A Process Modelling-Life Cycle Assessment-MultiObjective Optimization tool for the eco-design of conventional treatment processes of potable water

The eco-design of conventional potable water production processes is impeded by the high variability of operating conditions as a function of the inlet and the quality of outlet water, and by the large diversity of feasible technical solutions and treatment processes. Based on an already developed library of unit process modules which generates water treatment process inventories with integrated Life Cycle Assessment (EVALEAU tool), this work presents a new tool to combine Process Modelling, Life Cycle Assessment and MultiObjective Optimization (PM-LCA-MOO tool) for conventional potable water production processes, in order to solve the challenges of interconnecting LCA results and other conflicting process objectives. A hybrid (local-global) derivative-free algorithm was chosen for multiobjective optimization which includes: (1) the elitist Non-dominated Sorting Genetic Algorithm (NSGAII) for the first global search in the objective space to locate the most promising optimal zones; and (2) the COBYLA algorithm starting with final solutions of NSGAII, to improve accuracy within a reasonable calculation time. The PM-LCA-MOO tool was successfully applied to a test bed case of an existing drinking water plant located in the Paris region and resulted in a set of alternative solutions, known as the global Pareto-optimal front that trades different objectives against each other. Here, the objectives are the minimization of environmental impacts using the ReCiPe method (according to ISO 14040/44), the minimization of operation costs, and the maximization of produced water quality. According to conventional drinking water plants, the quality of water produced has to be optimized together with other environmental and economic objectives in order to grant fair design solutions. The quality of water produced incorporates multiple drinkability criteria that were aggregated in order to indicate water quality by way of a unique quality index the 'Average Water Quality Score'. (C) 2015 Elsevier Ltd. All rights reserved.