Unit Commitment With Probabilistic Reserve

This paper demonstrates how a probabilistic criterion based on the full capacity outage probability distribution (e.g., LOLP) could be integrated into the unit commitment (UC) optimization using simple statistical approximation. While this provides a direct and computationally cheaper means as compared to a recently published approach to locating the globally optimal SR allocation (and associated UC pattern), the usefulness of the results are contingent upon the accuracy of such approximation. A case study using the IEEE RTS 96 single area system discusses the implementation issues. It is shown that the approximation procedure provides a reasonably accurate and efficient means to integrate probabilistic reserve in UC, and such a criterion leads to a substantially improved risk profile as compared to the traditional deterministic criterion.