Professional Documents
Culture Documents
com
ScienceDirect
Solar Energy 120 (2015) 399–418
www.elsevier.com/locate/solener
Received 19 April 2015; received in revised form 24 July 2015; accepted 24 July 2015
Available online 15 August 2015
Abstract
The main purpose of this paper is to model, study and analyze the performance of different photovoltaic (PV) array configurations
under various partial shading conditions (PSC). The aim is to improve the efficiency of PV systems in general by reducing mismatch
losses and to select the most appropriate PV array configuration which provides the best performance i.e. the highest maximum power
and hence the lowest relative power losses under partial shading conditions. For this purpose, a comprehensive study which considers all
the available PV array configurations: Series (S), Parallel (P), Series–Parallel (SP), Total-Cross-Tied (TCT), Bridge-Linked (BL), and
Honey-Comb (HC), is carried out under all possible scenarios of shading. The performance and output characteristics of these array
configurations are analyzed and compared by using a 6 4 PV array size under various possible shading scenarios.
The Bishop model of a photovoltaic module which describes best the solar cell behavior at negative voltages is considered in this paper
for modeling the PV arrays, and it is implemented by using the Simulink and SimPower software.
On the whole, the obtained results prove the superiority of the TCT configuration which provides the best performances under most
cases of PSC: uneven or random distribution of partial shading. However, for some cases of PSC this is not the case because the
performance depends strongly on the patterns of shade. Therefore, this paper will enrich and complement previous studies in this area
through a detailed analysis and comprehensive study of the different PV array configurations under all possible shading scenarios which
may occur in reality, the thing that has not been carried out previously.
In addition, these results represent an interesting guide which provides useful and relevant information concerning the performance of
each PV array configuration under any pattern of shading. This will allow the estimation and prediction of maximum power and
consequently will help to select the most suitable PV array configuration.
Ó 2015 Elsevier Ltd. All rights reserved.
Keywords: Photovoltaic array configurations; Partial shading; Series–Parallel; Bridge-Linked; Honey-Comb; Total-Cross-Tied
http://dx.doi.org/10.1016/j.solener.2015.07.039
0038-092X/Ó 2015 Elsevier Ltd. All rights reserved.
400 F. Belhachat, C. Larbes / Solar Energy 120 (2015) 399–418
it. This is known as hot spot effect and fortunately, this (2014) simulated and discussed the SP, TCT and BL inter-
problem can be solved by using bypass diodes. Similarly, connection schemes for moving shade on 4 4 PV array.
partial shading of a single module in a module Series will The authors concluded that gaining more power could be
limit the power of the whole PV array. The effect of shad- achieved by choosing the right interconnection scheme
ing has been widely studied with respect to the PV module according to the shading pattern.
performances (Quaschning and Hanitsch, 1996; Hanitsch On the whole, the results obtained by the above cited
et al., 2001; Villa et al., 2012; Parlak et al., 2014). works showed the superiority of the TCT configuration
The losses associated to the shading effect can be reduced which provides the best performances in most cases of
by using several approaches: maximum power point track- partial shading conditions. The work proposed in this
ing (MPPT) techniques that allow the extraction of the glo- paper comes in its turn to confirm and prove these results.
bal maximum power point, system architectures, converter Furthermore, it provides additional results concerning all
topologies and PV array configurations (Bidram et al., the considered configurations which come to enrich and
2012). PV array configuration which is considered in this complement the obtained results through a detailed analy-
paper is one of the solutions that can significantly reduce sis and complete study of the different configurations by
mismatch losses. It is based on the PV array interconnec- testing several possible shading scenarios which may occur
tions of PV modules which are Series (S), Parallel (P), in reality, the thing that has not been carried out by the
Series–parallel (SP), Total-Cross-Tied (TCT), previous studies.
