Camelina-Derived Jet Fuel and

Diesel: Sustainable Advanced

Biofuels
David R. Shonnard,a Larry Williams,b and Tom N. Kalnesc
a
Department of Chemical Engineering and Sustainable Futures Institute, Michigan Technological University, Houghton, MI;
[email protected] (for correspondence)
b
Targeted Growth, Inc., 2815 Eastlake Ave E # 300, Seattle, WA
c
UOP LLC, a Honeywell Company, 25 E. Algonquin Rd, Des Plaines, IL 60017-5017

Published online 10 June 2010 in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/ep.10461

Recently, an isoparaffin-rich jet fuel derived from INTRODUCTION

camelina, a low-input nonfood oilseed crop, was The commercial airline industry has started evalu-
flight-tested by a commercial airline. To date, all test ating the use of biofuels as a means of reducing the
results indicate that this hydrotreated renewable jet carbon footprint of air travel. In 2009, both Japan Air-
fuel (HRJ) not only meets stringent engine fuel and lines (JAL) and KLM Royal Dutch Airlines successfully
performance specifications but also reduces environ- flight-tested a 50% blend of hydrotreated renewable
mental emissions. Several scenarios are now being jet fuel (HRJ) derived from the cultivation of camelina
considered for camelina as a sustainable feedstock (http://uop.com/pr/8050.html). Prior to commer-
for advanced biofuel production. For example, cial implementation, the production of new biofuels
growth of camelina in the Northern Plains of the such as camelina-derived jet fuel and diesel must be
United States on either marginal lands or as a rota- evaluated from both a technical and a sustainability
tion crop during fallow periods on existing lands al- perspective. Biomass-derived substitutes for transpor-
ready in food crop production can avoid the conflict tation fuel that primarily comprise isoparaffins have
with food cultivation and concerns with indirect material and chemical advantages compared to oxy-
land use change impacts. Updated estimates of camel- genated biofuels. Material compatibility issues, engine
ina cultivation requirements and commercial scale design, fuel economy, and fuel quality requirements
oil recovery and refining were used to calculate life strongly favor alternative bio-based fuels that more
cycle greenhouse gas (GHG) emissions and energy closely mimic current ultra low sulfur hydrocarbon
demand for both HRJ and renewable diesel (green fuels [1]. The process for converting fats and oils into
diesel, GD). GHG life cycle emissions for GD and HRJ high-quality transportation hydrocarbon fuel has
are 18.0 and 22.4 g CO2 equiv/MJ fuel, which repre- been demonstrated to be technically feasible [2]. Fur-
sent savings relative to petroleum counterparts of 80% thermore, isoparaffin rich fuels have been shown to
and 75%, respectively. Scenario analyses were con- reduce exhaust emissions of particulate matter and
ducted to determine response to model assumptions nitrogen oxides [3]. Additional details of the renew-
and data uncertainty, including allocation methodol- able oils refining process, refined oil product qual-
ogy, N fertilizer application rate, N2O emission factor, ities, production costs, and life cycle assessment
source of H2, and farm diesel consumption. Ó 2010 results can be found in another publication [4].
American Institute of Chemical Engineers Environ Prog, 29: Several scenarios are now being considered for
382–392, 2010 camelina as an energy crop for advanced biofuel pro-
Keywords: life cycle assessment, hydrotreated duction. For example, growth of camelina in the
renewable jet fuel Northern Plains of the United States on either mar-
ginal lands or as a rotation crop on fallow land can
Ó 2010 American Institute of Chemical Engineers avoid the conflict with food cultivation and concerns

