
volume increases from 1.0 m3 to 4.5 m3, the average chiller COP decreases from 1.00 to 0.82, and
11
ACCEPTED MANUSCRIPT the average collector efficiency increases from 43.4% to 47.9%. These variations are related to the
2
heat capacity of the tank. Larger tank volume means higher tank heat capacity and lower average
3
tank temperature. When the tank temperature is lower, the chiller has lower heat source
4
temperature and lower efficiency, the solar collector has lower working temperature and higher
5
efficiency. For the solar COP, it increases from 0.31 to 0.36 when the tank volume increases from
6
1.0 m3 to 2.5 m3, but it decreases from 0.36 to 0.30 when the tank volume increases from 2.5 m3 to
7
4.0 m3. This is affected by the thermal loss and will be discussed later.
8
Fig.12. Effects of tank volume on working time and cooling output
SC
9
RI PT
1
10
Fig. 12 shows the impact of tank volume on the working time and total cooling output. The
12
working time varies between 8.8 hours and 10.6 hours, and the total cooling output varies between
13
390.9kWh and 456.0kWh. The longest working time and largest total cooling output are achieved
14
with tank volume of 2.5m3 corresponding to the highest solar COP. The variations of solar COP,
15
working time and total cooling output can be explained as follows. When the storage volume
16
increases from 1.0 m3 to 2.5 m3, the heat storage is enhanced. The system works longer and
17
delivers more cooling output. Besides, the temperature of storage tank is decreased, and the
18
thermal loss is decreased due to smaller temperature difference with the ambient. When the tank
19
volume increases from 2.5 m3 to 4.5 m3, the heat capacity of water in tank increases. It takes
20
longer for the preheating of tank which reduces the working time. Besides, the heat loss increases
21
due to larger tank volume, which reduces the system efficiency.
TE D
EP
AC C
22
M AN U
11
When the tank volume is larger, the thermal loss will be increased by larger contact area with
23
the ambient. When the tank volume is smaller, the thermal loss will be increased by larger
24
temperature difference with the ambient. The optimal tank volume of 2.5m3 is a balance between
25
two kinds of thermal losses.
26
5.4 Effect of the chiller cut-off driving temperature
27 28
Fig. 13. Effect of cut-off driving temperature on system efficiencies
29 12
ACCEPTED MANUSCRIPT Fig. 13 shows the effect of cut-off temperature on the average collector efficiency, average
2
chiller COP and solar COP. The chiller starts to work when the hot water inlet temperature
3
exceeds the cut-off temperature. When the cut-off temperature increases, the average heat source
4
temperature increases, which increases the average chiller COP and decreases the average
5
collector efficiency. With cut-off temperature from 90oC to 115oC, the average collector efficiency
6
from 47.0% to 37.7% and the average chiller COP from 0.86 to 0.96 are obtained. The solar COP
7
decreases a little from 0.36 to 0.33.
RI PT
1
Fig. 14 shows the impact of cut-off temperature on the working time and total cooling output.
9
When the cut-off temperature increases from 90oC to 115oC, the working time decreases a little
10
from 10.8 hours to 9.7 hours, and the total cooling output decreases a little from 462.7 kWh to
11
424.5 kWh. When the cut-off temperature is lower, the chiller needs less time for preheating of
12
water tank, which increases the working time. Considering that the chiller COP is relatively low
13
when the heat source temperature is lower than 90oC as shown in Table 1, the cut-off temperature
14
should not be smaller than this to ensure good performance. In this case, a cut-off temperature
15
between 90oC and 100 oC is proper.
18 19
Fig. 14. Effects of cut-off driving temperature on working time and cooling output
5.5 Performance comparisons with other systems
EP
17
TE D
16
M AN U
SC
8
Table 3 shows the system configurations, efficiencies and working times of different solar
21
absorption cooling systems. As is shown in the table, the first system consists of low temperature
22
solar collector and single effect LiBr-water absorption chiller. It has short working time, large
23
area/load ratio and low chiller COP. The second system consists of medium temperature solar
24
collector and single effect LiBr-water absorption chiller. It has longer working time, smaller
25
area/load ratio but still low COP. The third system consists of medium temperature solar collector
26
and double effect LiBr-water absorption chiller. It has small area/load ratio, high chiller COP.
27
However, the working time is short.
AC C
20
28 29
Table 3- Comparison between different solar absorption cooling systems
13
ACCEPTED MANUSCRIPT 1 Compared with the first system, the system discussed in this paper has lower area/load ratio,
3
longer working time and higher efficiencies. Compared with the second system, this system has
4
shorter working time but higher efficiencies with only stationary solar collector. Compared with
5
the third systems, this system has lower efficiency but longer working time with only stationary
6
solar collector. From the comparison, it can be seen that this system can work long, efficiently
7
with small area of collector.
RI PT
2
8 6. Conclusion
SC
9
In this paper, the solar driving variable effect LiBr-water absorption cooling system is
11
theoretically investigated. A stationary non-imaging CPC is used as heat source. ANN model of
12
50kW variable effect absorption chiller is built based on experimental data. Good agreement
13
between the prediction and the experimental data is achieved.
14
developed based on the model, and the solar driving 50 kW variable effect absorption cooling
15
system is then simulated in TRNSYS.
M AN U
10
New TRNSYS modules are
The system performance is calculated with the weather data of Miami and single zone
17
building model. The building has cooling load about 50 kW and air conditioning temperature of
18
25oC. Daily performance is calculated for a sunny day in July. The system works 10.4 hours with
19
daily average chiller COP of 0.88 and daily solar COP of 0.35. The impact of different parameters
20
on system performance are analyzed. The following conclusions can be made. (1) Large solar
21
collector area increases the average chiller COP, solar COP and working time. However, the
22
increase slows down a lot when the solar collector area exceeds a certain area, with which the
23
system payback period reaches its minimum. (2) Large heat storage volume means large heat
24
capacity. It increases the preheating time and thermal loss caused by contact area with the ambient.
25
Small heat storage volume increases the heat loss caused by temperature difference with the
26
ambient. An optimal storage tank size can be obtained to balance the two aspects. (3) When the
27
cut-off temperature is lower, the variable effect chiller needs less time for preheating of water tank,
28
which increase the working time.
29
AC C
EP
TE D
16
Compared with other studies, this system has lower area/load ratio and higher efficiencies
14
ACCEPTED MANUSCRIPT than the low temperature collector driving single effect LiBr-water absorption cooling systems.
2
This system has similar efficiencies and longer working time than the medium temperature
3
collector driving double effect LiBr-water absorption cooling systems. It can be seen that the
4
system in this paper is a good option for high efficient solar cooling with small collector area and
5
long working time.