Bridge-Linked (BL) and Honey-Comb (HC) configurations. So, compared with previous research works in this field,
Several researchers proposed models to study and ana- the main objective of this paper is to present a comprehen-
lyze the effect of shading on different PV array configura- sive study that considers all the available PV array config-
tions in order to reduce mismatch losses and to find the urations under almost all the possible patterns of shading.
best PV array configuration which provides the highest per- The approach adopted for this purpose, involves first, the
formances. Shams El-Dein et al. (2011a,b) proposed novel modeling of all the PV array configurations with different
configurations for photovoltaic farms which reduce partial sizes and then the analysis of their performances under dif-
shading losses. They introduced a novel mathematical for- ferent scenarios of partial shading. This will provide a solu-
mulation for irradiance level mismatch that can be used for tion for the choice of the most optimal configuration under
comparative evaluation of different photovoltaic configura- a given partial shading patterns which improves the effi-
tions. Wang and Hsu (2011) analyzed the electrical perfor- ciency of the overall system.
mance of a PV module with S, SP, TCT, BL and HC The model of the PV array used in this work is based on
configurations by using the Newton–Raphson and the the PV cell model presented by Bishop (Bishop, 1988). It is
piecewise linear parallel branches (PLPB) model methods implemented by using the MATLAB Simulink and
of PV cells. The authors found, by using both methods, SimPower software. All scenarios were tested on different
that the TCT configuration has the maximum power value PV array sizes but only the 6 4 PV array size is presented
which is followed by the BL and HC configurations for the in this paper.
solar irradiance pattern considered. However, in this work This paper starts with the development of a PV module
the authors considered only one case of partial shading pat- equivalent circuit model which is based on the Bishop PV
tern which is inadequate to draw conclusions on the best cell model. This is followed by the application of the
configuration. (Ramabadran and Mathur, 2012) developed proposed PV module model for the modeling and analysis
a generalized Matlab program to study, analyze and com- of the different PV array configurations under partial
pare different PV array configurations for different random shading patterns: Series (S), Parallel (P), Series–Parallel
shading patterns to determine the configuration which is (SP), Total-Cross-Tied (TCT), Bridge-Linked (BL) and
less sensitive to power losses under partial shading. The Honey-Comb (HC) configurations. Section 4 is devoted
authors found that the TCT configuration is the best con- to the study and analysis of different PV array configura-
figuration for symmetrical array sizes and the HC configu- tions under different partial shading scenarios. All the
ration is the best for asymmetrical array sizes. The authors considered PV array configurations are analyzed and
analyzed the five configurations for different PV array sizes, compared using several scenarios of shading in order to
but their studies were focused on a single case of shading evaluate their performances and the ability of each config-
pattern which is the one with random distribution. uration to increase the power and reduce losses caused by
Jazayeri and Uysal (2014) have made comparative analysis partial shading. Finally, the obtained results are discussed
based on the performances of different PV array topologies and evaluated.
under various irradiance conditions. The SP, BL and TCT
PV array configurations were analyzed. Simulation models 2. Modeling and simulation of one PV module under partial
for PV array topologies were conducted using shading conditions using Bishop model
Matlab/Simulink and the performances were analyzed with
and without bypass diodes. The authors found that the The two diodes equivalent circuit is the most used model
TCT configuration outperforms the SP and BL configura- for PV cells. Shaded cells in a PV module can be driven into
tions under partial shading conditions. Pareek and Dahiya the negative voltage region. If there are no bypass diodes
F. Belhachat, C. Larbes / Solar Energy 120 (2015) 399–418 401
V þ Rs I
I ¼ I ph I s1 exp 1
m1 V t
V þ Rs I
I s2 exp 1
m2 V t
n
V þ Rs I V þ Rs I
1þa 1 ð1Þ
Rp Vb
configurations. Fig. 2 shows a Simulink/SimPower model Under partial shading conditions, the P–V characteristic
of one PV module. exhibits multiple peaks and the maximum output power is
Fig. 3 shows the Simulink/SimPower model of a 43.51 W. Table 1 gives maximum power, voltage and cur-
stand-alone photovoltaic system which consists of: PV rent of the module under uniform and partial shading
module, MPPT controller, a DC–DC boost converter, a conditions.
battery and a load.
