Comparative life cycle assessment of industrial multi-product processes

The demand for environmentally safe industrial processes is increasing. Therefore, environmental impacts of new processes have to be examined at an early stage. A method for analyzing environmental impacts is life cycle assessment (LCA). A major trouble of LCA are multi-functionality problems. Multi-functionality problems can be fixed using alternative methods such as system expansion, avoided burden and allocation. Each of the three methods requires choices by the LCA-practitioner. The choices made during fixing a multi-functionality problem increase uncertainty and reduce the acceptance of LCA. An international Standard and several guidebooks present recommendations for selecting an appropriate method to fix multi-functionality problems. But the recommendations are not yet specified for comparative LCA of alternative multi-product processes. Moreover, uncertainties due to fixing multi-functionality problems are rarely included in systematic uncertainty analysis. In some LCAs, alternative methods for fixing multi-functionality problems are compared in scenarios. Uncertainties in fixing multi-functionality problems are sometimes analyzed using stochastic methods. However, the approaches for analyzing uncertainties due to fixing multi-functionality problems are not yet standardized or even included in the computational structure of common LCA software tools. In this work, a systematic procedure is presented for fixing multi-functionality problems in comparative LCA of multi-product processes. Moreover, an analytical method is introduced for analyzing uncertainties due to fixing multi-functionality problems. The procedure for fixing multi-functionality problems in comparative LCA of multi-product processes requires distinguishing main products and by-products. The distinction is based on economic figures. A multi-functionality problem resulting from two alternative multi-product processes with differing main products should be fixed using system expansion. In contrast, a multi-functionality problem caused by different by-products should be fixed using avoided burden. If avoided burden cannot be applied, allocation should be used to fix the multi-functionality problem. An existing matrix formulation for LCA calculations is expanded to include methods for fixing multi-functionality problems. The expanded matrix formulation involves continuous parameters modelling the choices during fixing multi-functionality problems. The derived equations are used to calculate sensitivity coefficients with respect to the choices made during fixing a multi-functionality problem. Moreover, first-order error propagation is applied to systematically analyze uncertainties due to fixing multi-functionality problems. The derived methods are applied in a case study of two alternative processes for chlor-alkali electrolysis. Existing chlor-alkali electrolysis produce chlorine, caustic soda, and hydrogen. A novel process requires less electricity but only produces chlorine and caustic soda but no hydrogen anymore. A comparative LCA thus includes a multi-functionality problem. The procedure for fixing multi-functionality problems is applied in the case study. Two scenarios are analyzed depending on alternative use of hydrogen from existing chlor-alkali electrolysis. In both scenarios, the environmental impacts of the novel process are lower compared to the existing chlor-alkali electrolysis. The analytical method for analyzing uncertainties due to fixing multi-functionality problems is applied to the case study as well. The results of the uncertainty analysis show that uncertainties due to fixing multi-functionality problems cannot be ignored. In future work, the methods developed in this work can be integrated in existing software tools for LCA calculations