Multi-scenario-based optimization with CVaR risk control for renewable energy participation in electricity spot markets

Electricity spot-market participation by renewable producers is exposed to coupled uncertainty in generation, demand, and prices. This study develops a day-ahead bidding framework that combines multi-scenario stochastic optimization with Conditional Value-at-Risk (CVaR) control of downside revenue shortfall. A large pool of temporally correlated trajectories is reduced by fast-forward selection, and the resulting fixed bids are evaluated on an independently generated sample. A deterministic forecast bid and a risk-neutral stochastic model provide explicit benchmarks. The out-of-sample evaluation shows that stochastic bidding reduces deviation penalties relative to the deterministic baseline, while the CVaR model accepts a small expected-revenue concession relative to the risk-neutral solution in exchange for lower tail shortfall and less volatile profit. These findings support the internal revenue-risk logic of the proposed method under the constructed data-generating process; validation with measured market data remains necessary before practical deployment.