Hybrid Approach for Chlorine-Compliant Blow-Off Optimization in Dead-End Water Distribution Branches
Dead-end sections of water distribution systems often experience long residence times and stagnation, causing chlorine residuals to fall below regulatory standards and leading utilities to run continuous blow-offs that increase nonrevenue water. This study proposes a hybrid advection–reaction (AR) and advection–dispersion–reaction (ADR) screening–validation framework to identify cost-effective continuous blow-off configurations that satisfy minimum chlorine residual and pressure requirements across multiple operating scenarios. Candidate configurations are screened in EPANET, a public-domain hydraulic and water quality modeling software, using an AR formulation, within nondominated sorting genetic algorithm II (NSGA-II) to minimize worst-case chlorine deficit penalty while maximizing worst-case treatment-cost savings across multiple scenarios. Shortlisted nondominated configurations are then reevaluated in Washington University dead end simulator (WUDESIM), a dead-end water quality model that uses an ADR formulation and incorporates stochastic water demands, to confirm compliance in dispersion-dominated dead ends and to quantify AR–ADR discrepancy. The framework is tested on a multisource Canadian municipal system with 63 operational blow-offs supported by calibrated hydraulic and chlorine models and stochastic household demands. EPANET screening produced 14 configurations; ADR validation confirmed two configurations satisfied chlorine and pressure criteria, enabling closure of approximately 68%–73% of blow-offs while maintaining compliance, and yielding annual treatment-cost savings of approximately 64%–70% compared with baseline operation with all blow-offs open. AR screening underestimated the mean dead-end chlorine deficiency by up to 14% relative to ADR validation, indicating the importance of ADR confirmation for conservative closure decisions. Sensitivity analysis identifies nodal demand as the dominant source of uncertainty, followed by bulk decay, with wall decay having a smaller influence. Overall, the proposed hybrid workflow provides an efficient and practical approach to ensuring that safe chlorine levels are maintained.