Equilibrium simulation of the Nordic electricity spot price

Abstract

We develop a bottom-up, midterm model for the Nord Pool system price. Models for consumption, generation and exchange are combined using a market equilibrium approach. Scenarios for the most influential stochastic factors are created using statistical models. Forecasted price scenarios yield an indication on the expected future system price, and the uncertainty in future prices. Compared to stochastic dynamic programming based models, we use an empirical function to approximate the water value, with a possible benefit in terms of computational effort. In addition to providing a picture of the price distribution, the model can be applied to learn about the impact of fundamental variables. In particular, we study the relative influence of each of the stochastic factors. The performance of the model is evaluated in numerical examples over a two-year forecast period. Overall, we conclude the model provides realistic forecasts of the system price and its distribution. The performance under normal market conditions may deviate from our results, due to the extraordinary drop in fuel prices caused by the financial crisis.