A long-term operational scheme for hybrid hydro-PV systems that considers the uncertainties in reservoir inflow and solar radiation based on scenario trees

Abstract The majority of available long-term operation models of hydropower stations use deterministic historical data as inputs, and cannot update the decision scheme in real-time according to the actual solar radiation and inflow conditions. This results in a disconnect between the given plan and actual decision-making. To address existing challenges in the long-term operation of hydro-PV complementary power stations, a multi-stage rolling reservoir decision model considering the uncertainties of solar radiation and inflow is presented. This model can guide the formulation of long-term operation scheme of hydro-PV system. The typical wet, normal, and dry years are analyzed. We take the solar radiation series and inflow series generated by the scenario tree (ST) as the inputs of the reservoir optimal operation model, and use genetic algorithm (GA) to solve the model. During the solution process, the scheme is adjusted according to the actual solar radiation data and inflow data. The results illustrate that the model can better mentor the formulation of long-term operation rules for hydro-PV stations contrasted to the actual operation scheme and the traditional deterministic model.