RI PT
1
6 Acknowledgement
8
This work was supported by the key project of the Natural Science Foundation of China for
9
international academic exchanges (Grant No. 51561145012). The support from the China
10
Postdoctoral Science Foundation (Grant No. 2016M591670) and the Foundation for Innovative
11
Research Groups of the National Natural Science Foundation of China (Grant No. 51521004)
12
were also appreciated.
13
Reference
14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37
[1] Ullah K.R., Saidur R., Ping H.W., Akikur R.K., Shuvo N.H. A review of solar thermal refrigeration
M AN U
SC
7
and cooling methods. Renewable and Sustainable Energy Reviews. 2013;24:499-513. [2] Kim D.S., Ferreira C.A.I. Solar refrigeration options–a state-of-the-art review. International Journal
TE D
of refrigeration. 2008;31(1):3-15.
[3] Wang R.Z., Ge T.S., Chen C.J., Ma Q., Xiong Z.Q. Solar sorption cooling systems for residential applications: options and guidelines. International journal of refrigeration. 2009;32(4):638-60. [4] Zhai X.Q., Qu M., Li Y., Wang R.Z. A review for research and new design options of solar absorption cooling systems. Renewable and sustainable energy reviews. 2011;15(9):4416-23.
EP
[5] Assilzadeh F., Kalogirou S.A., Ali Y., Sopian K. Simulation and optimization of a LiBr solar absorption cooling system with evacuated tube collectors. Renewable Energy. 2005;30(8):1143-59. [6] Syed A., Izquierdo M., Rodriguez P., Maidment G., Missenden J., Lecuona A., et al. A novel
AC C
experimental investigation of a solar cooling system in Madrid. International Journal of refrigeration. 2005;28(6):859-71.
[7] Lizarte R., Izquierdo M., Marcos J.D., Palacios E. An innovative solar-driven directly air-cooled LiBr–H 2 O absorption chiller prototype for residential use. Energy and Buildings. 2012;47:1-11. [8] Mazloumi M., Naghashzadegan M., Javaherdeh K. Simulation of solar lithium bromide–water absorption cooling system with parabolic trough collector. Energy Conversion and Management. 2008;49(10):2820-32. [9] Bermejo P., Pino F.J., Rosa F. Solar absorption cooling plant in Seville. Solar Energy. 2010;84(8):1503-12. [10] Qu M., Yin H., Archer D.H. A solar thermal cooling and heating system for a building: experimental and model based performance analysis and design. Solar energy. 2010;84(2):166-82. [11] Velázquez N., García-Valladares O., Sauceda D., Beltrán R. Numerical simulation of a Linear Fresnel Reflector Concentrator used as direct generator in a Solar-GAX cycle. Energy Conversion and 15
ACCEPTED MANUSCRIPT [12] Herold K.E., Radermacher R., Klein S.A. Absorption chillers and heat pumps: CRC press, 1996. [13] Xu Z.Y., Wang R.Z. Absorption refrigeration cycles: categorized based on the cycle construction. International Journal of Refrigeration. 2016;62:114-36. [14] Xu Z.Y., Wang R.Z., Xia Z.Z. A novel variable effect LiBr-water absorption refrigeration cycle. Energy. 2013;60:457-63. [15] Xu Z.Y., Wang R.Z., Wang H.B. Experimental evaluation of a variable effect LiBr–water Refrigeration. 2015;59:135-43.
RI PT
absorption chiller designed for high-efficient solar cooling system. International Journal of [16] Al-Alili A., Islam M.D., Kubo I., Hwang Y., Radermacher R. Modeling of a solar powered absorption cycle for Abu Dhabi. Applied Energy. 2012;93:160-7.
[17] Florides G.A., Kalogirou S.A., Tassou S.A., Wrobel L.C. Modeling of the modern houses of
SC
Cyprus and energy consumption analysis. Energy. 2000;25(10):915-37.
[18] Fong K.F., Chow T.T., Lee C.K., Lin Z., Chan L.S. Comparative study of different solar cooling systems for buildings in subtropical city. Solar Energy. 2010;84(2):227-44.
[19] Lu Z.S., Wang R.Z., Xia Z.Z., Lu X.R., Yang C.B., Ma Y.C., et al. Study of a novel solar
M AN U
adsorption cooling system and a solar absorption cooling system with new CPC collectors. Renewable Energy. 2013;50:299-306.
[20] Labus J., Bruno J.C., Coronas A. Performance analysis of small capacity absorption chillers by using different modeling methods. Applied Thermal Engineering. 2013;58(1):305-13. [21] Hang Y., Qu M., Winston R., Jiang L., Widyolar B., Poiry H. Experimental based energy performance analysis and life cycle assessment for solar absorption cooling system at University of Californian, Merced. Energy and Buildings. 2014;82:746-57.
TE D
[22] Florides G.A., Kalogirou S.A., Tassou S.A., Wrobel L.C. Modelling and simulation of an absorption solar cooling system for Cyprus. Solar Energy. 2002;72(1):43-51.
EP
26
Management. 2010;51(3):434-45.
AC C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
16
ACCEPTED MANUSCRIPT Table 1- Experimental data of the variable effect absorption chiller [15] Tch1(oC)
Tch2(oC)
Tc1(oC)
Tc2 (oC)
Cooling power (kW)
COP
95.0
12.5
9.4
31.2
34.9
35.9
0.69
100.2
10.2
7.5
32.5
35.9
33.4
0.70
104.4
16.4
13.2
29.8
34.2
45.5
0.78
110.9
15.0
10.7
32.4
37.2
43.5
0.85
115.8
14.7
10.5
30.9
35.1
49.0
1.00
118.8
11.0
7.2
28.2
32.0
43.7
1.02
120.0
13.5
9.0
27.6
31.9
51.9
1.08
SC
RI PT
Tgen(oC)
AC C
EP
TE D
M AN U
1 – Steady state parameters of the chiller.
ACCEPTED MANUSCRIPT Table 2- ANN model of the variable effect absorption chiller Back propagation
Transfer function
tansig
Error
Mean square root error
Normalization
(-1,1)
Groups of data
450
Inputs
4
Outputs
2
Hidden layers
1
Hidden layer units
6
Correlation coefficient
AC C
EP
TE D
M AN U
SC
RI PT
Algorithm
0.994
ACCEPTED MANUSCRIPT Table 3- Comparison between different solar absorption cooling systems No.
