Large-Scale Multi-Period Optimal Asset Renewal Planning: A Comprehensive and Efficient MILP Approach

This is the authors' abstract. We don't add key points for this paper.

Many electric utilities worldwide are facing critical challenges of cost-effectively renewing aging assets to maintain power network reliability. Existing asset renewal planning (ARP) methods often have limited applicability in complex real-world scenarios, and may fail to capture the synergy between different asset types and the interplay between proactive and reactive renewals. To address these limitations, this paper develops a mixed- integer linear programming (MILP)-based approach for large-scale multi-period ARP. This approach formulates the multi-period ARP problem as an MILP model, which comprehensively accounts for the costs associated with both proactive and reactive asset renewals, regular asset maintenance, and system failures, thus capturing the interplay among them. Real-world limitations in manpower resources and requirements for system reliability can be conveniently integrated into the constraints of the MILP model. The approach integrates the renewal of all critical asset types, including transformers, switchgears, and cables, into one unified framework, maximizing synergies across them. Moreover, a candidate-selection-based iterative solution method is devised for the MILP model, which ensures computational tractability for large-scale ARP problems. The application of the approach to a real distribution network demonstrates its effectiveness and superiority over the traditional age-based asset renewal planning.