Two-level unit commitment and reserve level adjustment considering large-scale wind power integration

Integrating wind power generation into bulk power grids imposes significant challenging to system operations due to the stochastic variation and limited predictability of wind power generation. One of the key issues is how to optimally schedule various units and determine a proper spinning reserve level to guarantee a reliable operation of the integrated system in a robust and economic manner. In this paper, we propose a two-level unit commitment (TLUC) for power systems facing large uncertainties in the context of adjustable robust optimization. The commitment decision acts as first-stage decisions that need to be robust against uncertainty, whereas the output decision acts as adjustable decisions that is decided after uncertainty is revealed. This decision mechanism naturally decomposes a conventional unit commitment (UC) problem into two levels, the upper level provides unit working status, whereas the lower level checks if there is at least one dispatch solution under uncertainty subject to the status. Unlike conventional UC, where the reserve level is fixed from operating experiences, in TLUC, the reserve level also acts as a decision variable in the upper level that controls the robustness of UC decisions. A heuristic algorithm is developed to solve this TLUC problem, providing the robust UC decision against uncertainty as well as the minimal spinning reserve level required to maintain reliability. Numerical experiments are carried out on the IEEE 39-bus system and large-scale realistic Guangdong power grid in China and demonstrate the effectiveness of our approach. Copyright (c) 2013 John Wiley & Sons, Ltd.