STATISTICAL ANALYISIS OF THE PERFORMANCE LOSS RATE OF PV PLANTS DISTRIBUTED IN A REGION: A REAL-CASE STUDY IN SOUTH TYROL

: Information concerning the performance of photovoltaic (PV) plants in a region is of fundamental importance for many stakeholders of the sector. Nevertheless, often the lack of meteorological measurements – irradiance over all – prevent from correctly assessing the current health status of PV plants, i.e. the presence of ongoing mechanisms that gradually cause a performance degradation higher than expected and warranted. This study presents a methodology to infer probabilistic information on the performance loss rate (PLR) of a PV fleet in a region, when only few characteristics of PV plants (e.g. final yield, location, nominal power) are known. It makes use of satellite-retrieved values of global horizontal and diffuse horizontal irradiance, downloaded freely and only for an optimized number of representative locations. A fleet of 1899 PV plants located in South Tyrol (North-East of Italy) is used as a case study to validate the methodology, and results and possible applications are presented. In general, the presented methodology looks promising in extracting statistical information from a fleet of PV plants distributed in a region, and in inferring the probabilistic performance of future plants installed in the same area. Issues concerning the quality of the input data, related especially to the irradiance and the local skyline, are identified and discussed for a further improvement of the methodology.