Collector
Chiller type
Area/Load
Chiller COP
Solar COP
Daily working
2
(m /kW) 1
2
Evacuated
Single effect
tube
LiBr-water
PTC
Single effect
Reference
time
10.0
0.62~0.7
-
≈ 7 hours
[5]
3.3
0.67~0.76
-
11.5~12.3 hours
[8]
4.33
1.0~1.1
0.33~0.44
4.0
0.88
0.354
average
average
3
PTC
Double effect LiBr-water
CPC
Variable effect
4.6~5 hours
[10]
10.4 hours
This study
EP
TE D
M AN U
SC
LiBr-water
AC C
4
RI PT
LiBr-water
ACCEPTED MANUSCRIPT Pump-3
Solar collector
Pump-2
Cooling tower
Pump-1
Variable effect absorption chiller
Pump-4
AC C
EP
TE D
M AN U
SC
Auxiliary heat source
RI PT
Tank
Cooling load
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT A
P
A1 Th,in O
A2 Tc,in
Tc,out
A3 A4
Tch,in
A5 Qreq
Tch,out
AC C
EP
TE D
M AN U
A6
RI PT
O=tansig(A*lw+b2)
SC
A=tansig(P*iw+b1)
ACCEPTED MANUSCRIPT Weather Data Type 109
CPC Type 161
Type 2
Pump Type 3
Pump Type 3
Storage Tank Type 4
Auxiliary Heater Type 659
Pump Type 3
Chiller Type 151
Controller
Type 2 Controller
Type 2
Type 2
M AN U TE D EP AC C
Pump Type 3
SC
Controller
RI PT
Controller
Cooling Tower Type 510
Single Zone Load Type 12
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT Highlights ► The ANN model of the variable effect LiBr-water absorption chiller is built. ► The CPC driving variable effect absorption cooling system is simulated in TRNSYS.
AC C
EP
TE D
M AN U
SC
RI PT
► The system can work 10.4 hours a day with average COP of 0.88 and solar COP of 0.35.
S0960-1481(17)30574-8
DOI:
10.1016/j.renene.2017.06.069
Reference:
RENE 8936
To appear in:
Renewable Energy
Received Date: 7 April 2016 Revised Date:
27 April 2017
Accepted Date: 18 June 2017
Please cite this article as: Xu ZY, Wang RZ, Simulation of solar cooling system based on variable effect LiBr-water absorption chiller, Renewable Energy (2017), doi: 10.1016/j.renene.2017.06.069. 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.
ACCEPTED MANUSCRIPT
Simulation of solar cooling system based on variable effect LiBr-water absorption chiller Z.Y. Xu, R.Z. Wang*
RI PT
Institute of Refrigeration and Cryogenics, Shanghai Jiao Tong University, Shanghai 200240, China
* Corresponding author. Tel.: +86 21 34206548.
AC C
EP
TE D
M AN U
SC
E-mail address: [email protected] (R.Z. Wang).
ACCEPTED MANUSCRIPT 1
Simulation of solar cooling system based on variable effect
2
LiBr-water absorption chiller
3
Z.Y. Xu, R.Z. Wang*
4
Institute of Refrigeration and Cryogenics, Shanghai Jiao Tong University, Shanghai 200240, China
RI PT
5 Abstract
7
In solar absorption cooling system, the instability of solar power causes mismatch between the
8
solar collector and the absorption chiller. The variable effect absorption cycle was proposed to
9
improve this. In order to investigate its solar driving performance, a Compound Parabolic
10
Collector (CPC) driving variable effect LiBr-water absorption cooling system is simulated. Model
11
of the variable effect LiBr-water absorption chiller is built through artificial neural network (ANN)
12
modeling based on 450 groups of experimental data. Good agreement between the prediction and
13
experimental data is achieved with correlation coefficient of 0.994. The CPC driving absorption
14
cooling system is then built in TRaNsient SYstem Simulation program (TRNSYS) based on the
15
chiller model. The daily performance of this system is calculated and analyzed. The variable effect
16
chiller can work with low driving temperature, which guarantees a long working period. Besides,
17
the variable effect chiller has high COP under high driving temperature, which ensures a
18
competitive overall efficiency. The calculation shows that average chiller COP of 0.88 and solar
19
COP of 0.35 are obtained. The effects of solar collector area, storage tank volume and cut-off
20
driving temperature on the system performance are analyzed. The optimal solar collector area and
21
tank volume are obtained.
22
Keywords
23
Solar cooling, CPC, absorption cooling, simulation, TRNSYS
AC C
EP
TE D
M AN U
SC
6
24 25
Nomenclature
26
COP
coefficient of performance (-)
27
Q
heat transfer amount (kJ)
28
P
power (kW)
1
ACCEPTED MANUSCRIPT 1
η
collector efficiency (-)
2
A
collector area (m2)
3
I
irradiance (W/m2)
4
∆t
time step (s)
Subscripts and superscripts
7
ch1
chilling water inlet
8
ch2
chilling water outlet
9
c1
cooling water inlet
10
c2
cooling water outlet
11
col
collector
12
E
13
G
M AN U evaporation
generation
14 15
SC
6
RI PT
5
1. Introduction
The solar cooling system is an environmental friendly system which consumes renewable
17
energy. Among the potential options for solar cooling, solar powered absorption cooling system is
18
competitive due to the high efficiency and commercialization of absorption chiller. A solar
19
absorption cooling system consists of solar collector, heat storage, absorption chiller, cooling
20
tower, cooling load, pumps and control units. Solar absorption cooling systems with different solar
21
collectors and absorption chillers have been studied [1-4].
EP
AC C
22
TE D
16
In the past decades, solar absorption cooling system composed of low temperature solar
23
collector and single effect LiBr-water absorption chiller was popular in research. This system
24
adopts flat plate collector or evacuated tube collector as heat source and supplies hot water below
25
100oC to the absorption chiller. Assilzadeh et al. [5] studied an evacuated tube solar collector
26
driving single effect LiBr-water absorption cooling system in Malaysia. The system performance
27
was calculated in TRaNsient SYstem Simulation program (TRNSYS). The optimum design for
28
3.5 kW cooling output needed 35m2 collector, and worked about 5 hours a day. Syed et al. [6]
29
studied a 49.9m2 flat plate collector driving 35kW single effect LiBr-water absorption cooling
2
ACCEPTED MANUSCRIPT system in Madrid. Maximum instantaneous COP and daily average COP of 0.60 and 0.42 were
2
experimentally obtained. Lizarte et al. [7] studied a vacuum flat plate collector driving 4.5 kW
3
air-cooled single effect LiBr-water absorption cooling system in Madrid. Mean COP and solar
4
COP of 0.53 and 0.06 were experimentally obtained. These systems had short working time and
5
required large area of solar collector. Solar collector with higher working temperature can be
6
adopted to improve this. Mazloumi et al. [8] studied a Parabolic Trough Collector(PTC) driving
7
17.5kW single effect LiBr-water absorption cooling system by simulation. The chiller worked for
8
12 hours per day with COP from 0.67 to 0.76. The working time was longer but the system
9
efficiency was limited by the single effect absorption chiller.
