Comment Log Display

Note: The text of our comment follows below. I also append a pdf
file that contains the comments below with the original formatting
of the document because the formatting gets lost when inserted into
this message box.  

------------------------------------------------------------
Comments on “Detailed California-Modified GREET Pathway for
Denatured Corn Ethanol” (released April 21, 2008, version 1.0)
Draft—For Review

Kenneth G. Cassman1, Adam J. Liska1, and Virgil Bremer2
University of Nebraska-Lincoln, 
Department of Agronomy and Horticulture1, and Department of Animal
Science2
June 27, 2007

General

Life-cycle metrics are dependent upon numerous estimated
parameters that underpin the calculation. Appropriate references
must support all of the data used. The parameters and assumptions
used in the CA-GREET model and referenced in the Draft report lack
many necessary references and are not transparent. Although it is
difficult to tell because of the lack of transparency and adequate
citation, we believe the values for inputs and GHG performance of
corn-ethanol presented in the Draft are obsolete and are not
representative of current farming and ethanol industry practices.
Appropriate references are necessary to evaluate the assumptions
employed. For example, energy use on farm is from the 1990’s, and
a more recent value is available and should be employed. Also, the
source of the values for energy use at the ethanol plant is not
given, but we believe it is from a survey of ethanol plants taken
in 2001. Here again there more recent, and more representative
values for this parameter and they should be used. The methods
used to calculate the co-product credit is also outdated and
inaccurate.  By employing older, outdated data that so not
represent current farming practices, ethanol plant operation, or
co-product use, the proposed CA-GREET model does not accurately
represent the GHG emissions from the current corn-ethanol
industry. 

Corn Farming
1.	Energy use for farming is indicated in Btu per bushel, or unit
yield (Btu/bu) (Table 1.01). This is not an appropriate parameter
because this efficiency value changes overtime and is dependent on
grain yield and a number of known input rates. Changes in farming
practices, such as switching from conventional tillage to
no-tillage, may reduce energy inputs while having a minimal impact
on crop yield. We strongly believe that the underlying parameters
that determine the calculated Btu/bu, such as nitrogen and other
fertilizer application rates (e.g. lb N / ac), are given as
explicit input parameters. This will facilitate evaluation and
updating of the model by those interested in such activities. The
generic national averages also do not capture regional
variability, which are large.
2.	The references provided for farm input rates are from 1995-1999
(p. 18, footnotes), and cropping practices have become more
efficient since that time (Cassman et al., 2002). Changes in
practices have reduced petroleum use in corn production. This
increase in efficiency should not be estimated based on a general
estimate (e.g. +10%), but changes in cropping practices should be
calculated based on actual input rates and crop yields using the
most recent available data. For example, input rates for
fertilizer and pesticides are available for more recent years, and
energy inputs are available from 2001. A brief by Life Cycle
Associates indicates that the 2001 data reduces energy inputs by
33% compared to the estimates used by GREET.
3.	Fertilizer inputs are not generally directly proportional to
grain yield (e.g. g/bu) (Table F), and such parameters are also
not commonly used by crop producers. Fertilizer (e.g. nitrogen)
input is known on an area basis (e.g. lb/ac), it is associated
with regionally variable input rates and uptake efficiencies, and
is not accurately accounted for by the parameters employed as a
variable in calculations related to yield (e.g. g K2O/bu). Such
parameters should be given in units that are consistent with how
they are used—in this case in lb/ac.
4.	References for the energy intensity of fertilizer inputs sued
in the model are omitted in Table 2.01. The text indicates that
these intensities are relatively constant, but a report by G.
Kongshaug (Energy Consumption and Greenhouse Gas Emissions in
Fertilizer Production, 1998) documents substantial variability in
fertilizer production efficiency. Recent estimates based on
current practices, with appropriate references, are needed here.
Estimates in Table 2.02 lack appropriate references. The ethanol
yields in Table 2.03 are not referenced. 
5.	Nitrous oxide (N2O) emissions from N fertilizer are assumed to
be 2.0% of applied (Table 2.06). It lacks an appropriate
reference, and is inconsistent with current estimates. While
considering 9 parameters from the 2006 Guidelines for National
Greenhouse Gas Inventories (IPCC), the BESS model
(www.bess.unl.edu) estimates that direct and indirect N2O
emissions from fertilizer are approximately 1.8%--direct N2O
emissions from fertilizer are 1% of applied N is converted to N2O
(IPCC 2006).

