Life cycle inventory of Miscanthus production on a commercial farm in the US

Abstract

There has been considerable interest in use of Miscanthus (Miscanthus x giganteus) as a feedstock for bioenergy production due to its potential to reduce greenhouse gas emissions associated with cellulosic feedstock production and more recently for alternative uses as a biomass crop. To date, data on Miscanthus production in the US has been based on small scale research plots due to the lack of commercial scale production fields. Research plot yields are often much higher than commercial fields for a variety of reasons including reduced spatial variability and location on better quality farmland. The objectives of this study were to quantify the inputs for production of Miscanthus at the commercial farm scale, evaluating methods to characterize fuel use for establishment and management of Miscanthus production and using satellite data to characterize spatial yield variation of production fields. We logged energy use on agricultural machinery from Miscanthus production planted on more than 1000 ha of land and modeled N2O emissions and changes in soil carbon using DayCent. Although fuel use was higher for land preparation in fields with perennial vegetation, fuel to harvest Miscanthus dominated greenhouse gas (GHG) emissions (>90%) from agriculture machinery for crop management. The N2O emissions and changes in soil carbon were the largest source and sink of GHG emissions associated with Miscanthus production, respectively. Although ~ 50% of the established lands had Miscanthus yields < 5 Mg/ha, yields needed to be > 5 Mg/ha for ΔSOC to be positive. Given the large impact of yield on ΔSOC, net GHG for Miscanthus production with yields of 5 to 25 Mg/ha ranged ~130 to -260 kg CO2e/Mg biomass. Use of both energy use for Miscanthus harvest and satellite imagery were good methods to characterize spatial variability of commercial production fields. This demonstrates the potential to use this within field yield data to better understand factors driving subfield yield variability and use of satellite data to quantify early yield predictions.