SC
RI PT
1
In order to increase the system efficiency, absorption chillers with higher efficiencies including
11
double effect absorption chiller and Generator Absorber eXchange (GAX) absorption chiller can
12
be used. Bermejo et al. [9] experimentally studied a Linear Fresnel Reflector (LFR) driving double
13
effect LiBr-water absorption cooling system in Spain. The chiller was driven by pressurized hot
14
water powered by LFR with a backup natural gas burner. The chiller started to work at 12 a.m.
15
Daily average COP of the chiller was 1.1~1.25 and the solar cooling ratio was 0.44. Qu et al. [10]
16
experimentally studied a linear parabolic trough collector driving double effect LiBr-water
17
absorption cooling system. The chiller started to work after 12 a.m. The product of the chiller COP
18
and the collector efficiency was about 0.33~0.44. Velázquez et al. [11] studied a LFR direct
19
driving ammonia-water GAX absorption cooling system through simulation. The air-cooled GAX
20
cycle COP and solar COP reached 0.85 and 0.53 respectively. These systems required high heat
21
source temperature that they started to work late in the daytime.
TE D
EP
AC C
22
M AN U
10
The single effect and double effect LiBr-water absorption chillers obtain COPs about 0.7 and
23
1.2 under driving temperature around 90oC and 150oC respectively [12]. The chillers are not
24
flexible under variable driving temperature: their COPs vary little with the driving temperature
25
and their required driving temperatures are constrained in a small range. However, the solar heat
26
power has variable temperature. This makes the coupling between solar collector and absorption
27
chiller unstable. When the heat source has low temperature, it is unable to activate the absorption
28
chiller. When the heat source temperature is high, there will be a waste of energy grade. In order to
29
enhance the compatibility between solar collector and absorption chiller, various absorption cycles
30
have been developed [13] including the variable effect absorption refrigeration cycle previously 3
ACCEPTED MANUSCRIPT 1
proposed by the author [14]. The COP of variable effect cycle increases with driving temperature.
2
Later, a variable effect LiBr-water absorption chiller was built and tested based on this cycle [15]. As the variable effect absorption chiller is designed for high efficient solar cooling, the
4
characteristics and performance of solar powered variable effect absorption cooling system should
5
be investigated. In this paper, this solar cooling system is studied through simulation. Considering
6
the TRNSYS is a popular modeling software for solar absorption cooling system [5, 16-18], the
7
solar powered variable effect absorption cooling system is built in TRNSYS. An ANN model of
8
the variable effect absorption chiller is built from the experimental data and used in the TRNSYS
9
model. The performance of the solar driving variable effect absorption cooling system is studied
SC
10
RI PT
3
through simulation.
12
M AN U
11 2. Solar driving variable effect absorption cooling system
The system discussed in this paper is shown in Fig.1. It includes the solar collector, storage
14
tank, auxiliary heater, absorption chiller, cooling tower, cooling load and the pumps. Absorption
15
chiller integrated in the system is the variable effect LiBr-water absorption chiller. As the driving
16
temperature of the chiller is higher than 100oC, the medium temperature solar collector is essential
17
to supply heat source with sufficient temperature.
TE D
13
18 19
EP
Fig. 1 - The solar driving variable effect absorption cooling system
20
22
2.1. Variable effect LiBr-water absorption chiller
AC C
21
The chiller studied in this paper is built based on the variable effect absorption cycle.
23
According to the thermodynamic calculation, the variable effect LiBr-water absorption cycle
24
obtained COP from 0.8 to 1.06 under generation temperature from 95oC to 135oC, evaporation
25
temperature of 5oC, absorption temperature of 35oC and condensation temperature of 40oC
26
[14].The chiller COP was calculated from Eq.(1). COP =
(1)
27
The chiller prototype with rated power of 50kW was built based on this variable effect cycle.
28
The variable effect chiller prototype is also shown in Fig.1. Experimental results showed that the 4
ACCEPTED MANUSCRIPT 1
variable effect water-LiBr absorption chiller obtained COP from 0.69 to 1.08 under generation
2
temperature from 95.0oC to 120.0oC [15]. The COP increased as it is predicted by the calculation.
3
Table 1 shows some of the experimental results.
4 5
Table 1- Experimental data of the variable effect absorption chiller [15]
7
RI PT
6 2.2. Solar collector
Among different medium temperature solar collectors, the CPC is able to offer heat source of
9
enough temperature for the variable effect absorption chiller with simple stationary configuration.
10
A type of commercial CPC is used in this paper. It is an improved version of the collector used in
11
a previous study of solar driving cooling system [19]. In this stationary non-imaging CPC, copper
12
tube inside glass evacuated-tube is used for heat transfer. According to the tested data, the
13
collector achieves efficiencies of 50% and 48% under temperature of 130 oC and 150 oC
14
respectively, when solar irradiance is 900W/m2 and ambient temperature is 35 oC. The efficiency
15
of the solar collector can be calculated from the efficiency data from the manufacturer.
M AN U
SC
8
TE D
16 17
3. ANN model of the variable effect absorption chiller
18
3.1. Choice of modeling method
In order to precisely predict the performance of the variable effect absorption chiller, a proper
20
model is necessary. The commonly used modeling methods include the thermodynamic model
21
considering the heat transfer [12], the adapted characteristic equation model, the multivariate
22
polynomial regression model and the ANN model [20]. As parameters of the variable effect chiller
23
change a lot under different temperature conditions and cooling loads [14], the heat transfer
24
coefficients cannot be set as constant for simulation, and the thermodynamic model is not effective.
25
Among other modeling methods, the ANN model has better accuracy [20].
26
3.2. Preparation of experimental data
AC C
EP
19
27
In order to build the ANN model, the chiller was tested under different conditions, and 450
28
groups of data under steady state working condition were selected. Each group of data contained
29
the heat source temperature, generation temperatures of high pressure generator and low pressure 5
ACCEPTED MANUSCRIPT generator, inlet and outlet cooling water temperatures, condensation temperature, evaporation
2
temperature, cooling power and COP. The experimental parameters here were obtained under
3
constant flow rates. The heat source temperature varied between 85.34oC and 140.96 oC. The
4
cooling water inlet temperature varied between 24.74 oC and 34.25 oC. The cooling water outlet
5
temperature varied between 28.40 oC and 39.22 oC. The chilled water inlet temperature varied
6
between 9.05 oC and 20.83 oC. The chilled water outlet temperature varied between 4.43 oC and
7
16.25 oC. The cooling power varied between 30kW and 60kW, and the COP varied between 0.6
8
and 1.1.
