Water Consumption Footprint and Land
Requirements of Alternative Diesel and Jet Fuel
Production
by
MA SSACHUSETTS INWR
OF TECHNOLOGY
Mark Douglas Staples
JF TECHOLOG
B.Sc., Mechanical Engineering
University of Alberta, Canada, 2009
UBRARIES
Submitted to the Engineering Systems Division
in partial fulfillment of the requirements for the degree of
Master of Science in Technology and Policy
at the
MASSACHUSETTS INSTITUTE OF TECHNOLOGY
June 2013
@
Massachusetts Institute of Technology 2013. All rights reserved.
Author.... ............
...................
r. . ........
Sngineering Systems Division
1)
Certified by .........
May 10, 2013
w ..
Steven R. H. Barrett
Assistant Professor of Aeronautics and Astronautics
Thesis Supervisor
/ .
Certified by...
. . ..
..
. ... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
Robert Malina
Research Scientist
Thesis Supervisor
Accepted by.. . . . . . . . . . . . . . . . . . (
Dava J. Newman
Professor of Aeronautics and Astronautics and Engineering Systems
Director, Technology and Policy Program
2
Water Consumption Footprint and Land Requirements of
Alternative Diesel and Jet Fuel Production
by
Mark Douglas Staples
Submitted to the Engineering Systems Division
on May 10, 2013, in partial fulfillment of the
requirements for the degree of
Master of Science in Technology and Policy
Abstract
The Renewable Fuels Standard 2 (RFS2) is an important component of alternative
transportation fuels policy in the United States (US). By mandating the production
of alternative fuels, RFS2 attempts to address a number of imperfections in the
transportation fuels market: US economic vulnerability to volatile prices; security
and environmental externalities; and a lack of investment in alternatives to petroleumderived fuels. Although RFS2 aims to reduce the climate impact of transportation
fuels, the policy raises a number of additional environmental concerns, including the
water and land resource requirements of alternative fuel production. These factors
should be considered in order to determine the overall environmental viability of
alternatives to petroleum-derived transportation fuels. Middle distillate (MD) fuels,
including diesel and jet fuel, are of particular interest because they currently make
up almost 30% of liquid fuel consumption in the US, and alternative MD fuels could
potentially satisfy 21 of the 36 billion gallons of renewable fuels mandated by RFS2 in
2022. This thesis quantifies the lifecycle blue (surface and ground) water consumption
footprint of MD from conventional crude oil; Fischer-Tropsch (FT) MD from natural
gas and coal; fermentation and advanced fermentation (AF) MD from biomass; and
hydroprocessed esters and fatty acids (HEFA) MD and biodiesel from oilseed crops,
in the US. FT and rainfed biomass-derived MD have lifecycle blue water consumption
footprints between 1.4 and 18.1 lwater/lMD, comparable to conventional MD, between
4.1 and 7.5 lwater/lMD. Irrigated biomass-derived MD has a lifecycle blue water
consumption footprint potentially several orders of magnitude larger, between 2.5 and
5300 lwater/lMD. Results are geospatially disaggregated, and the trade-offs between
blue water consumption footprint and areal MD productivity, between 490 and 3710
lMD/ha, are quantified under assumptions of rainfed and irrigated biomass cultivation.
Thesis Supervisor: Steven R. H. Barrett
Title: Assistant Professor of Aeronautics and Astronautics
Thesis Supervisor: Robert Malina
Title: Research Scientist
3
4
Acknowledgments
I would first like to thank my thesis co-advisors, Professor Steven Barrett and Dr.
Robert Malina. Their academic guidance and mentorship not only made this work
possible, but has set me down an exciting and challenging career path. I would also
like to thank the Federal Aviation Administration, Air Force Research Laboratory and
Defense Logistics Agency Energy for supporting this work financially, and especially
Dr. James Hileman at the FAA for patiently offering his time and expertise.
I have been very fortunate to share an office space with many exceptional friends
and colleagues from the Laboratory for Aviation and the Environment over the past
2 years. In particular I would like to thank Dr. Hakan Olcay and Mr. Matthew
Pearlson for their earnest tutelage and good humor.
I can not imagine MIT without my TPP family. Thank you to my classmates
and dear friends, who never cease to amaze me with their intellect, drive, sense of
humor, and above all, good looks. A special thank you to all the Sciarappians, TPSS
e-boarders, Sciencewonks, and Jelly Badgers with whom I have had the pleasure of
living/working/writing/playing.
Finally, I would like to thank my parents and siblings for their support, and my
bride-to-be Barbara, for the type of love and encouragement that knows no borders.
5
THIS PAGE INTENTIONALLY LEFT BLANK
6
Contents
1 Introduction
15
2
19
19
20
22
23
25
28
30
Policy context
2.1 H istory . . . . . . . . . . . . . . . . . . . . . . . . . .
2.2 M echanics of RFS2 . . . . . . . . . . . . . . . . . . .
2.3 Transportation fuels market imperfections &RFS2 . .
2.3.1 Imperfect competition &economic vulnerability
2.3.2 Security and environmental externalities . . .
2.3.3 Innovation &industry . . . . . . . . . . . . . .
2.3.4 Conclusions . . . . . . . . . . . . . . . . . . .
. . . .
. . . .
. . . .
. . .
. . . .
. . . .
. . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
3
Analysis of water consumption and land resource requirements
3.1 M ethodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.1.1 Water use definitions . . . . . . . . . . . . . . . . . . . . . . .
3.1.2 Lifecycle methodology . . . . . . . . . . . . . . . . . . . . . .
3.2 GAEZ model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.2.1 Model validation . . . . . . . . . . . . . . . . . . . . . . . . .
3.2.2 Irrigated blue water consumption calculation . . . . . . . . . .
3.3 Fuel production pathways . . . . . . . . . . . . . . . . . . . . . . . .
3.3.1 Conventional MD . . . . . . . . . . . . . . . . . . . . . . . . .
3.3.2 FT MD from natural gas and coal . . . . . . . . . . . . . . . .
3.3.3 AF MD from sugarcane, corn and switchgrass . . . . . . . . .
3.3.4 HEFA MD and biodiesel from soybean, rapeseed and jatropha
33
33
34
34
36
37
37
38
38
45
47
54
4
Results and discussion
4.1 Results and geo-spatial disaggregation . . . . . . . . . . . . . . . . .
4.2 Areal productivity benefits of irrigated biomass cultivation . . . . . .
61
61
66
5
Conclusions and future work
71
7
A Blue water consumption footprint by lifecycle step
73
B Maps of water consumption footprint and areal productivity
77
C Marginal resource cost curves
85
D Maps of areal productivity benefits of irrigated biomass cultivation 93
8
List of Figures
2-1
RFS2 mandated production volumes. Adapted from McConnachie [71]
with the author's permission. . . . . . . . . . . . . . . . . . . . . . .
21
3-1
Lifecycle steps and system boundary of lifecycle analysis methodology.
35
3-2
A representative complex oil refinery, adapted from Chevron [30].
. .
42
3-3
Water flow in a typical North American refinery, adapted from Wu et
al. [12 3] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
43
Simplified flow diagram of the FT process, water consumption highlighted . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
47
Simplified process flow diagram of sugarcane milling pretreatment, water consumption highlighted. . . . . . . . . . . . . . . . . . . . . . . .
49
Simplified process flow diagram of corn grain milling pretreatment,
water consumption highlighted. . . . . . . . . . . . . . . . . . . . . .
50
Simplified process flow diagram of switchgrass dilute acid pretreatment, water consumption highlighted. . . . . . . . . . . . . . . . . . .
50
Simplified process flow diagram of monomer sugar metabolism and
upgrading to drop-in fuel, water consumption highlighted. . . . . . .
51
Simplified process flow diagram of HEFA MD process, water consumption highlighted. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
57
Lifecycle blue water consumption of a) FT and rainfed, and b) irrigated
MD production pathways. The conventional MD pathway is shown
for the purposes of comparison. Results shown are calculated by this
analysis unless otherwise cited. Calculated and literature results are
compared on the basis of liters of diesel equivalent. . . . . . . . . . .
63
Lifecycle blue water consumption footprint and areal productivity of
rainfed and irrigated corn AF MD production. . . . . . . . . . . . . .
65
3-4
3-5
3-6
3-7
3-8
3-9
4-1
4-2
9
4-3
4-4
4-5
4-6
a) Marginal blue water consumption of rainfed and irrigated corn AF
MD production, counties ranked to minimize water requirements. b)
Land requirements of rainfed and irrigated corn AF MD production,
counties ranked to minimize land requirements. . . . . . . . . . . . .
Areal productivity benefit of irrigation for corn AF MD production. .
Areal productivity benefit of irrigation for alternative MD production.
MD production pathway with the greatest areal productivity benefit
from irrigation in each county. . . . . . . . . . . . . . . . . . . . . . .
B-1 Lifecycle blue water
rainfed and irrigated
B-2 Lifecycle blue water
rainfed and irrigated
consumption footprint and areal
sugarcane AF MD production. .
consumption footprint and areal
corn AF MD production. . . . .
productivity
. . . . . . .
productivity
. . . . . . .
of
. .
of
. .
B-3 Lifecycle blue water
rainfed and irrigated
B-4 Lifecycle blue water
rainfed and irrigated
B-5 Lifecycle blue water
rainfed and irrigated
consumption footprint and areal
switchgrass AF MD production.
consumption footprint and areal
soybean HEFA MD production.
consumption footprint and areal
rapeseed HEFA MD production.
productivity
. . . . . . .
productivity
. . . . . . .
of
. .
of
. .
66
68
69
69
78
79
80
81
productivity of
. . . . . . . . .
82
B-6 Lifecycle blue water consumption footprint and areal productivity of
rainfed and irrigated jatropha HEFA MD production. . . . . . . . . .
83
C-1 a) Marginal blue water consumption of rainfed and irrigated sugarcane
AF MD production, counties ranked to minimize water requirements.
b) Land requirements of rainfed and irrigated sugarcane AF MD production, counties ranked to minimize land requirements. . . . . . . .
C-2 a) Marginal blue water consumption of rainfed and irrigated corn AF
MD production, counties ranked to minimize water requirements. b)
Land requirements of rainfed and irrigated corn AF MD production,
counties ranked to minimize land requirements. . . . . . . . . . . . .
C-3 a) Marginal blue water consumption of rainfed and irrigated switchgrass AF MD production, counties ranked to minimize water requirements. b) Land requirements of rainfed and irrigated switchgrass AF
MD production, counties ranked to minimize land requirements.
. . .
86
87
88
C-4 a) Marginal blue water consumption of rainfed and irrigated soybean
HEFA MD production, counties ranked to minimize water requirements. b) Land requirements of rainfed and irrigated soybean HEFA
MD production, counties ranked to minimize land requirements. . . .
10
89
C-5 a) Marginal blue water consumption of rainfed and irrigated rapeseed
HEFA MD production, counties ranked to minimize water requirements. b) Land requirements of rainfed and irrigated rapeseed HEFA
MD production, counties ranked to minimize land requirements. . . .
90
C-6 a) Marginal blue water consumption of rainfed and irrigated jatropha
HEFA MD production, counties ranked to minimize water requirements. b) Land requirements of rainfed and irrigated jatropha HEFA
MD production, counties ranked to minimize land requirements. . . .
D-1 Areal productivity benefit of irrigation for sugarcane AF MD production.
D-2 Areal productivity benefit of irrigation for corn AF MD production. .
D-3 Areal productivity benefit of irrigation for switchgrass AF MD productio n . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
D-4 Areal productivity benefit of irrigation for soybean HEFA MD production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
D-5 Areal productivity benefit of irrigation for rapeseed HEFA MD production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
D-6 Areal productivity benefit of irrigation for jatropha HEFA MD production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
11
91
94
95
96
97
98
99
THIS PAGE INTENTIONALLY LEFT BLANK
12
List of Tables
Oil production, water injection and technology share of various recovery technologies from Wu et al. [123] . . . . . . . . . . . . . . . . . .
39
3.2
Water use during oil extraction and recovery from Wu et al. [123] . .
40
3.3
Crude oil and finished fuel transportation assumptions for the conventional MD pathway from GREET 2011 [62]. . . . . . . . . . . . . . .
44
Indirect blue water consumption footprint of primary energy carriers
from G leick [49]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
44
Natural gas and coal feedstock transportation assumptions for the FT
MD pathway from GREET 2011 [62]. . . . . . . . . . . . . . . . . . .
45
3.6
FT MD transportation assumptions from GREET 2011 [62].
. . . . .
47
3.7
Variability of blue water consumption of FT MD pathways . . . . . .
48
3.8
AF feedstock transportation assumptions from GREET 2011 [62].
48
3.9
Data sources for AF MD feedstock-to-fuel process parameters. ....
