MEMORANDUM

The 2007 Energy Independence and Security Act significantly increases
the amount of renewable fuels required to be sold in the U.S. under
EPA’s Renewable Fuel Standard (RFS). This mandated increase in biofuel
consumption could have significant lifecycle greenhouse gas (GHG)
impacts due to global shifts in land use for the production of biofuel
crops due to changes in demand for energy, fertilizer and pesticides.
There is significant data available on domestic energy, fertilizer, and
pesticide used in agriculture. The work described here focuses on
compiling information on international (non-U.S.) agricultural inputs.  

This memorandum summarizes the methodology used to estimate GHG emission
factors and changes in GHG emissions related to international
agricultural energy use, fertilizer and pesticide consumption, and rice
cultivation due to potential changes in land use to meet increased
demand for biofuels. Data were obtained and analyzed from various
agricultural and energy datasets. Intergovernmental Panel on Climate
Change (IPCC) methodologies were used to estimate GHG emissions. 

Specifically, ICF developed the following:

Fertilizer and Pesticide Consumption Projections

N2O Emissions from Fertilizer Consumption and Crop Residues

GHG Emission Rates for Agricultural Energy Use

CH4 Emission Factors for Rice Cultivation

We provide a detailed explanation of the methodologies and data sources
used below.

Fertilizer and Pesticide Consumption Projections

Fertilizer Consumption Projections

Historical fertilizer application rates (kilograms of fertilizer applied
per hectare) and consumption (tonnes), as well as agricultural area,
were primarily obtained from the Food and Agriculture Organization’s
(FAO) Fertistat Dataset. Fertistat is a publicly available,
international fertilizer dataset containing consumption data by crop and
country. Fertistat data are available for nitrogen (N), phosphorus
(P2O5), and potassium (K2O) fertilizers, and are based on surveyed
observations for a single year or period between planting and
harvesting.  Survey data were collected between 1988 to 2004, with the
year 2000 being the most frequent year of survey data collection.
Application rates are calculated by Fertistat as total fertilizer
consumption per country divided by agricultural area fertilized.
Application rates for the baseline 2022 scenario are assumed to be equal
to rates reported by Fertistat. Tables 1 through 3 in the Appendix
present fertilizer application rates by country and crop. 

Fertistat did not report data for several crops of interest for certain
countries, including Russia wheat, China wheat, and India soybean.
Fertilizer application rates for wheat cultivation in Russia and China
were obtained from Harris, 1998. Application rates for soybean
production in India were obtained from the Fertilizer Association of
India. 

To determine the difference in fertilizer consumption between an
increased biofuels demand scenario and the 2022 baseline scenario,
application rates by country and crop were multiplied by projected
acreage changes for crop production from the Food and Agricultural
Policy Research Institute (FAPRI) agricultural models. Change in
fertilizer consumption was calculated for 33 individual countries and
for 8 regions (as shown in Table 1) to match the crop production change
data from the FAPRI model results. As FAPRI region definitions were
largely unavailable by crop, FAO regional definitions were used., 

Table   SEQ Table \* ARABIC  1 : Country and Region Definitions

Pesticide Consumption Projections

Pesticide consumption projections are calculated using the same
methodology as fertilizer projections. Historical data were taken from
FAO’s FAOSTAT database for pesticide consumption, including fungicides
and bactericides, herbicides, and insecticides. Data were available by
country, but not by crop. Data refers to the quantity of pesticides used
in or sold to the agricultural sector for crops and seeds. Rates of
pesticide application were determined by dividing FAOSTAT’s pesticide
data by “agricultural area,” a variable found in FAO’s
ResourceStat - Land dataset (see Appendix, Table 4). Agricultural area
is defined as arable land (land under temporary crops), land cultivated
with permanent crops (e.g. coffee), and permanent pastures (land used
for five or more years for herbaceous forage crops). Change in pesticide
consumption by country was calculated by multiplying pesticide
application rates by the change in crop production acreage due to
increased U.S. demand for biofuels. 

To ensure that pesticide application rates were representative of a
typical year, an average of pesticide consumption was calculated from
1995 through 2003. If data were not available during this period, data
were averaged from 1990 through 1995. 

N2O Emissions from Fertilizer Consumption and Crop Residues 

Change in fertilizer consumption and associated nitrous oxide (N2O)
emissions due to increased U.S. biofuel demand are projected based on
crop production acreage for priority crops and countries. GHG emissions
are calculated based on nitrogen (N) inputs from synthetic N fertilizer
consumption and crop residues, both of which cause direct and indirect
N2O emissions from agricultural soils. Emission estimates are based on
the IPCC 2006 default emissions factors and emissions equations for Tier
1 methodology. 

