A risk-averse security-constrained optimal power flow for a power grid subject to hurricanes

Abstract

During the course of a hurricane, many components in the power grid may be affected. In particular, loss of transmission lines and/or towers due to excess wind conditions may adversely impact the operation of the grid and force a re-dispatch of the generation units. However, large generation units have considerable ramp rates and usually are not able to vary their outputs fast enough. This might lead to temporary imbalances between load and generation that, if not resolved quickly, may result in more severe cascading failures. When a large scale disturbance such as a hurricane is forthcoming it is most beneficial to proactively dispatch the grid so as to minimize the likelihood of future failures. To assist the operator in proactively responding to an imminent event such as a hurricane, a risk-averse generation dispatch model is presented in this paper based on security-constrained AC optimal power flow. To perform (N–k) contingency analysis, a geospatial model of the power grid is developed that allows for the computation of outage probabilities of the transmission lines affected by the hurricane wind fields. Statistical analysis has been performed on the historical data on the past hurricane events in the US in order to simulate realistic hurricane scenarios. The IEEE 118-bus test system has been mapped onto the map of the state of Texas in order to provide a realistic test bed. The proposed algorithm takes into account the cost of operation, as well as the risks associated with overload and over/undervoltage conditions. Moreover, it allows for preventive as well as corrective dispatch of the power grid.