(104h) Identifying the Preferred Subset of Alternatives for Environmental Improvements Via an MILP Approach Based on the Analytic Hierarchy Process

The combined use of multi-objective optimization (MOO) and LCA has recently gained wider interest in process systems engineering. This approach provides as output a set of Pareto solutions that represent the optimal trade-off between the economic and environmental concerns considered in the analysis. From this set of optimal alternatives, decision-makers should identify the ones that better fulfill their preferences. Generating a large and representative enough subset of Pareto points to aid decision-making is challenging. This task is particularly difficult in problems with a large number of (environmental) objectives, as is the case when incorporating LCA principles in MOO.

In this work we present a mixed-integer linear programming method that simplifies MOO problems by concentrating on  determining only a reduced number of Pareto points that are particularly appealing. Our approach, which relies on the analytic hierarchy process (AHP), identifies a set of weights to be assigned to the environmental objectives so as to translate them into a single aggregated indicator that reflects to the maximum extent possible the decision-makers' preferences. For every such combination of weights, we solve a single objective problem that optimizes the corresponding weighted sum of objectives, thereby generating only a subset of alternatives (Pareto solutions) reflecting preferences with a large degree of consistency. We illustrate the capabilities of our approach through its application to the design of supply chains for bioethanol production.