Development of technical and statistical algorithm using Business Intelligence tools for energy yield assessment of large rooftop photovoltaic system ensembles

The aim of this paper is to develop a methodology for the energy yield assessment of a large ensemble of residential rooftop photovoltaic (PV) generators. The exploration and strategic analysis of data were applied for the first time, through Business Intelligence (BI) concepts combined with a technical-statistical algorithm for 1250 identical distributed rooftop PV systems installed in Santa Catarina, South of Brazil. Tilted solar irradiation maps were divided into four Ranges (statistical-quartiles) and, for the base year of 2019, the case-study results showed, with 95% reliability, an average annual daily energy yield of 3.35 kWh/kWp/day (Range 1); 3.74 kWh/kWp/day (Range 2); 3.82 kWh/kWp/day (Range 3); and 3.90 kWh/kWp/day (Range 4) for the four different quartiles into which the Santa Catarina state surface area was divided. The annual average PV generation of each PV rooftop analyzed was 3,353 kWh/year, which corresponds to an avoided CO2 emission of 268.25 kgCO(2)/year. For the combined 1250 PV rooftops ensemble, the average avoided CO2 emission was 335.30 tCO(2)/year. The implementation of BI concepts combined with the technical-statistical algorithm, provided relevant information about the performance of the PV systems, supporting fast decision-making strategies. The methodology applied can be used to predict the monthly/annual generation of future PV ensembles to be installed in any location for distributed generation asset allocation planning.

Automated summary

The aim of this paper is to develop a methodology for assessing the energy yield of a large ensemble of residential rooftop photovoltaic (PV) generators. Through the application of Business Intelligence concepts and a technical-statistical algorithm, data exploration and strategic analysis were conducted for 1250 identical rooftop PV systems in Santa Catarina, Brazil. The methodology can be used to predict the generation of future PV ensembles and support distributed generation asset allocation planning.