MEMORANDUM
TO: 		Air and Radiation Docket EPA-HQ-OAR-2014-0537
FROM: 	EPA Office of Transportation and Air Quality
DATE:		December 17, 2014
SUBJECT: 	Agricultural Inputs and Cellulosic Content of Biomass Sorghum

This memorandum provides information on sources of data used for EPA's analysis of the greenhouse gas emissions associated with the cultivation of biomass sorghum.  It also provides information on the data used to calculate the adjusted cellulosic content of biomass sorghum.  

A. AGRICULTURAL INPUTS AND EMISSIONS

To determine the greenhouse gas emissions associated with growing biomass sorghum, EPA used data from the petition, literature, and USDA.  The data and sources are summarized here, and details of how the data were used to calculate greenhouse gas emissions are provided in a spreadsheet available on the docket.  

   1. Yield

To estimate an average yield in 2022, we started with an assumed biomass sorghum yield of 11 dry tons/acre in 2012.  This value is based on an average of data from several sources, which showed results of field trials in Indiana, Texas, Kentucky, Mississippi, North Carolina, Kansas, Iowa, Michigan, and Illinois.[,][,][,][,][,][,][,]  Trials took place between 1979 and 2010, and used a range of growing conditions.  The average yield based on all of these studies was 10.77 dry tons/acre, with a range of 5.84 to 26.32 dry tons/acre.  Values in the higher end of this range came from a single study.[5]  If this study is excluded, the range in yield is 5.84-15.66 dry tons/acre.  

We assumed that the yield for biomass sorghum would increase by 2% per year, based on similar yield improvements that were projected for switchgrass.  Based on this yield improvement, we calculated a biomass sorghum yield of 13 dry tons/acre in 2022.  This value was used for subsequent emissions calculations.

   2. Fertilizers

We used fertilizer input values of 60 lbs nitrogen/acre, 15 lbs phosphorus/acre, and 7 lbs potassium/acre.  These values are based on data provided by the petitioner, which cited Forest and Agricultural Sector Optimization Model (FASOM) documentation as a source.  Using a yield of 13 dry tons/acre, the fertilizer inputs per dry ton are: 4.6 lbs nitrogen/dry ton, 1.2 lbs phosphorus/dry ton, and 0.5 lbs potassium/dry ton.  

   3. Other inputs

The petition provided two diesel input values: 9.2 gal/acre, based on FASOM documentation, and 7.35 gal/acre based on GREET 2011 and two reports.[,]  We used the more conservative value of 9.2 gal/acre, which is 0.7 gal/dry ton feedstock.  For herbicide inputs, we used USDA's estimate for energy cane, which is 5.5 lbs/acre, or 0.4 lbs/dry ton feedstock.  For pesticides, the petition provided the greenhouse gas emissions based on FASOM data, rather than the quantity of pesticide used.  We used this value (0.0006 tons CO2e/acre) to calculate the quantity of pesticide used, based on our emission factor for pesticides.  This calculation resulted in a pesticide input of 0.044 lbs/acre, or 0.003 lbs/dry ton feedstock.  

   4. N2O emissions

N2O is emitted from synthetic nitrogen fertilizer, as well as from the decay of crop residues.  The Intergovernmental Panel on Climate Change (IPCC) provides equations for direct emissions from these sources, and indirect emissions from the volatilization of synthetic fertilizer, and leaching/runoff of all nitrogen sources.  The IPCC also provides emission factors for these N2O sources.

To calculate the amount of crop residues, we used equations and factors in the IPCC.  With these equations, the crop yield is used to calculate the amount of above-ground residue, and the ratio of below-ground residue to above-ground biomass (including harvested material) is used to calculate the amount of below-ground residue.  The IPCC provides a table of factors to calculate crops residues and the nitrogen content of crop residues for a variety of crops.  Because factors for biomass sorghum are not included in this table, we used the factors provided for "Perennial grasses."  We chose this value because the harvesting of biomass sorghum is likely to be similar to the harvesting of perennial grasses  -  most of the plant will be harvested, with little residue remaining, in contrast to crops like corn or grain sorghum where only the grain is harvested.  However, because perennial grasses have a more developed root system than annual crops, it is likely that our calculations overestimated the below-ground biomass, and therefore overestimated the N2O emissions from below-ground biomass.  We calculated an above-ground residue of 3.9 tons/acre and a below-ground residue of 13.5 tons/acre for a yield of 13 dry tons/acre.  Total direct and indirect N2O emissions from nitrogen fertilizer and crop residues are 105 kgCO2e/dry ton.  Further details of these calculations can be found in the associated spreadsheet on the docket.  

   5. Total emissions from chemical and energy inputs and crop residues

We used the emission factors documented in the March 2010 rule to calculate the emissions from fertilizers, herbicides, pesticides, and diesel.  These results are shown in the spreadsheet posted on the docket.  Total emissions from agricultural inputs and crop residues (including N2O emissions) are 128 kgCO2e/dry ton feedstock.

B. CELLULOSIC CONTENT

For biomass sorghum-derived biofuels to qualify as cellulosic biofuel under the RFS program, the fuel must achieve a 60% lifecycle GHG reduction as compared to the 2005 baseline fuels, and must also be derived from cellulose, hemicellulose and lignin.  In the rule published on July 18, 2014 (the "July 2014 rule"), EPA determined that fuel generated from feedstocks with an average adjusted cellulosic content of 75% or more is eligible to generate cellulosic biofuel RINs for the entire fuel volume.  In that rulemaking, EPA also explained that we would apply the 75% threshold to feedstocks that we evaluated in the future, and finalized a definition of energy cane, which can have a wide range of cellulosic contents.  Consistent with that rulemaking, we have evaluated the cellulosic content of biomass sorghum, based on two scientific studies and industry data.  Data and calculations can be found in the spreadsheet posted to the docket.  