52
3.10 Feedstock-to-fuel process variability for blue water consumption of AF
M D pathways. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
53
3.1
3.4
3.5
.
3.11 Variability of blue water consumption of AF MD pathways due to
assumed irrigation practices and location of biomass and fuel production. 54
3.12 HEFA biomass transportation assumptions from GREET 2011 [62].
.
55
3.13 HEFA oil feedstock transportation assumptions from GREET 2011 [62]. 56
3.14 Blue water consumption of the HEFA MD feedstock-to-fuel process
from Pearlson et al. [94] . . . . . . . . . . . . . . . . . . . . . . . . .
57
3.15 Process variability of blue water consumption of HEFA MD and biodiesel
pathways. Biomass growth input, oil extraction input and oil yield data
from Stratton et al. [112] . . . . . . . . . . . . . . . . . . . . . . . . .
59
3.16 Variability of blue water consumption of HEFA MD and biodiesel pathways due to assumed location of biomass and fuel production. ....
60
4.1
Results for conventional and FT pathways [lwater/lfuel].
13
. . . . . . . .
62
4.2
4.3
Results for biomass feedstock pathways, rainfed and irrigated cultivation shown [iwater/lfuel]. . . . . . . . . . . . . . . . . . . . . . . . . .
Lifecycle blue water consumption footprint and areal productivity of
10 billion liters of MD production from each pathway, averaged over
three different assumptions. . . . . . . . . . . . . . . . . . . . . . . .
A. 1 Blue water consumption by lifecycle step for conventional MD, FT and
rainfed MD production pathways [lwater/lMD . . . . . . . . . . ..
A.2 Blue water consumption by lifecycle step for irrigated MD production
pathways [lwater/lMD] . . . . . . . . . . . . . . . . . . . . . . . . . . .
14
62
67
74
75
Chapter 1
Introduction
Transportation fuels are a fundamental part of the energy system: in 2011 the transportation sector accounted for 27 exajoules (28%) of total primary energy consumption in the United States (US) [8]. Under ideal market conditions the consequences
of that consumption should guide consumers' actions by being reflected in the price,
however imperfections in the transportation fuels market mean that this is not necessarily the case [118]. One attempt to alleviate the problems associated with these
imperfections is policy intended to encourage the production and consumption of
domestically produced renewable fuels [16]. In practice, policies may become mired
in political and institutional failure, leading to economically inefficient or social inequitable outcomes. Undesirable outcomes could include unforeseen environmental
externalities.
Alternative drop-in middle distillate (MD) fuels, including diesel and jet fuels
chemically similar to conventional petroleum-derived MD, and biodiesel, are of particular interest. Unlike ethanol, alternative drop-in diesel and biodiesel are compatible
with existing diesel trucks, automobiles, railroad locomotives and agricultural machinery, and alternative drop-in jet fuel is compatible with turbojet and turboprop
aircraft engines [40, 51]. The demand for diesel and jet fuel for these purposes makes
up almost 30% of liquid fuel consumption in the US [11].
Although alternative MD fuels hold promise in terms of reduced lifecycle greenhouse gas (GHG) emissions and air quality compared to conventional MD [25, 65, 112],
the overall environmental impact of commercial scale production, specifically in terms
of fresh water and land resource requirements, is not well understood [27, 37, 43, 59,
99, 103, 115, 124]. Despite uncertainty about the environmental impacts, production
of alternative MD fuels, encouraged by policy, is expected to grow: the International
Air Transport Association (IATA) has a goal of 10% alternative fuel use for avia15
tion by 2017 [55], and the US Federal Aviation Administration (FAA) has a goal
of 1 billion gallons of alternative fuel consumption by 2018 [12]. Additionally, the
Renewable Fuels Standard 2 (RFS2) mandates 36 billion gallons of alternative fuel
production by 2022, 21 billion gallons of which could be alternative MD or biodiesel
[17]. Therefore, in order to determine overall environmental sustainability, the water
consumption and land resource requirements of alternative MD production should be
considered.
This thesis begins by providing an analysis of US renewable fuels policy. In Chapter 2, the history of renewable fuels regulation and implementation in the US, and
the mechanics of current federal policy, RFS2, are reviewed. Next, RFS2 is critically
considered as an attempt to address a number of imperfections in the transportation
fuels market, including: economic vulnerability to dependence on oil from oligopolistic foreign sources; security and environmental externalities; and a lack of investment
in the development of alternative technologies [16]. These market imperfections, the
way in which policy attempts to address them, and the short-comings of RFS2 in
doing so, are all discussed through a political economy lens. The fact that RFS2 does
not consider environmental impacts beyond lifecycle GHG emissions, such as water
and land resource requirements, is of particular concern.
In order to more completely evaluate environmental sustainability, Chapter 3
quantifies the water consumption and areal land resource requirements of alternative
MD fuel production. The application of blue water for irrigation, which Falkenmark &
Rockstr6m [42] define as fresh water from surface and underground sources, increases
the areal productivity of alternative MD fuels derived from biomass. However, blue
water consumed during the MD lifecycle may compete with other uses, and could
limit the total production potential of alternative fuels [98, 115]. Quantification of
water and land resource requirements will help to inform resource allocation decisions
related to the development of alternative MD production technologies and facilities,
as well as the development of policy which considers a more complete definition of
environmental sustainability. These parameters are calculated for a number of MD
production pathways, under assumptions of rainfed and irrigated biomass cultivation,
including:
" Conventional MD production;
" Fischer-Tropsch (FT) MD from natural gas and coal;
" Fermentation and advanced fermentation (AF) MD from sugarcane, corn and
switchgrass;
16
"
Hydroprocessed esters and fatty acids (HEFA) MD from soybean, rapeseed and
jatropha; and
" Biodiesel from soybean, rapeseed and jatropha.
The results of this analysis are presented in Chapter 4. In addition to reporting a
range of values in order to capture variability in each pathway, results are geospatially
disaggregated, and the trade-offs between fresh water and land resource requirements
are quantified. Finally, Chapter 5 provides conclusions and suggestions for future
work.
17
THIS PAGE INTENTIONALLY LEFT BLANK
18
Chapter 2
Policy context
2.1
History
The use of non-petroleum transportation fuels is not a new concept in the US. The
first vehicle built by Henry Ford in 1896 ran on pure ethanol, and the Model T,
first produced in 1908, was designed as a flex-fuel vehicle capable of running on
both gasoline and ethanol [4]. At the time, ethanol was considered to be a significant
potential source of fuel for transportation, but as the market for liquid transportation
fuels grew with the proliferation of personal vehicles, petroleum-derived fuels emerged
as the dominant technology [87].
US renewable fuel production was minimal until World War I, when the use of
ethanol as a petroleum substitute spiked due to acute petroleum fuel shortages. From
1914 to 1918 ethanol production quintupled from 10 to 50 million gallons per year,
and following an interwar lull, peaked again at 600 million gallons per year during
World War II [79]. Ethanol production declined after the end of World War II until
the oil embargo by the Organization of Petroleum Exporting Countries (OPEC) in
1973 and the interruption of Iranian oil imports in 1978 exposed US dependence on
crude oil imports, and gave new life to interest in domestically produced renewable
fuels [87]. In 1978, the federal government passed the Energy Tax Act, which exempted ethanol from the $0.04 gasoline excise tax [87]. This policy was introduced
in response to the sudden increase in the price of oil and came from a desire to reduce dependence on imported oil by meeting a portion of demand for transportation
fuels with domestically produced ethanol [74]. In effect, the tax exemption was an
attempt to lower costs for domestic producers in order to support fuels that reduced
dependence on imported oil. Since 1978 the federal government has enacted a number
19
of additional legislative policies to encourage renewable fuel production. Most have
taken the form of tax credits, tax exemptions, or loan guarantees specific to ethanol.
For example, the Energy Security Act of 1980 provided loan guarantees and insurance for construction of new ethanol facilities [1]; and the Tax Reform Act of 1984
increased ethanol subsidies to $0.60 per gallon [28]. More recently, US renewable fuels
policy has taken the form of mandated production volumes. The Energy Policy Act
of 2005 included the Renewable Fuels Standard (RFS), which mandated the blending
of cornstarch ethanol with gasoline [16].
Up until this point in the history of renewable fuels, the primary motivation
for policy development was its potential to reduce dependence on imported oil [74].
However, as concerns about climate change have grown it has been suggested that
renewable fuels, such as ethanol and biodiesel, could offer an improvement in environmental performance over conventional petroleum fuels [33]. While the combustion of
conventional petroleum fuels produces CO 2 which had been removed from the atmosphere and geologically sequestered millions of years ago, renewable fuel combustion
produces CO 2 only recently removed from the atmosphere during photosynthesis. It
follows from this logic that renewable fuels could result in fewer net CO 2 emissions.
One means of quantifying the actual environmental performance of renewable
fuels is the use of lifecycle analysis (LCA). LCA is a "cradle-to-grave" accounting
methodology, applied in this case to determine the entire GHG footprint of a fuel by
adding up the energy and material inputs at each step of the fuel lifecycle, including
feedstock cultivation, feedstock transportation, conversion of feedstock to fuel, fuel
transportation and distribution, and fuel combustion. LCA and net energy analyses
have demonstrated that the GHG impact of renewable fuels is highly dependent upon
the technologies employed for production [122, 96], and US renewable fuels policies,
up to and including RFS, did not take this into account.
Current US renewable fuels policy is largely defined by the Energy Independence
and Security Act (EISA) of 2007, which expanded the RFS program to what is now
RFS2.
2.2
Mechanics of RFS2
RFS2 expands upon RFS in a number of ways. First, it defines two broad categories
of mandated renewable fuels: conventional renewable fuel (predominantly cornstarch
ethanol) and advanced biofuels. Within advanced biofuels, three types of fuel are
20
distinguished: biomass-based diesel, cellulosic biofuel and undifferentiated advanced
biofuel [97].
In order to qualify under RFS2, a renewable fuel must provide a GHG benefit
over the petroleum fuel that it replaces [16]. Using LCA, the US Environmental
Protection Agency (EPA) must determine if a renewable fuel satisfies a threshold of
GHG performance, defined as a required reduction in lifecycle emissions compared to
petroleum fuels [16]:
" Biomass based diesel - 50% reduction
" Cellulosic biofuel - 60% reduction
" Undifferentiated advanced biofuel - 50% reduction
" Conventional renewable fuel - 20% reduction
RFS2 also increases the mandated volumes for blending with transportation fuels
from a total of 9 billion gallons in 2008 to 36 billion gallons in 2022 [16]. 15 billion
gallons are to come from conventional renewable fuels, and the remaining 21 billion
gallons are to come from advanced biofuels. Almost all of the mandated increase in
renewable fuel production must come from advanced biofuels. Figure 2-1 shows the
mandated RFS2 production volumes, broken out by type of renewable fuel.
40
35
U)
C
-
30
25
20
C
15
0 Biomass Based Diesel
-
'eli
C
Advanced biofuels
---------------------------------
-
-
I
* Undifferentiated Advanced:
Biofuel
14
10
j Cellulosic Biofuel
----------------------------------*Conventional Renewable
Fuel (Corn Ethanol)
0
YV
ea
Year
Figure 2-1: RFS2 mandated production volumes. Adapted from McConnachie [71]
with the author's permission.
21
In order to implement the RFS2 production volume mandates, the EPA defines
obligated parties that must meet a renewable volume obligation (RVO). An obligated
party is any producer or importer of gasoline or diesel fuel [14]. Obligated parties'
RVOs are calculated on the basis of EPA's renewable fuel standard, or blend ratio,
for a given year, and the volume of non-renewable gasoline or diesel that is produced
or imported by the obligated party in a given year. Each obligated party's RVO is
calculated as follows [72]:
RVOi = (RFStdi - GVi) + Di_ 1 , where:
RVOi = The RVO for an obligated party for calendar year i
RFStdi = The renewable fuel standard for calendar year i, determined by EPA (percent).
GVi = The nonrenewable gasoline and diesel volume, which is produced or imported
by the obligated party in calendar year i (gallons).
Di_1 = Renewable fuel deficit or carryover from the previous year (gallons).
In order to monitor obligated parties' compliance with their RVO, the EPA developed the concept of Renewable Identification Numbers (RINs) [97]. The EPA assigns
a unique RIN to each gallon of renewable fuel produced that qualifies under RFS2.
Each RIN is tied to a specific gallon of renewable fuel until that gallon is blended
with non-renewable fuel, at which point the fuel and the RIN are separated, and can
be bought and sold separately. At the end of each reporting period the obligated
parties must report RINs corresponding to their RVO to the EPA, and will be fined
if they fail to do so. Obligated parties that do not physically blend sufficient volumes of renewable fuels have the option of purchasing separated RINs to meet their
RVO. Obligated parties with excess RINs may trade them for a profit. In theory,
obligated parties should be willing to pay up to the value of the EPA fine for RVO
non-compliance for any shortfall in RINs at the end of the reporting period [97].
