Scheduling of Repair Work Crew Routing for Power Network Restoration under Uncertainty

The power network plays a vital role in modern society, providing essential services, driving economic growth, and ensuring energy security, relying on its uninterrupted and efficient operation. This network is prone to unavoidable disruptive events, which can unpredictably impact its performance and lead to significant social implications. The main topic of this thesis is the recovery operation of power networks after natural disasters, with a specific emphasis on coordinating crews and optimizing power network operations. In the aftermath of significant disaster events, one of the principal challenges for operators is effectively managing crews required to repair damaged lines in the network, thereby reestablishing normal network operations. The objectives of this thesis are to enhance current performance and assist utilities in their decision-making processes, enabling the efficient restoration of the system quickly and safely. Therefore, this study proposes optimizing repair scheduling and operation of power networks by developing a mathematical model for optimizing an integrated framework that schedules repair crews and manages the recovery of the power network. Two distinct approaches are presented in this thesis. Firstly, a deterministic mixed-integer linear program (MILP) is developed to facilitate the restructuring of the network and the routing of repair crews, aiding the efficient scheduling of repair activities for damaged lines between the generation facility and distribution nodes in a power network. Secondly, the mathematical model is adapted by introducing a parameter of uncertainty, converting the original MILP model into a two-stage Stochastic Mixed-Integer Linear Program (SMILP) incorporating a scenario-based optimization technique. This modified model addresses uncertainties in repair times within the repair and restoration problem context. The first stage of the SMILP involves the dispatch of repair crews to damaged lines based on the expected required time for repair. The outcome of the first stage decisions provides routes for the crews, while the second stage is dedicated to restoring the power network under all realized scenarios. The integrated model aims to minimize penalty costs incurred due to unmet demand and the costs associated with inoperable lines caused by link disruptions. A numerical example is utilized to validate the model, with extensive sensitivity, uncertainty, and complexity analyses demonstrating the model's performance. Finally, comprehensive discussions are drawn from the results obtained.