Optimization of long-term renewal investment plan for drinking water conveyance networks using dynamic programming

The increasing issue of aging water pipes highlights the need for a long-term renewal investment plan to respond effectively within constrained resources. This study proposes an optimization method designed to minimize life cycle costs while considering the structural safety of pipes and economic efficiency in pipe renewals. The method consists of two prediction models: artificial neural network (ANN)based corrosion prediction models and ANN-based failure rate prediction models. This study uses these models to develop a dynamic programming (DP)-based optimization method that meets annual budget constraints. It applies successfully to a target area of conveyance pipe networks. Moreover, the findings demonstrate that adhering to annual budget constraints allows water utilities to allocate their budgets efficiently, facilitating strategic investments within set yearly limits. These findings offer substantial promise as foundational technologies for the asset management of water supply infrastructures, highlighting their potential for future applicability and value.