382 October 2010 Environmental Progress & Sustainable Energy (Vol.29, No.3) DOI 10.1002/ep
with indirect land use change (LUC) impacts. The for growing camelina are in Montana. Therefore, we
goal of this article is to introduce camelina as a sus- will focus on the specifics of how this practice in
tainable energy crop and to summarize the estimated Montana would result in the production of large
life cycle greenhouse gas (GHG) emissions for HRJ amounts of camelina with no reduction in the
and green diesel (GD) produced by the UOP process amount of land devoted to food production. The
when using camelina oil as feedstock. These results same basic logic would apply as camelina cultivation
will be compared to conventional petroleum-derived spreads to other states.
jet fuel and diesel. According to a recent United States Department of
Agriculture (USDA) census [8], there are approxi-
mately 2600 farm owners in Montana who classify
Introduction to Sustainable Camelina Cultivation themselves as wheat producers. These farmers report
Camelina has been cultivated in Europe since the that the average size of their operation is 2540 acres
Bronze Age [5]. For the past 60 yrs, camelina has not (72% of the farms acreage) is planted to wheat
been important commercially as Europe and the annually. The other 28% of their land is fallowed or
United States provided subsidies for higher yielding rotated to other crops. Allowing the ground to be
major commodity grains and oilseed crops. It is still a rotated to camelina during a fallow year provides sev-
relatively common weed in much of Europe and in eral key benefits for the subsequent wheat crop, as
dryland regions of North America with more than 120 shown in Figure 1. First, soil moisture increases pro-
different varieties found in Montana alone. Camelina, viding a significant boost to crop yields the following
also known as false flax, is a member of the Brassicas year. Second, breaking a crop cycle and not growing
family. Recent interest in camelina arises mainly from the same crop every year reduces pest problems and
the need for low-input oilseed crops with the poten- disease potential. Third, the nutrient profile changes
tial for nonfood use within a sustainable agricultural positively through complex soil biochemical mecha-
system [6]. Camelina is a short-season crop (85–100 nisms not fully understood. The soil moisture factor
days) well adapted to cultivation in the temperate cli- is very important to Montana farmers (and dryland
mate zone. It germinates at low temperatures and is farmers in other areas) because the 10–15 inches of
very frost tolerant. Camelina is relatively easy to annual rainfall is not sufficient for maximum crop
grow, needing fewer inputs than most other crops. It yields. Thus, a fallow-cropping system increases
grows on marginal land and responds well under farmer revenue over a continuous cropping
drought stress conditions. Thus, it is better suited to approach. Growing camelina during the fallow year
low rainfall regions on marginal agricultural land than provides the same yield improvements to subsequent
other oilseed crops [7]. wheat crops and provides additional revenue that
Agronomic and crop production improvements are year as well.
just beginning to be applied to camelina cultivation. There are approximately 1,800,000 acres in the
Important yield improvements have already been operations of these Montana farmers that need to be
realized from modern plant breeding and agricultural fallow or rotate annually. Plant breeders from Tar-
techniques. Field tests beginning in 2006, at over 85 geted Growth Inc. using modern agronomic techni-
locations in the United States and Canada have ques have selected and bred camelina to allow farm-
resulted in yield improvements of approximately ers an opportunity to grow a rotation crop on this
33%. In addition, the application of yield improving fallow land. Because camelina has shallow roots with
genes (via mutation breeding, a recent non-GMO drought resistant characteristics, the land can be
innovation) is expected to improve yields more rap- returned to wheat cultivation the following year with
idly than those produced by traditional plant-breed- moisture and soil nutrients intact quantitatively similar
ing approaches. For sustainable camelina cultivation, to a fallow year. Because the economics of camelina
it is important to understand that these yield improve- cultivation favors this dryland crop rotation practice,
ments are being developed in the context of dryland it is reasonable to conclude that there will be no
farming techniques using low applications of fertilizer food producing land lost as a result of cultivating
and other inputs. camelina.
The fallow-cropping system and the use of mar- Beyond the wheat farmers of Montana and other
ginal agricultural lands for camelina cultivation con- states, there is additional camelina acreage potential
tribute to a decreased concern about the ‘‘food versus in other states and beyond those farmers who iden-
fuel’’ issue. The choice of fallow-cropping versus tify themselves primarily as wheat farmers. Some
continuous cropping of wheat is determined by the farms have a primary focus on other cash sources
economic returns to the farmer. The fallow-cropping such as livestock, and engage in dryland cereals culti-
approach is currently favored by dry-land farmers vation (wheat, oats, and barley) to diversify their
because the extra soil moisture and nutrients provide farms and to produce feed. Other farms may be
a significant boost to wheat yields (and farmer reve- growing peas, lentils, alfalfa, or several other crops
nue) following a fallow year. Camelina fits into the and would find benefits in rotation to camelina to
fallow-cropping approach because of its low moisture restore soil characteristics. Altogether more than 5
and nutrient requirements. Thus, current agricultural million U.S. acres have the potential to grow camel-
practice favors the growing of camelina as a rotation ina in a sustainable manner with no impact on food
crop with wheat during fallow periods because it supply. Thus, with expected future gains in yields/
maximizes farmer revenue. Most of near term markets acre, camelina could be providing 800 million gallons
Environmental Progress & Sustainable Energy (Vol.29, No.3) DOI 10.1002/ep October 2010 383
Figure 1. Benefits of camelina as a rotation energy crop with wheat or other dryland crops.

