Risk-constrained strategic bidding of a hydro producer under price uncertainty

Abstract

Price uncertainty faced by generation companies in electricity spot markets is a relevant source of risk. The inherent volatility of price series can have a direct influence on short term operational profits, and therefore, there is a natural trend to incorporate risk aversion criteria into the operational decisions. This presentation shows the mathematical formulation of the strategic bidding problem faced by a hydroelectric company, taking into account both price uncertainty and risk aversion constraints. The objective of the model is to find the optimal supply functions to be submitted to the market operator. The company is assumed to be price-taker, and therefore, prices are considered exogenous variables. Uncertainty is modeled via scenarios, which are generated by an inputoutput hidden Markov model (IOHMM). Risk aversion is introduced by means of coherent risk measures, such as the conditional value-at-risk (CVaR) and its generalized version (GCVaR) that can be handled easily in the context of LP and MILP optimization models.

Price uncertainty faced by generation companies in electricity spot markets is a relevant source of risk. The inherent volatility of price series can have a direct influence on short term operational profits, and therefore, there is a natural trend to incorporate risk aversion criteria into the operational decisions. This presentation shows the mathematical formulation of the strategic bidding problem faced by a hydroelectric company, taking into account both price uncertainty and risk aversion constraints. The objective of the model is to find the optimal supply functions to be submitted to the market operator. The company is assumed to be price-taker, and therefore, prices are considered exogenous variables. Uncertainty is modeled via scenarios, which are generated by an inputoutput hidden Markov model (IOHMM). Risk aversion is introduced by means of coherent risk measures, such as the conditional value-at-risk (CVaR) and its generalized version (GCVaR) that can be handled easily in the context of LP and MILP optimization models.