As an example of simulation, Figs. 4 and 5 show respec- 3. Modeling, presentation and simulation of different PV
tively the P–V and I–V characteristics of one PV module of array configurations under partial shading conditions
36 cells obtained with Simulink and SimPower using the
model described above. The simulation was carried out The aim of this part is to present all the PV array con-
under uniform condition i.e. with no shading (at 25 °C and figurations used. For all the simulation tests and for all
1000 W/m2) and under PSC at 25 °C and different irradiance analyzed configurations, we have used 24 PV modules
levels: 18 cells are exposed to an irradiance of 1000 W/m2, 12 and each module consists of 36 cells connected in Series
cells are exposed to 800 W/m2, 4 cells are exposed to and protected by anti-parallel bypass diodes. The tempera-
600 W/m2 and 2 cells are exposed to an irradiance of ture was fixed at 25 °C and the irradiance levels of the
200 W/m2. The solar cell parameters used in this simulation shading patterns on the PV array are illustrated in Table 2.
are: m1 = 1, Is1 = 1.26 109 A, m2 = 2, Is2 = 2.53 106
A, Rs = 1 m O, Rp = 1 k O; Iph = 3.798 A. Bishop’s term: 3.1. Parallel (P) configuration
Vb = 15 V, a = 2 103, n = 3 (Silvestre et al., 2009).
The Simulink/SimPower model of a parallel configura-
tion of PV array is shown in Fig. 6. The P–V and I–V char-
acteristics under uniform and partial shading conditions
Uniform condition
60
Partial shading conditions for parallel configuration are shown in Figs. 7 and 8
50 respectively.
40
We notice that the obtained characteristics exhibit a sin-
Power (W)
2
shown in Fig. 12. The PV array has four parallel strings
1.5
and each string contains six modules connected in Series.
The P–V and I–V characteristics under uniform and partial
1
shading conditions for Series–Parallel configuration are
0.5
shown in Figs. 13 and 14 respectively.
0
0 2 4 6 8 10 12 14 16 18 20 22
Voltage (V)
1500
Uniform condition 100
Partial shading conditions
80
1000
Uniform condition
Power (W)
Current (A)
500 40
20
0
0 5 10 15 20 25 0
0 5 10 15 20 25
Voltage (V)
Voltage (V)
Fig. 7. Simulated P–V characteristics for 6 4 parallel PV array
configuration under uniform and PSC. Fig. 8. Simulated I–V characteristics for 6 4 parallel PV array config-
uration under uniform and PSC.
Fig. 15 shows a TCT architecture where each PV mod- 3.5. Bridge-Linked (BL) configuration
ule is connected to its adjacent PV module. All the PV
modules are then cross connected. The P–V and I–V char- Fig. 18 shows the model of the BL configuration. The P–
acteristics under uniform and partial shading conditions V and I–V characteristics under uniform and partial shad-
for TCT configuration are shown in Figs. 16 and 17 ing conditions for BL configuration are shown in Figs. 19
respectively. and 20 respectively.
404 F. Belhachat, C. Larbes / Solar Energy 120 (2015) 399–418
1500
Uniform condition
3.6. Honey-Comb (HC) configuration
Partial shading conditions
3.5
Uniform condition where PM refers to the maximum power of the PV array
3 Partial shading conditions under uniform condition and PMPS is the maximum power
Current (A)
1.5
4. Study and analysis of different PV array configurations
1
under different shading scenarios
0.5
0
0 50 100 150 200 250 300 350 400 450 500 In order to evaluate the behavior of each PV array con-
Voltage (V)
figuration under mismatch caused by non-uniform solar
Fig. 11. Simulated I–V characteristics for 6 4 Series PV array config- irradiance and in order to choose the most optimal config-
uration under uniform and PSC. uration that provides the highest performances, we carried
F. Belhachat, C. Larbes / Solar Energy 120 (2015) 399–418 405
10 right.
8 PS-H: The same pattern of shading is moving randomly
6 in the PV array.
4
1500 16
Uniform condition
Partial shading conditions 14
12
Uniform condition
1000
Power (W)
Current (A)
6
500
4
0 0
0 20 40 60 80 100 120 0 20 40 60 80 100 120
Voltage (V) Voltage (V)
Fig. 16. Simulated P–V characteristics for 6 4 TCT PV array config- Fig. 17. Simulated I–V characteristics for 6 4 TCT PV array configu-
uration under uniform and PSC. ration under uniform and PSC.
