APPROPRIATE Life Cycle Assessment: <b>A</b> <b>PRO</b>cess-Specific, <b>PR</b>edictive <b>I</b>mpact <b>A</b>ssessmen<b>T</b> Method for Emerging Chemical Processes

A sustainable chemical industry needs to quantify its emissions and resource consumption by life cycle assessment (LCA). However, LCA requires detailed mass energy balances, which are usually not available at early process development stages. Here, we introduce a framework (A PROcess-specific, PRedictive impact AssessmenT method for Emerging processes, APPROPRIATE) provide fully automated, predictive the phases of development. Based on Gaussian Process Regression, is already applicable Technology Readiness Level 2. To overcome limited data availability, employ an encoder–decoder network in combination with transfer learning achieve latent representation as condensed molecular descriptor. We further propose integrate only but also descriptors, e.g., stoichiometric sum reactants’ impacts. Thereby, can distinguish between alternatives incorporate changes background systems. The compared state-of-the-art approaches shows increased prediction accuracy terms coefficient determination R2 = 0.61 global warming 0.3 former studies. Highly relevant features impacts descriptors. APPROPRIATE supports decision making stages allowing distinction quantifying predictions’ uncertainty.