Multiobjective\nSpatial Optimization Framework for\nDetermining the Optimal Degree of Decentralization for Nonpotable\nWater Reuse in Existing Cities: A Case Study of Hong Kong

Investing\nin nonpotable water reuse (NPWR) is essential for circular\nurban water management. Existing research lacks methods to determine\nthe number and capacities of NPWR plants (i.e., degree of decentralization)\nfor large-scale applications in existing cities. We developed a spatial\noptimization framework in which the degree of decentralization emerges\nfrom the collective decisions of urban districts regarding where to\nsend wastewater for reclamation and where to source water for nonpotable\nuses. We modified the genetic algorithm to optimize collective decisions\nwith objectives including minimizing freshwater withdrawal, electricity\nconsumption, and the cost of NPWR plants. Optimization results suggest\nan optimal number from one to eight among Pareto optimal solutions,\nwith two to three being most common in Hong Kong. The cost-effective\nsolutions suggest locations of NPWR plants in Kowloon and Hong Kong\nIsland where NPWR demand is significant, while the electricity use\nfor freshwater and seawater is high. The city could save about 6%\nfreshwater and 29.4% seawater while consuming 20.7% more electricity.\nOverall, our spatial optimization framework provides a holistic evaluation\nof the optimal degree of decentralization for NPWR at the water-energy-cost\nnexus on an urban scale. Our findings serve as a benchmark to explore\nmore energy-conscious planning strategies in Hong Kong.