Capacity response sets for security-constrained unit commitment with wind uncertainty

Abstract

Power systems face growing uncertainty from intermittent renewable resources. This uncertainty makes it harder to schedule generators efficiently because the economics depend on unknown outcomes. This research modifies security-constrained unit commitment to anticipate the cost of dispatching backup capacity if and when it is needed. The model uses capacity constraints to cover individual scenarios by defining scenario response sets, which are revised iteratively using a mixed-integer program. The approach provides a mathematical way to derive capacity requirements like those used today in CAISO and ISO-NE. Testing on the IEEE 73-bus test case demonstrates the capacity requirements are economical compared to other deterministic policies from the literature.