Multi-objective Spatial Optimization of Decentralized Water Reuse Implementation and Service Allocation in Hong Kong

Hong Kong is one of the fast-urbanized cities in the world with a population of more than 7.4 million, consuming about 21% more freshwater per capita than the global average. However, local yields only account for 30% of the city’s total water supply due to its mountainous terrain, making it hard to collect or store rainwater. Considering its high demand but low supply, this city adopted a dual water supply system to extend the use of seawater and lower-grade water for non-potable purposes; and has been actively pursuing water reclamation as a valuable alternative source which is more calculable in quantity. Decentralized water reuse (WR) emerges as a potential option that can enhance urban water security and sustainability by mitigating the reliance on freshwater imports and energy consumption for water transmission and distribution. Despite technological developments, the implementation and guidance for water reuse applications are still lacking. There are minimal spatial planning concepts or practices to drive water reuse deployments across different scales. To fill the gaps, we developed an integrated spatial water-energy modeling and multi-objective optimization framework to support the citywide implementation of WR facilities using Hong Kong as a testbed. The framework starts with calculating daily freshwater & seawater demands and wastewater production of each urban community based on water consumption surveys of residential, commercial, and industrial uses. Based on the estimation, we calculated the hydraulic flows and energy consumptions at different water transmission stages, from water sourcing, treatment, and distribution to wastewater collection, treatment, and discharge. The spatial water-energy accounting highlights regions with intensive water and wastewater services and serves as a benchmark for further optimizing WR deployments and their impacts. In the optimization phase, we used Genetic Algorithm to evaluate and optimize the implementation of WR facilities from the perspectives of minimizing the freshwater import, electricity use, and investment costs. Afterward, we simulated the water age of the freshwater supply network as an external constraint to eliminate infeasible solutions from the optimal ones (i.e., Pareto-fronts), including those of which the water age would either double or exceed 28 calendar days in over 5% of total urban communities. Our optimization results spatially identify the optimal location and treatment capacity for designing each WR facility and the service allocations between WR facilities and urban communities as investment increases. The reduction in freshwater & seawater withdrawal and electricity use was evaluated as the impacts/benefits on urban water systems. Overall, our framework can provide a systematic view of spatial electricity intensities for the urban water system and help cities adaptively integrate the water reuse concepts into urban water infrastructural planning to realize holistic-integrated water resource management in a more sustainable and cost-effective way.