Multi-Level Optimization of Hydro Power Plants

Stochastic optimization techniques are applied for the medium-term planning and the short-term unit commitment of hydro power plants. In contrast to deterministic approaches which rely only on point forecasts of uncertain exogenous factors – like spot prices for electricity, rainfalls into reservoirs etc. –, the employed stochastic techniques take into account the essential dynamic behavior of the relevant uncertain factors (including variance, correlation, trend, and seasonality). In doing so, increased planning risk is taken into account which is due to liberalized and increasing volatile electricity markets, decomposed into intraday-, day-ahead- and futures trading platforms. Time series analyses determine the necessary statistical parameters of the uncertain factors on a periodical basis, thus ensuring that the stochastic evolutions used within the tool reflect the observed real-life dynamics.