of oil per year for use as a secure and climate indication that this topic is still an active area of
friendly biofuel feedstock. research.
With 20 million acres of dry-land cereals (wheat, Regulators in Europe and California have reacted
oats, and barley) and fallowing every 3 or 4 yrs there to the LUC and iLUC research differently. In Decem-
would be 5–7 million acres available for camelina ber 2008, the European Parliament declared that the
cultivation annually as a rotation crop. Planned yield iLUC of a biofuel ‘‘is not currently expressed in a
improvements are expected to produce 3000 lbs of form that is immediately usable by economic opera-
seed per acre by 2016. The seeds are 36% oil with tors.’’ On April 23, 2009, the California Air Resources
the remainder meal. Thus, the oil potential is Board passed the low carbon fuel standard (LCFS).
between 750 and 1000 million gallons/year. The LCFS requires the state’s transportation system to
reach 90% of present carbon intensity by 2020. The
LCSF encompasses multiple actions. The one affecting
Discussion of Land Use Change Impacts biofuels will be the requirement on its producers to
Direct LUC (e.g. conversion of forest into agricul- use a Life Cycle Analysis that includes iLUC to certify
tural land) is a large source of GHGs and thus con- their reductions. The California regulatory action on
tributes to climate change. Indirect land use change biofuels has impacted federal action. The U.S. EPA
(iLUC) is an implied change, outside the production followed the California lead on iLUC while develop-
boundary of a feedstock, due to economic market ing the second version of the federal Renewable Fuel
forces initiated by energy crop production on existing Standards (RFS2) mandated by the Energy Independ-
agricultural lands. Displaced demand for food is ence and Security Act of 2007.
likely to produce LUC in places not directly associ- The EPA issued the final rule implementing RFS2
ated with biofuel production. In early 2008, two stud- on February 3, 2010. RFS2 assigns various combina-
ies appeared in Science [9, 10] indicating that biofuel tions of feedstock, fuel type, and production process
production may emit more GHGs than the fossil- to one of four categories of fuels eligible for compli-
based fuels displaced as a result of direct LUC and ance. For the purposes of GHG LCAs camelina, HRJ
iLUC. In the Fargione paper, GHG release from LUC and GD would be considered the same as soybean-
was termed a ‘‘carbon debt’’ and was identified as a based biodiesel. According to lifecycle analysis car-
potentially significant contributor to the life cycle ried out by the EPA biodiesel produced from domes-
analyses of biofuels. The time required to overcome tic soybean oil reduces GHG emissions by 57% com-
the carbon debt was termed the ‘‘payback period.’’ pared to petroleum diesel fuel and thus qualifies for
The original paper estimated ‘‘payback periods’’ in the biomass-based diesel program. Accordingly, cam-
the range of 100–1000 yrs. A more recent paper elina HRJ and GD also qualify for the program. The
arrived at payback periods between 3 and 14 yrs, a EPA uncertainty analysis indicates that GHG reduc-
difference of several orders of magnitude [11] and an tions could be as high as 85%. The EPA lifecycle cal-
384 October 2010 Environmental Progress & Sustainable Energy (Vol.29, No.3) DOI 10.1002/ep
Figure 2. Life cycle product diagrams for petroleum diesel and jet, GD, and HRJ. [Color figure can be viewed in
the online issue, which is available at wileyonlinelibrary.com.]

culations implemented in the final rule include iLUC and harvesting of the camelina crop. Inputs of seed,
but differ considerably from calculations published fuel, fertilizer, and chemical pesticides are all used for
with the draft RFS2 regulations issued in May 2009. cultivation of the camelina plant for oil production.
The EPA indicated that significant new scientific data After harvesting, the camelina seeds are transported
was available to the agency. This is another indica- to a processing facility, where the oils can be
tion that biofuel lifecycle analysis remains an active extracted. A crusher and solvent-based process
research area. extracts the oil and recovers the co-product seed
meal. The oil is processed to remove oil impurities,
METHODOLOGY OF LIFE CYCLE ASSESSMENT FOR CAMELINA-DERIVED and then transported by rail (from Culberston, Mon-
JET AND DIESEL tana to Seattle, WA) to a site for GD or HRJ produc-
tion. The UOP conversion process requires hydrogen
Fuel Pathways to produce HRJ and GD. This hydrogen can come
The life cycle diagrams for petroleum diesel, petro- from the steam reforming of natural gas or from a
leum jet, GD, and HRJ are shown in Figure 2 with naphtha-to-gasoline reformer in an existing refinery,
important inputs and co-product outputs indicated. or a combination of these inputs. In this study, it is
HRJ and GD life cycles originate with the growing assumed that hydrogen is obtained 60% from a gaso-
Environmental Progress & Sustainable Energy (Vol.29, No.3) DOI 10.1002/ep October 2010 385
Table 1. Cultivation inputs for camelina seed harvested (per kg seed).

Camelina seed yield (kg/ha) Farmer Farmer Forward

Resources 2007 (664) 2008 (1123) 3000 (3368)
Energy content of seed (MJ) 30.4 30.4 30.4
Potassium chloride, as K2O (kg) 0.0169 0.010 0.0033
Thomas meal, as P2O5 (kg) 0.0253 0.015 0.0050
Urea, as N (kg) 0.0169 0.0370 0.0249
Diesel, low-sulfur (kg) 0.0402 0.0238 0.0079
Emissions to air
Methane (g) 0.0024 0.0024 0.0024
Dinitrogen monoxide (g) 0.352 0.77 0.518
Carbon dioxide (diesel) (kg) 0.125 0.0738 .0245
Carbon dioxide (urea) (kg) 0.0265 0.0581 .0391