2.3
Transportation fuels market imperfections &
RFS2
In general, justification of renewable fuels policy is discussed in terms of a number of imperfections of the global transportation fuels market, including: economic
vulnerability to dependence on oil from oligopolistic foreign sources; security and
22
environmental externalities; and a lack of innovation and investment in alternative
technologies [33, 74]. Indeed, the EPA's website states that "RFS2 lays the foundation for achieving significant reductions of greenhouse gas emissions from the use of
renewable fuels, for reducing imported petroleum, and encouraging the development
and expansion of our nation's renewable fuels sector" [16]. The following section
of this chapter discusses the imperfections in the transportation fuels market which
RFS2 purports to correct, and critically assesses the policy's effectiveness at doing
so.
2.3.1
Imperfect competition & economic vulnerability
RFS2 is justified in part by a desire wean the US off of imported crude oil, thereby
limiting the economy's vulnerability to oligopolistic behavior, and to high, volatile
prices in the global transportation fuels market [16]. In 2011, the US consumed an
average of 18.8 million barrels of oil per day and domestic production averaged 10.1
million barrels per day. The shortfall of 7.7 million barrels per day, over 40% of US
crude oil consumption, was met by the import of crude oil and refined fuel products
[6]. OPEC countries account for approximately 40% of global crude oil production,
and such a significant concentration of market power raises the possibility of collusive
or oligopolistic behavior to manipulate crude oil prices [6]. EIA describes a number
of ways in which higher oil prices, potentially due to collusion of OPEC member
countries, can adversely affect the US economy [7]:
" Consumers must spend more on fuel, and have less disposable income to spend
on other goods and services. A large proportion of the additional money spent
on fuel flows out of the US economy to foreign oil producers.
" The cost of production goes up for many sectors of the economy, which has the
effect of reducing overall economic output.
" A reduction in real wages due to the increased cost of goods, combined with
"stickiness" or friction in firms' willingness to reduce prices in light of this, slows
adjustment to the new economic conditions, and results in welfare loss.
" The price of other forms of energy increases because transportation fuels are
an input for their production, and substitution effects drive up the price of
petroleum alternatives.
23
The US Department of Energy (DOE) estimates that between 2004 and 2008,
oil price shocks, price manipulation, and hedging against price fluctuation by the
US to avoid these effects, cost the economy approximately $1.9 trillion [86]. This is
equivalent to approximately 10% of gross domestic product annually [13]. Proponents
of RFS2 argue that imperfect competition in the global transportation fuels market,
due to OPEC's market power and collusive behavior, is harmful for the US economy,
and therefore the policy seeks to substitute a portion of foreign sourced petroleum
consumption with mandated renewable fuel RVOs [48, 16].
However, it is not clearly demonstrable that OPEC has been able to use market
power to manipulate energy prices [18, 50, 105, 20]. Casting of RFS2 as a means to
reduce dependence on anti-competitive markets is problematic because crude oil and
transportation fuels markets may actually exhibit competitive behavior.
Regardless of the reasons for high, volatile crude oil prices, RFS2 may not be
effective at reducing the US economy's vulnerability to these effects in any case. In
2011 the US produced approximately 1 million barrels of renewable fuels per day,
equivalent to approximately 5% of crude oil consumption by volume [15]. If the 2022
RFS2 production mandates are met and US crude oil consumption stays constant, renewable fuels will offset only 12% of consumption by volume. Because renewable fuels
represent a relatively small proportion of total transportation fuel demand, the US
economy will ultimately remain vulnerable to high, volatile crude oil prices. Furthermore, 1-2 million barrels of crude oil consumption, offset by renewable fuels, would
represent a relatively small change in demand for the 90 million barrel per day global
crude oil market [9]. It is unlikely that a 1-2% reduction in demand would have a
significant effect on prices set in a global market.
In reality, RFS2 may actually have the effect of raising transportation fuel prices.
Instead of a tax exemption or loan guarantee, which implicitly prescribes the value
of renewable fuel production to society and offers the producer a compensation in
that amount, the EPA has set up the RIN market. This encourages the lowest cost
producers to profitably produce renewable fuels, up to the point at which the RFS2
volume mandates are met. In theory, obligated parties should be willing to pay any
price for RINs, up to the point at which it would be less expensive to pay the EPA fine
for non-compliance. Assuming that renewable fuel producers are rational economic
actors, the cost of compliance, despite being set at the lowest possible cost by the
RIN market, must be greater than zero or else the renewable fuel would have already
been in production. The result, as obligated parties spread the additional cost of
RINs over their entire renewable/non-renewable fuel blend product, is that RFS2 has
24
the effect of subsidizing renewable fuels with higher prices for transportation fuels in
general.
In addition to the economic efficiency implications of higher transportation fuel
prices, there are equity consequences tied to policy instruments such as RFS2 [89].
Increased fuel prices could drive up the costs of production, consumer goods, and food
through increased demand for agricultural commodities for fuel production. These
types of impacts are manifest in small price increases for basic goods, spread over
the entire population. Because price increases in basic goods represent a larger proportion of income for the economically disadvantaged in society, and may amplify
wealth disparity, RFS2 is a potentially regressive policy. The concept of Olsonian
selective interests is useful here [88]. Because the US population's individual interests are diffuse in the context of an increase in transportation fuel prices, a collective
action problem arises. If the private costs of organization to lobby against higher
prices out-strip the costs imposed by RFS2, the general population's interests may
be underrepresented in the development and implementation of the policy.
Justification of RFS2 on the basis of reducing US economic vulnerability to imperfect competition may not be warranted. Oligopolistic behavior is not definitively
observable in the global crude oil market, and RFS2 does relatively little to insulate
the US economy from high, volatile prices. Rather, RFS2 may have the effect of
raising prices, thereby exacerbating the economic effects of expensive transportation
fuels that the policy aims to alleviate.
2.3.2
Security and environmental externalities
US dependence on imported oil incurs significant security costs external to the price
of transportation fuels. Although partially internalized by US society as a whole,
this externality raises concerns about the equitable distribution of the security costs
associated with transportation fuels [32].
Energy security is clearly incorporated into military objectives, and the US military expends significant effort to secure access to foreign crude oil resources [85].
Dancs et al. [34] estimate this expenditure at $90.2 billion in 2010, or $166.3 billion
if the Iraq war is included. The RAND Corporation estimates that between $75.5
billion and $91 billion could be cut from the annual US defense budget by removal of
missions to defend oil supplies and sea lines of communication from the Persian Gulf
[32]. $75.5 billion annually corresponds to an increase of $1.80, or approximately 51%
of the price of a gallon of gasoline. Although there is uncertainty about the precision
25
of these estimates, the magnitude of the costs are clearly significant when compared
to the current price of transportation fuels. From the consumer's perspective, however, there is no distinction between domestic and imported fuel products. A gallon
of gasoline produced in Texas will command the same price as a gallon produced in
Saudi Arabia.
These issues raise questions of both negative externalities and the equitable distribution of costs. Instead of being paid directly by consumers, the security cost of
ensuring access to imported oil is borne by society at large through taxes used to
fund defense budgets, and may therefore be considered a negative externality. However, since practically all US taxpayers consume transportation fuels either directly
or indirectly, the costs are at least partially borne by the end user. The equity issue
comes into play because the value of taxes paid by individuals are not a function of
consumption (excepting those placed directly on transportation fuels), and therefore
taxpayers who use more transportation fuel are subsidized by those who use less. This
is another example of an Olsonian collective action problem [88]. The diffuse interests
of taxpayers who use less transportation fuel mean that the inequitable distribution
of security costs will persist over time.
This is a situation that invites policy intervention, however RFS2 does little to
address these concerns despite a stated goal of "reducing imported petroleum" [16].
As shown above, RFS2 has the potential to offset a relatively small proportion of
total petroleum consumption, meaning the US will likely continue to import crude
oil and transportation fuels.
Additionally, environmental costs are not reflected in the price of transportation
fuels. For example, the production of crude oil in Nigeria often involves venting of
natural gas to the atmosphere. Natural gas is a more potent GHG than C0 2 , meaning that the purchase and combustion of transportation fuels derived from Nigerian
crude has a larger GHG footprint than fuels derived from other sources of crude oil:
gasoline derived from Nigerian crude has a well-to-wheels GHG footprint of 105.6
gCO2e/MJgaso1ine, compared to the 2005 US EPA average of 91.2 gCO2e/MJgasoline
[64]. Despite this, there is no price differentiation between gasoline derived from
Nigerian and US-average crude oil.
A number of other important environmental externalities are not reflected in the
price of transportation fuels. These include the risk of ecological damage due to
tanker spills; fresh water use, especially in water-intensive production operations such
as steam assisted gravity drainage (SAGD) [123]; and land disturbance [57]. The environmental externalities of transportation fuels are another example of an Olsonian
26
collective action problem, as diffuse costs are spread amongst a diverse set of individuals. Climate change is an extreme example of an Olsonian collective action problem,
because the costs of GHG emissions are global in nature. Other environmental costs,
although more localized, may also be spread amongst individuals unable to organize
to lobby in their own interests.
RFS2 attempts to address the collective action problem of climate change by
mandating the production of fuels with a lifecycle reduction in GHGs from conventional fuels. However, RFS2 does not consider any other environmental externalities
associated with transportation fuels. This is particularly concerning because the commercial scale production of renewable fuels can have non-GHG related environmental
effects far more severe than conventional petroleum fuels. For example, renewable
fuel production is potentially many times more water and land intense than conventional petroleum fuel production [123]. By considering only the climate change externalities of transportation fuels, RFS2 may have the effect of stimulating renewable
fuel production that is, in aggregate, more environmentally costly than conventional
petroleum-derived fuel production. Quantification of the water and land resource
requirements of transportation fuels is the subject of Chapters 3 and 4.
From an economic efficiency point of view, RFS2 is not the optimal policy design to
deal with the security and environmental externalities associated with transportation
fuels [58]. As a command and control type regulation, RFS2 defines RVO standards
and subsidizes renewable fuel production through the RIN market. Rather than subsidize the technology option with smaller security and environmental externalities, it
would be more efficient to internalize the externalities associated with all transportation fuel production options. That way, by including all of the costs in the price, the
market could determine the lowest cost production option that results in the greatest
welfare creation. In the case of security costs this could take the form of a tax on
petroleum fuels dedicated to covering the costs associated with their secure supply.
For climate change externalities, the most economically efficient policy would likely
be a carbon tax or cap-and-trade system for GHGs that, as a market-oriented form
of regulation, would allow trading between sectors. Despite potentially being the
lowest cost options, taxation and cap-and-trade policies have lacked enough stakeholder support due to non-economic considerations [58]. RFS2 may simply represent
a practicably implementable policy in the near-term [58].
27
2.3.3
Innovation & industry
An incumbent technology may inhibit development of alternative technologies due to
the externalities and large, risky investments associated with technological innovation.
This can lead to a non-optimal diversity of research and development efforts [2]. This
section considers RFS2 as an attempt to subsidize the development of a renewable
fuels industry to overcome these barriers to innovation, and discusses the political
and institutional challenges that entails.
Investment in innovation creates positive externalities by allowing private actors,
other than those who made the initial investment, to profit from the fruits of others' labor. The patent system exists precisely for the purpose of internalizing these
externalities: innovators are more willing to invest significant resources to innovate,
secure in the knowledge that they will receive a legally sanctioned monopoly on their
innovation for 20 years. Although the patent system makes great gains towards internalizing the positive externalities associated with innovation, Acemoglu argues that
it fails to do so completely [2]. First, subsequent innovators may be able to build
upon the publicly available patent information of the initial innovation, and being
able to meet the "new, useful, and non-obvious" criteria of the US Patent & Trademark Office, may not even be required to license the innovation that was the source
of inspiration. Second, the 20 year life of patents means that any profits to be realized after that point are likely to be external to the innovator who made the initial
investment. While Acemoglu recognizes that these features of the patent system are
both intentional and desirable, his contention is that they lead to a "lack of diversity
in investment" nonetheless [2].
If a profit-maximizing firm, subject to resource constraints and with a fiduciary
duty to it's shareholders, must decide to invest in either innovation of fossil fuels or of
renewable fuels, it will weigh the profits to be realized from each type of investment
against each other. Considered in this calculation will be the fact that profits on
innovation of the incumbent technology (fossil fuels in this case) can be realized
immediately, while profits on innovation of an "alternative" technology (renewable
fuels) may have to wait until consumer preferences or stable government policy evolves
to create an attractive investment environment. In addition, the firm will likely take
into account the fact that subsequent innovations in the "alternative" technology may
render their innovation obsolete before any profits can be realized. In aggregate, these
effects create barriers to innovation.
