Uncovering relationships between environmental metrics in the multi-objective optimization of energy systems: A case study of a thermal solar Rankine reverse osmosis desalination plant

Multi-objective optimization (MOO) is increasingly being used in a wide variety of applications to identify alternatives that balance several criteria. The energy sector is not an exception to this trend. Unfortunately, the complexity of MOO grows with the number of environmental objectives. This limitation is critical in energy systems, in which several environmental criteria are typically used to assess the merits of a given technology. In this paper, we investigate the use of a rigorous dimensionality reduction method for reducing the complexity of MOO as applied to an energy system (i.e., a solar Rankine cycle coupled with reverse osmosis and thermal storage). Instead of using an aggregated environmental metric, a common approach for reducing the number of environmental objectives in MOO, we propose to optimize the system in a reduced search space of objectives that fully describe its performance and which results from eliminating redundant criteria from the analysis. Numerical results show that it is possible to reduce the problem complexity by omitting redundant environmental indicators from the optimization.

► The application of a rigorous dimensionality reduction method in the multiobjective optimization of a reverse osmosis plant was studied. ► Individual environmental metrics were used instead of an aggregated one. ► Redundant objectives were identified and eliminated for the analysis. ► The computational task was simplified and the computational burden reduced.