Professional Documents
Culture Documents
com
ScienceDirect
Solar Energy 117 (2015) 192–202
www.elsevier.com/locate/solener
Received 26 November 2014; received in revised form 28 March 2015; accepted 21 April 2015
Available online 16 May 2015
Abstract
Based on finite volume method, the steady-state thermal performances of the flat-plate solar collector are studied by taking account of
absorber plate thickness, collector tube spacing, collector tube length, collector tube diameter and insulating layer thickness. A physical
model of gilled flat-plate solar collector is built, then the numerical simulation of the model is carried out and the numerical simulation
results are compared and analyzed with experimental results. The results show that: Either increasing the absorber plate thickness or
reducing the collector tube spacing can significantly improve the instantaneous efficiency of the collector. Setting the solar radiation
intensity of 700 W/m2 and the environmental speed of 4 m/s, when the absorber plate thickness increases from 0.1 mm to 2.1 mm,
the collector instantaneous efficiency increases from 46.57% to 64.03%. When the collector tube spacing decreases from 170 mm to
50 mm, the collector instantaneous efficiency increases from 52.81% to 66.01%. Reducing the collector tube length and increasing col-
lector tube diameter are both conducive to improve the instantaneous efficiency of the collector. When the collector tube length decreases
from 2800 mm to 1200 mm, the collector instantaneous efficiency increases from 57.50% to 60.12%. When the collector tube diameters
increases from 8 mm to 20 mm, the collector instantaneous efficiency increases from 56.18% to 63.97%. When the thickness of insulating
layer is 30 mm or more, increasing its thickness has no significant effect on improving the instantaneous efficiency of the collector. The
research results are helpful to optimize the design parameters of the flat-plate solar collector.
Ó 2015 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.solener.2015.04.027
0038-092X/Ó 2015 Elsevier Ltd. All rights reserved.
Z. Jiandong et al. / Solar Energy 117 (2015) 192–202 193
Nomenclature
standards (ANSI/ASHRAE, 2003; EN 12975-2, 2006) and from glass which is easy to burst of tube, fouling and
articles (Norton, 1992; Garg, 1985) about solar thermal unable to building integration of solar energy systems.
application have been published. The flat-plate solar collectors have the advantages of sim-
Flat-plate and vacuum tube solar collectors are the most ple structure, high pressure bearing, durable, low mainte-
widely used heaters among the solar collector systems. nance rate, high heat efficiency and low production costs.
Although vacuum tube solar collector enjoys advantages It will become the main trend in the future for the high
of vacuum insulation and mature technology, it is made demand of building integration of solar energy systems
194 Z. Jiandong et al. / Solar Energy 117 (2015) 192–202
Table 2
Physical parameters of the materials.
Aluminum Copper Rockwool Glass cover
3
Density (kg m ) 2700 8920 140 2220
Specific heat capacity J kg1 °C1) 900 385 1220 830
Thermal conductivity (W m1 k1) 120 398 0.039 0.76
Transmissivity 0 0 0 0.9
Emissivity 0.1 0.1 – 0.9
Absorptivity 0.9 0.9 – 0.1
2.3. Mathematical model Fig. 2. Unit structure diagram of a flat-plate solar collector (a and b are
an enlarged view of a dashed box in Fig. 1).
Based on the above assumptions, the mathematical
models required in solving equations include:
(1) flow and heat transfer equations for general control Which Gk represents the average velocity gradient gener-
(Patankar, 1980): ated due to entry of the turbulent kinetic energy caused by
k, the calculation formula:
divðqU /Þ ¼ divðC/ grad/Þ þ S / ð1Þ ( "
2 2 2 # 2
(2) The solid heat conduction equations (Tao, 2001): @u @v @w @u @v
Gk ¼ lt 2 þ þ þ þ
@x @y @z @y @x
@2T @2T @2T
þ þ ¼0 ð2Þ 2 2 )
@x2 @y 2 @z2 @u @w @v @w
þ þ þ þ ð5Þ
(3) k equation and e equations were (Tao, 2001): @z @x @z @y
@ @ l @k k2
ðqkui Þ ¼ lþ t þ Gk qe ð3Þ lt ¼ qcl ð6Þ
@xi @xj rk @xj e
(4) Boussinesq assumption equation (FLUENT Inc,
2003):
@ @ lt @e e e2
ðqkui Þ ¼ lþ þ c1 Gk c2 q ð4Þ
@xi @xj re @xj k k ðq q0 Þg q0 bðT T 0 Þg ð7Þ
196 Z. Jiandong et al. / Solar Energy 117 (2015) 192–202
(5) For a scattering, absorption and emission properties method. For partitions solving boundary coupling calcula-
of the medium, at the position ~ r in the ~
f direction tion method, the speed of the iterative process of conver-
at the radiation transfer equation (Tao, 2001) is: gence is depends on the information transfer of the
coupling boundary. For the entire solution method is an
r; ~
r ðIð~ f Þ~ r; ~
f Þ þ ða0 þ bf ÞIð~ fÞ effective way to solve the coupled problem, it also can sig-
