Evaluating Agricultural Water Use with Crop Life Cycle Assessments
United Soybean Board
Efforts to evaluate resource use by major row crops have advanced significantly in recent years thanks to the development of more robust and sophisticated life cycle assessments (LCA). However, life cycle modelers frequently wrestle with challenges that are unique to production agriculture. Modelers must consider the impacts of co-products, additional utilization of natural resources from non-agricultural elements within an ecosystem, and the ongoing challenge of obtaining high quality data. These challenges are often addressed by trying out different system boundaries and allocation models, but they can have varying effects on the results of the LCA. The United Soybean Board (USB) encountered several of these challenges when it initiated a life cycle assessment of soybean production in the U.S. three years ago. The now-complete LCA evaluated the consumptive water use for soybean production in the U.S., with results that were similar to other recent studies (The Keystone Center, Environmental Resource Indicators for Measuring Outcomes of On-Farm Agricultural Production in the United States, First Report, January 2009), showing that soybean water use efficiency has increased by 20 percent since 1987. The LCA also evaluated the impact of soybean production on more than 50 water effluents, and consolidated several of these values into a eutrophication potential for soybeans, measured in kg Nitrogen equivalent, per 1000kg soybeans. To address several of the modeling challenges mentioned earlier, the USB LCA incorporated ISO 14044 mass criterion in setting system boundaries (cut-off goal of 99percent material inputs) and used mass allocation to account for co-products, with economic allocation being used to assess the sensitivity of the modeling decision. When the results were run with the economic allocation, for the most part, the overall outcomes did not change too dramatically between the economic allocation and the mass allocation results. However, the actual ratio differential was almost 50 percent in some cases. This provides further evidence of the impact that modeling decisions have on the results of LCAs. While the results of the soybean LCA are certainly of interest, this presentation will primarily focus on the implications of the soybean modeling decisions and the impact they had on the final LCA results. This broader discussion has much more significant relevance to the evaluation of water use and effluents in other crops, and how those evaluations could be affected by other modeling decisions. Through a review of the results of the USB study, we will discuss key learnings and best practices that could help guide similar water use evaluations in other crops.