1500 16
Uniform condition
Partial shading conditions 14
12
1000 Uniform condition
Current (A)
10
Power (W)
6
500
4
0
0 0 20 40 60 80 100 120
0 20 40 60 80 100 120
Voltage (V) Voltage (V)
Fig. 19. Simulated P–V characteristics for 6 4 BL PV array configura- Fig. 20. Simulated I–V characteristics for 6 4 BL PV array configura-
tion under uniform and PSC. tion under uniform and PSC.
4.2. Under partial shading conditions - We tested the case where two rows are completely and
evenly shaded. The first and the second row receive
4.2.1. Scenario PS-A: one row is completely and uniformly the same irradiance value of 300 W/m2. The simulations
shaded give the same results as for scenario PS-A. The configu-
The shading patterns and the corresponding simulation rations S, SP, TCT, BL and HC provide the same max-
results for scenario PS-A are shown in Tables 4 and 5 imum power (988.85 W) and thus the same relative
respectively. power losses (33.33%).
Fig. 25 shows the obtained maximum powers and rela- - We tested as well the case where two rows are com-
tive power losses with scenario PS-A. pletely and unevenly shaded. The first and the second
We notice that the configurations S, SP, TCT, BL and row receive different irradiance values of 300 W/m2
HC present the same maximum power and hence the same and 800 W/m2 respectively. The simulations give the
relative power losses. same results as for scenario PS-A. The configurations
408 F. Belhachat, C. Larbes / Solar Energy 120 (2015) 399–418
1500 16
Uniform condittion
Partial shading conditions 14
12
1000
Power (W)
10 Uniform condition
Current (A)
500 6
2
0 0
0 20 40 60 80 100 120 0 20 40 60 80 100 120
Voltage (V) Voltage (V)
Fig. 22. Simulated PV characteristics for 6 4 HC PV array configu- Fig. 23. Simulated IV characteristics for 6 4 HC PV array configu-
ration under uniform and PSC. ration under uniform and PSC.
4.2.2. Scenario PS-B: one row is completely and unevenly - We tested the case where two rows are completely and not
shaded uniformly shaded. The first row receives different irradi-
The shading patterns and the corresponding simulation ance values of 100, 200, 300 and 400 W/m2 respectively
results for scenario PS-B are shown in Tables 6 and 7 and the second row receives 500, 600, 700 and
respectively. 800 W/m2 respectively. The simulations give the same
F. Belhachat, C. Larbes / Solar Energy 120 (2015) 399–418 409
Table 3
Maximum power and relative power losses for different topologies and 6 4 PV array.
Configuration Uniform condition Partial shading conditions
PM (W) VM (V) IM (A) PMPS (W) VMPS (V) IMPS (A) DPL (%)
P 1483.30 17.26 85.95 1301.70 17.10 76.14 12.22
S 1483.30 413.62 3.59 1009.80 73.54 13.73 31.92
SP 1481.50 104.57 14.17 1009.80 73.54 13.73 31.92
TCT 1483.30 103.26 14.36 1014.90 90.66 11.19 31.58
BL 1481.50 104.57 14.17 993.17 89.50 11.10 33.04
HC 1481.50 104.57 14.17 998.45 89.92 11.10 32.69
Power (W)
1600 Table 5
1400 Simulation results for different configurations under PSC for scenario PS-
1200 A.
1000 Configuration Scenario PS-A
800 PMPS (W) VMPS (V) IMPS (A) DPL (%)
600
400
P 1301.70 17.18 75.77 12.24
S 1236.10 345.07 3.58 16.67
200
SP 1236.10 86.14 14.35 16.67
0
TCT 1236.10 86.14 14.35 16.67
P S SP TCT BL HC
BL 1236.10 86.14 14.35 16.67
Fig. 24. The obtained power for different configurations under uniform HC 1236.10 86.21 14.34 16.67
irradiance condition.
Fig. 25. The obtained maximum powers and relative power losses with Table 7
scenario PS-A. Simulation results for different configurations under PSC for scenario PS-
B.
Table 4 Configuration Scenario PS-B
Shading patterns in [W/m2] for scenario PS-A and for 6 4 PV array.