9
3.3. ANN model
SC
RI PT
1
The commonly used ANN models include feed-forward neural network, feedback neural
11
network, etc. The Back Propagation (BP) network is a simple and effective feed-forward neural
12
network. Considering that the chiller modeling in this paper is simple and doesn’t need feedback,
13
the BP network is chosen.
M AN U
10
In the chiller model, the input parameters include the heat source inlet temperature, cooling
15
water inlet temperature, chilled water inlet temperature and the required cooling output. With the
16
constant water flow rates, these four parameters are enough to decide the cooling water outlet
17
temperature and chilled water outlet temperature. The cooling power and COP can be further
18
calculated from the temperatures and flow rates. In this case, the network needs 4 inputs and 2
19
outputs. According to the calculation, one hidden layer with 6 units is enough for the precise
20
prediction. Fig.2 shows the construction of the network including input layer (P), hidden layer (A),
21
output layer (O) and the transfer functions. Considering that the calculation from inlet
22
temperatures and required cooling output to the outlet temperatures is nonlinear, the hyperbolic
23
tangent sigmoid function (tansig) is used both for the hidden layer calculation and output layer
24
calculation. The mean square root error is minimized during the iteration.
AC C
EP
TE D
14
25 26
Fig. 2 - Construction of the ANN model
27 28
The ANN model is built in the MATLAB neural network tool box. Normalization of the
29
experimental data is carried out in the range of (-1, 1). In the model building, part of the
6
ACCEPTED MANUSCRIPT 1
experimental data are used for model training, and the rest data are used for model validation.
2
Table 2 shows the summary of this ANN model. A correlation coefficient (R2) of 0.994 is obtained
3
which proves the effectiveness of the ANN model.
4 5 6
4. Simulation of the solar cooling system in TRNSYS
SC
7
RI PT
Table 2- ANN model of the variable effect absorption chiller
The system performance is studied in the software TRNSYS. New modules (named as “type”
9
in TRNSYS) are built both for the variable effect absorption chiller and the commercial CPC. In
10
the chiller module, the flow rates of cooling water and chilled water are set as 20m3/h and 10m3/h
11
respectively, which are the same with experimental values [15]. Considering the chiller was tested
12
under limited temperature range, the ANN model can be invalid when its inputs deviate too much
13
from the tested range. In the simulation, this kind of invalid inputs are avoided to ensure
14
reasonable prediction. Module of the CPC is built from the manufacturer supplying data. The
15
variable effect absorption chiller module and CPC module are named as “type 151” and “type 161”
16
respectively.
19 20
TE D
EP
18
Fig. 3 - CPC driving variable effect LiBr-water absorption cooling system in TRNSYS
AC C
17
M AN U
8
Fig.3 shows the system in TRNSYS. The blue blocks, grey blocks and white blocks represent
21
the self-built modules, component modules and controller modules respectively. The component
22
modules are connected according to Fig.1. The controllers control the operation of solar collector
23
pump, auxiliary heater, hot water pump, cooling water pump and chilled water pump. The
24
controlling logics are as follows: the solar collector pump operates when the collector working
25
temperature is higher than the tank temperature; the auxiliary heater operates when its inlet
26
temperature is lower than the setting temperature; the chiller works when the hot water
27
temperature and room temperature are both higher than their setting values. The weather data
7
ACCEPTED MANUSCRIPT 1
supplied by the software are used for simulation. The TMY2 file named “US-FL-Miami-12839” is
2
selected. For building simulation, the single zone model of Type 12c is used. The air conditioning
3
temperature of the building is set as 25oC. Several efficiencies are used to evaluate the system performance. The average chiller COP is
5
defined as the ratio of total cooling output to total heat input in Eq. (2). The average collector
6
efficiency is defined as the ratio of total heat output to total irradiance input when the collector
7
works in Eq. (3). The solar COP is defined as the ratio of total cooling output to total irradiance
8
input in Eq. (4).
COP =
∑ Δ ∑ Δ
η =
SC
∑ Δ ∑ Δ
∑ Δ ∑ Δ
9 10
5. Results and discussions
M AN U
η =
RI PT
4
(2)
(3) (4)
The system is simulated with a time step of 0.1 hour. The performance in a sunny day in July
12
is studied. In order to study the solar driving performance of the system, the auxiliary heater is not
13
activated during the simulation. The daily performance is calculated based on the following
14
conditions: the rated power of variable effect absorption chiller is 50kW; the collector area is
15
200m2; the storage tank has volume of 3m3; the chiller starts to work when the heat source
16
temperature exceeds 100 oC. The daily performance is presented, and the effects of some key
17
parameters on the system performance are studied.
18
5.1. Daily performance
EP
AC C
19
TE D
11
Fig. 4 shows the solar irradiance and ambient temperature. The irradiance is higher than zero
20
from 6:12 to 19:00. It reaches the maximum of 989.4W/m2 at 13:12. The ambient temperature
21
varies between 25.6 oC and 30.6oC.
22 23
Fig.4 – Solar irradiance and ambient temperature
24 25
Fig. 5. Daily performance of the solar collector 8
ACCEPTED MANUSCRIPT 1 Fig.5 shows the solar collector temperatures and efficiency. The following points are
3
essential to be mentioned. (1) The collector efficiency varies between 0.3 and 0.51 from 9:00 to
4
16:00. (2) From 7:48, the collector starts to deliver thermal power to the tank, when the collector
5
temperature reaches 95.1oC and exceeds the tank temperature. The tank temperature stays stable
6
with a high temperature before 7:48 because the heat loss coefficient of the tank is small. (3) The
7
collector temperature reaches its maximum value of 148.3oC at 15:30, which is not at the time of
8
maximum irradiance. This is caused by the consumption by chiller.
RI PT
2
Fig. 6 shows the temperature and COP of the variable effect absorption chiller. The chiller
10
works between 8:54 and 19:24 when the hot water temperature is higher than its cut-off driving
11
temperature. During this period, the inlet and outlet heat source temperatures vary in the range of
12
100oC~146.6 oC and 94.5 oC~143.3 oC respectively. The inlet and outlet cooling temperatures vary
13
in the range of 27.2 oC ~29.8 oC and 31.0 oC~ 34.7oC respectively. The inlet and outlet chilled
14
water temperatures vary in the range of 13.4oC~14.6oC and 9.9oC~10.3oC respectively. The COP
15
varies between 0.78 and 1.1. The variations of heating source temperature and COP are similar.