RFS2 seeks to overcome these barriers to innovation by encouraging and stimulating a diversity of investment in innovation in the transportation fuels industry. This
28
is evidenced by the RVO production mandates. The majority of growth in renewable
fuel production volumes is to come from advanced biofuels, and in particular cellulosic biofuels. By defining distinct classes of renewable fuels, and capping corn ethanol
production specifically, RFS2 legally mandates investment in innovation of at least
four diverse technologies. Furthermore, RFS2 subsidizes innovation in transportation
fuels through the RIN market. From the governmental perspective, RFS2 should aid
the renewable fuels industry in achieving efficiency gains though learning-by-doing
effects and economies of scale. In theory, the support required by industry should
taper off over time and the industry should become self-sustaining.
Under RFS2, however, the advanced biofuels industry has struggled to become
self-sustaining. The original RFS2 mandated volumes of cellulosic biofuels were 100,
250 and 500 million gallons in 2010, 2011 and 2012, respectively. Due to lack of production, the EPA was forced to reduce the required volumes to 6.5, 6.6 and 8.65 million gallons in those years, respectively [97]. Inflated RFS2 production mandates, as
evidenced by the need for the EPA to drastically reduce the mandated volumes, may
be an example of Stiglerian regulatory capture, where the renewable fuels and agriculture industries have influenced regulation for their own economic interests [111]. A
number of organized lobby groups, such as the Biotechnology Industry Organization,
the National Biodiesel Board, the Renewable Fuels Association, and the Alliance for
Abundant Food and Energy represent the interests of the biofuels and agriculture industries in Washington, DC. Inflated production volume mandates, which guarantee
a market for their products, could indicate that their lobbying efforts had an effect
on the final form of RFS2 that became law with the passage of EISA.
The American Petroleum Institute (API), on the other hand, represents the interests of the petroleum industry and is a particularly well funded and influential
lobby group [47]. API has taken considerable effort to undermine the validity of
RFS2, including legal action [68]. Renewable fuels represent a fungible product that
may reduce demand for petroleum fuels, and the additional costs associated with
the purchase of RINs and fuel blending operations may squeeze profit margins. On
top of this, petroleum refining (the obligated party under RFS2) is one of the least
profitable sectors of the conventional fuel value chain: from 2006 to 2009 the eight
firms which account for 50% of US refining capacity saw their net income shrink by
79% due to high crude prices, weak demand due to large inventories of finished fuels,
and a shrinking price spread between light and heavy crude [21]. In order to legally
reduce the mandated volumes of RFS2 cellulosic biofuel in 2010, 2011, and 2012, the
EPA had to find evidence that compliance with the mandates would cause "severe
economic harm" to the obligated parties. API's ability to influence the EPA in rec29
ognizing "severe economic harm" to their economic interests through lobbying and
legal action may represent another example of Stiglerian capture.
If it is true that Stiglerian capture has influenced the way in which RFS2 stimulates innovation in the renewable fuels industry, the organized interests on opposing
sides of the policy debate are an interesting study in the Olsonian concept of selective
interests [88]. The common, concentrated selective interests of firms in the renewable
fuels and agriculture industries mean that they have been able to guarantee a market
for their products through a sustained lobbying effort. The API, a lobbying effort diametrically opposed to the interests of the renewable fuels and agriculture industries,
has also emerged. Interestingly, the organization of groups with selective interests in
opposition to each other may have a positive outcome in this case. The two groups
may effectively offset one another's lobbying efforts, and allow the EPA the latitude
to develop a more socially beneficial policy. For RFS2, this is only likely if the diffuse
interests of the general public are sufficiently represented to address the collective
action problems described above, such as economic vulnerability, and security and
environmental externalities.
Finally, repeated reduction of the mandates has contributed to regulatory uncertainty, undermining RFS2's stated purpose of fostering innovation and development
of the renewable fuels industry. RFS2 was intended to provide renewable fuels producers with a guaranteed market for their product, subsidized by the RIN market.
However, the continual reduction of the mandates sends a signal that erodes firms'
and investors' confidence in the long-term existence of that market, and hinders development of the industry.
RFS2 aims to stimulate investment in new, alternative technologies, and has had
moderate success in doing so. Unfortunately, the development of a self-sustaining renewable fuels industry may be hindered by Stiglerian capture in the development and
implementation of regulation, and by regulatory uncertainty caused by production
mandate reductions.
2.3.4
Conclusions
The chapter critically discusses the effectiveness of RFS2 at addressing: US economic
vulnerability to energy from oligopolistic foreign sources; the security and environmental externalities of transportation fuels; and a lack of investment in the development of alternatives to petroleum-derived fuels. Although RFS2 defines thresholds for
lifecycle GHG emissions, it is the only metric of environmental performance that the
30
policy considers. By doing so, RFS2 incentivizes the production of alternative fuels
irrespective of their non-GHG environmental impacts, even though the production
of alternative fuels may be resource-intense, particularly in terms of fresh water and
arable land resources. The remainder of this thesis quantifies these metrics in order
to develop a more complete picture of the environmental performance of alternative
fuel production.
31
THIS PAGE INTENTIONALLY LEFT BLANK
32
Chapter 3
Analysis of water consumption and
land resource requirements
Chapter 2 establishes the policy context for US alternative fuel production by describing the ways that RFS2 seeks to address imperfections in the transportation
fuels market, and by discussing the policy's short-comings in doing so. One deficiency of RFS2 is that it does not fully consider the environmental sustainability of
alternative fuel production. Although RFS2 mandates the production of alternative
fuels with lifecycle GHG emissions lower than conventional fuels, there is no consideration of non-climate related environmental issues. Chapter 3 turns to an analysis of
two factors key to developing a more complete picture of the environmental sustainability of alternative fuel production: blue water consumption and areal land resource
requirements. Specifically, the analysis focuses on alternative MD fuel technologies,
due to expected growth in their production.
3.1
Methodology
The blue water consumption footprint of each feedstock-to-fuel pathway is calculated
on a lifecycle basis, taking into account the consumption of blue water in each step.
In order to understand the drivers of variability in these results, the AF and HEFA
pathway results are geo-spatially disaggregated at a county resolution for the contiguous US. Biodiesel results are not geo-spatially disaggregated because they agree
with the HEFA results within ±6%for each county. Marginal resource cost curves are
constructed for the AF and HEFA pathways, ranked by the blue water consumption
footprints and land requirements of MD production, and three different assumptions
33
are tested to quantify the trade-offs between blue water consumption and land use
requirements for 10 billion liters of MD production from each pathway. 10 billion
liters was selected for comparison because, given the constraining assumptions, it is a
volume of fuel that could be produced by each AF and HEFA MD pathway. Finally,
the areal productivity benefit of irrigation is calculated on a county basis to quantify which regions of the contiguous US will realize the greatest benefit from biomass
irrigation for AF and HEFA MD production.
3.1.1
Water use definitions
Water used within the MD production lifecycle comes from precipitation, surface and
underground sources. Fresh water from precipitation is categorized as green water,
fresh water from surface and underground sources is categorized as blue water, and
polluted water is categorized as grey water [42]. Water that exits a defined system
boundary via direct consumption, evaporation or evapotranspiration, and is no longer
available for use, is considered consumed and may be either blue or green water [49].
This analysis quantifies blue water consumed during the MD lifecycle.
3.1.2
Lifecycle methodology
In order to compare blue water consumption across pathways, a methodology consistent with the principles of the GHG lifecycle methodology described in Allen et
al. was adopted [19]. This analysis includes, when applicable: biomass cultivation, recovery and transportation; feedstock extraction, recovery and transportation;
feedstock-to-fuel conversion; and transportation and distribution of MD. Water vapor
released to the atmosphere during MD combustion is beyond the system boundary.
Within the system boundary two types of blue water consumption are accounted for
in this analysis: direct and indirect. Direct blue water consumption is water that
exits the system boundary during MD production. Indirect blue water consumption
is due to the blue water consumption footprints of the material and energy inputs to
MD production. The sum of the direct and indirect values is the total blue water
consumption footprint of the MD lifecycle. The system boundary considered in this
analysis is shown in Figure 3-1.
The material and energy outputs of each MD production pathway include products
(e.g. diesel, jet fuel and biodiesel), and co-products (e.g. animal feed, electricity
and non-MD fuels) or wastes. Total blue water consumption is allocated amongst
34
Land Use
R
Soybean
Rapeseed
Jatropha
Corn
Switchgrass
Sugarcane
Biomass
Transportation
Feedstock
Extraction
Feedstock
Transportation
MD Fuel
Production
MD Fuel
Transportation
MD Fuel
Combustion
Natural Gas
Coal
Crude Oil
System Boundary
Figure 3-1: Lifecycle steps and system boundary of lifecycle analysis methodology.
non-fuel co-products according to market allocation: at the point where the fueldestined product stream is physically separated from the co-product streams, the blue
water consumption is allocated amongst the process streams in proportion to their
relative market values [120]. The jatropha HEFA MD pathway is an exception to this
methodology, as blue water consumption is allocated to the electricity co-product by
energy allocation. The remaining blue water consumption is allocated amongst all fuel
products according to their relative energy contents [120], and results are reported in
terms of liters of blue water consumed per liter of MD produced [lwater/lMD]. These
allocation methods are consistent with previous lifecycle GHG emissions studies on
alternative MD production pathways [112].
For the purposes of this analysis, it is assumed that no significant direct blue water consumption is associated with the transportation of the biomass and fossil fuel
feedstocks, or finished fuel products of any pathway. This is consistent with previous studies on the water footprint of transportation fuels [99]. In order to quantify
indirect blue water consumption, the primary energy carriers associated with transportation, such as coal, natural gas and refinery products, were obtained from the
default assumptions in GREET 2011 [62]. The blue water consumption associated
with these primary energy carriers are from Gleick [49].
35
For the AF, HEFA and biodiesel pathways, blue water is consumed during biomass
cultivation due to evaporation and evapotranspiration of water applied to crops for
irrigation. Data on the blue water consumption of biomass cultivation was obtained
from the Global Agro-Ecological Zones (GAEZ) model [44], discussed in section 3.2
of this chapter.
A range of results is calculated for each pathway, reported as low, mid and high values. This is done to capture variability in each feedstock-to-fuel pathway, and depends
upon assumptions regarding the location of biomass cultivation, fossil fuel extraction
methods, feedstock-to-fuel conversion efficiencies, and process water requirements.
The low and high values represent the assumptions that generate the lowest and
highest results. The mid value is calculated from the combination of assumptions
most representative of the technology on the basis of engineering assumptions and
empirical data. The definitions of these of assumptions for each feedstock-to-fuel
pathway are described in section 3.3 of this chapter.
3.2
GAEZ model
For the pathways that use biomass feedstocks, the blue water consumption footprints
of feedstock cultivation were taken from the GAEZ v3.0 model. GAEZ calculates
geography and climate specific maximum attainable biomass yields. Crop water balances are used to estimate evapotranspiration, crop water deficit during the growth
cycle, irrigation water requirements to maximize yields, and the rainfed and irrigated
biomass yields. Crop water balances are used to estimate actual crop evapotranspiration, accumulated crop water deficit during the growth cycle, irrigation water
requirements for irrigated conditions, and the corresponding rainfed and irrigated
biomass yields. GAEZ takes into account yield reductions due to agro-climatic constraints, soil and terrain limitations; climate data; [77] soil data; [45] elevation, terrain
slope and aspect data; [90] and land cover data [29, 56, 114, 82, 91, 92, 104, 26, 63].
GAEZ also accounts for year-to-year average climatic and soil moisture variability
and yield losses due to disease, water stress, soil workability, and early or late frosts.
The blue water consumption and biomass yields for rainfed and irrigated conditions were extracted from GAEZ for all six biomass feedstocks of interest, at a 5
arc-minute and 30 arc-second resolution. This data was used to determine rainfed
yields, blue water consumption requirements to maximize yield, and maximized irrigated yield. This was calculated for each county in the contiguous US which GAEZ
36
determines is suitable, given soil and climatic conditions, for cultivation of the crop
of interest [44, 46].
3.2.1
Model validation
The GAEZ model has been verified both internally by IIASA, and by FAO's Economic, Social, and Agricultural Departments [93]. The crop yields and distributions
have been verified against national agricultural statistics [44]. Masutomi et al. [70]
demonstrate that simulated crop yields calculated by M-GAEZ, a derivative of the
GAEZ model, agree with actual historical crop yields within ±50%, which is considered an acceptable level of accuracy for crop models. The GAEZ model was used by
the Intergovernmental Panel on Climate Change (IPCC) for the Fourth Assessment
Report (AR4) for assessing food security under soil, terrain and climate constraints.
