European Power Market Model
Abstract
We develop and present a stochastic power market model to study the short-term spot market in Northern Europe. The model is formulated as a Mixed Linear Complementarity Problem using Conjectural Variations for the market power representation. The power producers and the transmission system operator simultaneously solve a profit maximization problem to reach a Nash-Equilibrium. The model results show dispatch, transmission and prices in 2010 and 2020 for different stochastic wind forecasts. We see a clear distinction between technologies used for base load and for balancing the market. Prices in Norway and Denmark will increase towards 2020 and decrease in Germany and the Netherlands. We study how our market power representation can be adjusted to be as realistic as possible, the effect of wind uncertainty, how robust the power system is with respect to volatile prices and the effect of the carbon price and bio power feed-in tariffs. Because of the uncertain wind there is an increase in natural gas dispatch compared to a deterministic case. The model is also decomposed with Dantzig-Wolfe decomposition technique. Unfortunately the decomposed model is not able to solve as big problems as the non-decomposed model. However we show significant algorithmic improvements and are able to reduce the number of iterations on a small problem from 183 to 17.