One study presented compositional data for 22 varieties of forage sorghum.  The composition was measured using National Renewable Energy Laboratory (NREL) Laboratory Analytical Procedures.  The samples included photoperiod sensitive (PS) varieties as well as non-PS varieties.  Data in the paper showed that the PS varieties had higher yields than the other varieties, and that these higher yields were similar to the yields of other bioenergy crops.  For our calculations, we separated the varieties into "high-yield" (the PS varieties), "sorghum/sudangrass" (which includes two of the PS varieties), and "forage."  Although the paper did not specifically categorize any of the samples as biomass sorghum, we assumed that the high-yield, photoperiod sensitive varieties are most similar to biomass sorghum, based on information from the petitioners.

This paper showed data for ash, protein, soluble sucrose, soluble starch, total solubles, lignin, glucan, xylan, galactan, arabinan, and total structurals.  In addition, we received supplementary data from the authors for water extractives, ethanol extractives, and acetyl.  These data (published and supplementary) are included in the spreadsheet on the docket.  Using methods detailed in a separate memorandum, we calculated the adjusted cellulosic content of the samples (Table 1).  We found that "high-yield" varieties had an average adjusted cellulosic content of 75.4% (with a range of 68.9-82.8%), sorghum/sudangrass had an average adjusted cellulosic content of 70.0% (with a range of 61.2-75.8%), and forage sorghum had an average adjusted cellulosic content of 60.5% (with a range of 50.5-84.4%).

Another study presented data from 152 sorghum samples.  51 of these samples were categorized as "biomass sorghum."  These samples were "photoperiod sensitive and produce[d] little to no grain."  This study also presented data from 41 forage sorghum samples, 54 sweet sorghum samples, and 6 sorghum/sudangrass samples.  The composition of the samples was analyzed using NREL Laboratory Analytical Procedures.  This paper presented data on ash, protein, sucrose, lignin, xylans, glucans, and starch.  Using these data, we found that the biomass sorghum samples had an average adjusted cellulosic content of 63.2%, the forage sorghum samples had an adjusted cellulosic content of 70.1%, the sweet sorghum samples had an adjusted cellulosic content of 61.8%, and the sorghum/sudangrass samples had an adjusted cellulosic content of 72.5% (Table 1).

We also received data from NexSteppe, a developer of sorghum hybrids that have been recently released or are under development, and for which there is not yet published data available.  NexSteppe provided data from seven samples grown in four locations.  NREL Laboratory Analytical Procedures were used to measure the composition of the samples.  NexSteppe provided data on the free sugar content, starch content, total cellulosic content, and adjusted cellulosic content for each sample.  The average adjusted cellulosic content was 83.7%, with a range of 77.4-88.6% (Table 1).   

Table 1 -- Chemical composition of different types of sorghum samples, as determined by two research studies and from industry data.  The adjusted cellulosic composition was calculated by adjusting the reported content of cellulose, hemicellulose and lignin for the ash content and for the total yields.
 
 
                           Chemical Composition (%)
                                       
 
                                                                         Source
                            Dahlberg et al.  (2011)
                           Stefaniak et al.  (2012)
                                   NexSteppe
 
                                                                Sorghum variety
                                  High-Yield
                                Sudan/ Sorghum
                                    Forage
                                    Biomass
                                Sudan/ Sorghum
                                    Forage
                                     Sweet
                                    Biomass
 
                               Number of samples
                                       5
                                       4
                                      15
                                      51
                                       6
                                      41
                                      54
                                       7
Sucrose (sugar)
                                    Average
                                      2.9
                                      2.7
                                      1.0
                                      9.0
                                      2.4
                                      1.1
                                      9.8
                                      4.5

                                     Range
                                    1.6-4.6
                                    0.4-3.5
                                    0.2-1.7
                                    0.3-19
                                    0.4-4.6
                                    0.2-3.0
                                   0.2-23.9
                                    1.2-8.5

                                    Average
                                      0.8
                                      5.6
                                     18.1
                                      5.6
                                      1.1
                                      1.8
                                      7.3
                                      3.4
Starch
                                     Range
                                      0-4
                                     0-15
                                    0-25.2
                                    0-12.0
                                     0-4.0
                                     0-8.9
                                    0-16.6
                                    0.3-8.1
Cellulosic Components
                                    Average
                                     66.7
                                     62.0
                                     54.9
                                     59.2
                                     63.9
                                     66.4
                                     58.3
                                     77.5

                                     Range
                                   61.3-72.3
                                   53.8-67.5
                                   46.8-73.6
                                       -
                                       -
                                       -
                                       -
                                   75.3-80.5
Adjusted Cellulosic Composition
                                    Average
                                     75.4
                                     70.0
                                     60.5
                                     63.2
                                     72.5
                                     70.1
                                     61.8
                                     83.7

                                     Range
                                   68.9-82.8
                                   61.2-75.8
                                   50.5-84.4
                                       -
                                       -
                                       -
                                       -
                                   77.4-88.6

In the July 2014 rule, EPA considered the cellulosic content of energy cane.  Like biomass sorghum, cane can be bred for a wide range of cellulosic and sugar contents.  In that rule, EPA defined "energy cane" as cultivars containing at least 75% adjusted cellulosic content.  EPA also indicated that in the future, feedstocks that could be bred for a wide range of uses and fiber content would have registration requirements similar to energy cane, in order to demonstrate that the adjusted cellulosic content of varieties used is at least 75%.  Therefore, for the purposes of the cellulosic content issue, EPA intends to treat biomass sorghum similar to energy cane.  For purposes of the Notice, we consider biomass sorghum to include varieties containing at least 75% adjusted cellulosic content.  

