Optimal wind farm bids under different balancing market arrangements

If a wind power producer must pay the costs of imbalances, the question arises of what is the optimal bid, given the market rules and the statistical properties of the wind forecasts and the imbalance prices. In this paper we derive optimal wind power bids for two sets of market rules, reflecting previous and new rules in the Nordic power system. The optimal bids are based on the evaluation of a large number of scenarios for the realizations of the wind forecasts and the balancing market prices respectively. Scenario aggregation is used to limit the total number of scenarios. The wind forecast errors are described by a traditional ARMA model. The balancing market prices are described by a model that uses time series and statistical models of the volumes and prices on the balancing market. The optimal bid in the so-called 1-price system will normally be to either bid zero or maximum production. In the alternative 2-price system, the optimal bid will be close to expected production. For the particular case study in this paper, the optimal bid for the 1-price system gives an improvement in total revenues of 2.5%, while there is no observable improvement for the 2-price system. As such, the new system does give incentives to wind power producers to bid “correctly”. However, the total revenues for the wind park are reduced with 2% in the 2-price system, compared with the 1-price system.