line reformer and 40% from steam reforming of meth- ate because energy content is a fundamental charac-
ane. Renewable fuel co-products such as naphtha, teristic for all transportation fuels. Inventory data for
fuel gas and propane/butane (renewable LPG) are camelina cultivation, transportation, and oil extraction
also produced from camelina oil as shown in Figure were provided by Targeted Growth, Inc. for all cam-
2. A portion of the environmental burdens of the fuel elina cultivation inputs and oil extraction require-
production processes are allocated to these co-prod- ments. Data for conversion of oil feedstock to HRJ
ucts, and a similar allocation occurs at the stage of and GD were obtained from basic engineering design
camelina oil extraction. data, and was supplied by UOP. Energy allocation
Petroleum-based jet fuel and diesel were the refer- was applied to distribute GHG impacts among the
ence fuels for comparison. The GHG profile for pe- various products and co-products along the fuels life
troleum fuel is derived from a composite mix of U.S. cycle. Energy allocation was selected because a
production, which is likewise dependent on feed- recent guidance report recommends this form of allo-
stocks from different regions of the world. This com- cation for fuel products [13]. At the camelina oil
posite mix includes both domestic and foreign pro- extraction stage, inventory data from camelina culti-
duction of crude oil and tar sands oil. Transportation vation up to and including oil extraction were allo-
of crude oil to domestic refiners includes transoceanic cated to camelina oil and co-product seed meal using
tanker and pipeline modes and from refineries jet an energy balance. Allocation to oil 5 (A 3 LHVA)/
fuel or diesel is transported by barge, pipeline, rail, (A 3 LHVA 1 B 3 LHVB), where A is the mass flow
and semi truck. Impacts of refining are also derived rate of output oil from the seed extraction step, B is
from a mix of domestic and foreign facilities. GHG mass flow rate of meal, LHV is lower heating value,
emissions/MJ jet fuel and diesel were obtained from a and subscripts A and B are camelina oil and seed,
recent U.S. Department of Energy (DOE) study [12]. respectively. A similar calculation was performed at
the GD and HRJ production stages considering the
Goal and Scope co-products produced at that stage. Energy allocation
factors are presented in the next section.
The goal of this LCA is to determine the GHG
emissions, cumulative energy demand, and fossil
energy demand of camelina-derived HRJ and GD Life Cycle Inventory Analysis
over the entire life cycle and compute the GHG sav-
This section covers the input data and assumptions
ings per MJ of fuel compared to petroleum-based die-
used for procuring the camelina seed, extraction of
sel and jet fuel. A second goal is to evaluate impacts
crude camelina oil, and production of degummed
of fertilizer application rates and fossil fuel use in
(refined) camelina oil. Tables 1–3 show key inputs to
farm equipment employing data provided by Mon-
the camelina biofuel life cycle for these stages and
tana farmers and predict impacts of crop yield
for feedstock transportation. Cultivation inputs were
improvements expected in the future. A final goal is
based on actual field application rates as in the Farmer
to explore the impacts of model parameter uncer-
2007 and 2008 data, where increased N fertilizer appli-
tainty in the calculation of GHG emissions. The scope
cation rates for 2008 were designed to boost camelina
of this study encompasses the entire life cycle from
yields. The values reported as Forward 3000 represent
acquisition of raw materials from the environment
projected gains from crop improvement research.
through the production and use of the fuels in vehi-
These gains increased per acre seed productivity
cle and aircraft operations.
(3000 lb/ac; 3360 kg/ha) and decreased per acre N
application rates (75 lb N in fertilizer/ac; 84 kg/ha;
Functional Unit and Energy Allocation rate is equivalent to 55% of N content in seed, which
The functional unit chosen in this analysis is 1 MJ is 0.045 kg N/kg seed) and diesel consumption (33.3
of energy content in the fuel. This choice is appropri- L/ha) on the farm. Seed oil extraction inputs were also
386 October 2010 Environmental Progress & Sustainable Energy (Vol.29, No.3) DOI 10.1002/ep
Table 2. Processing inputs for camelina seed Table 3. Processing Inputs for of camelina refined oil,
processed to degummed oil (per kg seed). per MJ.

