Towards a Stochastic Operation of Switzerland’s Power Grid

Abstract: As Europe’s power production becomes increasingly reliant on intermittent renewable energy sources, uncertainties are likely to arise in power generation plans. Similarly, with the growing prevalence of electric vehicles, electric demand is also becoming more uncertain. These uncertainties in both production and demand can lead to challenges for European power systems. This thesis proposes the use of Monte-Carlo simulations to translate uncertainties in power generation and demand into uncertainties in the power grid. To integrate stochasticity in the forecasts, this thesis separates the multivariate probabilistic forecasting problem by first forecasting the marginal loads individually and probabilistically. Copula theory is then used to integrate spatial correlations and create realistic scenarios. These scenarios serve as inputs for Monte-Carlo simulations to estimate uncertainties in the power system. The methodology is tested using power injection data and the power system model of Switzerland. The results demonstrate that integrating stochasticity in forecasts improves the reliability of the power system. The proposed approach effectively models the uncertainty in both production and demand and provides valuable information for decision-making.