04 Z 4p nificantly reduce the computation time.
rT bf ! !
¼ a0 n2 þ r; f 0 ÞUð~
Ið~ f f 0 ÞdX ð8Þ At the state of turbulence entrance, using the standard
p 4p 0
k–e two-equation model calculations. Select Discrete
Ordinates radiation model to calculate. In the solving pro-
cess, when the variables residuals of continuity equation,
2.4. Numerical model momentum equation, k equation, e equation is less than
103 and the variables residual of energy equations and
Velocity inlet boundary conditions are adopted in radiation equation is less than 107 that it is considered
numerical calculations. The velocity value of water inlet convergence of calculation.
and air inlet for each boundary node is specified, namely:
u ¼ uin ; v ¼ 0; w¼0 2.5. Grid independence test
Outlet is pressure outlet boundary conditions, namely:
In this paper, it is divided of using the hexahedron struc-
P ¼ P out ture grid. Take model parameters in Table 3 for example,
Wall using no-slip boundary condition, namely close to the ambient temperature Ta is 15 °C, the working fluid inlet
the wall of the fluid: temperature is 20 °C, the solar radiation intensity I is
700 W/m2, computing results under different numbers of
u¼v¼w¼0 grids are shown in Table 4. The increasing number of grids
The symmetry boundary conditions are set to both sides will have little impact on the calculation results when the
of the model (as shown in Fig. 2(b)). grids are 2.983 million, the calculation results are shown
During the simulation, using the Boussinesq hypothesis, in Table 4. The instantaneous efficiency of error between
which can help reduce the computation amount under con- the calculation results of the grid number of 3.681 million
dition of small temperature difference and natural convec- and the grid number of 2.983 million is only 0.084% by
tion, thus accelerating the convergence. Using finite volume the Eq. (9). Therefore, it was found that when the number
method for discrete control equations, discrete equations is of grids are 2.983 million, it can satisfy the requirement of
selected the second-order upwind scheme, and SIMPLE accuracy, in other words, the solution of grid independence
algorithm to deal with the coupling of pressure and has been acquired. As a result, the grid is divided in the
velocity. same way in the following calculations. Similar indepen-
Flow-solid and solid–solid contact surface is set to cou- dence analysis will also be applied to the models with dif-
pling wall boundary. In this paper, thermal boundary con- ferent absorber plate thickness, collector tube diameter,
ditions cannot be predetermined for the flow and heat tube collector spacing and so on. Meanwhile, the process
transfer problems of the glass cover in contact surface with will not be explained further (see Fig. 3).
air, absorber plate in contact surface with air, collector jgMþ1 gM j
tube in contact surface with air, collector tube in contact w¼ ; ðM ¼ 1; 2; 3; Þ ð9Þ
gMþ1
surface with water and the insulation layer in contact sur-
face with the air, and the heat conduction problem of the
absorber plate in contact surface with collector tube, the Table 3
absorber plate in contact surface with the insulation layer, Model parameters (unit: mm).
the collector tubes in contact surface with the insulation L W D s t e b c
layer, whereas they are constrained by the interaction 1800 110 12 1 0.9 30 4 30
between the fluid or solid with the wall. At this point,
regardless of the temperature or heat flux density on the
interface should be either seen as part of the calculation Table 4
result, rather than the known conditions. The problem of Diagram of instantaneous efficiency changes with the grid number.
the thermal boundary conditions have to be dynamically M Grid number (million) g (%) Deviation w (%)
decided by the heat exchange process instead of being pre-
1 49.8 70.85 –
determined, known as coupled heat transfer problems. For 2 113.6 68.43 3.54
most meaningful coupling problems are unable to obtain 3 175.3 63.29 8.12
analytical solutions, while the use of numerical solution. 4 232.7 61.30 3.25
Numerical solution can be divided into partitions solving 5 298.3 59.84 2.44
6 368.1 59.79 0.084
boundary coupling methods and the entire field solving
Z. Jiandong et al. / Solar Energy 117 (2015) 192–202 197
72 56
70 52
68
48
66
T / °C
44
η/%
64 Tin
40
62 Tout, exp[22]
36 Tout, sim
60
Tout, sim[22]
32
58
9 10 11 12 13
0 50 100 150 200 250 300 350 400 Time / hr
The number of grid / unit: ten thousand Fig. 4. Comparison chart of simulation and experimental values of the
working fluid outlet temperature.