Ethanol Production
1.	Appropriate references are not provided to support the values
associated with the energy use in the ethanol plant (Table 4.01).
The numbers used are likely to be obsolete and not representative
of the current ethanol industry. These numbers have a large impact
on the GHG emissions totals from corn-ethanol systems and therefore
the source of these data must be fully documented with acceptable
citations. Without citations, our best guess is that these values
come from an EPA estimate obtained from consulting engineering
firms. More recent industry surveys using data from state
regulatory agencies and other industry surveys suggest that the
values cited in the Draft are too high and that the current
ethanol industry is considerably more energy efficient. The
efficiencies from these surveys were presented in a recent memo to
CARB (March 26, 2008) from Ken Cassman and Adam Liska, and are also
used in the BESS model (www.bess.unl.edu). 

Co-product credits
1.	The co-product credits are inaccurate as designated in Table
6.05 for a dry mill biorefinery. Our group has recently
recalculated co-product credits based on Klopfenstein et al.
(2008). The method for calculating these credits is based on
current feeding practices and is described in the User’s Guide of
the BESS model (www.bess.unl.edu). One manuscript is submitted and
another is in progress to describe the GHG credit due to distillers
grains based on current feeding practices. Distillers grains plus
solubles (DGS) do not replace soybean meal in the majority of
cattle diets. The replacement materials for DGS are primarily corn
and urea, not corn, oil, and soybean meal (see point 4 below). The
displacement method used by GREET model ignores the most accurate
and current biological data (e.g. the BESS co-product crediting
system based off of extensive biological data and environmental
factors) for cattle performance and DGS inclusion level being fed
by the feedlot industry.
2.	GREET 1.8b, like the other GREET versions, discounts the total
co-product credit by 15% since it was originally believed that
there would be an oversupply of DGS and therefore the beef
industry would have to grow to use up all the DGS. The thought was
that this "new beef industry growth" caused by DGS could not be
credited. The number they calculated was that a 15% growth was
needed to use all the DGS (Table 6.02). This assumption is
incorrect because the beef industry has not grown with the DGS
boom. DGS is being used to replace corn that has been diverted to
the ethanol industry from the cattle feeding industry. This means
that the 15% discount should be eliminated from the GREET model
calculations.
3.	Some of the GREET 1.8b calculations for soybean transportation
are based on the wrong weight of soybeans per bushel. The cells in
columns I and J of sheet BD use a 56 lb bushel weight of soybeans.
This number should be 60 lb per bushel. The 56 lb/bu number is
correct for corn but not for soybeans. This number is an important
part of converting energy values per ton of soybeans to per bushel
of soybeans to be compatible with the rest of the model. These
calculations are not used directly for co-product calculations,
but appear to have been the basis for some of the co-product
calculation inputs. 
4.	The co-product feeding substitution scheme provided by the
Draft is underdeveloped and unrepresentative of current feeding
practices. The references for Table 6.02 are for brief, non-peer
reviewed, largely undocumented conference presentations
(http://www.mncpoe.org/Previous_events/mar13_energy%20forum/Cellulosic%20Ethanol-Tiffany.Mar.13.07.print.pdf,
http://www.ddgs.umn.edu/ppt-swine/2005-Shurson-%20High%20quality%20corn%20ddgs.pdf).
The EPA document (ref. 11, p.65) does not appear to contain any
text on co-product substitution rates (the Draft suggests that 1
ton of DGS substitutes 0.5 ton of corn and 0.5 ton of soybean meal
in cattle diets—this could not be found in the document:
http://www.epa.gov/EPA-AIR/2007/May/Day-01/a7140a.htm). This
substitution assumption is also not supported by a recent USDA
survey of use of DGS and related animal feeding studies, as
described below.
The Renewable Fuels Association calculated that 82% of
biorefineries were dry-mills in 2006 (RFA 2006; this percentage
has increased due to recent industry expansion). Dry-mills produce
distillers grains co-products instead of corn gluten feed from
wet-mills. The National Agricultural Statistics Service (NASS) has
released a 2006 survey of beef, dairy, and swine operations on
ethanol co-product use for livestock feed (USDA-NASS 2007). The
survey was conducted in the Corn Belt for a region that contains
50%, 33%, and 70% of the United States 2006 beef, dairy, and pork
production, respectively (USDA-NASS, 2008).  In 2006, this area
represented 3.2 million head of dairy cattle, 11.3 million head of
cattle in 1,000+ head feedlots, and 64.1 million pigs, and a large
portion of these animals are fed co-product. Moreover, the larger
