A parametric study on the impact of open-source inventory variability and uncertainty for the life cycle assessment of road bituminous pavements

Abstract

Current asphalt pavement life cycle assessment implementation limitations are in part linked to data availability. On top of this, there is currently a trend towards the creation of open-source accessible data for pavement LCA. Therefore, the objective of this study is to undertake a comprehensive assessment of open-source data for the “cradle-to-laid” LCA of asphalt pavements to better understand its potential risks. This study assesses the variability and uncertainty associated with open-access life cycle inventory data, collated from multiple open-source case study scenario results, project reports and inventories in order to assess its reliability. Six impact categories were considered. The collected data was applied to three different case study scenarios for a pavement surface course, which include a conventional hot-mix, a half-warm mixture and a half-warm mixture with 50% reclaimed asphalt pavement. The results of this study indicated the variabilities and uncertainties found at both the process level and those propagated to the final LCA results. The results found that fine aggregate production, truck transportation and construction machinery were the most troublesome processes for LCA uncertainty, given their large variability and low impact on the final LCA results. Meanwhile, the binders and cement production and plant operations, being the largest contributors to the environmental impacts (33, 23, 32% average, respectively), also provided large result uncertainties (12, 31, 21% average, respectively) across all impact categories for the cradle-to-laid assessment. Overall, total acidification potential and particulate matter formation potential were the two impact categories most likely to increase due to source variabilities (483.0 and 335.3%, respectively), whilst eutrophication potential was the most likely to decrease (61.7%) for the case studies assessed. The results of this work show that open source data could be used for the LCA of asphalt mixtures and pavements. Nonetheless, justified data sources should be used and users must be aware of the potential uncertainty areas within the system, where this paper has aimed to contribute to the identification and understanding of these uncertainties in order for users minimise them in their work and help identify targets for LCI development. Future work should address the variabilities between open-source and commercial tools, the exploration of the use, maintenance and end-of-life stages and the better quantification of binder, RAP and low-temperature manufacturing processes.