Fig. 3. Grid independence analysis.
In order to verify the accuracy of the calculation result, 3.1. Flow characteristics and heat transfer performance
the simulation value is compared with experimental and analysis of flat-plate solar collectors
simulated values from the Ref. (Selmi et al., 2008), as
shown in Fig. 4. It can be seen that the outlet temperature Fig. 5 shows the internal air speed field of a solar collec-
of working fluid tends to increase with the increase of inlet tor, for which the ambient temperature is 15 °C, working
temperature of working fluid and vice versa. According to fluid inlet temperature is 20 °C, collector tube spacing of
198 Z. Jiandong et al. / Solar Energy 117 (2015) 192–202
110 mm, collector tube diameter of 12 mm, and it is temperature of the collector is increased gradually along
1700 mm from the cross-section of the working fluid inlet. the flow direction of working fluid. From the cross-section
What we can see from Fig. 5 is that: a swirl movement in where L is 200 mm away from the entrance of working
the opposite direction started in the air on both sides of the fluid to the cross-section where L is 1700 mm away from
collector tube, because the temperature of the absorber the entrance of working fluid, the maximum temperature
plate in the middle of the two adjacent collector tubes is of absorber plate increases from about 299 K to about
higher, and the surrounding temperature close to the col- 302 K, the working fluid temperature increases from about
lector tube is lower. It makes the air near the middle of 294 K to 297.4 K and the inner surface temperature of the
the two adjacent collector tubes move upward. After the glass cover increases from about 293 K to about 295 K.
air experiencing the glass cover, its movement converts to
both side of the glass cover. The air cools down when it 3.2. Effect of the absorber plate thickness t on the
moves above the collector tube, and then moves down to instantaneous efficiency g of flat-plate solar collector
the absorber plate. Later the re-heated air continues such
movements, eventually the vortex is produced as shown Fig. 7 shows how the collector instantaneous efficiency g
in Fig. 5. changes with the absorber plate thickness t under different
Also from the Fig. 5, we can see the air flow rate reach- solar radiation intensities I. As can be seen from the figure,
ing the maximum occurs about 25 mm away from collector when the solar radiation intensity I keeps constant, the col-
tube in the horizontal direction and close to the surface of lector instantaneous efficiency g increases with the increase
absorber plate and the glass cover. The maximum air flow of the absorber plate thickness t. This is duo to the increase
rate is close to 3:25 104 m=s. And a strong vortex is of the absorber plate thickness increasing the contact area
formed, which indicates strong air convection in the inter- between the absorber plate and collector tubes, and then
nal collector and increasing heat loss at the top of collector. reducing the thermal-conduction resistance, so that the
From the above analysis: in order to reduce the heat loss of heat of the absorber plate becomes easier to be transferred
the top of collector, some certain measures need to be to the working fluid within the collector tubes, and then the
taken to suppress the natural convection of air inside the collector instantaneous efficiency g is improved. When the
collector. absorber plate is relatively thin (t < 0.9 mm), increasing the
Fig. 6 is the internal temperature distribution in the absorber plate thickness can significantly improve the col-
cross-section of the collector along length of the collector lector instantaneous efficiency. When the absorber plate
tube when L is 200 mm, 700 mm, 1200 mm and 1700 mm, is relatively thick (t > 0.9 mm), increasing the absorber
for which the ambient temperature is 15 °C, the inlet tem- plate thickness cannot improve the collector instantaneous
perature of working fluid is 20 °C, the collector tube spac- efficiency significantly. When the solar radiation intensity I
ing is 110 mm, and the collector tube diameter is 12 mm. equals to 700 W/m2, the absorber plate thickness t
It can be seen from Fig. 6 that the maximum tempera- increases from 0.1 mm to 0.9 mm, the collector instanta-
ture inside the collector occurs in the absorber plate in neous efficiency g is increased by 13.27%. When the absor-
the middle of two adjacent collector tubes. The minimum ber plate thickness t increases from 0.9 mm to 2.1 mm, the
temperature inside the collector occurs just above the collector instantaneous efficiency g is increased by 4.19%.