Resources Resources
Camelina seed input per kg 84.5 MJ Camelina crude oil per MJ 1.0417 MJ
oil output refined oil
Electricity, medium voltage, 0.0083 kWh Steam 0.00282 kg
average U.S. mix Electricity, medium voltage, 0.0006 MJ
Heat, natural gas, at industrial 0.421 MJ average U.S. mix
furnace >100 kWh
Diesel, low-sulfur, at regional 0.0064 L
storage (seed transport) elina. GHG results for petroleum kerosene-based jet
Hexane (oil extraction solvent) 0.00125 L fuel was obtained from a recent DOE study [12].
Rail transport of crude 2 tkm
camelina oil
Impact Assessment and Comprehensive
Assumptions
The software used for this LCA was SimaPro 7.1
[16], which contains a large database of inventory
provided by TGI. Emission of CO2 from atmospheric data for material, chemical, and energy inputs. Inven-
oxidation of emitted oil extraction solvent hexane was tory data is from the Ecoinvent database, which is
also included. Crude camelina oil refining inputs were comprised of mostly European data that has close
obtained from a recent study [14]. Energy allocation technology relevance to U.S. production, but when-
factors for assigning impacts to the product crude ever possible, these ecoprofiles were adjusted for
camelina oil extraction were calculated using energy U.S. conditions; for example for electricity generation
balances around the process, as described above. was modeled as an average U.S. mix. The GHG
Energy allocation factor for the crude camelina oil impact assessment method used in SimaPro was IPCC
extraction processes as listed in Table 2 is 0.445, and 2007 GWP 100a V1.01 method whose output is in
for refined camelina oil conversion to biofuel, energy CO2 equivalents for all of the GHG emissions. The
allocation were calculated from data provided by UOP three primary gases that make up the CO2 equiva-
LLC. While included in the LCA modeling, the UOP lence are CO2, methane (CH4), and nitrous oxide
data describing the detailed hydroprocessing inputs (N2O), but the minor contributions from solvents and
are not included in these inventory tables. However, refrigerants were also included. In the IPCC 2007
key inputs to the GD and HRJ conversion process GWP 100a V1.01 method, each of these gases has dif-
include not only hydrogen, as shown in Figure 2, but ferent GHG potentials; for example, N2O has the
also electricity, steam, and water for boiler feed and GHG potential of 298 compared to CO2, and methane
process cooling as well as input rates of hydrogen has a GHG potential of 25. For the end use of the
compared to camelina oil and renewable co-product fuels, notably combustion in a vehicle, it was
production rates (CO2, H2O, fuel gas, LPG, and naph- assumed that the fuels would completely combust,
tha) are similar to those listed in a recent publication forming maximum yields of water and CO2. This
[4]. Steam is also generated as a result of biofuel pro- assumption is very close to reality in terms of the car-
duction through hydroprocessing, and steam was gen- bon dioxide emissions, and therefore stoichiometric
erated in sufficient amount to credit 4.2 g CO2 equiv/ emission factors were utilized based on the carbon
MJ GD and 2.3 g CO2 equiv/MJ HRJ. content of each fuel (e.g. 3.17 kg CO2/kg petroleum
Emissions of N2O from nitrogen fertilizers applied diesel and jet). Emission of CO2 from combustion of
to soils during plant cultivation was estimated using biofuels does not count toward the GHG calculation
both direct and indirect emission factors from the because biogenic carbon is sequestered during cam-
Intergovernmental Panel for Climate Change [15] in elina growth, and combustion simply returns this car-
which 1.325% of applied N in fertilizer is emitted as bon to the atmosphere in a closed cycle.
N2O nitrogen. No accounting for N in camelina resi- Not included within the scope of this study are the
due left on farm is included in this N2O emission construction impacts of farming equipment and proc-
analysis due to lack of data on residues. CO2 emis- essing facilities, which are typically neglected in LCA
sion from urea mineralization in the soil was esti- studies of biofuels, as they constitute a relatively
mated to be 1.57 kg CO2/kg N fertilizer based on the small fraction of life cycle impacts. It is important to
fraction of carbon and nitrogen in urea. No effects of note here that no wastewater treatment or solid waste
LUC on possible GHG emissions were included in disposal was factored into the study, which typically
this analysis, although the anticipated cultivation as a constitute less than 1% to the total. Biomass feedstock
rotation energy crop with wheat that displaces fallow transportation was included in this study for a camel-
weeds suggests negligible adverse direct and indirect ina seed to oil processing facility based in Montana.
LUC impacts, since no food production is displaced In all cases, transport of the final fuel product to mar-
by camelina seed cultivation. The geographic context ket was included over a distance of 150 km. This is a
for the production of green jet is the U.S., with culti- reasonable and representative value for a U.S. facility
vation occurring in the U.S. Northern Plains for cam- serving a local market.
Environmental Progress & Sustainable Energy (Vol.29, No.3) DOI 10.1002/ep October 2010 387
Figure 3. Cumulative energy demand by fuel type for each camelina biofuel product.

Figure 4. GHG emissions for all fuels in this study.

RESULTS AND DISCUSSION: GHG EMISSIONS FOR CAMELINA-DERIVED The emission of GHG over the life cycle for each
JET AND DIESEL transportation fuel alternative is shown in Figure 4
Energy consumption for each product over the life for the three camelina cultivation scenarios. The larg-
cycle is another important characteristic to judge est contribution to HRJ and GD GHG emissions is, in
the comparative advantages of camelina biofuels. roughly descending order: feedstock production >
Figure 3 shows the cumulative energy demand for biofuel production > feedstock chemicals  feed-
each biofuel and cultivation method combination. stock transport. Emissions from fuel product transpor-
Total energy consumption for each camelina biofuel tation are negligible and combustion emissions of
product is comparable to petroleum fuels, but most CO2 do not contribute to climate warming because
energy is renewable biomass as opposed to non biofuels such as HRJ and GD contain no fossil car-
renewable fossil, and consumption of petroleum over bons, only carbon atoms sequestered from the atmos-
the life cycle for each biofuel product is <0.15 MJ phere as CO2 during plant growth. Feedstock produc-
(data not shown in Figure 3). Except for the use of tion involves all unit processes in camelina seed culti-
diesel on the farm and during transportation steps, vation, including emissions of urea-related N2O and
petroleum resources are nearly absent from the CO2 (high majority) as well as CO2 released from die-
camelina biofuel life cycles. sel use on the farm. Fuel production emissions are

388 October 2010 Environmental Progress & Sustainable Energy (Vol.29, No.3) DOI 10.1002/ep
Figure 5. Contributions of GHG gases to total emissions.

Table 4. Effect of change in nitrogen fertilization rate, IPCC N2O emission factor, hydrogen production
method, and farm diesel input rate.