The Global Environmental Change and Food Systems (GECAFS) project compared
the GAEZ model against others with similar objectives. Notable projects and work
containing AEZ assessments include Wackernagel et al. [119], and the Comprehensive Assessment of Water Management in Agriculture (CAWMA) [78]. The main
limitation of the GAEZ model is that the quality and reliability of data sets are geographically heterogeneous. Furthermore, the data sets contain generalizations due
to the scale of the analysis, and the inclusion of socioeconomic trends to account for
land resource allocation is limited [44].
3.2.2
Irrigated blue water consumption calculation
Blue water consumption for irrigated biomass cultivation is calculated in the following
manner:
= Irrigated ETa -
* Blue water consumed by evapotranspiration (ETa)
Iyear_
Rainfed ETa;
" It is assumed that an average of 71% of applied irrigation water is consumed
by evapotranspiration [52], and that 5% of applied irrigation water is consumed
through evaporation [125]. Therefore, an additional 7% of blue water consumption from evapotranspiration is consumed directly by evaporation;
" Total blue water consumption = ETa + 0.07. ETa
37
3.3
Fuel production pathways
The following pathways were selected for the feasibility of commercial scale production
in the near-term. In 2011, almost 1 billion gallons of biodiesel were consumed for
ground transportation in the US [5]. An indicator of the feasibility of emerging
alternative MD technologies is certification by ASTM International. FT and HEFA
fuels have already been certified under ASTM D4054, and AF (specifically alcoholto-jet fuel) is expected to be the next to be certified [39].
3.3.1
Conventional MD
The conventional MD pathway lifecycle includes crude oil recovery, crude oil transportation, refining of crude oil to MD, and MD transportation and distribution.
Crude oil is a mixture of hydrocarbons and other organic compounds that is
extracted by drilling wells into underground geological reservoirs. The crude is drawn
from the wells in the form of liquid, and is accompanied by gas and produced water
(PW). Different technologies are used to extract crude oil from wells as the wells age.
Generally, a new oil well has sufficient reservoir pressure to carry the mixture of oil,
gas and PW up to the surface. This naturally occurring type of extraction is called
primary recovery. Over time, as material is removed from the reservoir, the reservoir
pressure, and the efficacy of primary recovery, drops. Secondary recovery techniques
are then applied, which involve injecting water (recycled PW, saline or fresh water)
into the reservoir to maintain reservoir pressure and continue to push crude oil to
surface. This technique, also known as water flooding, is only efficient for a certain
period of time, as the less viscous water and surface tension eventually causes the
more viscous oil to be trapped in the reservoir rock. Tertiary oil recovery, which
is also called enhanced oil recovery (EOR), typically makes use of either CO 2 and
surfactant injection to reduce the surface tension, or of steam and micellar polymer (a
type of surfactant) injection to reduce viscosity contrast. A third EOR technique is
called forward combustion, during which a flame front created by combustion of the
deposits with continuous air injection propagates towards the well, which decreases
the viscosity of the oil to be extracted due to high temperatures [106]. Forward
combustion and other EOR technologies account for only 2% of total EOR.
67% of US oil production relies on crude oil extracted from onshore wells, and this
analysis assumes that all secondary and tertiary recovery is taking place in onshore
wells [123]. If secondary recovery was performed in offshore wells, seawater would
38
most likely be used, therefore the assumption of secondary extraction taking place
onshore results in a higher estimate of fresh water use.
The amount of water used during extraction depends on the technology used.
The values range from 0.21 [lwater/lcrude] recovered for the case of primary recovery,
to 343 [lwater/lcrude] for EOR using micellar polymer injection [123]. For secondary
recovery and EOR, water consumption is primarily associated with injected water that
cannot be recycled or re-used. For primary recovery, however, water is used during
drilling for mixing the drilling mud, and during recycled water (RW) treatment. Table
3.1 shows the amount of water used by each technology along with the technology
shares for oil extraction. A technology-weighted average water consumption value
of 8 [lwater/lcrudel, excluding re-injection, is obtained from all the major primary,
secondary and tertiary recovery systems [123].
Table 3.1: Oil production, water injection and technology share of various recovery
technologies from Wu et al. [123]
Recovery
technology
Oil prod.
[bpd]
Oil prod.
[Mgal/d]
Tech. share
Water inj.
[Mgal/d]
Spec. water consump.
[%]
CO 2 miscible
CO 2 immiscible
Steam
Combustion
Other EOR'
Sec. water flood
Primary recovery
Total
234 315
2 698
286 668
13 260
112 276
2 589 000
227 783
3 466 000
9.8
0.1
12.0
0.6
4.7
108.7
9.6
145.6
10.9
0.1
5.5
0.1
3.5
79.7
0.2
100
127.9
1.5
65
1.1
40.9
933
2
1 171
13
13
5.4
1.9
8.7
8.6
0.2
-
Weighted av.
8.0
[lwater/lcrude]
aData on water use are not publicly available for "other EOR" technologies, including hydrocarbon miscible/immiscible, hot-water flooding, and nitrogen injection. Average values for C0 2 ,
steam and air combustion EOR is assumed for the "other EOR" technologies for which data is not
available.
The blue water consumption of crude oil extraction varies mainly according to the
produced water re-injection technologies employed in each Petroleum Administration
for Defense District (PADD), as does the amount of oil produced and the number
of wells being operated. CONUS is divided into five PADDs, three of which (PADD
II, III and V) account for 81% of US refinery products and 90% of onshore crude oil
production [123]. This analysis uses a weighted average of the values for PADDs II,
III and V as a proxy for the US average. Table 3.2 presents the average volume of
PW that is re-injected during oil recovery for PADDs II, III and V. The net water
needed, also given in the table, indicates the net amount of water consumed during
39
oil extraction and recovery. An average blue water consumption of 3.3 [iwater/lcrude
is estimated for the US [123].
Table 3.2: Water use during oil extraction and recovery from Wu et al. [123]
Technologyweighted average
water injection
PADD
II
III
V
PW re-injected
Net water req'd.
[lwater/lcrudel
5.9
5.7
2.6
8.0
2.1
2.3
5.4
The refining process separates crude oil into its constituent hydrocarbons; removes impurities such as sulfur and nitrogen through hydrotreatment; and increases
marketable fuel yields and fuel quality via hydrocracking and catalytic conversion
techniques. Oil refineries consist of a number of processing units, and based on the
presence of units that are used to process heavy fractions of crude oil, such as vacuum
distillation unit, they can be classified as simple or complex. Simple refineries use
crude oil feedstock that is light (low density) and sweet (containing little sulfur). This
section provides an overview of the refinery process flows, and offers insight into the
refinery units producing MD. Figure 3-2 shows a representative process flow diagram
for a complex oil refinery.
The first step in refining is fractional distillation, which takes advantage of the
different boiling points of different hydrocarbons. The products directly obtained
from the distillation unit are called straight-run products, such as the straight-run
diesel and jet fuel shown in Figure 3-2. Before entering the blending pool, sour
products are sweetened by reducing sulfur-containing mercaptan compounds:
Merox:
2R-SH +
2
-+ R-SS-R + H2 0
Hydrodesulfurization:
R-SH + H2
-
R-H + H25
where R- indicates an alkyl group.
40
Mercaptans are corrosive compounds that have a very strong odor. Figure 3-2
shows two examples to carry out this sweetening process. In the Merox (mercaptan
oxidation) process that has been developed by Honeywell-UOP, the mercaptans are
oxidized into hydrocarbon disulfides. The hydrotreatment process, on the other hand,
removes the sulfur content of the MD stream in the form of hydrogen sulfide (H2 S)
through hydrodesulfurization reactions.
Hydrotreatment processes also help to remove the nitrogen content of the MD
stream, saturate olefins which cause fuel instability during fuel storage, and, at more
severe conditions, saturate aromatic compounds which are associated with particulate
matter and higher NOx and SOx emissions during fuel combustion. The third jet fuel
stream that is sent to the blending pool, as shown in Figure 3-2, is obtained by
cracking the very heavy crude oil fractions in the presence of a catalyst and hydrogen
gas. This analysis assumes a product slate that is 22.9% MD fuels [121].
The processes that use water in a typical refinery include the cooling tower, crude
distillation unit and fluid catalytic cracker (FCC). Steam and cooling operations in
a refinery make up about 96% of the refinery water consumption [123]. Figure 3-3
shows the water flow in a typical North American refinery. Approximately 1.5 liters
of water are consumed for every liter of crude oil processed in an oil refinery [123].
GREET 2011 [62] is used as the reference for assumptions associated with transportation of crude oil, residual oil, diesel fuel and conventional jet fuel. These assumptions include the transportation modes, fuel types, energy intensities, and distances
transported with each mode, as shown in Table 3.3. Crude oil is transported from
the well to the refinery by ocean tanker, barge, pipeline, rail and truck, and refined
MD products are transported and distributed by rail and truck [62]. GREET 2011
is also used to calculate the energy used during the crude oil extraction, recovery
and refining processes. GREET 2011 energy assumptions are aggregated for each
step and related to the primary energy sources of coal, natural gas and petroleum
products. Other primary energy sources, such as wind or nuclear, are neglected due
to their small contribution to overall water footprint. Figure 3.4 shows the indirect
blue water consumption associated with each primary energy carrier from Gleick [49].
These values were used for the calculation of the indirect blue water consumption of
the transportation steps for all pathways.
The transportation modes are fuelled by petroleum refinery products, including
residual oil and diesel, and in the case of pipelines by natural gas and electricity.
During the oil extraction, recovery and refining steps, petroleum refinery products
41
H2
H2
CokerLight Gasoline
m CokerHeavyGasoline
I
CohorLight
Gs61n1*
Hydrotmeter
Straight-Run Light Gasoline
lmerizatein
Straight-Run Light Gasoline
Straight-Run
or
Hydrocracked
Light Gasoline
and/orgIeene
Satsation
or
Isomerate
tH2
Mmhede
oistllatift
Straight-Run HeavyGasoline
Heavy
-Reformate
ftormer
001901l00.
Hydretreater
Hydrocracked
Crud
Oil1
Straight-Run Jet
(Kerosene)
Straight-Run Diesel
(Heating Fuel)
Hydrocracked Light Gasoline
PTo Distillate
MydrecavkerHydrocracised
(Diesel,
Jet, andHeatingOil)
Hyrre Hra FuelBlending
Propylene (C3)
isobutane
(C4)
Propylene/Butenes (C3/C4)
Butylenes/Amylenes
0
Straight-Run
HeavyGasoli u
Airr heric Wa
Bottos i- that
Moter
Polymerized
Gasoline
Geelne
-+
menw
Alkylate
(C4/Cs)
Butylenes
Light Vacuu
Gasoli
H2
HeavyVacuumc
Dinerization
roe Feed
Hydrtrater
.
-I
Gasoil
FCC
IFCC
Gaone
Hyretol
Dimate
FCCLight Gasoline
FCCHeavyGasoline
FCCLight Gasoll
Merex
Idrx
Sweetened
Jet
Heavy Gasoil
Straght-RunStraight-Run
Light Gasoil
Vacuum
Caker
Jet
Jet
HeavyGasoline ,
Light Gasoline
-+
of
Hydrocracked Jet
Coke
H2
-
Diesel
Straight-Run
Dines
Hretreetw
Hydrocracked
Diesel
Figure 3-2: A representative complex oil refinery, adapted from Chevron [30].
are used along with other energy sources. Therefore, estimation of the indirect water
use associated with the production of all the fuels utilized during these transportation and processing steps requires iterative calculations. Direct water consumption
from transportation has not been considered in this analysis due to its negligible
contribution [99].
The iterative procedure used in this analysis to estimate the direct and indirect
water consumption from the processing and transportation steps is as follows:
1. The direct water consumption values given in Table 3.2 for the three PADDs
are used as initial values in this analysis. The analysis is carried out for each
PADD separately.
2. The direct water consumption values in Table 3.2 are allocated among a representative oil refinery product slate based on the energy contents of each fuel
cut [121]. Lower heating values are taken from GREET 2011 [62].
42
BIowdown/Re cycle
Breakdown/Recycle
Figure 3-3: Water flow in a typical North American refinery, adapted from Wu et al.
[123]
3. Crude oil transportation uses diesel, residual oil, natural gas and electricity.
Step 2 values for diesel and residual oil are used as inputs to determine the
contribution of crude oil transportation, and the water intensity of natural gas
is taken from Gleick [49]. The water intensity of electricity is calculated for
each PADD using data from Wu et al. [123] The results are allocated amongst
the assumed product slate and added to the values calculated in step 2.
4. The values calculated in step 3 are used as petroleum refinery inputs to calculate
the contribution from crude oil extraction and recovery. Water intensities for
coal and natural gas are taken from Gleick [49]. The results are allocated
amongst the assumed product slate and added to the values calculated in step
3.