Projections of Changes in GHG Emissions from Fertilizer Consumption

Changes in fertilizer consumption cause changes in the amount of N added
to soils, which change the amount of N2O eventually emitted to the
atmosphere.  For this analysis, we estimate changes in N2O emissions due
to changes in synthetic fertilizer consumption and changes in crop
residue application for certain crops.  As Fertistat reports only
mineral (or synthetic) fertilizer consumption data, we did not estimate
changes in GHG emissions from organic fertilizer consumption. Emissions
from organic fertilizer were handled separately by EPA through analysis
of manure management changes from livestock operations.  Calculations
are based on Tier 1 methodologies for managed soils from the IPCC 2006
Guidelines. Tier 1 methodologies do not consider different land cover,
soil type, climatic conditions, or management practices, and also do not
consider any lag time for direct emissions from crop residues. 

The pathways of N in the soil are complex, but can be summarized as
follows: N2O emissions from soils occur either directly or indirectly. 
Direct emissions occur when N is applied to soil (from fertilizer, crop
residues, or other sources), and eventually N2O is emitted through the
processes of nitrification and denitrification.  Indirect emissions
occur in two ways: (1) N applied to soils can be volatilized in a non-
N2O form, and redeposited in another location, where N2O emissions will
occur and (2) applied N can be leached by water in a non- N2O form, and
the N transported in the runoff will emit N2O in a different location
from that where the N was applied. This analysis looks at direct and
indirect emissions from synthetic N fertilizer application and crop
residues. GHG emissions from fertilizer consumption are determined based
on the annual amount of synthetic fertilizer applied. Crop residue
emissions are based on the N content of above- and below-ground crop
residues (including N-fixing crops) that are returned to soils. Indirect
crop residue emissions only include leaching/runoff emissions, as crop
residue N is not thought to volatilize. In summary, GHG emission
pathways include: 

Direct emissions from N additions to soils from synthetic fertilizers

Indirect emissions from N additions to soils from synthetic fertilizers
from volatilization and leaching/runoff.

Direct emissions from N in crop residues

Indirect emissions from N in crop residues due to leaching and runoff.

Emissions were estimated using IPCC default emission factors and default
crop residue parameters (see Appendix, Tables 5 and 6). Emissions were
calculated using the following Tier 1 equations: 

Direct Emissions: 

Direct N2O emissions from synthetic fertilizers:  

Emissions  = FSN  × EF1 × 44/28

Where: 

FSN = the annual amount of synthetic fertilizer N applied to soils (kg
N)

EF1 = emission factor,(equal to 0.1 kg N2O-N/kg N input)

44/28 = conversion of N2O -N to N2O

Direct N2O emissions from crop residues:  

Emissions = FCR × EF1 × 44/28

Where

FCR = the annual amount of N in crop residues and forage/pasture
renewal, kg N2O-N. 

EF1 = emission factor,(equal to 0.1 kg N2O-N/kg N input)

44/28 = conversion of N2O -N to N2O

= ∑ (Yield FreshT × DRYT × ST + IT) × AreaT × (Nag(T) + Rbg(T) ×
Nbg(T))

Where: 

T = crop or forage type

	Yield Fresh = fresh weight yield of crop (kg fresh weight/ha)

	DRY = dry matter fraction of harvested crop (kg dry matter/kg fresh
weight)

	S = Slope for above-ground residue dry matter 

	I = Intercept for above-ground residue dry matter

	Area = total annual area harvested (ha)

	Nag = N content of above-ground residues (kg N/kg dry matter)

Rbg = ratio of below-ground residues to harvested yield 

Nbg = N content of below-ground residues (kg N/kg dry matter)

Table 6 in the Appendix presents crop residue factors by crop type. If
default factors were not available for a particular crop, then proxies
were used based on “major crop type” categories. Rice and sorghum
estimates are based on default factors for grains; and peanut and
sugarbeet estimates are based on root crops default factors. IPCC
default factors were not available for cotton, palm oil, rapeseed,
sugarcane and sunflower. As a result, direct and indirect emissions from
crop residues for these crops are not included in total N2O emissions
estimates. 

To determine the fresh weight yield for priority crops and countries,
the change in crop production provided from the FAPRI model results in
the year 2022 was divided by the change in crop production acreage. Area
in the FCR equation refers to the change in crop production acreage.
Changes in crop acreage were obtained from the FAPRI model forecasts for
key crops, countries, and regions.  

