Assessing Risk Aversion and Two-Stage Uncertainty Representation in Strategic Wind Power Investment

In this paper, we investigate strategic wind power investment in competitive electricity markets, considering participation in both the day-ahead and balancing markets. A strategic wind power investor determines the location and capacity of new wind power units. The upper-level objective minimizes the investor’s net cost by accounting for investment expenditure and revenues from participation in the day-ahead and balancing markets. The problem is formulated as a bilevel optimization model, where the upper level represents the investor’s strategic decisions and the lower level captures the electricity market-clearing process. The bilevel formulation is transformed into a Mathematical Program with Equilibrium Constraints (MPEC) and subsequently reformulated as an equivalent Mixed-Integer Linear Programming (MILP) problem. The main contributions of this paper are two-fold. The first contribution is the introduction of risk aversion through Conditional Value-at-Risk (CVaR), applied to balancing costs to limit exposure to severe losses caused by wind production deviations. The second contribution is the replacement of the static uncertainty representation with a two-stage scenario tree, allowing balancing decisions to adapt as information becomes available. The proposed formulations are evaluated on 3-, 24-, and 118-bus test systems. Results show that stronger risk aversion leads to more conservative investment decisions and reduces exposure to extreme balancing losses, while the two-stage scenario-tree formulation improves economic performance through sequential adaptation of balancing decisions.