5. Using the water intensities calculated in step 4, along with the water intensities
for natural gas and coal, the contribution from producing residual oil, ULS
diesel, diesel, and conventional jet fuel is estimated. The results, specific to
each fuel (and hence, already allocated), are then added to the values in step 4.
6. The step 5 values, along with water intensities of natural gas and electricity,
are used to calculate the contribution of transporting residual oil, ULS diesel,
diesel, and conventional jet fuel. The calculated fuel-specific values are added
to the step 5 results.
7. The water intensities calculated for residual oil and ULS diesel in step 6, along
with those of coal and natural gas, are used to estimate the impacts of MD
43
Table 3.3: Crude oil and finished fuel transportation assumptions for the conventional
MD pathway from GREET 2011 [62].
Crude oil transp.
Residual oil transp.
Conventional MD
Transp. & Dist.
Model
Shareb
Distance [km]
Ocean tanker
Barge
Pipeline
Ocean tanker
Barge
Pipeline
Ocean tanker
Barge
Pipeline
Rail
Truck
57%
1%
100%
24%
40%
60%
16%
6%
75%
7%
100%
8179
805
1207
4828
547
644
2333
837
644
1287
48
'Barge runs on residual oil; truck and rail run on diesel; and pipelines run on 20% diesel, 50%
residual oil, 24% natural gas and 6% electricity. The energy intensities are 0.513, 1.49, 0.267 and
0.183 MJ/tonne-km, respectively.
bMass-based share of a feedstock that relies on a certain transportation mode. The total can
exceed 100% because feedstock may be moved from location to location by different transportation
modes until it reaches its final destination.
transportation. The calculated contributions for each PADD are then summed
with the step 6 results.
Variability in the results for the conventional MD pathway is due to assumptions
regarding the blue water consumption of crude oil recovery and extraction in the
different PADDs from Wu et al. [123] This analysis expands on the blue water
consumption footprint of conventional MD reported by Wu et al. [123] by including
indirect blue water consumption from transportation, and material and energy inputs,
and by allocating results amongst refinery fuel products.
Table 3.4: Indirect blue water consumption footprint of primary energy carriers from
Gleick [49].
Primary
energy carrier
Coal
Natural gas
Petroleum
Diesel
Residual oil
Indirect blue water consumption footprint
[lwater/MJenergy]
0.164
0.109
0.139
0.143
0.146
44
3.3.2
FT MD from natural gas and coal
The FT process converts any carbon-containing feedstock, such as natural gas, coal
or biomass, to paraffinic hydrocarbons. The process involves steam reforming of
natural gas, or gasification of a solid feedstock, into a mixture of carbon monoxide
and hydrogren called synthesis gas, or syngas, which is then purified and converted to
fuel via FT synthesis. Longer hydrocarbon chains are then cracked down to maximize
the MD cut of the product slate. This analysis assumes that the energy requirements
to produce FT MD are the same as those used in Stratton et al. [112] In addition,
it is assumed that 70% of the product slate, by energy content, is MD, and that the
plant is able to produce enough electricity for its own requirements [112].
In this analysis two FT pathways are considered for fuel production: coal-toliquids (CTL) and natural gas-to-liquids (GTL). For coal feedstock extraction, water
is consumed during open pit and underground mining operations, and for washing
to remove contaminants. For natural gas feedstock, water is consumed primarily
during treatment to remove H2 S and CO 2 . Underground mining of coal is more
water intense than surface mining [491, and natural gas production from shale is
a more water intense process than conventional natural gas production [75]. The
blue water consumption footprint for the extraction and treatment of these primary
energy carriers are between 0.161 and 0.169 [lwater/MJcoal] for coal, and 0.109 and
0.110 [lwater/MJNG] for natural gas [49].
After the feedstock has been extracted it is transported to an FT refinery. The
GREET 2011 [62] assumptions for the energy intensity of coal and natural gas transportation to an FT facility are used to estimate the indirect water consumption of
this step. These assumptions are shown in Table 3.5.
Table 3.5: Natural gas and coal feedstock transportation assumptions for the FT MD
pathway from GREET 2011 [62].
Feedstock
Natural gas
Coal
Fuel
Energy intensity
[J/MJ]
Coal
Natural gas
Petroleum
Coal
Natural gas
Petroleum
50
921
4
42
187
837
Once at the refinery, the first step in the FT process is steam reforming or gasi45
fication, during which the feedstock is partially oxidized into syngas. Purification
of the syngas to remove impurities, such as sulfur, is crucial in order to prevent
catalyst poisoning during the downstream FT synthesis. This is accomplished in a
hydrotreatment process. Another important parameter in the FT synthesis is the
CO-to-H 2 ratio that needs to be adjusted to minimize coking and maximize straightchain hydrocarbon production. This ratio can be tuned through the water-gas shift
(WGS) reaction, which requires fresh water input:
H2 0 + CO -
H2 + CO 2
Following gasification and purification, FT synthesis takes place, which is a polymerization reaction of carbon monoxide in the presence of hydrogen and an ironor cobalt-based catalyst [112]. The synthesis process is described by the following
reaction:
mCO + n
2m
-+
2
CmHn + mH 2 0
Water produced as a by-product of FT synthesis can be treated for use within
the FT process or other industrial processes. The gasification and synthesis processes
both require electrical power, and most commercial facilities choose to produce electricity on-site. On-site electricity production requires additional fresh water for steam
production and cooling.
In addition to the cooling system, water may be used directly in the FT process
in a number of other ways, such as the WGS reaction. Steam methane reforming also
requires fresh water, as it uses steam to convert methane into syngas. The reaction
for steam methane reforming is as follows:
H2 0 + CH 4 -
CO + 3H2
A simplified process flow diagram of the FT fuel production process is shown in
Figure 3-4, with water consumption highlighted. The transportation and distribution
of FT MD is subject to the same assumptions as the other alternative MD pathways,
for which the default GREET 2011 assumptions are shown in Table 3.6, including
barge, truck and rail modes of transport.
In order to calculate the lifecycle blue water consumption of the CTL and GTL
FT pathways, a range of process parameters are considered for each of the feedstockto-fuel conversion processes. This is done in order to capture the effects of variability
46
Water consumption:
Evaporation & cooling losses
Onsite electricity
production
nal
gas
Cakn
Gasirfication &Synthesis
Dfe
Hydrogen
production
Water consumption:
Water gas shift or SMR
Figure 3-4: Simplified flow diagram of the FT process, water consumption highlighted.
Table 3.6: FT MD transportation assumptions from GREET 2011 [62].
Alternative MD
Transp. & Dist.
Modea
Shareb
Distance [km]
Truck
Barge
Rail
Truck (dist.)
63%
8%
29%
100%
80.5
837
1 287
48
aBarge runs on residual oil, and truck and rail run on diesel. The energy intensities are 0.513
and 1.49 MJ/tonne-km, respectively.
bMass-based share of a feedstock that relies on a certain transportation mode. The total can
exceed 100% because feedstock may be moved from location to location by different transportation
modes until it reaches its final destination.
in technology implementation on the results. Table 3.7 shows the assumptions and
associated references that define low, mid and high scenarios for each feedstock-tofuel conversion process, including variability due to the feedstock extraction process,
lower heating value (LHV) conversion efficiency, and direct process water use, from
Bao et al. [23], Matripragada [69], and Stratton et al. [112]
3.3.3
AF MD from sugarcane, corn and switchgrass
The AF MD pathway lifecycle includes biomass cultivation and transportation, feedstockto-fuel conversion, and MD transportation and distribution. Three types of AF
47
Table 3.7: Variability of blue water consumption of FT MD pathways
Process water consump.
Process
CTL
GTL
Case
Feedstock extraction
Low
Mid
High
Low
Mid
High
Surface mining [49]
50% surface, 50% underground [49]
Underground mining [49]
Conventional gas [49]
Conventional gas [49]
Shale gas [49]
LHV conv. eff.
53%
50%
47%
65%
63%
60%
[112]
[112]
[112]
[112]
[112]
[112]
[Iwater/Ifuel]
7.5 [69]
9.4 [69]
11.4 [69]
0.0 [23]
biomass feedstock are considered in this analysis. Sugary feedstocks include biomass
in which mono- or disaccharide sugars, such as glucose and sucrose, are present in
significant quantities; starchy feedstocks include biomass in which a polysaccharide,
such as starch, is the main sugar component; and lignocellulosic feedstocks include
biomass in which sugar is stored in complex polysaccharides, such as cellulose and
hemicellulose. One representative feedstock from each class was selected for this
analysis: sugarcane, corn and switchgrass, respectively. Water is consumed during
cultivation of these feedstocks by evaporation and evapotranspiration. Estimation of
blue water consumption associated with feedstock cultivation is discussed in detail in
Section 3.2.
After the biomass has been harvested from the field it is transported to a biorefinery. The AF feedstock transportation assumptions, shown in Table 3.8, are
based on the default assumptions in GREET 2011 [62].
Table 3.8: AF feedstock transportation assumptions from GREET 2011 [62].
Sugarcane
Corn
grain
Switchgrass
Modea
Shareb
Distance [km]
Truck
Truck
Truck
Truck
100%
100%
100%
100%
19
16
64
64
aTruck runs on diesel, with an energy intensity of 1.49 MJ/tonne-km.
bMass-based share of a feedstock that relies on a certain transportation mode. The total can
exceed 100% because feedstock may be moved from location to location by different transportation
modes until it reaches its final destination.
Following cultivation and transportation to a processing facility, the feedstock undergoes pretreatment to extract the complex carbohydrate sugars, or polysaccharides,
from the biomass. This typically involves mechanical size reduction and physical or
chemical processing to extract the sugars from the structure of the feedstock.
48
Figure 3-5: Simplified process flow diagram of sugarcane milling pretreatment, water
consumption highlighted.
Water consumption:
BFW make-up & cooling losses
Bagasse
combustion
(CHP)
|Washing'
Sugacan
csrdig
Juice treatment
Extracted
& cncentration
sugar solutio~n
Water consumption:
Water consumption:
Sugarcane washing,
Lime solution &
imbibition & bearing cooling
blowdown losses
In the case of sugarcane, the energy and water consumption intensity of sucrose
extraction was estimated from a review of sugar-mill technologies [36, 41, 66, 67].
Water consumption during sugarcane pretreatment is primarily due to losses during
sugarcane washing, steam production and turbo-generator cooling during electricity
co-production [67]. A simplified process flow diagram of the sugarcane pretreatment
process is shown in Figure 3-5.
Starch is extracted from corn grain via milling pretreatment. The energy and
water requirements of corn grain milling were estimated from a survey of the corn
ethanol literature [76, 61, 73, 80, 95, 101, 122]. Water is consumed during evaporation
from cooling and boiler feed water (BFW) make-up for cooking and liquefaction
processes. Drying of distillers dried grains and solubles (DDGS) is also a major
source of consumptive water use [123]. A simplified process flow diagram of the corn
pretreatment process is shown in Figure 3-6.
In this analysis it is assumed that sugars are extracted from switchgrass using
dilute acid pretreatment, and the associated energy and water consumption is estimated from the literature [3, 53, 60]. Water consumption during the pretreatment of
switchgrass is due to evaporation from cooling processes, and from steam production
and feedstock combustion during electricity co-production [53]. A simplified process
flow diagram of the switchgrass pretreatment process is shown in Figure 3-7.
The steps following feedstock pretreatment are common to all of the AF pathways considered in this analysis. Saccharification is used to break down the polysac49
Figure 3-6: Simplified process flow diagram of corn grain milling pretreatment, water
consumption highlighted.
Water consumption:
Evaporation losses
Water consumption:
BFW make-up &
cooling losses
Figure 3-7: Simplified process flow diagram of switchgrass dilute acid pretreatment,
water consumption highlighted.
Water consumption:
BFW make-up & cooling losses
Water consumption:
BFW make-up &
blowdown losses
50
Figure 3-8: Simplified process flow diagram of monomer sugar metabolism and upgrading to drop-in fuel, water consumption highlighted.
Extracted
sugar solution
Saccharification &
microbial metabolism
Water consumption:
Cooling losses
Oe:eEnergy
sprto..upgrading
1
Water consumption:
Blowdown &
cooling losses
carrier
MDfe
Water consumption:
BFW make-up &
cooling losses
charide molecules to monomeric C5 and C6 sugars. The sugar monomers are fed
to a genetically engineered micro-organism that metabolizes the sugar to a platform molecule and CO2 . The energy and water consumption of saccharification
and metabolism by the engineered micro-organism was estimated from ethanol plant
data [76, 61, 73, 80, 95, 101, 122], characteristic bioreactor values [81], and consultation with researchers and industry [108, 109, 107, 110]. During saccharification
and metabolism, water is consumed primarily for cooling of the bioreactor [67]. It is
assumed that the micro-organism employed metabolizes monomer sugars to alkanes,
fatty acid, ethanol or isobutanol.
