Midpoint vs single score in multi-criteria optimization under life cycle assessment constraints: the case of potable water treatment chains

Abstract

This paper addresses the multi-objective constrained optimization of a drinking water production plant. It reports the successful coupling between the Strength Pareto Evolutionary Algorithm (SPEA2), which is a well-established multi-goal elitist metaheuristic global optimizer, and EVALEAU, which is a state-of-the-art process modelling – life cycle assessment (LCA) tool for simulation of potable water treatment chains. The paper assesses as well the pros and cons of using midpoint versus single score as environmental goal metrics in a decision making process based on very challenging multi-goal optimization of potable water treatment chains. The proposed approach is illustrated using the model of a real-world drinking water production plant. Keywords— drinking water production plant; life cycle assessment; multi-objective optimization; meta-heuristic algorithms; ReCiPe single score; ReCiPe midpoint level impact.