GD Base case (g CO2 eq./MJ) GD Scenarios (g CO2 eq./MJ)

Energy allocation, 13 N fertilizer 18.04 Energy allocation, 23 N fertilizer 26.99
Energy allocation, IPCC EF Energy allocation, 23 IPCC EF 23.06
Energy allocation, H2 40% SRM Energy allocation, H2 100% SRM 23.40
Energy allocation, 13 farm diesel Energy allocation, 23 farm diesel 18.98

EF, emission factor for N2O released from soil; SRM, steam reforming of methane.

dominated by generation of hydrogen, the use of Table 5. Effect of N fertilizer choice.

process heat and power, and solvent use for camelina
oil extraction. Feedstock transportation is a notable Fertilizer scenarios (GD) (g CO2 eq./MJ)
contributor to GHG emissions, mostly from long dis-
tance rail transport of refined camelina oil from Mon- Energy allocation, 13 urea 18.04
tana to Washington state. Emissions of GHGs are larg- Energy allocation, 13 NH3 (liq) 16.17
est for Farmer 08 cultivation because of the elevated Energy allocation, 13 NH4NO3 21.02
N fertilizer application to boost yield. The GHG emis-
sions for camelina GD and HRJ exhibit savings com-
pared to petroleum diesel and jet of 80% and 75%,
respectively, for the Forward 3000 case. GHG savings The emission of N2O is subject to environmental
are always greater than 67% for the Farmer 2007 and conditions and farm practices, and is dependent on
2008 cases for GD and HRJ. soil moisture, soil type, climatic conditions and culti-
Figure 5 shows the contributions of each GHG to vation method. As a result, there is significant uncer-
the total GHG profile for all fuel alternatives. In gen- tainty as to the actual emission rate of N2O from culti-
eral, the highest contribution to total GHGs is from vated lands. The application rate of N fertilizer and
CO2, followed by N2O and the least is from CH4. The diesel use on farm for camelina cultivation and har-
total contribution of refrigerants and solvents having vesting are also sources of uncertainty given that
global warming potentials were negligible and so are long-term field trials are just underway. Other sources
not shown in this figure. A portion of the GHG emis- of uncertainty are the selection of N fertilizer (the
sions from plant cultivation are from emission of N2O choice has fertilizer production GHG implications)
from soils as a result of application of nitrogen fertil- and the source of hydrogen used in the UOP process
izer. The importance of soil emissions of N2O on the for production of GD and HRJ. All of these sources of
GHG profiles for both biofuels is evident from this uncertainty, as well as effects of allocation method, are
figure. explored in the next section (Tables 4–6).
Environmental Progress & Sustainable Energy (Vol.29, No.3) DOI 10.1002/ep October 2010 389
Table 6. Effect of allocation method. are shown in Table 6. These results are slightly lower
than for the energy allocation case. A second alloca-
GHG emissions tion scenario was analyzed assuming that all environ-
(g CO2 eq./MJ) mental burdens of camelina biofuel production are
allocated to the biofuel, and that co-products substi-
Scenarios GD HRJ tute for and avoid the production and use of prod-
Energy allocation base case 18.04 22.36 ucts in the market. Camelina seed meal is assumed to
Mass allocation 15.80 20.18 substitute for soybean meal as an animal feed, whose
Displacement allocation 4.23 217.03 GHG emissions were estimated to be 405 g CO2
equiv/kg soybean meal produced using inventory
inputs from U.S. average cultivation of soybeans [18]
and for soybean transportation and oil processing
Scenario Analyses: Inventory Uncertainty and from a DOE study [19], and using mass allocation for
Effect of Allocation Method (Table 4) the meal impacts. Ecoprofiles in SimaPro7.1 were uti-
Nitrogen fertilizer application rate and the emis- lized to convert inventory inputs into GHG impacts
sion of N2O from the farm soil are both critical fac- using the IPCC GWP 100a method. For renewable co-
tors in this biofuels GHG analysis. The application products produced at the GD and HRJ production
rate of N fertilizer was doubled from the value given stage (fuel gas, liquid petroleum gas (LPG), and
in Table 1 for the Forward 3000 case to 150 lb/ac, a naphtha), GHG credits were estimated using ecopro-
value slightly higher than the N requirement for cam- files in SimaPro7.1 and avoided combustion emis-
elina seeds, and the resulting GHG impacts increased sions were estimated to be 3.0 kg CO2 equiv/kg fuel
from 18.0 to 27.0 g CO2 equiv/MJ GD. This increased gas, 2.75 kg CO2 equiv /kg LPG, and 3.17 kg CO2
value has a GHG savings compared to petroleum die- equiv/kg naphtha, assuming stoichiometric conver-
sel of 70% compared to 80% for original case. When sion of each fuel to CO2. GHG emissions for GD and
the emission factor for N2O release from farm soil is HRJ for displacement allocation are 4.23 and 217.03 g
doubled from 1.325% of applied fertilizer N to 2.65% CO2 equiv/MJ, respectively (Table 6). The reason for
and using the same application rate as in Table 1 for the relatively low and even negative GHG impact
