Linear integer programming for power system recovery following outages

摘要

System reliability and resilience are crucial for electric power networks. Outages caused by extreme events or equipment failures are a threat to the reliability of a power system. An important step to deal with this problem is to manage power recovery effectively. To this end, it is necessary to determine the outage scenario accurately and restore the affected grid efficiently. This dissertation is concerned with the development of methodologies for decision-support of outage management for both distribution and transmission subsystems of the power grid.;In this dissertation, a new algorithm based on multiple hypotheses is proposed to quickly determine outage scenarios in distribution systems. The result is a decision-support tool to enhance outage management of distribution feeders. By incorporating information from smart meters, the faulted line section and activated protective device(s) are located. Complex outage scenarios with multiple faults, failures of fault indicators, and protection miscoordination are identified. Simulation results using the Avista-Pullman distribution system with around 13,000 smart meters validate the effectiveness of this algorithm for enhanced outage management of distribution systems.;This dissertation also proposes a method for system recovery of the transmission system. When a blackout occurs, the primary task for system operators is to restore power grids safely and efficiently. In this research, a computational algorithm is proposed for system restoration planning. The result is a sequence of actions to energize generators, transmission substations and lines, and critical loads. Based on the proposed method, a software tool has been developed to assist system operators and planners in decision-making of system restoration. Options for a proposed installation of new blackstart capability are evaluated based on the contribution of each option to the overall available system generation capability. Real-world power system test cases from New England Independent System Operator (ISO) and Duke-Energy are used to demonstrate the optimality and efficiency of the developed tool for power system restoration planning.;By solving the above power system recovery problems with advanced linear integer programming techniques, the results of this dissertation provide new computational strategies for enhanced outage management of distribution systems and blackstart capability planning for transmission systems.