M AN U
SC
9
17 18
TE D
16
Fig. 6. Daily performance of the absorption chiller
From 8:54 to 9:18, the heat input from solar collector is less than the consumption of chiller,
20
the tank temperature cannot keep higher than the cut-off driving temperature of absorption chiller,
21
and fluctuations happens. From 9:18 to 11:00, the COP rises and falls due to the change of solar
22
irradiance. Later, the COP grows with the heat source temperature and achieves the highest value
23
from 15:18 to 15:54. The COP and heat source temperature increase slowly from 8:54 to 15:18 but
24
fall quickly from 15:54 to 19:24. The reasons are as follows. (1) From 8:54 to 15:18, the chiller
25
COP is low which means higher heat consumption rate. The large heat consumption rate slows
26
down the increase of tank temperature. (2) When the heat input from solar collector becomes
27
smaller than the heat consumption of absorption chiller at 15:18, the heat source temperature and
28
chiller COP start to decrease. (3) From 15:54 to 19:24, solar irradiance keeps decreasing. On the
29
other hand, the chiller consumes more thermal energy due to lower COP. In this case, the heat
30
source temperature and COP decrease quickly.
AC C
EP
19
9
ACCEPTED MANUSCRIPT 1 2
Fig. 7. Daily energy flows
3 Fig. 7 shows the energy flows in the whole day. The received solar power, collector gained
5
thermal power, heat consumption power, cooling output power and cooling load are presented.
6
The received solar power and collector gained thermal power have similar variations. The heat
7
consumption power varies between 38.2kW and 68.4kW. The cooling load varies between 39.7kW
8
and 49.8kW. Both the heat consumption power and cooling load are high at the beginning of
9
operation due to the high initial room temperature. The average chiller COP and solar COP are
SC
10
RI PT
4
0.88 and 0.35 respectively.
In this section, the daily variations of system parameters are presented. In the following parts,
12
the overall system performance will be evaluated. The analyzed parameters include the solar
13
collector area, storage tank volume and cut-off driving temperature.
14
5.2 Effect of solar collector area
17 18 19
efficiency and solar COP.
TE D
16
Fig. 8 shows the impact of solar collector area on the average chiller COP, average collector
Fig. 8 - Effect of solar collector area on system efficiencies
EP
15
M AN U
11
When the solar collector area increases from 150m2 to 275m2, the overall collector efficiency
21
falls from 49.1% to 43.7% and the chiller average COP increases from 0.82 to 1.00. The reason is
22
that large collector area increases the average tank temperatures. This further decreases the
23
average collector efficiency and increases the average chiller COP. The increase of solar COP
24
from 0.24 to 0.37 indicates that the large collector area is beneficial to the solar COP.
AC C
20
25
Fig. 9 shows the impact of solar collector area on the chiller working time. The working
26
time increases from 6.8 hours to 10.8 hours when the area increases from 150m2 to 275m2. At
27
the same time, the total cooling output increases from 309.1 kWh to 461.3kWh. Both the solar
28
COP and total cooling output increase with larger solar collector area. However, they increase
10
ACCEPTED MANUSCRIPT 1
slowly when the collector area exceeds 200m2.
2 3
Fig. 9 - Effects of solar collector area on working time and cooling output
4 Considering that the collector area is the key parameter affecting the economic performance
6
[16], economic performance under different collector areas is calculated. There are many criteria
7
for economic analysis including the annualized operation cost, lift cycle cost, payback period and
8
net present value [16, 21, 22]. Here, the payback period is used which refers to the time needed for
9
the cumulative fuel savings to equal the initial investment[16]. The investment of absorption
10
chiller is about 22000$ according to our experiment. The investment for CPC is 165$/m2 [21]. The
11
investment for a 100kW cooling tower and 3m3 storage tank are 6760$ and 2700$ respectively[16].
12
The fuel prices of 0.08$/kWh, 0.1$/kWh and 0.12$/kWh are used for calculation [16]. The annual
13
simulation is carried out. When the chiller is driven by fuel, it obtains the highest COP about 1.1.
14
In this case, the saved fuel consumption is calculated from annual cooling output divided by COP
15
of 1.1. Fig. 10 shows the payback period under different solar collector areas and fuel prices. The
16
payback period decreases a little when the area increases from 150m2 to 200m2. The reason is that
17
larger area of solar collector increases the working time and average system efficiency, and more
18
fuel consumption is saved. The payback period becomes longer when the area is larger than 225m2.
19
This means that the solar collector area of 225m2 is enough for the system. Further increase of
20
solar collector area mainly increases the initial investment. Areas of 200m2 and 225m2 are the
21
optimal areas for solar collector.
23 24
SC
M AN U
TE D
EP
AC C
22
RI PT
5
Fig. 10. Effect of solar collector area on payback period
5.3 Effect of storage tank volume
25 26
Fig. 11. Effect of tank volume on system efficiencies
27 28
Fig. 11 shows the impact of storage tank volume on the system efficiencies. When the tank
29
volume increases from 1.0 m3 to 4.5 m3, the average chiller COP decreases from 1.00 to 0.82, and
11
ACCEPTED MANUSCRIPT the average collector efficiency increases from 43.4% to 47.9%. These variations are related to the
2
heat capacity of the tank. Larger tank volume means higher tank heat capacity and lower average
3
tank temperature. When the tank temperature is lower, the chiller has lower heat source
4
temperature and lower efficiency, the solar collector has lower working temperature and higher
5
efficiency. For the solar COP, it increases from 0.31 to 0.36 when the tank volume increases from
6
1.0 m3 to 2.5 m3, but it decreases from 0.36 to 0.30 when the tank volume increases from 2.5 m3 to
7
4.0 m3. This is affected by the thermal loss and will be discussed later.
8
Fig.12. Effects of tank volume on working time and cooling output
SC
9
RI PT
1
10
Fig. 12 shows the impact of tank volume on the working time and total cooling output. The
12
working time varies between 8.8 hours and 10.6 hours, and the total cooling output varies between
13
390.9kWh and 456.0kWh. The longest working time and largest total cooling output are achieved
14
with tank volume of 2.5m3 corresponding to the highest solar COP. The variations of solar COP,
15
working time and total cooling output can be explained as follows. When the storage volume
16
increases from 1.0 m3 to 2.5 m3, the heat storage is enhanced. The system works longer and
17
delivers more cooling output. Besides, the temperature of storage tank is decreased, and the
18
thermal loss is decreased due to smaller temperature difference with the ambient. When the tank
19
volume increases from 2.5 m3 to 4.5 m3, the heat capacity of water in tank increases. It takes
20
longer for the preheating of tank which reduces the working time. Besides, the heat loss increases
21
due to larger tank volume, which reduces the system efficiency.
