Life cycle assessment of chemicals: data generation and uncertainty

Abstract

Chemicals play an important role in many products and industries, but there is a lack of data on the production processes and environmental impacts of the different types of chemicals. Based on this problem, this paper presents a review of methods for the generation of life cycle inventory data and life cycle impact assessment results for chemicals not yet included in life cycle assessment databases. 31 methods in peer-reviewed papers, mainly from 2012 to 2021, were evaluated and methodologies on how to gain the data were identified and clustered. The identified subcategories include knowledge engineering and data mining, process simulation, predictive LCA using machine learning, and multivariate statistics as well as the identification of chemicals for special purposes using computer-aided molecular design. The areas and literature were summarized by application and according to advantages or disadvantages of the methods. Additionally, the possibilities of including model and data uncertainties were discussed and open science aspects of software and programming code were evaluated, as these provide a good basis for other researchers to further expand the solutions developed.