Abstract
Purpose
Recently, an update of the Life Cycle Impact Assessment (LCIA) method ReCiPe was released: ReCiPe 2016. The aim of this study was to analyse the effect of using this update instead of the previous version: ReCiPe 2008. Do the absolute outcomes change significantly and if so, does this lead to different conclusions and result-based recommendations?
Methods
Life cycle assessments (LCAs) were conducted for 152 foods for which cradle-to-plate inventories were available and that together are estimated to account for 80% of the total greenhouse gas emissions, land use and fossil resource depletion of food consumption in the Netherlands. The LCIA was performed on midpoint and endpoint level, with both ReCiPe 2008 and 2016, and using the three perspectives provided by ReCiPe. Both the uses of the global-average characterisation factors (CFs) and the Dutch-specific CFs were explored.
Results and discussion
Results showed a strong correlation between LCAs performed with ReCiPe 2008 and with 2016 on midpoint and endpoint level, with Spearman’s rank correlation between 0.85 and 0.99. Ranking of foods related to their overall environmental impact did not differ significantly between methods when using the default hierarchist perspective. Differences on endpoint level were largest when using the individualist perspective. The predicted average absolute impact of the foods studied did change significantly when using the new ReCiPe, regardless of which perspective was used: a larger impact was found for climate change, freshwater eutrophication and water consumption and a lower impact for acidification and land use. The use of Dutch CFs in ReCiPe 2016 leads to significant differences in LCA results compared with the use of the global-average CFs. When looking at the average Dutch diet, ReCiPe 2016 predicted a larger impact from greenhouse gas emissions and freshwater eutrophication, and a lower impact from acidification and land use than ReCiPe 2008.
Conclusions
The update of ReCiPe leads to other LCIA results but to comparable conclusions on hotspots and ranking of food product consumption in the Netherlands. Looking at the changes per product due to the update, we recommend updating endpoint-level LCAs conducted with ReCiPe 2008, especially for products that emit large amounts of PM2.5 or consume large amounts of water within their life cycle. As new and updated methods reflect the scientific state of art better and therefore include less model uncertainty, we recommend to always use the most recent and up-to-date methodology in new LCAs.