The platform molecules are then separated from the other products of metabolism
and sent to post-processing for upgrading to a drop-in fuel product slate, including
some MD fraction. Three technologies were considered for the separation and concentration of alkanes and fatty acids after metabolism; centrifugation [24], hexane
extraction [112], and KOH steam lysing [117]. It is assumed that distillation is used
to separate and concentrate ethanol and isobutanol following metabolism. Finally,
the energy and water consumption requirements of upgrading platform molecules to a
drop-in fuel slate were estimated through consultation with industry for ethanol and
isobutanol platform molecules [109, 107], and by using the process parameters for hydroprocessing of esters and fatty acids, for alkane and fatty acid platform molecules
[94, 102, 117]. The monomer sugar metabolism, platform molecule extraction, and
post-processing steps are shown in Figure 3-8, and the data sources for the AF pathways are shown in Table 3.9. Transportation and distribution of AF MD fuel is
subject to the same assumptions as the other alternative MD pathways, shown in
Table 3.6.
The overall water footprints of the AF pathways are dependent upon the allocation
methodology applied to the analysis. For all of the AF pathways, market allocation
is used to allocated the water footprint between the co-products and the platform
molecule produced [120]. The free-on-board prices for the co-products of the AF
51
Table 3.9: Data sources for AF MD feedstock-to-fuel process parameters.
Process step
References
Sugarcane pretreatment
Corn pretreatment
Switchgrass pretreatment
Advanced fermentation
Platform molecule upgrading
[36, 41, 66, 67]
[76, 61, 73, 80, 95, 101, 122]
[3, 53, 60]
[108, 109, 107, 110]
[109, 107, 94, 102, 117]
pathway, such as DDGS, wet distillers grain, corn gluten feed and corn gluten meal,
are from the US Grain Council's weekly report [31]. The sugarcane and switchgrass
AF pathways co-produce electricity, and the nationwide average electricity price is
taken from EIA [10]. Prices for the platform molecules are from the Independent
Chemical Information Service (ISIC) [100]. All prices are for August, 2012. Blue
water consumption is allocated amongst the constituent fuel slate products using
energy allocation. These allocation methods are -consistent with previous lifecycle
GHG emissions studies on alternative MD production pathways [112].
The parameters contributing to variability due to feedstock-to-fuel conversion process parameters are shown in Table 3.10. The assumptions regarding irrigation practices and the assumed locations of biomass and fuel production are shown in Table
3.11. Low and high scenario counties were selected on the basis of being blue water consumption extremes for each feedstock and irrigation assumption pairing. The
sugarcane and corn mid scenario counties were selected as they reported the largest
yields for those crops in the 2007 USDA Census of Agriculture [83]. Due to a lack
of historical switchgrass production, Robertson, TN was selected as the mid scenario
county for switchgrass, due the region's suitability for cultivation of that crop [35].
52
Table 3.10: Feedstock-to-fuel process variability for blue water consumption of AF MD pathways.
Q1
Pathway
Scenario
Platform
molecule
Power inputs
[kWh/lMDI
NG inputs
[MJ/lMDI
Make-up water
[lwater/lMD]
Feed-to-fuel
[kgfeed/IMD
Co-prod.
.alloc.
Fuel
alloc.
Sugarcane
AF
(50% moist.)
Corn
AF
(15.5% moist.)
Switchgrass
AF
(0% moist.)
Low
Mid
High
Low
Mid
High
Low
Mid
High
Alkanes
Fatty acid
Ethanol
Isobutanol
Fatty acid
Ethanol
Alkanes
Fatty acid
Ethanol
0
0
0
0.47
0.75
1.8
0
0.73
1.5
0
0
0
8.4
8.6
22.4
0
0
0
6.2
10.9
14.9
3.8
4.6
6.4
2.5
4.4
7.2
18.4
21.3
25.9
3.6
5.8
7.6
8.2
8.3
11.4
79%
91%
94%
85%
75%
90%
92%
100%
100%
81%
81%
78%
100%
81%
78%
81%
81%
78%
Table 3.11: Variability of blue water consumption of AF MD pathways due to assumed
irrigation practices and location of biomass and fuel production.
Pathway
Rainfed sugarcane AF
Irrigated sugarcane AF
Rainfed corn AF
Irrigated corn AF
Rainfed switchgrass AF
Irrigated switchgrass AF
3.3.4
Scenario
County
Low
Mid
High
Low
Mid
High
Low
Mid
High
Low
Mid
High
Low
Mid
High
Low
Mid
High
Dixie, FL
Palm Beach, FL
St. Charles, LA
Wakulla, FL
Palm Beach, FL
Caldwell, TX
Northampton, VA
La Salle, IL
Suffolk, NY
McDowell, WV
La Salle, IL
Nueces, TX
Crawford, GA
Robertson, TN
Suffolk, NY
Habersham, GA
Robertson, TN
Young, TX
HEFA MD and biodiesel from soybean, rapeseed and
jatropha
The HEFA MD and biodiesel pathway lifecycles include biomass cultivation and transportation, vegetable oil extraction and transportation, vegetable oil to MD or biodiesel
conversion, and MD fuel transportation and distribution.
Triglycerides or triacylglycerol (TAG), which consists of one glycerol and three
fatty acid molecules, are the primary components of animal fats, vegetable and algal
oils. These molecules can be hydrotreated into straight-chain alkanes, known as
hydroprocessed esters and fatty acid (HEFA) fuels. Based on the type of feedstock,
fatty acids can vary in size, and this will have a direct effect on the final product
distribution. These fuels can then be isomerized to achieve better fuel properties,
and catalytically cracked to maximize jet and naphtha production. Alternatively,
TAGs can be transesterified into biodiesel.
This analysis considers soybean, rapeseed and jatropha for HEFA MD and biodiesel
production. After the biomass has been harvested it is transported to an oil extraction
54
mill by diesel-powered trucks. The biomass transportation assumptions, consistent
with the default assumptions in GREET 2011 [62], are shown in Table 3.12.
Table 3.12: HEFA biomass transportation assumptions from GREET 2011 [62].
Soybean
Rapeseed
Jatropha
Transportation step
Modea
Shareb
Distance [km]
To stacks
Stacks to plant
To stacks
Stacks to plant
To stacks
Stacks to plant
Truck
Truck
Truck
Truck
Truck
Truck
100%
100%
100%
100%
100%
100%
16
64
16
64
16
64
'Trucks run on diesel, with an energy intensity of 3.18 and 2.48 MJ/tonne-km for oilseed and oil
transportation, respectively.
bMass-based share of a feedstock that relies on a certain transportation mode. The total can
exceed 100% because feedstock may be moved from location to location by different transportation
modes until it reaches its final destination.
Vegetable oil is extracted from the biomass by pressing the oilseeds and introducing an organic solvent, such as hexane [102]. In the case of soybean and rapeseed, the
meal separated from the oil has a high protein content with commercial value as an
animal feed, and market value allocation is used to allocate the water consumption
to the meal co-product of oil extraction.
Jatropha has a structure that is quite different than the other oil seeds, resulting
in co-products other than just the meal. The jatropha fruit is essentially a capsule
containing a husk and two or three seeds. Each of the seeds has a shell and an
oil-containing kernel. After the oil has been extracted from the kernel, the meal is
leftover. Most varieties of jatropha fruit are toxic to humans and therefore this analysis assumes combustion of all the co-products (husks, shells and meal) for electricity
generation [112]. Energy allocation is used to allocate water consumption to the
electricity co-product of jatropha oil extraction.
The direct fresh water consumption associated with oil extraction from oil crops is
due to BFW steam generation and cooling water make-up. The indirect fresh water
consumption comes from the production of the material and energy inputs used in
the extraction process, such as natural gas, electricity and hexane. The extracted
oil is then transported to the HEFA plant to be converted into fuel via the HEFA
process. The assumptions for transportation of oil to the HEFA facility, are shown in
Table 3.13.
55
Table 3.13: HEFA oil feedstock transportation assumptions from GREET 2011 [62].
Soybean oil
Jatropha oil
Rapeseed oil
Modea
Shareb
Distance [km]
Truck
Barge
Rail
Truck
Rail
40%
40%
20%
67%
33%
129
837
1127
129
1127
aBarges use residual oil, and trucks and rail run on diesel fuel. The energy intensities are 1.49
and 0.513 MJ/tonne-km, respectively.
bMass-based share of a feedstock that relies on a certain transportation mode. The total can
exceed 100% because feedstock may be moved from location to location by different transportation
modes until it reaches its final destination.
Figure 3-9 shows a simplified process flow diagram of the HEFA MD process
considered in this study. Vegetable oil is taken from feed storage and fed into a hydrodeoxygenation reactor where the olefinic double bonds of the TAGs are saturated
in the presence of hydrogen, and the oxygen content is removed in the form of water
and CO2 . During this reaction glycerol is separated from the rest of the TAG structure in the form of propane. The hydrodeoxygenation reaction generates water, which
is treated for reuse in the boiler or cooling water system, rather than discharged to
the sewer. The hydrogen required for hydrodeoxygenation is obtained from steam
reforming (SMR) of natural gas. Water is used as a reactant in the SMR reactor.
This is an endothermic reaction, and the heat required is supplied by high-pressure
steam generated from BFW by burning natural gas.
The effluent from the hydrodeoxygenation reactor is then cooled down as steam
is generated, and sent to an isomerization unit where cracking also takes place. The
isomerized product is later cooled with cooling water and phase-separated into gases
and liquids. The liquids are separated into different fuel products in a distillation
unit, where steam is used for heating the boiler section. Gases are sent to a gasprocessing unit where hydrogen and CO 2 are separated in a pressure swing absorption
(PSA) unit. The gas-processing unit uses cooling water to facilitate the separation of
methane, ethane and propane from water and other impurities to produce a dry gas
suitable for use as a fuel. These gases are further purified, and can be used as process
fuel in the process, or in the SMR unit, which is assumed to use natural gas in this
analysis. Unreacted hydrogen is recycled back to the hydrodeoxygenation reactor.
The liquid products from this process include liquefied petroleum gas (LPG),
naphtha, jet fuel and diesel fuel. 80.9% of this product slate is composed of middle
distillate fuels [22, 38, 54, 116]. The design used in this analysis integrates the BFW
56
Figure 3-9: Simplified process flow diagram of HEFA MD process, water consumption
highlighted.
Hydrogen
,Ga
prodution
|Vegetable)
oil
Water recycle
Hydro-
Heat
lisomerization
Heat
dleoxygenation
exchanger
&tcat. cracking
exchanger
Water consumption:
Cooling losses
Water consumption:
BFW make-up
Drop-in
Sprtr
product sl~ate
Steam
with the process, such that make-up water demand may be reduced by generating
steam instead of using cooling water. Hence, the direct fresh water consumed during
the HEFA process is primarily due to the losses in the BFW that is used for steam
generation and as cooling water, as shown in Figure 3-9. 89% of direct fresh water
consumption is for boiler feed water makeup due to steam generation and cooling
losses [94]. The indirect water consumed by the HEFA process is primarily due to the
electricity and natural gas requirements of the process. Electricity is used to power
pumps, compressors and other electrical controls around the refinery, and natural
gas is used as a process fuel that is burned to supply heat in various units around
the refinery, such as in the boiler of the SMR unit. Energy based allocation is used
to allocate the fresh water consumption within the HEFA process amongst the fuel
co-products. Table 3.14 summarizes the consumptive fresh water associated with the
HEFA process.
Table 3.14: Blue water consumption of the HEFA MD feedstock-to-fuel process from
Pearlson et al. [94]
Direct blue water consumption [iwater/loil]
BFW make-up
SMR
Produced water
Total
0.8
0.2
-0.1
0.9
Alternatively TAGs can be transesterified to biodiesel, during which TAGs react with an alcohol in the presence of a catalyst. The glycerol backbone separates
from the TAG leaving three fatty acid molecules, which form fatty acid alkyl esters
(or biodiesel) by including alcohol into their structures. When methanol is used in
this process, fatty acid methyl esters (FAME) are formed. In this analysis, default
57
values and assumptions from GREET 2011 are employed for the processing steps
[62]. The HEFA MD and biodiesel produced by these processes is then transported
and distributed to its final destination, subject to the same assumptions as the other
alternative MD pathways shown in Table 3.6.