TE D
EP
AC C
22
M AN U
11
When the tank volume is larger, the thermal loss will be increased by larger contact area with
23
the ambient. When the tank volume is smaller, the thermal loss will be increased by larger
24
temperature difference with the ambient. The optimal tank volume of 2.5m3 is a balance between
25
two kinds of thermal losses.
26
5.4 Effect of the chiller cut-off driving temperature
27 28
Fig. 13. Effect of cut-off driving temperature on system efficiencies
29 12
ACCEPTED MANUSCRIPT Fig. 13 shows the effect of cut-off temperature on the average collector efficiency, average
2
chiller COP and solar COP. The chiller starts to work when the hot water inlet temperature
3
exceeds the cut-off temperature. When the cut-off temperature increases, the average heat source
4
temperature increases, which increases the average chiller COP and decreases the average
5
collector efficiency. With cut-off temperature from 90oC to 115oC, the average collector efficiency
6
from 47.0% to 37.7% and the average chiller COP from 0.86 to 0.96 are obtained. The solar COP
7
decreases a little from 0.36 to 0.33.
RI PT
1
Fig. 14 shows the impact of cut-off temperature on the working time and total cooling output.
9
When the cut-off temperature increases from 90oC to 115oC, the working time decreases a little
10
from 10.8 hours to 9.7 hours, and the total cooling output decreases a little from 462.7 kWh to
11
424.5 kWh. When the cut-off temperature is lower, the chiller needs less time for preheating of
12
water tank, which increases the working time. Considering that the chiller COP is relatively low
13
when the heat source temperature is lower than 90oC as shown in Table 1, the cut-off temperature
14
should not be smaller than this to ensure good performance. In this case, a cut-off temperature
15
between 90oC and 100 oC is proper.
18 19
Fig. 14. Effects of cut-off driving temperature on working time and cooling output
5.5 Performance comparisons with other systems
EP
17
TE D
16
M AN U
SC
8
Table 3 shows the system configurations, efficiencies and working times of different solar
21
absorption cooling systems. As is shown in the table, the first system consists of low temperature
22
solar collector and single effect LiBr-water absorption chiller. It has short working time, large
23
area/load ratio and low chiller COP. The second system consists of medium temperature solar
24
collector and single effect LiBr-water absorption chiller. It has longer working time, smaller
25
area/load ratio but still low COP. The third system consists of medium temperature solar collector
26
and double effect LiBr-water absorption chiller. It has small area/load ratio, high chiller COP.
27
However, the working time is short.
AC C
20
28 29
Table 3- Comparison between different solar absorption cooling systems
13
ACCEPTED MANUSCRIPT 1 Compared with the first system, the system discussed in this paper has lower area/load ratio,
3
longer working time and higher efficiencies. Compared with the second system, this system has
4
shorter working time but higher efficiencies with only stationary solar collector. Compared with
5
the third systems, this system has lower efficiency but longer working time with only stationary
6
solar collector. From the comparison, it can be seen that this system can work long, efficiently
7
with small area of collector.
RI PT
2
8 6. Conclusion
SC
9
In this paper, the solar driving variable effect LiBr-water absorption cooling system is
11
theoretically investigated. A stationary non-imaging CPC is used as heat source. ANN model of
12
50kW variable effect absorption chiller is built based on experimental data. Good agreement
13
between the prediction and the experimental data is achieved.
14
developed based on the model, and the solar driving 50 kW variable effect absorption cooling
15
system is then simulated in TRNSYS.
M AN U
10
New TRNSYS modules are
The system performance is calculated with the weather data of Miami and single zone
17
building model. The building has cooling load about 50 kW and air conditioning temperature of
18
25oC. Daily performance is calculated for a sunny day in July. The system works 10.4 hours with
19
daily average chiller COP of 0.88 and daily solar COP of 0.35. The impact of different parameters
20
on system performance are analyzed. The following conclusions can be made. (1) Large solar
21
collector area increases the average chiller COP, solar COP and working time. However, the
22
increase slows down a lot when the solar collector area exceeds a certain area, with which the
23
system payback period reaches its minimum. (2) Large heat storage volume means large heat
24
capacity. It increases the preheating time and thermal loss caused by contact area with the ambient.
25
Small heat storage volume increases the heat loss caused by temperature difference with the
26
ambient. An optimal storage tank size can be obtained to balance the two aspects. (3) When the
27
cut-off temperature is lower, the variable effect chiller needs less time for preheating of water tank,
28
which increase the working time.
29
AC C
EP
TE D
16
Compared with other studies, this system has lower area/load ratio and higher efficiencies
14
ACCEPTED MANUSCRIPT than the low temperature collector driving single effect LiBr-water absorption cooling systems.
2
This system has similar efficiencies and longer working time than the medium temperature
3
collector driving double effect LiBr-water absorption cooling systems. It can be seen that the
4
system in this paper is a good option for high efficient solar cooling with small collector area and
5
long working time.
RI PT
1
6 Acknowledgement
8
This work was supported by the key project of the Natural Science Foundation of China for
9
international academic exchanges (Grant No. 51561145012). The support from the China
10
Postdoctoral Science Foundation (Grant No. 2016M591670) and the Foundation for Innovative
11
Research Groups of the National Natural Science Foundation of China (Grant No. 51521004)
12
were also appreciated.
13
Reference
14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37
[1] Ullah K.R., Saidur R., Ping H.W., Akikur R.K., Shuvo N.H. A review of solar thermal refrigeration
M AN U
SC
7
and cooling methods. Renewable and Sustainable Energy Reviews. 2013;24:499-513. [2] Kim D.S., Ferreira C.A.I. Solar refrigeration options–a state-of-the-art review. International Journal
TE D
of refrigeration. 2008;31(1):3-15.
[3] Wang R.Z., Ge T.S., Chen C.J., Ma Q., Xiong Z.Q. Solar sorption cooling systems for residential applications: options and guidelines. International journal of refrigeration. 2009;32(4):638-60. [4] Zhai X.Q., Qu M., Li Y., Wang R.Z. A review for research and new design options of solar absorption cooling systems. Renewable and sustainable energy reviews. 2011;15(9):4416-23.
EP
[5] Assilzadeh F., Kalogirou S.A., Ali Y., Sopian K. Simulation and optimization of a LiBr solar absorption cooling system with evacuated tube collectors. Renewable Energy. 2005;30(8):1143-59. [6] Syed A., Izquierdo M., Rodriguez P., Maidment G., Missenden J., Lecuona A., et al. A novel
AC C
experimental investigation of a solar cooling system in Madrid. International Journal of refrigeration. 2005;28(6):859-71.
