Stochastic Capacity Expansion Model Accounting for Uncertainties in Fuel Prices, Renewable Generation, and Demand

Capacity expansion models for electricity grids typically use deterministic optimization, addressing uncertainty through ex-post analysis by varying input parameters. This paper presents a stochastic capacity expansion model that integrates uncertainty directly into optimization, enabling the selection of a single strategy robust across a defined range of uncertainties. Two cost-based risk objectives are explored: “risk-neutral” minimizes expected total system cost, and “risk-averse” minimizes the most expensive 5% of the cost distribution. The model is applied to the U.S. Midwest grid, accounting for uncertainties in electricity demand, natural gas prices, and wind generation patterns. While uncertain gas prices lead to wind additions, wind variability leads to reduced adoption when explicitly accounted for. The risk-averse objective produces a more diverse generation portfolio, including six GW more solar, three GW more biomass, along with lower current fleet retirements. Stochastic objectives reduce mean system costs by 4.5% (risk-neutral) and 4.3% (risk-averse) compared to the deterministic case. Carbon emissions decrease by 1.5% under the risk-neutral objective, but increase by 3.0% under the risk-averse objective due to portfolio differences.