Simultaneous Finding the Optimal Location, Process and Planning of Wastewater Treatment Plants Using Meta-heuristic Algorithms Equipping with GIS

• Simultaneous determination of the number, location, process, and capacity of WWTPs • Determining suitable areas for locating WWTPs using Boolean analysis in GIS • Modeling WWTP construction, operating, and conveyance costs • Defining the optimization model with a Reduced-Dimension Formulation Approach • Deriving cost functions for three wastewater treatment processes Amid water scarcity, wastewater treatment and reuse are essential to augment resources and protect ecosystems. This research proposes an integrated Mixed-Integer Nonlinear Programming (MINLP) optimization framework that simultaneously optimizes the number and locations of wastewater treatment plants (WWTPs), the biological treatment process, and the capacity to minimize total system cost. The model integrates a Geographic Information System (GIS) with meta-heuristic optimization algorithms, namely Particle Swarm Optimization (PSO) and a Genetic Algorithm (GA). Three widely applied biological treatment processes, extended aeration (EA), sequencing batch reactor (SBR), and moving bed biofilm reactor (MBBR), were considered. Process-specific construction and operating cost functions were developed using statistical analysis of real-world data. These were used in the optimization framework, along with wastewater conveyance costs. A conventional Full Variable Formulation Approach (FVFA) was first implemented, but its high dimensionality made it computationally infeasible. To overcome this limitation, a Reduced Dimension Formulation Approach (RDFA) was developed, improving convergence and enabling the model to obtain an optimal, feasible solution. The RDFA reduces dimensionality by encoding spatial decisions using integer index variables. The proposed framework was assessed using Isfahan, Iran, as a case study. Results identified an optimal configuration of two WWTPs with EA selected at both locations. The minimum total cost was estimated at 222.10 million USD using PSO, which was 0.23% lower than the GA results. This demonstrated the robustness and reliability of the proposed RDFA-based optimization model. These findings highlight the applicability of the proposed framework for large-scale urban wastewater planning, offering a computationally efficient approach.