Abstract
A photovoltaic power generation system (PVS) is one of the promising measures to develop a low carbon society. Because of the unstable power output characteristics, a robust forecast method must be employed for realizing the high penetration of PVS into an electric power system. Considering the difference in power output patterns among PVSs dispersed in the service area of electric power system, the forecast error would vary among locations, resulting in the reduced forecast error of the ensemble average power output of high penetration PVS. In this paper, by using the multi-point data of insolation observed in Chubu area during four months, we evaluated the forecast error of the ensemble average insolation of 11 districts, and compared it with the forecast error of individual district. As the results, the number of periods with the forecast error larger than the average insolation during four months is reduced by 16 hours for the ensemble average insolation compared with the average value of individual forecast. The largest forecast error during four months is also reduced to 0.45 kWh/m2 for the ensemble average insolation from 0.68 kWh/m2 on average of 11 districts.