collector tube and close to the inner surface of glass cover. Therefore, it is reasonable to select the absorber plate
The closer to the collector tube, the lower temperature of thickness of about 0.9 mm.
the absorber plate is, because the heat of absorber plate When the absorber plate thickness t increases from
closer to the collector tube is more easily be transmitted 0.1 mm to 2.1 mm, the collector instantaneous efficiency g
to the working fluid in the collector tube. The temperature under different solar radiation intensities I (I equals to
of air interlayer in the middle of two adjacent collector 400 W/m2, 700 W/m2, 1000 W/m2 respectively) increases
tubes is significantly higher than that right above the collec- from 40.07% to 59.13%, from 46.57% to 64.03%, from
tor tube. It can also be seen from the Fig. 6 that the internal 52.67% to 67.58% respectively.
Z. Jiandong et al. / Solar Energy 117 (2015) 192–202 199
70
70 I=400W/m2
65 I=700W/m2
65 I=1000W/m2
60
η/%
60
55
η/%
50 55
I=400W/m 2
45 I=700W/m 2 50
I=1000W/m 2
40
45
40 60 80 100 120 14 0 16 0 180
0.0 0.5 1.0 1.5 2.0 2.5
t / mm W /mm
Fig. 7. Diagram of the collector instantaneous efficiency changes with the Fig. 8. Diagram of the collector instantaneous efficiency changes with the
absorber plate thickness. collector tube spacing.
η/%
when the solar radiation intensity I keeps constant, the col- I=400W/m2
lector instantaneous efficiency g decreases with the increase 56
I=700W/m2
of the collector tube length L. This is due to when the col-
54
I=1000W/m 2
lector tube is longer, the residence time of the working fluid
in the collector tube is longer, so the working fluid can
52
obtain more heat from the tube wall, resulting higher aver-
age temperature of the working fluid in the collector tube. 50
And the temperature difference between the working fluid 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0
and the absorber plate is smaller, which weakens the heat L/m
transfer. Thus, the collector instantaneous efficiency g is
Fig. 9. Diagram of the collector instantaneous efficiency changes with the
reduced. When the collector tube length L is relatively
collector tube length.
short (L < 2000 mm), increasing the collector tube length
has little effect on the collector instantaneous efficiency g.
When the collector tube length L is relatively long
66
(t > 2000 mm), increasing the collector tube length has
greater effect on the collector instantaneous efficiency g. 64
and the collector tube spacing W reduces from ANSI/ASHRAE, 2003. ASHRAE Standard 93 Methods of Testing to
170 mm to 50 mm, the collector instantaneous effi- Determine Thermal Performance of Solar Collectors.
Banos, R., Manzano-Agugliaro, F., Montoya, F.G., et al., 2011.
ciency g is increased from 52.81% to 66.01%. Optimization methods applied to renewable and sustainable energy:
Therefore, reducing the collector tube spacing is an a review. Renew. Sustain. Energy Rev. 15, 1753–1766.
effective measure to improve the collector instanta- Bei, Ni, Yun, Ding, Qing, Yang, 2011. Numerical simulation study on flat-
neous efficiency. plate solar collector in series-connected system. Energy Conserv.
(3) Reducing the collector tube length L will help Technol. 29 (1), 20–23 (in China).
Dagdougui, Hanane, Ouammi, Ahmed., Robba, Michela, et al., 2011.
improve the collector instantaneous efficiency. Thermal analysis and performance optimization of a solar water heater
When the solar radiation intensity I equals to flat plate collector: application to Tétouan (Morocco). Renew. Sustain.
700 W/m2, and the collector tube length L reduces Energy Rev. 15 (1), 630–638.
from 2800 mm to 2000 mm, the collector instanta- Dhariwal, S.R., Mirdha, U.S., 2005. Analytical expressions for the
neous efficiency g is increased by 2.11%. When the response of flat-plate collector to various transient conditions. Energy
Convers. Manage. 46, 1809–1836.
collector tube length L reduces from 2000 mm to EN 12975-2, 2006. European Standards. Thermal solar systems and
1200 mm, the collector instantaneous efficiency g is components-Solar collectors – Part 2: Test methods.
increased by 0.51%. Therefore, it is reasonable to FLUENT Inc, 2003. FLUENT Theory Guide FLUENT Inc.
select the collector tube length of about 2000 mm. Garg, H.P., 1985. Solar Water Heating Systems[M]. Kluwer Academic
(4) Increasing the collector tube diameter D will help Publishers.