PMPS (W) VMPS (V) IMPS (A) DPL (%)
String 1 String 2 String 3 String 4
P 1353.50 17.17 78.81 8.75
Row 1 1000 1000 1000 1000 S 1236.10 86.14 14.35 16.67
Row 2 1000 1000 1000 1000 SP 1236.10 86.21 14.34 16.67
Row 3 1000 1000 1000 1000 TCT 1236.10 86.14 14.35 16.67
Row 4 1000 1000 1000 1000 BL 1236.10 86.21 14.34 16.67
Row 5 1000 1000 1000 1000 HC 1236.10 86.14 14.35 16.67
Row 6 300 300 300 300
results as for scenario PS-B. The configurations S, SP, Fig. 27 shows the obtained maximum powers and rela-
TCT, BL, and HC provide the same maximum powers tive power losses with scenario PS-C.
(988.85 W) and therefore the same relative power losses For the partial shading scenario PS-C, we notice that the
(33.33%). configurations SP, TCT, BL and HC have the same best
performance. They show higher maximum powers and
hence lower relative power losses with respect to the
4.2.3. Scenario PS-C: a column is completely and uniformly Series configuration.
shaded
The shading patterns and the corresponding simulation - We tested the case where two columns (strings) are com-
results for scenario PS-C are shown in Tables 8 and 9 pletely and evenly shaded. The first and the second col-
respectively. umn receive the same value of irradiance of 300 W/m2.
410 F. Belhachat, C. Larbes / Solar Energy 120 (2015) 399–418
Fig. 30. The obtained maximum powers and relative power losses with Table 19
scenario PS-E-2. Simulation results for different configurations under PSC for scenario PS-
F-2.
Configuration Scenario PS-F-2
Table 16
Shading patterns in [W/m2] for scenario PS-F-1 and for 6 4 PV array. PMPS (W) VMPS (V) IMPS (A) DPL (%)
Table 17 1500 10
Simulation results for different configurations under PSC for scenario PS- 8
1450
F-1.
6
Configuration Scenario PS-F-1 1400
4
PMPS (W) VMPS (V) IMPS (A) DPL (%)
1350
P 1470.30 17.21 85.42 0.88 2
S 1421.50 396.24 3.59 4.17 1300 0
SP 1427.50 105.35 13.55 3.76 P S SP TCT BL HC
TCT 1462.40 103.74 14.10 1.41
BL 1448.40 105.31 13.75 2.35 Fig. 32. The obtained maximum powers and relative power losses with
HC 1456.20 104.07 13.99 1.83 scenario PS-F-2.
Fig. 31. The obtained maximum powers and relative power losses with
scenario PS-F-1. Fig. 33 shows the obtained maximum powers and
relative power losses with scenario PS-F-3.
4.2.6.2. Scenario PS-F-2. The shading patterns and the 4.2.6.4. Scenario PS-F-4. The shading patterns and the
corresponding simulation results for scenario PS-F-2 are corresponding simulation results for scenario PS-F-4 are
shown in Tables 18 and 19 respectively. shown in Tables 22 and 23 respectively.
Fig. 32 shows the obtained maximum powers and Fig. 34 shows the obtained maximum powers and
relative power losses with scenario PS-F-2. relative power losses with scenario PS-F-4.
4.2.6.3. Scenario PS-F-3. The shading patterns and the 4.2.6.5. Scenario PS-F-5. The shading patterns and the
corresponding simulation results for scenario PS-F-3 are corresponding simulation results for scenario PS-F-5 are
shown in Tables 20 and 21 respectively. shown in Tables 24 and 25 respectively.
F. Belhachat, C. Larbes / Solar Energy 120 (2015) 399–418 413
Row 1 800 1000 1000 1000 PMPS (W) VMPS (V) IMPS (A) DPL (%)
Row 2 800 1000 1000 1000 P 1418.40 17.23 82.34 4.38
Row 3 800 1000 1000 1000 S 1289.50 431.32 2.99 13.07
Row 4 800 1000 1000 1000 SP 1409.80 102.61 13.74 4.96
Row 5 1000 1000 1000 1000 TCT 1414.60 103.10 13.72 4.63
Row 6 1000 1000 1000 1000 BL 1409.10 102.34 13.77 5.00
HC 1412.00 104.19 13.55 4.81
Table 23
Simulation results for different configurations under PSC for scenario PS-
F-4.
