A day-ahead market model for power systems: benchmarking and security implications

A new AI study reveals that standard models used to secure power grids dangerously overestimate their resilience. By simulating a more realistic, market-driven day-ahead dispatch on the IEEE-118 bus system, researchers found the potential for unmet demand could be up to 80% higher than traditional optimal power flow models predict. This exposes a significant security gap, as current assessments ignore complex market interactions that push systems toward cost-minimizing but riskier configurations.