scale, more innovative producers are the ones adopting co-product
feeding (USDA-NASS, 2007; Waterbury et al., 2009). An example of
co-product use comes from the Nebraska beef industry. A Nebraska
state survey found that 59% of feedlot operations were feeding
co-products in 2007 (Waterbury et al., 2009).  However, on an
animal basis, 91% of cattle on feed were fed co-products. A Texas,
Midwest, and Western states feedlot nutritionist survey conducted
by Vasconcelos and Galyean (2007) agrees with the Nebraska study
by showing 83% of the feedlots used co-products. The respondents
in both the consultant study and Nebraska study indicated that
distillers grains was the most common co-product used.  The
nutritionist survey indicated 69% of the 29 nutritionists
(consulting for about 69% of cattle on feed in the United States)
were feeding distillers grains as the primary co-product in the
diet.      
Feeding studies have demonstrated that up to 50% of diet dry
matter can be replaced with DGS in feedlot diets and improve
cattle performance (Klopfenstein et al., 2008).  NASS survey data
suggests that Corn Belt feedlots feeding DGS have average dietary
inclusion of distillers grains at 22% to 31% of the diet (as-is,
wet basis).  Waterbury et al. (2009) has shown that feedlots are
feeding 37% of the diet (as-is) as co-product in Nebraska.
Vasconcelos and Galyean (2007) suggest the average co-product
inclusion rate on a dry matter basis is 20% with a range of 5 to
50% of diet dry matter.
Research has shown that 20% of dairy diet dry matter can be
provided as DGS without hurting performance (Anderson et al.,
2006). NASS survey data suggests that the average inclusion of DGS
in dairy diets is 10 to 22 percent of the diet (as-is). When the
water in the as-is weight is discounted, this amount is about 10%
of diet dry matter. The dairy industry has been using DGS as a
protein supplement to replace corn and soybean meal in the diet
(Anderson et al., 2006). As the inclusion level increases, the
corn energy will be replaced with distillers grains for milk
production energy.
The swine industry can efficiently use up to 20% of diet dry
matter as dry DGS without hurting pig performance (Stein, 2007). 
NASS data suggest that few swine operations have been feeding DGS,
and the average as-is inclusion is about 10 to 11% of the diet for
those operations that do feed DGS. 
Cumulatively these data suggest that the beef and dairy industries
have been the major consumers of DGS produced by dry mills. The
beef industry feeds greater inclusions of DGS to more cattle than
the dairy industry, even accounting for two steer finishing
periods per dairy cow year. However, dairy cattle eat roughly two
times the amount of dry matter each day that feedlot cattle eat.
This suggests that the dairy industry may be utilizing about the
same amount of distillers grains as the feedlot industry. The
feedlot industry may have more potential for future increased use
of 
DGS than the dairy industry (Klopfenstein et al., 2008), because
the dairy industry does not have as much potential without
decreasing animal performance. Although the swine industry has the
potential to utilize DGS, the industry has been feeding low
inclusion levels and has not been a major consumer of the
co-product.         
These findings indicate that the beef and dairy industries are the
primary systems to model co-product use.  While the initial use of
DGS was for protein replacement in both beef and dairy diets when
the amount of corn used for ethanol was small, with large amounts
of corn used for ethanol as is now the case, DGS are used
primarily as an energy source in cattle and dairy diets
(Klopfenstein et al., 2008; Anderson et al., 2006). Therefore, the
DGS can not be completely credited as a protein source as they are
in the GREET model. Distillers grains use has been studied more
extensively in feedlot cattle than in dairy production
(Klopfenstein et al., 2008).  Therefore, we can accurately
evaluate the feedlot industry, but the dairy industry needs
further analysis.
Historical developments in the cattle feeding industry show that
part of the DGS co-product credit is the replacement of urea
(nitrogen) in feedlot diets and does not include the replacement
of soybean meal. By the mid 1960’s the ruminant feeding industry
recognized that urea was as effective as soybean meal for feedlot
cattle protein supplements (Perry et al., 1967; White et al.,
1975). Urea supplied dietary protein (nitrogen) less expensively
than did plant protein supplements such as soybean meal and
therefore became the main nitrogen supplement for feedlot cattle,
but co-products can replace urea and a 2007 subsequent survey
found wide spread use of ethanol co-products as protein sources
(Vasconcelos and Galyean, 2007). Therefore, the BESS model assumes
that co-products are used to replace corn and urea in cattle diets
and are given a GHG credit for the emissions saved by making this
replacement. Details are provided in the BESS model User’s Guide
(www.bess.unl.edu). 