From scenarios PS-F-1 to PS-F-5, we can observe that
the TCT has the best performance, with the highest maxi-
Configuration Scenario PS-F-4
mum power and the lowest relative power losses. Whereas
PMPS (W) VMPS (V) IMPS (A) DPL (%) for scenario PS-F-6, which is in fact similar to scenario
P 1431.20 17.15 83.45 3.51 PS-C with a column completely and uniformly shaded,
S 1298.50 433.09 3.00 12.46 we have a confirmation of the result obtained with this
SP 1415.40 103.06 13.73 4.58
TCT 1424.60 103.28 13.79 3.96
latter. The configurations SP, TCT, BL and HC have the
BL 1421.80 103.19 13.78 4.15 same best performance with respect to the Series
HC 1421.00 103.71 13.70 4.19 configuration.
The above tests were carried out with a moving shading
pattern having an irradiance of 800 W/m2. However, when
Fig. 35 shows the obtained maximum powers and the moving shading pattern is dense with an irradiance
relative power losses with scenario PS-F-5. value of 300 W/m2, the Series configuration provides the
best performance with the highest maximum power and
4.2.6.6. Scenario PS-F-6. The shading patterns and the hence the lowest relative power losses for scenarios
corresponding simulation results for scenario PS-F-6 are PS-F-1 to PS-F-5. For scenario PS-F-6, we have the same
shown in Tables 26 and 27 respectively. results as above and as for scenario PS-C. The configura-
Fig. 36 shows the obtained maximum powers and tions SP, TCT, BL and HC have the same best perfor-
relative power losses with scenario PS-F-6. mance with respect to the Series configuration.
414 F. Belhachat, C. Larbes / Solar Energy 120 (2015) 399–418
Fig. 35. The obtained maximum powers and relative power losses with
scenario PS-F-5. Table 29
Simulation results for different configurations under PSC for scenario PS-
G-2.
Table 26 Configuration Scenario PS-G-2
Shading patterns in [W/m2] for scenario PS-F-6 and for 6 4 PV array.
PMPS (W) VMPS (V) IMPS (A) DPL (%)
String 1 String 2 String 3 String 4
P 1457.30 17.22 84.62 1.75
Row 1 800 1000 1000 1000 S 1359.70 379.26 3.59 8.33
Row 2 800 1000 1000 1000 SP 1378.30 106.02 13.00 7.08
Row 3 800 1000 1000 1000 TCT 1419.70 106.23 13.36 4.29
Row 4 800 1000 1000 1000 BL 1412.50 106.18 13.30 4.77
Row 5 800 1000 1000 1000 HC 1401.60 106.25 13.19 5.51
Row 6 800 1000 1000 1000
1300 4.2.7.3. Scenario PS-G-3. The shading patterns and the cor-
5
responding simulation results for scenario PS-G-3 are
1250
shown in Tables 30 and 31 respectively.
1200 0 Fig. 38 shows the obtained maximum powers and rela-
P S SP TCT BL HC
tive power losses with scenario PS-G-3.
Fig. 36. The obtained maximum powers and relative power losses with
scenario PS-F-6.
4.2.7.4. Scenario PS-G-4. The shading patterns and the cor-
responding simulation results for scenario PS-G-4 are
4.2.7. Scenario PS-G: spread of shading in the same row shown in Tables 32 and 33 respectively.
from left to right Fig. 39 shows the obtained maximum powers and rela-
4.2.7.1. Scenario PS-G-1. This scenario is similar to sce- tive power losses with scenario PS-G-4.
nario PS-F-1 and the corresponding results are given in From scenarios PS-G-1 to PS-G-3, we notice that the
Tables 14, 15 and Fig. 30. TCT has the best performance, with the highest maximum
power and the lowest relative power losses. Whereas for
4.2.7.2. Scenario PS-G-2. The shading patterns and the cor- scenario PS-G-4, which is in fact similar to scenario PS-A
responding simulation results for scenario PS-G-2 are with a row completely and uniformly shaded, we have a
shown in Tables 28 and 29 respectively. confirmation of the result obtained with this latter. The
F. Belhachat, C. Larbes / Solar Energy 120 (2015) 399–418 415
Table 30 Table 33
Shading patterns in [W/m2] for scenario PS-G-3 and for 6 4 PV array. Simulation results for different configurations under PSC for scenario PS-
G-4.
