The Clemson University/SRNL Bioethanol collaborative project developed a comprehensive process to convert switchgrass (Panicum virgatum L.) and sweet sorghum (Sorghum bicolor L. including Dale and M81E) to fuel grade ethanol. The project was conducted by Clemson University and the Savannah River National Laboratory (SRNL) originally funded in October of 2008 with continued research through September 2012. In 2010 the research team developed a complementary sweet sorghum process to convert the syrup and the bagasse to ethanol. The commonality of the two processes supports separate or combined feedstock facilities designed to produce ethanol in agricultural regions across the Southeastern US and represents potential processes for similar soils around the world including southern

Africa and Australia. The commercial production cost for switchgrass to ethanol is projected at $2.32 per gallon with sweet sorghum at $1.79 per gallon without consideration of potential co-products. In 2011, the team concentrated on the expansion of the application of the switchgrass/sweet sorghum technology to include sugarcane bagasse and coastal loblolly pine. The pretreatment and hydrolysis concepts utilized in switchgrass and sweet sorghum proved directly transferable to bagasse.

The process, with pilot scale confirmation, represents a viable source of green carbon for ethanol production. The switchgrass grows on marginal land, requires little inputs, is drought resistant and can be bailed twice annually using hay bailing equipment common to most farms. Our field research suggests that a single acre can yield approximately 80 gallons of ethanol per ton. Our technology review suggests that the process may be optimized, assuming high levels of xylose conversion to yield 120 gallons per acre. Further, the switchgrass can be stored in the field for several months eliminating most plant related storage requirements.

Figure 7 represents the developed process flow diagram for conversion of switchgrass (Panicum virgatum L.) to ethanol. The process entails milling switchgrass to 1 mm average particle size followed by pretreatment with ammonium hydroxide (SAA) concentrated to 8% by weight and recovered for further reuse. Lignin is separated from the bottoms of the steam stripper section and recovered using sulfuric acid precipitation (pH 2-3) followed by a leaf filter and drying for potential use of high-value products (biomaterials) or for energy use within the biorefinery setting. Enzyme hydrolysis is conducted on the resulting pretreated fibers after a series of water wash steps by using cocktails of cellulases ranging from novel Dyadic and Genecor products. The resulting sugars are then separated with centrifugal decanter and pumped to the fermentation vessels operated in fed-batch mode with the first vessel fermenting glucose fractions using high-ethanol tolerant strains of Saccharomyces and the second vessel fermenting xylose fractions with novel bacteria. Resulting ethanol is then distilled to 92-95% in a series of distillation columns before being dried using conventional swing-bed zeolite absorption.

The following Table 7 presents a comparison of production costs by feedstock comparing the 2007 projections of NREL stover costs. In this analysis the report identifies the cost based on assumed technical capabilities of the industry. During the switchgrass process development capital estimates and manufacturing models were developed. ASPEN II models generated by SRNL provided the process inputs, outputs, energy and mass balances. The conceptual numbers illustrate that higher ethanol yield was attainable through the processing of all available sugars from the initial cane squeezing and the residual bagasse which is suggested to be 30% or greater than switchgrass. The following numbers represent production targets for pilot phase scale up.