LCA of tomato greenhouse production using spatially differentiated life cycle impact assessment indicators: an Albanian case study

The increasing attention to agricultural exports and sustainability issues is driving a surge of interest in the life cycle assessment (LCAs) of greenhouse crop production in Albania. Meanwhile, most of the reported agricultural LCAs tend to be generic without considering regionalized environmental sensitivities. In this study, ReCiPe 2016, covering 18 midpoint indicators and 3 endpoint indicators was used to generate a full-fledged cradle-to-farm gate LCA of greenhouse tomatoes in a typical Albanian farm including spatial differentiation and indicators not covered by contemporary LCAs. The most important midpoint categories per 1 ha identified from foreground-background analysis were global warming (2660.4 kg CO2-eq), stratospheric ozone depletion (0.0308 kg CFC11-eq), particulate matter formation (7.99 kg PM2.5-eq), human health and ecosystem ozone formation (8.47 and 14.95 kg NOx-eq), water consumption (2293.23 m(3)), and terrestrial acidification (42.28 kg SO2-eq). The application of spatial differentiation resulted in higher impacts with about 21% for particulate matter formation, 12% for human health ozone formation, 134% for ecosystem ozone formation, 19% for terrestrial acidification, and 13% for water consumption. The impacts primarily originated from nitrogen-based fertilizer emissions and diesel fuel with the origin of the impact from nitrous oxide (N2O), ammonia volatilization (NH3), nitrogen oxides (NOx), and non-methane volatile organic compounds (NMVOCs). Water consumption was dominated by irrigation water use. Overall, at the endpoint level, 9% and 24% less cumulative damage to human health and ecosystem quality were calculated with respect to the site-generic analysis primarily from the cause-and-effect chain of water consumption (mainly lower water stress index). This affirms the importance of regional considerations in LCA calculations to reflect the impacts accordingly (i.e., the magnitude of impacts, the most relevant midpoint categories, and their relevance on endpoint level) and increase the possibility of making correct conclusions and sub-optimizations, i.e., increase the discriminating power of LCA.