Indirect emissions: 

The two pathways for indirect emissions from managed soils are: (1)
volatilization of N as NH3 and oxides of N (NOx), and the deposition of
these gases and their products NH4 and NO3 onto soils and the surface of
lakes and other waters; and (2) the leaching and runoff from land of N
from synthetic fertilizer and crop residues. Leaching and runoff refers
to the inorganic N in or on soils which bypasses biological retention
mechanisms by transport in runoff, or overland water flow, and through
flow through soil macropores or pipe drains.

Indirect emissions from synthetic fertilizer consumption: 

Emissions = [(FSN × FracGASF × EF2) + (FSN × Fracleach × EF3)] ×
44/28

Where: 

FSN = annual amount of synthetic fertilizer N applied to soils (kg N)

FracGASF = fraction of synthetic fertilizer N that volatilizes as NH3
and NOx (equal to 0.10 kg N 

	volatilized/kg N applied)

EF2 = emission factor for N2O emissions from N volatilization (equal to
0.01 kg N2O-N/(kg NH3-N + NOx-N volatilized))

Fracleach = N lost from leaching and runoff (equal to 0.30 kg N/kg N
applied) 

EF3 = emission factor for N2O emissions from N leaching and runoff
(equal to 0.0075 kg N2O-N/kg N leached or runoff)

44/28 = conversion of N2O -N to N2O

4. Indirect emissions from crop residues: 

Emissions = FCR × Fracleach × EF3 × 44/28

Where: 

Fracleach = N lost from leaching and runoff (equal to 0.30 kg N/kg N
applied) 

EF3 = emission factor for N2O emissions from N leaching and runoff
(equal to 0.0075 kg N2O-N/kg N leached or runoff)	

44/28 = conversion of N2O -N to N2O

Total Emissions

To derive total N2O emissions from changes in fertilizer consumption due
to changes in crop acreage, emissions were summed across these four
pathways.

GHG Emission Rates for Agricultural Energy Use

We estimated GHG emissions per area of agricultural land by country, due
to agricultural energy inputs in the form of direct emissions from fuel
consumption and indirect emissions from electricity and heat. 

Total CO2 emissions from fuel combustion in the
agriculture/forestry/fishing sector of each country for 2005 and 2006
were taken from the International Energy Agency’s (IEA) CO2 Emissions
from Fuel Combustion 2007. As these estimates also include forestry and
fishing activities, using these estimates for agriculture results in an
overestimate of emissions. However, we believe this overestimate to be
small, as agriculture is by far the largest consumer of energy use of
these sectors.  Furthermore, emissions were determined per acre of
cropland, no distinction was made between different types of crops. 
Emissions from the use of the following fuel types are minimal and were
therefore not included in each country’s total CO2 emissions: Biogas,
Charcoal, Gas Works Gas, Geothermal, Other liquid biofuels, Primary
Solid Biomass, Solar thermal.

To estimate indirect emissions from the generation of electricity and
heat, 2005/2006 data on electricity and heat consumption in the
agriculture/forestry/fishing sector were obtained from IEA’s Energy
Statistics of Non-OECD Countries, 2008 and Energy Statistics of OECD
Countries, 2008. ,   CO2 emissions were estimated by multiplying
consumption by the average rate of CO2 produced per kWh of electricity
or heat generated (provided by IEA’s CO2 Emissions from Fuel
Combustion, 2007). 

Lifecycle GHG emission factors were applied to all calculated emissions
to estimate GHG emissions from fuel exploration, production,
transportation, and distribution (see Appendix, Table 7).  To estimate
CO2 emissions per agricultural area, emissions estimates were divided by
total agricultural area for each country (see Appendix, Table 8). 

CH4 Emission Factors for Rice Cultivation

For this analysis, we developed country- and region specific emission
factors for rice cultivation. We also provided rice growing season
lengths.

Calculating emissions from rice cultivation, per the IPCC 2006
guidelines, requires the following data: area of rice harvested, an
emissions factor, and planting to harvesting season length.  Changes in
area of rice harvested were provided from the FAPRI model results. 
Emissions from rice cultivation can be affected by a number of factors,
namely water regimes during the cultivation period, water regimes before
the cultivation period, and organic amendments.  If country-specific
data are available on these, the data can be used to scale the IPCC
default emission factor.  

For countries in this analysis, country-specific data on organic
amendments and the water regime before the cultivation period were not
available. Data were available, however, for the water regimes used
during the cultivation period. Therefore, the default IPCC emission
factor was scaled for each cropping regime: irrigated, rainfed lowland,
upland and deepwater. Default factors are presented in the Appendix,
Table 9. Rice cultivation season lengths were available from the
International Rice Research Institute (IRRI) for priority countries.

