Ethylene and propylene production from steam cracking in Europe: a consequential perspective

PurposeEthylene and propylene production is currently included in life cycle assessment (LCA) databases using an attributional life cycle inventory (LCI) modelling approach. This approach entails modelling both products as co-products of the steam cracking process. In this article, we provide what, to the best of our knowledge, constitutes the first attempt to construct a consequential LCI model for ethylene and propylene production, focusing on European conditions and using publicly available data.MethodsA market analysis for the European market was conducted, showing that steam cracking of naphtha is the marginal production route for ethylene, while for propylene the marginal production route is propane dehydrogenation (PDH). This market analysis also identifies propane as a constrained feedstock and suggests that an increase in demand for propane will induce a shift to gasoline consumption by ‘autogas’ vehicles. Following these findings, we develop a consequential LCI model including PDH, steam cracking, treatment of steam cracking by-products (butadiene extraction, pyrolysis gasoline hydrotreatment, benzene-toluene-xylenes extraction), and marginal production routes for all substituted products, such as hydrogen, butadiene, benzene, toluene, and mixed xylenes, among others. The LCI model was linked to the background ecoinvent database.Results and discussionThe model was evaluated at the impact assessment level, focusing only on greenhouse-gas (GHG) emissions per kg product, showing that they are larger for ethylene (1.83 kg CO2e) compared to propylene (1.35 kg CO2e). In both cases, the main contribution is the supply of feedstock, namely naphtha and propane, respectively, although total emissions are highly influenced by substitutions associated with by-products. For ethylene, several substituted products occur, especially benzene, hydrogen, and propylene, while for propylene the only relevant substitution is for hydrogen. A sensitivity analysis shows that the results for both propylene and ethylene are highly sensitive to how the propane constraint is addressed. In particular, GHG emissions for propylene drop by 48% and those for ethylene increase by 22% when the marginal propane user shifts to natural gas instead of gasoline.ConclusionsA consequential approach shows that a demand for ethylene and propylene, respectively, triggers and affects different production processes, thereby yielding distinct cradle-to-gate environmental impacts for each product. This stands in contrast to typical attributional models employing mass allocation (partitioning), which results in identical impacts per kilogram. Future research efforts should be aimed at validating the presented model, as well as expanding it to cover regions other than Europe, where marginal propylene production routes may vary.