the Forward 3000 case, the GHG emissions for GD values for displacement allocation is that relatively
are increased from 18.0 to 23.1 g CO2 equiv/MJ GD GHG intensive fuels are replaced by fuels (GD, HRJ,
(GHG savings of 74% compared to 80% for original and co-product fuels) that exhibit much low GHG
case). The effect of hydrogen source was tested by impacts to produce and use. The HRJ impacts are
assuming that all hydrogen was generated from steam negative for displacement allocation (meaning an
reforming of methane. The result of this sensitivity improvement compared to production and use of
analysis increased GHG emissions from 18.0 to 23.4 g fossil fuels) because of a larger fuel co-product
CO2 equiv/MJ GD (GHG savings of 74% compared to production than for GD.
80% for original case). Diesel fuel use on the farm for
camelina seed cultivation and harvesting was doubled
Land Use Change Discussion
from the value in Table 1 (from 33.3 to 66.6 liters die-
How might camelina logically fit into the LUC
sel/ha) and this change increased GHG emissions
framework? Referring back to the earlier discussion in
from 18.0 to 19.0 g CO2 equiv/MJ GD (GHG savings
of 79% compared to 80% for original case). These Section ‘‘Discussion of Land Use Change Impacts’’
about the actual cultivation practices under which cam-
uncertainty analyses confirm the importance of N fer-
elina is grown today and will be grown in the future it
tilizer rate, N2O emissions from soil, source of fuel
seems logical that ‘‘light footprint’’ and rotation crop
for H2 product, and farm diesel requirements; how-
ever, savings of GHG compared to petroleum diesel cultivation practices will result in negligible additional
GHG emissions due to indirect and direct LUC.
did not change very much.
The most commonly applied N fertilizers in the
United States are urea, ammonium nitrate, and am- CONCLUSIONS
monia [17]. The effect of choice of N fertilizer is Synthetic isoparaffin-rich fuels produced by hydro-
shown in Table 5. Anhydrous ammonia exhibits the processing camelina oil in the UOP conversion pro-
lowest GHG impacts, whereas ammonium nitrate cess show great promise as ‘‘drop-in’’ alternatives to
causes the highest GHG emissions. The differences in petroleum jet and diesel fuel. Recent evaluations of
GHG impacts shown in Table 5 are due to both the HRJ fuel derived from camelina oil indicate the fuel
manufacturing of these fertilizers and their emissions performed just as well as petroleum based fuel but
of N2O and CO2 in the field. with lower exhaust and GHG emissions. Using
Allocation method can have a significant effect on updated estimates of camelina cultivation and com-
results of life cycle assessments. The results presented mercial scale estimates of oil recovery and refining
previously were generated using energy allocation. requirements a life cycle GHG savings of 75 and
When mass allocation is used, mass allocation factor 80.0% was estimated for camelina-derived HRJ and
for the stages of camelina seed oil extraction is 0.36 GD relative to their petroleum counterparts for the
(versus 0.447 for energy allocation), and for biofuel future scenario, Forward 3000.Using data from recent
production essentially the same as energy allocation. field trials (Farmer 2007 and 2008), GHG savings of
GHG emissions for GD and HRJ for mass allocation >67% are achieved. Although the impacts of LUC
390 October 2010 Environmental Progress & Sustainable Energy (Vol.29, No.3) DOI 10.1002/ep
were not included in this analysis, the authors expect program results, In 9th AIAA Aviation Technology,
these impacts to be negligible for cultivation on mar- Integration, and Operations Conference (ATIO) (p
ginal agricultural land and rotation with grain crops 19), Hilton Head, South Carolina: American Institute
during fallow periods (where no food production is of Aeronautics and Astronautics.
displaced by camelina seed cultivation). 3. Hileman, J., Wong, H.M., Donohoo, P., Stratton,
Several scenarios are now being considered for R.W., Weiss, M., & Waitz, I. (2009). Environmental
camelina as an energy crop for advanced biofuel pro- feasibility of alternative jet fuels, In UC Davis
duction. For example, growth of camelina in the Symposium on Aviation Noise & Air Quality,
Northern Plains of the United States on either mar- Palm springs, CA.
ginal lands or as a rotation crop on existing lands al- 4. Koers, K.P., Kalnes, T.N., Marker, T., & Shonnard,
ready in food crop production can avoid the conflict D.R. (2009). Green diesel: A technoeconomic and
with food crops and concerns with indirect LUC environmental life cycle comparison to biodiesel
impacts. UOP is currently in the final stages of and syndiesel, Environmental Progress & Sustain-
designing the refining process that will enable the able Energy, 28, 11–120.
flexible conversion of camelina oil to isoparaffin-rich 5. Schultze-Motel, J. (1979). Die anbaugeschichte
HRJ and GD. It is expected that HRJ will be certified des leindotters, Camelina sativa (L.), Crantz
for use by commercial jet engines and that commer- Archaeo-Physika, 8, 267–281.