String 1 String 2 String 3 String 4
Configuration Scenario PS-G-4
Row 1 800 800 800 1000
Row 2 1000 1000 1000 1000 PMPS (W) VMPS (V) IMPS (A) DPL (%)
Row 3 1000 1000 1000 1000 P 1431.20 17.15 83.45 3.51
Row 4 1000 1000 1000 1000
S 1298.50 433.13 3.00 12.46
Row 5 1000 1000 1000 1000
SP 1295.30 107.35 12.07 12.67
Row 6 1000 1000 1000 1000
TCT 1297.50 107.78 12.04 12.53
BL 1297.90 108.68 11.94 12.50
HC 1295.30 107.34 12.07 12.67
Table 31
Simulation results for different configurations under PSC for scenario PS-
G-3. Power (W) Power Losses
Configuration Scenario PS-G-3 1450 14
PMPS (W) VMPS (V) IMPS (A) DPL (%) 12
1400
10
P 1444.30 17.23 83.83 2.63 1350 8
S 1308.30 435.31 3.01 11.68
SP 1335.60 107.33 12.44 9.96 1300 6
TCT 1363.70 107.16 12.73 8.06 4
1250
BL 1355.10 107.06 12.66 8.64 2
HC 1358.90 106.15 12.80 8.39 1200 0
P S SP TCT BL HC
Fig. 39. The obtained maximum powers and relative power losses with
Power (W) Power Losses scenario PS-G-4.
1500 14
1450 12
10 Table 34
1400
8 Shading patterns in [W/m2] for scenario PS-H-1 and for 6 4 PV array.
1350
6 String 1 String 2 String 3 String 4
1300
4
Row 1 500 1000 1000 1000
1250 2
Row 2 500 1000 1000 1000
1200 0 Row 3 1000 1000 1000 1000
P S SP TCT BL HC Row 4 1000 1000 1000 1000
Row 5 1000 1000 1000 1000
Fig. 38. The obtained maximum powers and relative power losses with
Row 6 1000 1000 1000 1000
scenario PS-G-3.
Fig. 40. The obtained maximum powers and relative power losses with
scenario PS-H-1.
Table 39
Simulation results for different configurations under PSC for scenario PS-
Table 36 H-3.
Shading patterns in [W/m2] for scenario PS-H-2 and for 6 4 PV array. Configuration Scenario PS-H-3
String 1 String 2 String 3 String 4 PMPS (W) VMPS (V) IMPS (A) DPL (%)
Row 1 1000 500 1000 1000 P 1418.30 17.23 82.34 4.38
Row 2 1000 500 1000 1000 S 1359.70 379.26 3.59 8.33
Row 3 1000 1000 1000 1000 SP 1307.90 105.01 12.46 11.82
Row 4 1000 1000 1000 1000 TCT 1365.10 105.94 12.88 7.97
Row 5 1000 1000 1000 1000 BL 1341.20 105.07 12.76 9.47
Row 6 1000 1000 1000 1000 HC 1339.70 105.69 12.68 9.57
Table 37
Simulation results for different configurations under PSC for scenario PS- Power (W) Power Losses
H-2. 1450 14
12
Configuration Scenario PS-H-2 1400 10
PMPS (W) VMPS (V) IMPS (A) DPL (%) 8
1350
P 1418.30 17.23 82.34 4.38 6
S 1359.70 379.21 3.59 8.33 1300 4
SP 1307.80 105.04 12.45 11.83 2
TCT 1365.10 105.94 12.88 7.97 1250 0
BL 1351.10 105.81 12.77 8.91 P S SP TCT BL HC
HC 1346.20 105.67 12.74 9.24
Fig. 42. The obtained maximum powers and relative power losses with
scenario PS-H-3.
Fig. 42 shows the obtained maximum powers and rela-
tive power losses with scenario PS-H-3.
highest maximum power and the lowest relative power
4.2.8.4. Scenario PS-H-4. The shading patterns and the losses. The S configuration comes second best whereas
corresponding simulation results for scenario PS-H-4 are the SP configuration has the lowest performances.
shown in Tables 40 and 41 respectively. The results show as well that the performances of the BL
Fig. 43 shows the obtained maximum powers and rela- and HC configurations are sensitive to changes in position
tive power losses with scenario PS-H-4. of the same shading pattern whereas the performances of
The results for the entire scenario PS-H show that the the TCT, S and SP configurations are unaffected by these
TCT configuration provides the best performances i.e. the changes.
