Forecasting of SPP Generation at Different Stages of its Existence Using the Example of the 2.3 MW Plant in the Kharkiv Region of Ukraine

Abstract

The results of studies of the generation of a solar power plant (SPP) with a capacity of 2.336 MW, which is located in the Kharkiv region of Ukraine, are presented. The results of the forecast generation, which were determined using the PHOTOVOLTAIC GEOGRAPHICAL INFORMATION SYSTEM (PV-GIS) program. Analysis of the forecast and actual generation data showed that the actual annual deviation from the forecast ranged from +5.47% to −5.24%, which is slightly higher than the annual generation variability of 4.87% determined by PV-GIS. The economic performance of a solar power plant is affected by the accuracy of forecasting hourly generation a day in advance. In order to improve the accuracy of forecasting the hourly generation of the SPP, an analysis of data from the guaranteed buyer, who buys all the generated electricity, and the plant's SmartLogger1000 monitoring system was performed. According to the results of the analysis, it was found that there is a difference in the data of the guaranteed buyer and the monitoring system, which arose as a result of the non-synchronization of the SmartLogger1000 monitoring system with the guaranteed buyer. The analysis of the hourly forecasts of the guaranteed buyer and the forecasting service showed that the sum of the hourly deviations of the forecast of the SPP service from the actual generation for December 2022 was 71.4%, and the sum of the hourly deviations of the forecast of the guaranteed buyer was 117.9%. An increase in the accuracy of forecasting was achieved due to the accumulation of statistical data and the study of factors influencing the generation of SPP.