Citations
Anderson J. L., D. J. Schingoethe, K. F. Kalscheur, and A. R.
Hippen. 2006. Evaluation of Dried and Wet Distillers Grains
Included at Two Concentrations in the Diets of Lactating Dairy
Cows. J Dairy Sci. 89: 3133-3142.
Cassman, K.G., Dobermann, A.D., and Walters, D.T. 2002. 
Agroecosystems, N-Use Efficiency, and N Management.  AMBIO
31:132-140.
Culbertson C. C., John M. Evvard, W. E. Hammond, and Q. W.
Wallace. 1924. Protein Supplements with Different Roughages for
Fattening Cattle. J Anim Sci 1924 1924: 13-24.
Galyean, M. L., and J. F. Gleghorn. 2001. Summary of the 2000
Texas Tech University Consulting Nutritionist Survey. Texas Tech
University, Dept. of Anim. and Food Sci., Burnett Center Internet
Progress Report No. 12.
http://www.asft.ttu.edu/burnett_center/progress_reports/bc12.pdf. 
Accessed June 26, 2008.
Klopfenstein T.J., G. E. Erickson, and V. R. Bremer BOARD-INVITED
REVIEW: Use of distillers by-products in the beef cattle feeding
industry, J Anim Sci 2008 86: 1223-1231.
Perry T. W, W. M. Beeson, and M. T. Mohler. 1967. A Comparison of
High-Urea Supplements with Natural Protein Supplements for Growing
and Fattening Beef Cattle. J Anim Sci 1967 26: 1434-1437.
RFA (Renewable Fuels Association). 2008. Changing the Climate:
Ethanol Industry Outlook 2008. Washington, DC.
Stein, HH. 2007. Distillers dried grains with solubles (DDGS) in
diets fed to swine.  Department of Animal Sciences University of
Illinois at Urbana-Champaign HHS-SwineFocus-001.
USDA-NASS (National Agricultural Statistics Service). 2007.
Ethanol Co-Products Used for Livestock Feed. Washington, DC.
USDA-NASS (National Agricultural Statistics Service). 2008.
www.nass.usda.gov. Accessed June 26, 2008.
Vasconcelos J. T. and M. L. Gallean. 2007. Nutritional
recommendations of feedlot consulting nutritionists: The 2007
Texas Tech University survey. J Anim Sci. 85: 2772-2781.
Waterbury, JA, DR Mark, RK Perrin, SM Thoms, GE Erickson, and TK
Klopfenstein. 2009. Results of the Ethanol Co-Product Survey: An
Economic Overview of Ethanol Co-Product Utilization in Nebraska. 
Nebraska Beef Cattle Report 2009. (Submitted).
White T. W., W. L. Reynolds, and F. G. Hembry. 1975. Comparison of
Supplements Containing Soybean Meal and Urea Fed with Whole or
Ground Shelled Corn to Beef Cattle. J Anim Sci. 40: 1-5