Effect of geographical location and stochastic weather variation on life cycle assessment of biodiesel production from camelina in the northwestern USA

PURPOSE: The effect of regional factors on life cycle assessment (LCA) of camelina seed production and camelina methyl ester production was assessed in this study. While general conclusions from LCA studies point to lower environmental impacts of biofuels, it has been shown in many studies that the environmental impacts are dependent on location, production practices, and even local weather variations. METHODS: A cradle-to-farm gate and well-to-pump approaches were used to conduct the LCA. To demonstrate the impact of agro-climatic and management factors (weather condition, soil characteristics, and management practices) on the overall emissions for four different regions including Corvallis, OR, Pendleton, OR, Pullman, WA, and Sheridan, WY, field emissions were simulated using the DeNitrification-DeComposition (DNDC) model. openLCA v.1.4.2 software was used to quantify the environmental impacts of camelina seed and camelina methyl ester production. RESULTS AND DISCUSSION: The results showed that greenhouse gas (GHG) emissions during camelina production in different regions vary between 49.39 and 472.51 kg CO₂-eq./ha due to differences in agro-climatic and weather variations. The GHG emissions for 1 kg of camelina produced in Corvallis, Pendleton, Pullman, and Sheridan were 0.76 ± 11, 0.55 ± 10, 0.47 ± 18, and 1.26 ± 6 % kg CO₂-eq., respectively. The GHG emissions for 1000 MJ of camelina biodiesel using camelina produced in Corvallis, Pendleton, Pullman, and Sheridan were 53.60 ± 5, 48.87 ± 5, 44.33 ± 7, and 78.88 ± 4 % kg CO₂-eq., respectively. Other impact categories such as acidification and ecotoxicity for 1000 MJ of camelina biodiesel varied across the regions by 43 and 103 %, respectively. CONCLUSIONS: It can be concluded that process-based crop models such as DNDC in conjunction with Monte Carlo analysis are helpful tools to quantitatively estimate the influence of regional factors on field emissions which consequently can provide information about the expected variability in LCA results.