IMPROVING ENERGY GENERATION FORECASTS OF SMALL HYDROPOWER PLANTS BY SPATIAL DEPENDENCE ANALYSIS

Electric power generation forecasting has an important role for distribution utilities when planning their operations. The difficulty increases when involving Small Hydropower Plants, in general, run-of-river plants with a strong dependence on hydrological regimes. But that is not the only concern, these plants in many cases do not have inflow or precipitation data available. Joining inflow data as exogenous variable from neighboring basins seems to be an interesting approach to solve data unavailability. This study aims to present a spatial dependence analysis to identify the best inflow series to obtain good prediction closer to reality. In this study, SARIMA and Transfer Function models were used.