To be able to apply the cropping practice-specific emission factor, the
area harvested under each cropping regime must be known. Data were
collected from IRRI regarding the cropping practices in the major rice
growing countries of the world. Data covers the percentage of area
cultivated under each cropping regime (irrigated, rainfed lowland,
upland and deepwater) in each country (see Appendix, Table 10). To
calculate emissions from rice cultivation, the IRRI cropping regime
percentages for each country can be applied to area harvested to
determine the area grown under each regime in the country. Then, using
the season length for the country and the scaled emission factors for
each cropping regime, emissions can be calculated for each cropping
regime, and then summed to produce the total emission estimate for each
country. These country totals were multiplied by the changes in rice
production acres from the FAPRI model results to determine overall rice
methane emission changes.  

 Food and Agricultural Organization. Fertistat Database.   HYPERLINK
"http://www.fao.org/ag/agl/fertistat/" 
http://www.fao.org/ag/agl/fertistat/ . Last accessed October 10, 2008. 

 Personal correspondence. Wolfgang Prante, Information Management
Officer. Food and Agriculture Organization. July 15th, 2008.

 For several countries, Fertistat reported two or three years of data
collection to represent growing periods for crops or yearly averages for
a range of years. Data does not refer to fertilizer consumption over a
two or three year timespan. For crop data with two years mentioned, the
data refer to the period between planting and harvesting, and the latter
year is used in the analysis. For crop data with three years mentioned,
values reported are yearly averages, and the middle year was used in the
analysis. 

 Where the percentage of total agricultural area fertilized is not
available, application rates are calculated by Fertistat as consumption
divided by agricultural area planted (i.e. this assumes all area is
fertilized). Personal correspondence. Jan Poulisse, Senior Manager. Food
and Agriculture Organization. 

 Harris, Gene. 1998. An Analysis of Global Fertilizer Application Rates
for Major Crops. Agro-Economics Committee. Fertilizer Demand Meeting.
Toronto, Canada.  

 Fertilizer Association of India. “Usage of Fertilisers by Various
Crops: 1996-97.” 

 Region definitions: Switzerland and Norway are included in the “Rest
of World” category for land use change. Sudan is included in “Other
Middle East.” “Russia and Ukraine” category applies only to
Russia. The United States is excluded from the dataset. 

 FAPRI regions vary by crop type, so “Other” categories (e.g. Other
Latin America”) could potentially differ in the countries they
include.

 FAO. FAOSTAT: Pesticide Consumption.   HYPERLINK
"http://faostat.fao.org/site/424/default.aspx#ancor" 
http://faostat.fao.org/site/424/default.aspx#ancor . Last accessed:
October 9, 2008. 

 FAO. ResourceStat-Land.   HYPERLINK
"http://faostat.fao.org/site/377/default.aspx#ancor" 
http://faostat.fao.org/site/377/default.aspx#ancor . Last accessed:
October 15, 2008.

 FAO. FAOSTAT: Glossary.   HYPERLINK
"http://faostat.fao.org/site/379/DesktopDefault.aspx?PageID=379" 
http://faostat.fao.org/site/379/DesktopDefault.aspx?PageID=379 . Last
accessed: October 10, 2008. 

 Intergovernmental Panel on Climate Change. 2006 IPCC Guidelines for
National Greenhouse Gas Inventories. Volume 4: Agriculture, Forestry,
and Other Land Use. 

 IPCC. op. cit., pg. 11.6

 IPCC, op. cit., pg. 11.19

 International Energy Agency. 2007. CO2 Emissions from Fuel Combustion:
1971-2005. IEA Statistics. IEA reports 2006 data for OECD countries and
2005 data for OECD countries.

 International Energy Agency. 2008. Energy Statistics of Non-OECD
Countries. IEA Statistics.

 International Energy Agency. 2008. Energy Statistics of OECD Countries.
IEA Statistics.

 Factors provided by Vincent Camobreco, EPA, based on “Greenhouse
Gases, Regulated Emissions, and Energy Use in Transportation” (GREET)
model results for different fuels and scaling combustion vs. upstream
GHG emissions.  

 FAO: ResourceStat-Land. Dec. 2007.   HYPERLINK
"http://faostat.fao.org/site/377/default.aspx#ancor" 
http://faostat.fao.org/site/377/default.aspx#ancor . Last accessed:
October 17, 2008. 

e

f

A

B

m

ž

Ú

d

g

k

u

…

‘

™

 

¡

¢

ß

í

ú

	



@

2@

m

|

ž

Ÿ

 

Ç

Ì

Í

Ñ

Ò

Ù

Ú

 h*

 h*

h

h

h

h*

 hœ

. Last accessed: October 15, 2008.

 PAGE   

  PAGE  \* Arabic  7 

