Skip to Main Content
This paper presents a stochastic mixed-integer linear programming (SMILP) approach to maximize total expected profit of one price-maker hydro producer in a pool-based electricity market. Head dependence, commitment decisions, discharge ramping, startup costs and forbidden zones are all effectively handled in our approach. Uncertainty about the competitors' offers is adequately represented by residual demand curves (RDCs) scenarios. The management of risk is suitably addressed by conditional value-at-risk (CVaR) to provide the efficient frontier, i.e., the solutions set for which the expected profit may not be augmented without enlarging the variance of profit. Appropriate offering strategies to the pool are also developed, consisting of supply functions built for different risk levels. A representative cascaded hydro system with 7 reservoirs is considered to analyze and compare risk-neutral versus risk-averse results.