Effective multi-objective selection of inter-subnetwork power shifts to mitigate cascading failures

Abstract

Electricity market forces and renewable energy integration promote the exchange of large power volumes over long distances, push the grid to operate closer to its limits, and make it vulnerable to cascading outages, e.g., triggered by disconnections of overloaded lines. During such emergencies, system operators need to redistribute power quickly and adequately. This paper presents a heuristic decision support system for alleviating line overloads and mitigating the consequences of cascading events. A graph-partitioning algorithm is developed to identify the subnetworks for which the overloaded lines act as power gates. Inter-subnetwork power shifts by generation rescheduling and, possibly, load shedding are applied to relieve the overloaded lines which bridge these subnetworks. The power shifts are selected based on their potential to simultaneously reduce overloads and prevent new ones. This multi-objective decision making is based on the sensitivities of the overloaded lines with respect to power shifts and on the time required to complete them. Line overloads are eliminated by guiding the formation of self-sufficient subnetworks. The framework has been tested on the modified RTS-96 and on the IEEE 118 systems to illustrate its capability of load shedding reduction using a developed cascading outages simulation model. A novel optimization method which minimizes the redispatch time is formulated to validate the proposed framework, and the impact of considering ramping rates is shown with comparison to standard optimization based on minimizing redispatch power. The sensitivity-based method identifies compromise solutions which minimize the redispatch power and the redispatch time, and support the dispatcher's decision making by providing further corrective actions alternative to normal OPF.