APPLICATION

The proposed method is used to assess the performance and determine the feasible configurations of the PV system, which provides energy for a national demonstration project of water purification process in Singapore. According to electrical energy requirement of the process, it is estimated that 60 pieces of 175 Wp (peak power) solar panels from SolarWorld are needed. For these 60 solar panels, all feasible configurations of the system are identified based on the input voltage ranges, the maximum input cur­rents and the nominal capacities of the inverters commercially available in the market, as shown in Table 1 and Figs. 4-7. Figure 4 shows one con­figuration of the system which consists of 12 strings each having 5 panels in series, Fig. 5 illustrates one configuration which consists of 10 strings each having 6 panels in series, Fig. 6 shows three configurations of the systems which consist of 6 strings each having 10 panels in series, and Fig. 7 shows one configuration of the system which consists of 5 strings each having 12 panels in series. It should be noted that the special “text string” is defined to represent each configuration, for example, the “c12p12s05” in Fig. 4 represents the configuration that includes 12 PV inverters, 12 strings and each having 5 solar panels in series.

image156 image157

Grid connection

Подпись: Stnng 10Stnng 1 String 2

FIGURE 5: Ten string configurations with 10 inverters

TABLE 1: Parameters of inverters for evaluation of YEEP and system cost

Configuration

Capacity

(W)

Voltage

(V)

Current

(A)

Prices

(S$)

O & M (S$)

Peak efficiency

(%)

c12p12s05

1,100

139-400

10

2,559

256

93.0

c10p10s06

1,100

139-400

10

2,559

256

93.0

c06p06s10

2,500

224-600

12

4,444

444

94.1

c03p06s10

3,800

200-500

20

5,083

508

95.6

c02p06s10

5,500

246-600

26

8,156

816

96.1

c05p05s12

2,500

224-600

12

4,444

444

94.1

image159Note: one solar panel cost is S$1,400 and its monthly O&M cost is assumed to 1% of solar panel cost, namely S$14 each panel

Подпись: Economical Evaluation of Large-Scale Photovoltaic Systems 235

Grid connection

image161

String 1 Siring 2 String 6

 

Grid connection

image162

String 1 String 2 Siring 3 String 4 String 5 String 6

 

Grid connection

 

image163 image164

(a)c06p06sl 0

 

(b)c03p06sl0

 

(c) c02p06s!0

 

FIGURE 6: Six string configurations with different numbers of inverters

 

Подпись: String inverter 2
Подпись: String inverter 1
Подпись: DC/AC Подпись: DC/AC image169

image170Grid connection

Подпись: String SString 1 String 2

FIGURE 7: Five string configuration with 5 inverters

The basic parameters of the PV inverters including nominal capacity, input voltage range, purchasing cost, O&M cost and peak efficiency for different configurations are shown in Table 1 [19].

For evaluating the failure rate of the IGBT of the inverter shown in Fig. 3, the base failure rate kb is set as 0.060, the temperature factor nT is computed from nT = exp[-1,925([1/(T.+273)] • 1/298)] [20], where T is the junction operating temperature of the device and set as 40°C, nQ = 5.0 and = 1.0 for the other conditions.

The repair rate of the IGBT equals to 0.0017 per hour. The failure rates of the solar panel and the string diode are set as 0.2068 and 0.0198 per million hours, respectively; the repair rates of the solar panel and the string diode are 4.0556 per year [20]. The current of the solar panel string is 4.89 A at the maximum power and the voltage drop of the blocking diode is set as 2.0 V. The yearly peak sun hours (PSHs) equals to 1,721.7 hours in Sin-

image172

FIGURE 8: Yearly expected energy production of PV systems

gapore, which is the average yearly PSHs over the period of 1993 to 2007 [21]. The inflation rate and the interest rate are assumed as 2.1% and 1%, respectively. The system life cycle is assumed to be 20 years.

With all these parameters, the YEEP for each configuration is evalu­ated and the results are shown in Fig. 8. It can be observed from Fig. 8 that the configuration of “c02p06s10” has the highest expected energy production (15,898.6 kWh/year), which consists of 2 inverters, 6 strings, each of the strings having 10 solar panels in series. The YEEP for differ­ent configurations ranges from 15,255.7 kWh/year to 15,898.6 kWh/year, with the difference of 4.21%. It represents that the difference of expected energy production for different configurations in the system life cycle can be 12,860 kWh.

The differences in the YEEP are mainly caused by the reliability dif­ferences of PV arrays and the connected PV inverters for various con­figurations. The assessment of the ALCC for each configuration of the PV systems is also conducted. Consequently, the EUCE is easily calculated as the ratio of the ALCC to the YEEP, as shown in Fig. 9.

It can be seen that the EUCE for different configurations ranges from 0.434-0.598 S$/kWh. The configuration of “c03p06s10” has the lowest EUCE (0.434 S$/kWh), which also has the second highest YEEP. The configuration of “c02p06s10” has the second lowest EUCE (0.441 S$/ kWh) with the highest YEEP. The configuration of “c12p12s05” has the highest EUCE (0.598 S$/kWh). Therefore, the “c03p06s10” is the opti­mal configuration with the lowest EUCE. The comparison results show that simply increasing the system cost by using relatively large number of low capacity inverters in the PV systems cannot guarantee high expected energy production, instead the configuration with high reliable PV arrays, and high capacity inverters can achieve the lowest unit cost of electricity.

image173

FIGURE 9: Unit cost of PV systems

8.6 CONCLUSION

In this paper, the UGF technique is used to represent reliability models of solar panel arrays, PV inverters and EPUs in a large-scale PV system. Based on the developed probabilistic performance distribution models, the expected energy production for PV systems is evaluated with respect to the reliability of system elements. The reliability based cost analysis of PV systems is conducted for providing informative metrics to stake­holders for making the optimal decision. A new economical index for PV systems—EUCE is also developed in this paper. The proposed method is used to identify the feasible configurations of PV systems and determine the economically optimal one.