Chemical product design integrating MCDA: Performance prediction and human preferences modelling

Computation‐based techniques and modelling of human knowledge and preferences by using multi‐criteria decision aid (MCDA) methods are integrated in a multi‐scale and multi‐disciplinary approach for the chemical product and process design. The proposed methodology has four main stages: (a) construction of a molecular model for predicting product performance, (b) validation of product performance, (c) selection of alternatives integrating preferences of manufacturers and consumers, and (d) process optimization implementing MCDA methods. The methodology is oriented to find new products that can replace components of formulations, whose performance is already known. It was applied to an exploratory study about the use of glycerol as raw material to produce plasticizers for polyvinyl chloride (PVC), replacing 2‐ethylhexyl phthalate (DEHP). The results of the case study and the proposed process design offer promising prospects regarding their application in other chemical products. Scheme for a chemical product design methodology using multi‐criteria decision aid.