Bulk power systems, commonly referred to as power grids, need to be safely operated under uncertainty in load demand and power generation as well as unplanned loss of system elements. In recent years, increasing participation of variable generators, like wind and solar generators, has significantly increased the system uncertainty, and therefore, increased grid vulnerability to inadequate or insufficiently flexible power supply. In this work, we present a reliability and risk assessment framework to evaluate the adequacy and flexibility associated with a generator unit commitment and economic dispatch decision. These decisions may be made using traditional, deterministic security-constrained unit commitment and security-constrained economic dispatch optimization algorithms or using advanced, stochastic optimization algorithms that have been proposed in the literature. We define risk and reliability metrics at three levels: conditional expectation, probability of failure, and risk. The first two metrics can be used for reliability assessment, whereas the third metric considers the (monetary) consequence to evaluate the risk. The proposed framework could be used to evaluate and communicate the reliability/risk of a proposed generator portfolio, enabling day-ahead or hours-ahead risk assessment and risk-versus-cost trade-off analysis. It can also be used to assess the suitability of various operational optimization algorithms for maintaining the desired risk tolerance. We demonstrate the computation of these metrics for a 200-bus synthetic grid. We also show how these metrics can be used for performing day- or hours-ahead risk assessment, as well as for selecting a suitable decision-making algorithm.
展开▼