F. Belhachat, C. Larbes / Solar Energy 120 (2015) 399–418 417
Power (W) Power Losses When the shading covers completely the PV array and
1450 14 symmetrically with respect to the columns, the SP, TCT,
12 BL and HC configurations provide the best performance.
1400 10
When the shading covers completely the PV array and
8
1350 symmetrically with respect to the rows, the S, SP,
6
4 TCT, BL and HC configurations provide the best
1300
2 performance.
1250 0 When the shading covers partially and unevenly the PV
P S SP TCT BL HC array, the TCT configuration provides, in almost all
Fig. 43. The obtained maximum powers and relative power losses with cases, the best performance in terms of highest maxi-
scenario PS-H-4. mum power and lowest relative power losses. It is usu-
ally followed by the BL or HC configurations.
For high irradiance values, when the shading pattern
Table 40
Shading patterns in [W/m2] for scenario PS-H-4 and for 6 4 PV array.
spread in the same row or in the same column of an
array, the TCT configuration outperforms all the other
String 1 String 2 String 3 String 4
configurations.
Row 1 1000 1000 1000 1000 For low irradiance values, when the shading pattern
Row 2 1000 1000 1000 1000
Row 3 1000 1000 1000 1000
spread in the same row or in the same column of an
Row 4 1000 1000 1000 1000 array, the S configuration outperforms all the other
Row 5 1000 1000 1000 500 configurations.
Row 6 1000 1000 1000 500 When the same shading pattern is moving randomly in
the array, the performances of the BL and HC configu-
Table 41 rations are sensitive to the changes in position whereas
Simulation results for different configurations under PSC for scenario PS- the performances of the TCT, S and SP configurations
H-4. are unaffected by these changes.
Configuration Scenario PS-H-4
PMPS (W) VMPS (V) IMPS (A) DPL (%)
P 1418.30 17.23 82.34 4.38 6. Conclusion
S 1359.70 379.28 3.58 8.33
SP 1307.80 105.04 12.45 11.83 In this paper, a comprehensive study which considers all
TCT 1365.10 105.94 12.88 7.97 the available PV array configurations: Series (S), Parallel
BL 1351.40 105.20 12.85 8.89
(P), Series–Parallel (SP), Total-Cross-Tied (TCT),
HC 1349.20 105.63 12.77 9.04
Bridge-Linked (BL), and Honey-Comb (HC), has been car-
ried out under all possible scenarios of shading. For this
5. Results and performance evaluation purpose, the modeling of PV arrays was performed using
the model of Bishop and was implemented using
From the above obtained results it can be concluded Simulink/SimPower software.
that: Under partial shading conditions, the results clearly
show that the performances of the PV array configurations
The S, SP, TCT, BL and HC configurations provide the are variables and depend strongly on the scenario of shad-
same maximum powers and hence the same relative ing used. The detailed analysis of each configuration
power losses when one or two rows are completely through multiple shading scenarios show that the choice
and either evenly or unevenly shaded. of the most optimal and appropriate configuration depends
The SP, TCT, BL and HC configurations provide the strongly on the intensity of shading, the shading pattern,
same maximum powers and hence the same relative the location of the shading pattern, and the type of shading
power losses when one or two columns are completely affecting the PV array (uniform or not).
and uniformly shaded. Finally, this study gives useful information on how and
The TCT configuration presents the best performance when to use each PV array configuration in the case of
i.e. the highest maximum power and hence the lowest shading; it helps to predict and therefore to choose the
relative power losses when one or two columns are com- optimal PV array configuration, which gives the best per-
pletely and unevenly shaded. formance i.e. highest maximum power and lowest relative
When the shading covers completely and unevenly the power losses, according to the shading pattern. The
PV array, the TCT configuration provides, in all cases, obtained results could also be used in an intelligent algo-
the best performance in terms of highest maximum rithm for the tracking of the global maximum power point
power and lowest relative power losses. in the case of shading.
418 F. Belhachat, C. Larbes / Solar Energy 120 (2015) 399–418