GB/T 4271, 2007. China Standards. Test method for the thermal
improve the collector instantaneous efficiency. performance of solar collectors.
When the solar radiation intensity I equals to Hestnes, A.G., 1999. Building integration of solar energy systems. Sol.
700 W/m2, and the collector tube diameters D Energy 67, 181–187.
increases from 8 mm to 20 mm, the collector instanta- Khoukhi, Maatouk, Maruyama, Shigenao, 2006. Theoretical approach of
neous efficiency g is increased from 56.18% to a flat-plate solar collector taking into account the absorption and
emission within glass cover layer. Sol. Energy 80, 787–794.
63.97%. Therefore, increasing the collector tube Kong, Xiangqiang, Lin, Lin, Li, Ying, et al., 2013. Simulation of thermal
diameter is an effective measure to improve the collec- performance of flat-plate solar collector. Acta Energiae Solaris Sincia
tor instantaneous efficiency. 38 (8), 1404–1408 (in China).
(5) When the insulation thickness e is relatively small Lu, Yu, Hongwen, Yu., Haicheng, Ding., et al., 2003. Mathematical
(e < 30 mm), increasing the insulation thickness has modeling and simulation of thermal properties of flat-plate solar
collector. J. Univ. Jinan (Nat. Sci.) 27 (3), 293–297 (in China).
great effect on the collector instantaneous efficiency. Molero Villar, N., Cejudo López, J.M., Domı´nguez Mu~ noz, F., et al.,
When the insulation thickness e is relatively large 2009. Numerical 3-D heat flux simulations on flat plate solar
(e > 30 mm), increasing the insulation thickness has collectors. Sol. Energy 83, 1086–1092.
little effect on the collector instantaneous efficiency. Norton, Brian., BSc, MSc, PhD, F Inst E, C Eng, 1992. Solar Energy
Therefore, it is reasonable to select the insulation Thermal Technology[M]. Springer-Verlag London Limited.
Osorio, Tiago, Carvalho, Maria Joao, 2014. Testing of solar thermal
thickness of about 30 mm. collectors under transient conditions. Sol. Energy 104, 71–81.
(6) According to the two cases of collector tube spacing Patankar, S.V., 1980. Numerical Heat Transfer and Fluid Flow[M].
W and collector tube diameter D discussed above, McGraw-Hill, New York.
we can see that their limit forms are flat-cassette col- Rojas, D., Beermann, J., Klein, S.A., Reindl, D.T., 2008. Thermal
lectors. It shows that the flat-cassette collectors own performance testing of flat-plate collectors. Sol. Energy 82, 746–
757.
the higher instantaneous efficiency. Saito, A., Utaka, Y., Tsuchio, T., Katayama, K., 1984. Transient response
of flat plate solar collector for periodic solar intensity variation. Sol.
Energy 32 (1), 17–23.
Appendix A. Supplementary data Selmi, Mohamed, Al-Khawaja, Mohammed J., Marafia, Abdulhamid,
2008. Validation of CFD simulation for flat plate solar energy
collector. Renew. Energy 33, 383–387.
Supplementary data associated with this article can be Tao, Wenquan, 2001. Numerical Heat Transfer [M]. Xi’an Jiaotong
found, in the online version, at http://dx.doi.org/10.1016/ University Press, Xi’an (in China).
j.solener.2015.04.027. Yang, Yuguo, Wang, Qichun, Xiu, Dapeng, 2013. A building
integrated solar collector: all-ceramic solar collector. Energy
References Buildings 62, 15–17.
Zambolin, E., Del Col, D., 2010. Experimental analysis of thermal
performance of flat plate and evacuated tube solar collectors in
Akhtar, N., Mullick, S.C., 2007. Computation of glass-cover temperatures
stationary standard and daily conditions. Sol. Energy 84, 1382–
and top heat loss coefficient of flat-plate solar collectors with double
1396.
glazing. Energy 32, 1067–1074.
Zhai, X.Q., Wang, R.Z., Dai, Y.J., Wu, J.Y., Xu, Y.X., Ma, Q., 2007.
Alvarez, A., Cabeza, O., Muñiz, M.C., et al., 2010. Experimental and
Solar integrated energy system for a green building. Energy Buildings
numerical investigation of a flat-plate solar collector. Energy 35 (9),
39, 985–993.
3707–3716.