cial volumes will be available by 2012. With expected 6. Putnam, D.H., Budin, J.T., Field, L.A., & Breene,
future gains in yields/acre, camelina oil production W.M. (1993). Camelina: A promising low-input
and hydroprocessing has the potential to provide the oilseed. In J. Janick & J.E. Simon (Eds.), New
United States an estimated 800 million gallons per Crops: Exploration, Research, Commercialization.
year of high quality, climate friendly renewable fuel. Proc. New Crops, Oct. 6–9, 1991, Indianapolis, IN
(pp. 314–322), New York, USA: Wiley.
ACKNOWLEDGMENTS 7. Vollmann, J., Damboeck, A., Eckl, A., Schrems,
We appreciate the contributions of Scott Johnson, H., & Ruckenbauer, P. (1996). Improvement of
Fernando Guillen-Portal, Lew Rubin, and Christine Camelina sativa, an underexploited oilseed. In
Shewmaker for their expertise and assistance in J. Janick (Ed.), Progress in new crops (pp. 357–
assembling the agricultural inputs for this paper. 362), Alexandria, VA, USA: ASHA Press.
8. NASS. (2009). 2007 Census publication, Vol. 1,
Abbreviations Chapter1: State level data, Montana, National Ag-
ricultural Statistics Service, U.S. Department of
ACES American Clean Energy Security Act Agriculture.
DOE U.S. Department of Energy 9. Searchinger, T., Heimlich, R., Houghton, R.A.,
EF emission factor for N2O released from Dong, F., Elobeid, A., Fabiosa, J., Tokgoz, S.,
soil Hayes, D., & Yu, T.-H. (2008). Use of U.S. crop-
GD green diesel
GHG greenhouse gas lands for biofuels increases greenhouse gases
GMO genetically modified organism through emissions from land use change, Science,
GWP global warming potential 319, 1238–1240.
HRJ hydrotreated renewable jet 10. Fargione, J., Hill, J., Tilman, D., Polasky, S., &
iLUC indirect land use change Hawthome, P. (2008). Land clearing and the bio-
IPCC Intergovernmental Panel on Climate
fuel debt, Science, 319, 1235–1238.
Change
LCA life cycle assessment 11. Kim, H., Kim, S., & Dale, B.E. (2009). Biofuels,
LCFS low carbon fuel standard (California) land use change, and greenhouse gas emissions:
LHV lower heating value Some unexplored variables, Environmental Sci-
LPG liquid petroleum gas ence and Technology, 43, 961–967.
LUC land use change 12. NETL. (2008). Development of baseline data and
RFS renewable fuel standard (United States)
SRM steam reforming of methane analysis of life cycle greenhouse gas emissions of
USDA United States Department of Agriculture petroleum-based fuels, National Energy Technol-
ogy Laboratory, U.S. Department of Energy.
LITERATURE CITED 13. Allen, D.T., Allport, C., Atkins, K., Cooper, J.S., Dil-
1. NSF. (2008). Breaking the chemical and engineer- more, R.M., Draucker, L.C., Eickmann, K.E., Gillen,
ing barriers to lignocellulosic biofuels: Next gen- J.C., Gillette, W., Griffin, W.M., Harrison III, W. E.,
eration hydrocarbon biorefineries (180 p), Wash- Hileman, J.I., Ingham, J.R., Kimler III, F.A., Levy, A.,
ington D.C.: University of Massachusetts Amherst, Murphy, C.F., O’Donnell, M.J., Pamplin, D., Schivley,
National Science Foundation, Chemical-Bioengin- G., Skone, T.J., Strank, S.M., Stratton, R.W., Taylor,
eering-Environmental and Transport Systems P.H., Thomas, V.M., Wang, M., & Zidow, T. (2009).
Division. Framework and guidance for estimating greenhouse
2. Rahmes, T., Kinder, J.D., Henry, T.M., Crenfeldt, gas footprints for aviation fuels (108 p), U.S. Air
G., LeDuc, G.F., Zombanakis, G.P., Abe, Y., Lambert, Force.
D.M., Lewis, C., Juenger, J.A., Andac, M.G., Reilly, 14. Edwards, R., Larivé, J.-F., Mahieu, V., & Rouveir-
K.R., Holmgren, J.R., McCall, M.J., & Bozzano, A.G. olles, P. (2007). Well-to-wheels analysis of future
(2009). Sustainable bio-derived synthetic paraffinic automobile fuels and powertrains in a European
kerosene (Bio-SPK) jet fuel flights and engine tests context: Well-to-tank report version 2c, WTT Ap-
Environmental Progress & Sustainable Energy (Vol.29, No.3) DOI 10.1002/ep October 2010 391
 pendix 1 (p. 80), CONCAWE, EUCAR, European 16. PRe‘, C., SimaPro version 7.1.2008.
Commission Joint Research Council. 17. USDA. (2009). US fertilizer use and price. Available
15. De Klein, C., Novoam, R., Ogle, S., Smith, K.A., at: http://www.ers.usda.gov/Data/FertilizerUse/.
Rochette, P., Wirth, T.C., McConkey, B.G., Mosier, A., 18. NASS. (2007). Agricultural chemical usage 2006
Rypdal, K., Walsh, M., & Williams, S.A. (2006). N2O field crops summary, National Agricultural Statis-
emissions from managed soils, and CO2 emissions tics Service, U.S. Department of Agriculture.
from lime and urea application. In S. Eggelston, L. 19. Sheehan, J., Camobreco, V., Duffield, J., Gra-
Buendia, K. Miwa, T. Ngara, K. Tanabe (Eds.), IPCC boski, M., & Shapouri, H. (1998). A life cycle
guidlines for national greenhouse gas inventories Vol- inventory of biodiesel and petroleum diesel
ume 4, Agriculture, Foresty, and Other Land Use for use in an urban bus, National Renewable
(Chapter 11, p. 54), Intergovernmental Panel for Cli- Energy Laboratory, U.S. Department of
mate Change (IPCC), UN and WMO. Energy.

392 October 2010 Environmental Progress & Sustainable Energy (Vol.29, No.3) DOI 10.1002/ep