[7] Lizarte R., Izquierdo M., Marcos J.D., Palacios E. An innovative solar-driven directly air-cooled LiBr–H 2 O absorption chiller prototype for residential use. Energy and Buildings. 2012;47:1-11. [8] Mazloumi M., Naghashzadegan M., Javaherdeh K. Simulation of solar lithium bromide–water absorption cooling system with parabolic trough collector. Energy Conversion and Management. 2008;49(10):2820-32. [9] Bermejo P., Pino F.J., Rosa F. Solar absorption cooling plant in Seville. Solar Energy. 2010;84(8):1503-12. [10] Qu M., Yin H., Archer D.H. A solar thermal cooling and heating system for a building: experimental and model based performance analysis and design. Solar energy. 2010;84(2):166-82. [11] Velázquez N., García-Valladares O., Sauceda D., Beltrán R. Numerical simulation of a Linear Fresnel Reflector Concentrator used as direct generator in a Solar-GAX cycle. Energy Conversion and 15
ACCEPTED MANUSCRIPT [12] Herold K.E., Radermacher R., Klein S.A. Absorption chillers and heat pumps: CRC press, 1996. [13] Xu Z.Y., Wang R.Z. Absorption refrigeration cycles: categorized based on the cycle construction. International Journal of Refrigeration. 2016;62:114-36. [14] Xu Z.Y., Wang R.Z., Xia Z.Z. A novel variable effect LiBr-water absorption refrigeration cycle. Energy. 2013;60:457-63. [15] Xu Z.Y., Wang R.Z., Wang H.B. Experimental evaluation of a variable effect LiBr–water Refrigeration. 2015;59:135-43.
RI PT
absorption chiller designed for high-efficient solar cooling system. International Journal of [16] Al-Alili A., Islam M.D., Kubo I., Hwang Y., Radermacher R. Modeling of a solar powered absorption cycle for Abu Dhabi. Applied Energy. 2012;93:160-7.
[17] Florides G.A., Kalogirou S.A., Tassou S.A., Wrobel L.C. Modeling of the modern houses of
SC
Cyprus and energy consumption analysis. Energy. 2000;25(10):915-37.
[18] Fong K.F., Chow T.T., Lee C.K., Lin Z., Chan L.S. Comparative study of different solar cooling systems for buildings in subtropical city. Solar Energy. 2010;84(2):227-44.
[19] Lu Z.S., Wang R.Z., Xia Z.Z., Lu X.R., Yang C.B., Ma Y.C., et al. Study of a novel solar
M AN U
adsorption cooling system and a solar absorption cooling system with new CPC collectors. Renewable Energy. 2013;50:299-306.
[20] Labus J., Bruno J.C., Coronas A. Performance analysis of small capacity absorption chillers by using different modeling methods. Applied Thermal Engineering. 2013;58(1):305-13. [21] Hang Y., Qu M., Winston R., Jiang L., Widyolar B., Poiry H. Experimental based energy performance analysis and life cycle assessment for solar absorption cooling system at University of Californian, Merced. Energy and Buildings. 2014;82:746-57.
TE D
[22] Florides G.A., Kalogirou S.A., Tassou S.A., Wrobel L.C. Modelling and simulation of an absorption solar cooling system for Cyprus. Solar Energy. 2002;72(1):43-51.
EP
26
Management. 2010;51(3):434-45.
AC C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
16
ACCEPTED MANUSCRIPT Table 1- Experimental data of the variable effect absorption chiller [15] Tch1(oC)
Tch2(oC)
Tc1(oC)
Tc2 (oC)
Cooling power (kW)
COP
95.0
12.5
9.4
31.2
34.9
35.9
0.69
100.2
10.2
7.5
32.5
35.9
33.4
0.70
104.4
16.4
13.2
29.8
34.2
45.5
0.78
110.9
15.0
10.7
32.4
37.2
43.5
0.85
115.8
14.7
10.5
30.9
35.1
49.0
1.00
118.8
11.0
7.2
28.2
32.0
43.7
1.02
120.0
13.5
9.0
27.6
31.9
51.9
1.08
SC
RI PT
Tgen(oC)
AC C
EP
TE D
M AN U
1 – Steady state parameters of the chiller.
ACCEPTED MANUSCRIPT Table 2- ANN model of the variable effect absorption chiller Back propagation
Transfer function
tansig
Error
Mean square root error
Normalization
(-1,1)
Groups of data
450
Inputs
4
Outputs
2
Hidden layers
1
Hidden layer units
6
Correlation coefficient
AC C
EP
TE D
M AN U
SC
RI PT
Algorithm
0.994
ACCEPTED MANUSCRIPT Table 3- Comparison between different solar absorption cooling systems No.
Collector
Chiller type
Area/Load
Chiller COP
Solar COP
Daily working
2
(m /kW) 1
2
Evacuated
Single effect
tube
LiBr-water
PTC
Single effect
Reference
time
10.0
0.62~0.7
-
≈ 7 hours
[5]
3.3
0.67~0.76
-
11.5~12.3 hours
[8]
4.33
1.0~1.1
0.33~0.44
4.0
0.88
0.354
average
average
3
PTC
Double effect LiBr-water
CPC
Variable effect
4.6~5 hours
[10]
10.4 hours
This study
EP
TE D
M AN U
SC
LiBr-water
AC C
4
RI PT
LiBr-water
ACCEPTED MANUSCRIPT Pump-3
Solar collector
Pump-2
Cooling tower
Pump-1
Variable effect absorption chiller
Pump-4
AC C
EP
TE D
M AN U
SC
Auxiliary heat source
RI PT
Tank
Cooling load
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT A
P
A1 Th,in O
A2 Tc,in
Tc,out
A3 A4
Tch,in
A5 Qreq
Tch,out
AC C
EP
TE D
M AN U
A6
RI PT
O=tansig(A*lw+b2)
SC
A=tansig(P*iw+b1)
ACCEPTED MANUSCRIPT Weather Data Type 109
CPC Type 161
Type 2
Pump Type 3
Pump Type 3
Storage Tank Type 4
Auxiliary Heater Type 659
Pump Type 3
Chiller Type 151
Controller
Type 2 Controller
Type 2
Type 2
M AN U TE D EP AC C
Pump Type 3
SC
Controller
RI PT
Controller
Cooling Tower Type 510
Single Zone Load Type 12
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT Highlights ► The ANN model of the variable effect LiBr-water absorption chiller is built. ► The CPC driving variable effect absorption cooling system is simulated in TRNSYS.
AC C
EP
TE D
M AN U
SC
RI PT
► The system can work 10.4 hours a day with average COP of 0.88 and solar COP of 0.35.