Variability in the lifecycle results for the HEFA MD and biodiesel pathways is due
to the choice of feedstock; biomass growth, oil extraction and oil yield assumptions;
whether biomass is rainfed or irrigated, and the assumed location of biomass and fuel
production. The parameters contributing to this variability due to process inputs and
yields are shown in Table 3.15. The assumptions regarding irrigation practices and the
assumed locations of biomass and fuel production are shown in Table 3.16. Low and
high scenario counties were selected on the basis of being blue water consumption
extremes for each feedstock and irrigation assumption pairing. The soybean and
rapeseed mid scenario counties were selected as they reported the largest yields for
those crops in the 2007 USDA Census of Agriculture [83]. Due to a lack of historical
jatropha production, Palm Beach, FL was selected as the mid scenario county for
jatropha. Palm Beach reported the largest yields for sugarcane in the 2007 USDA
Census of Agriculture, and jatropha and sugarcane require similar soil and climatic
conditions [83].
58
Table 3.15: Process variability of blue water consumption of HEFA MD and biodiesel pathways. Biomass growth input, oil
extraction input and oil yield data from Stratton et al. [112]
Pathway
Scenario
Low
Soybean
HEFA
Mid
High
Low
V1
Rapeseed
HEFA
Mid
High
Biomass growth
[MJ/kgfeed]
Jatropha
Mid
High
Direct =
Indirect
Direct
Indirect
Direct
8.22
0.17
8.22
0.17
8.22
Direct
Indirect
Direct
Indirect
Direct
0.65
0.30
0.76
0.39
1.11
Indirect
0.39
Indirect = 0.17
0.28
0.91
0.42
1.25
0.91
Direct = 2.75
0.30
Indirect
2.84
Direct
Indirect = 0.27
Direct = 3.02
1.71
Indirect = 0.32
Direct
Indirect
Direct
Indirect
Direct =
Indirect
Low
Oil extract.
[MJ/kgji]
-
Direct = 1.23
Indirect = 1.98
Direct = 1.39
Indirect = 2.14
Direct
=
1.56
Indirect = 2.21
2.45
Direct
Indirect = 0.22
Direct = 2.51
Indirect = 0.21
Direct
=
2.58
Indirect = 0.22
Oil yield
[kgfeed/kgoil]
Co-prod.
alloc.
MD
alloc.
Biodiesel
alloc.
4.7
47%
81%
90%
4.7
47%
81%
90%
4.7
47%
81%
90%
2.3
77%
81%
90%
2.4
76%
81%
90%
2.5
74%
81%
90%
3.0
65%
81%
90%
3.0
65%
81%
90%
3.0
65%
81%
90%
Table 3.16: Variability of blue water consumption of HEFA MD and biodiesel pathways due to assumed location of biomass and fuel production.
Pathway
Rainfed soybean HEFA & biodiesel
Irrigated soybean HEFA & biodiesel
Rainfed rapeseed HEFA & biodiesel
Irrigated rapeseed HEFA & biodiesel
Rainfed jatropha HEFA & biodiesel
Irrigated jatropha HEFA & biodiesel
60
Scenario
County
Low
Mid
High
Low
Mid
High
Low
Mid
High
Low
Mid
High
Low
Mid
High
Low
Mid
High
Greene, VA
Cass, ND
Nassau, NY
Washington, VA
Cass, ND
Knox, TX
Franklin, PA
Latah, ID
Nassau, NY
Essex, NY
Latah, ID
Merced, CA
Kemper, MS
Palm Beach, FL
St. Charles, LA
Levy, FL
Palm Beach, FL
Nueces, TX
Chapter 4
Results and discussion
Results are reported as a range of low, mid and high values for the lifecycle blue water
consumption footprint of each MD production pathway. The results are geo-spatially
disaggregated for the AF and HEFA pathways to quantify the effect of the assumed
location of biomass feedstock cultivation and MD production on the results. The
trade-offs between blue water consumption footprint and areal productivity are also
quantified.
This is the first analysis to calculate the blue water consumption footprint of
alternative MD pathways. Previous studies that quantify the lifecycle water footprint
of alternative fuels have focused on ethanol and biodiesel, and present results as a
range to capture variability and uncertainty in the input parameters[59, 99, 115, 123].
With the exception of Chiu & Wu [27], results are not geo-spatially disaggregated,
and no previous studies have quantified the trade-offs between the water and land
resource requirements of alternative fuels production.
4.1
Results and geo-spatial disaggregation
Lifecycle blue water consumption footprints are calculated for each MD production
pathway, including variability in each pathway according to the assumptions defined
in the previous chapter. The results for all pathways studied in this analysis are shown
in Table 4.1 and Table 4.2, and are compared to the literature under assumptions of
rainfed and irrigated biomass consumption in Figure 4-1[27, 59, 99, 123]. Because this
is the first analysis to calculate the blue water consumption footprint of alternative
MD pathways, and previous studies have focused primarily on conventional petroleum
fuels and ethanol, the results are compared to the literature on the basis of the
61
energy equivalent of a liter of conventional diesel. The whiskers in Figure 4-1 indicate
variability in the results: for example, the low and high results for the irrigated AF
switchgrass pathway are 2.9 and 5272.5 lwater/lMD respectively. These correspond to
the combinations of biomass feedstock cultivation, and feedstock-to-fuel conversion
parameter assumptions that yield the lowest and highest results. Comparison with
the literature shows that the results of this analysis are congruent with the range of
values previously reported. For example, Figure 4-1 shows that the low, mid and high
results for conventional MD (4.1, 4.9 and 7.5) lie in the midst of results calculated by
King & Webber [59], Scown et al. [99] and Wu et al. [123]
Table 4.1: Results for conventional and FT pathways [lwater/lfuellPathway
Low
Mid
High
Conventional MD
Coal FT MD
NG FT MD
4.10
14.2
5.69
4.89
16.1
5.88
7.51
18.1
6.33
Table 4.2: Results for biomass feedstock pathways, rainfed and irrigated cultivation
shown [lwater/lfuellPathway
Low
Rainfed
Mid
Sugarcane AF MD
Corn AF MD
Switchgrass AF MD
Soybean HEFA MD
Rapeseed HEFA MD
Jatropha HEFA MD
Soybean biodiesel
Rapeseed biodiesel
Jatropha biodiesel
4.81
4.66
2.57
1.90
1.58
1.56
1.84
1.43
1.58
11.0
9.00
7.71
1.98
1.89
2.02
1.95
1.79
1.86
62
High
15.0
17.1
13.8
2.61
2.11
2.34
2.40
1.99
2.11
Low
Irrigated
Mid
High
6.3
6.1
2.90
3.04
2.60
2.91
2.82
2.48
2.71
523
330
511
1510
1740
279
1430
1690
265
2300
2360
5270
3050
3480
1750
2890
3370
1670
MD from conv. crude
Dieselfrom conv.crude (King & Webber 2008)
R
Gasolinefrom conv. crude (Scownet al 2011)
Gasolinefrom conv. crude (Wu et al. 2009)
NG FTMD
NG FTdiesel (King & Webber 2008)
CoalFT MD
CoalFTdiesel (King & Webber 2008)
Sugarcane AF MD
Corn grain AF MD
Corn grain E85(King & Webber 2008)
Corn grain and stover EtOH(Scown et al. 2011)
SwitchgrassAF MD
Corn stover E85 (King & Webber 2008)
SwitchgrassEtOH(Wu 2009)
-
r--
Soybean HEFAMD
RapeseedHEFAMD
Jatropha HEFAMD
Soybeanbiodiesel
Soybean biodiesel (King& Webber 2008)
+
Rapeseed biodiesell
Jatropha biodiesel
0
6
4
2
8
a)
10
12
14
18
16
200
'blue water consumed diesel equivalent
ConventionalMD
SugarcaneAF MD
Com AF MD
Corn grain EtOH(Chui & Wu 2012)
Corn grain E85(King & Webber 2008)
Corn grain and stover EtOH(Scown et al. 2011)
Corn grain EtOH (Wu 2009)
SwitchgrassAFMD
Cornstover EtOH (Chui& Wu 2012)
Wheat straw EtOH (Chui& Wu 2012)
--
4-
Cornstover E85(King & Webber 2008)
Miscanthus EtOH(Scown et al. 2011)
Soybean HEFAMD
Rapeseed HEFAMD
Jatropha HEFAMD
Soybean biodiesel
Soybean biodiesel (Chui & Wu 2012)
Soybean biodiesel (King & Webber 2008)
Rapeseedbiodiesel
Jatropha biodiesel
b)
0
1000
2000
3000
4000
5000
6000
Iblue water consumed/'diesel equivalent
Figure 4-1: Lifecycle blue water consumption of a) FT and rainfed, and b) irrigated
MD production pathways. The conventional MD pathway is shown for the purposes
of comparison. Results shown are calculated by this analysis unless otherwise cited.
Calculated and literature results are compared on the basis of liters of diesel equivalent.
63
Figure 4-la) shows that the blue water consumption footprint of FT and rainfed
MD production ranges between 1.4 and 18.1 lwater/lMD. This range depends upon
the MD production pathway considered: the biodiesel and HEFA MD production
pathways have the lowest blue water consumption footprints, and FT MD from coal
has the highest. With the possible exception of FT MD from coal, alternative MD
production has a blue water consumption footprint comparable to MD from conventional crude under an assumption of rainfed biomass cultivation. In contrast, Figure
4-1b) demonstrates that under an assumption of irrigated biomass cultivation, alternative MD pathways that use biomass feedstocks have blue water consumption
footprints several orders of magnitude greater than MD from conventional crude. For
example, rainfed soybean HEFA MD has a blue water consumption footprint of 2.0
lwater/lMD, compared to conventional MD with 4.6 lwater/lMD under mid assumptions. However, irrigated soybean HEFA MD has a blue water consumption footprint
of 1510.7 lwater/lMD under mid assumptions. The large blue water consumption
footprints of the pathways in Figure 4-ib) are due to the blue water consumption of
irrigation. For example, for soybean HEFA MD the biomass cultivation step accounts
for 1508.7 lwater/lMD of the total lifecycle blue water consumption footprint of 1510.7
lwater/lMD. Results are broken out by lifecycle step for each fuel production pathway
in Appendix A.
Figure 4-1 shows a range of values in order to capture variability in the assumptions related to blue water irrigation requirements, feedstock-to-fuel process water
requirements, and conversion efficiency. The results are sensitive to these assumptions, therefore the AF and HEFA pathway results are further disaggregated to investigate the largest source of variability: the assumed location of biomass feedstock
cultivation and MD production. Geo-spatial location contributes to variability in the
results because of the irrigation requirements of maximized crop production specific
to local climate and soil conditions and, to a lesser extent, the indirect blue water
consumption footprint of electricity in each county. The blue water consumption of
irrigated biomass cultivation was extracted from the GAEZ model [44], and the indirect water footprint of electricity was taken from Strzepek et al [113]. The blue water
consumption requirements of feedstock irrigation are associated to counties, and a
lifecycle result is calculated for each pathway under the assumptions of rainfed and
irrigated biomass cultivation. In addition, the areal productivity of MD production
is calculated using biomass feedstock yield data from GAEZ. The feedstock-to-fuel
process water requirements and conversion efficiencies are held constant at the mid
assumption parameters for each pathway, shown in the previous chapter. An example
of the results is shown in Figure 4-2 for the corn AF MD pathway. The blue water
64
-
-----___- -
--
_____
consumption footprint of MD production is higher under an assumption of irrigated
biomass, and increases as irrigated biomass cultivation occurs in the drier climates
in the western states. For example, the blue water consumption footprint of irrigated corn AF MD production is 357.5 lwater/lMD in Robertson, TN, and increases
to 2124.8 lwater/lMD in Dewey, OK. The corresponding increase in areal productivity
of corn AF MD, due to irrigated biomass cultivation, is also shown in Figure 4-2.
Under an assumption of rainfed corn AF MD production areal productivity is 1607
lMD/ha/year, and increases to 1800 lMD/ha/year under an assumption of irrigated
corn AF MD production in Robertson, TN. Results maps for the other AF and HEFA
MD pathways are reported in Appendix B.
Blue Water Consumption Footprint [1,1/,0D]
**i
Areal productivity [I /ha/year]
,-T
Figure 4-2: Lifecycle blue water consumption footprint and areal productivity of
rainfed and irrigated corn AF MD production.
In order to quantify the cumulative effects of commercial scale alternative MD
production, marginal resource cost curves are constructed for the AF and HEFA
pathways. All counties in the contiguous US are ranked by the blue water consumption footprint and land requirements of MD production, and plotted against the
cumulative MD production capacity of that county for each AF and HEFA production pathway. The MD production capacity of each county is calculated on the basis
of areal biomass feedstock productivity from GAEZ and the calculated feedstock-tofuel conversion efficiency. Each county's MD production capacity is constrained by
available harvested cropland from the 2007 USDA Census of Agriculture [83], and
65
available fresh water resources, without inducing a water stress index above 0.4, from
the USDA Forest Service Water Supply Stress Index (WaSSI) [84]. Examples of the
marginal water and land resource cost curves for the corn AF MD pathway are shown
in Figure 4-3. Results for the other AF and HEFA pathways are reported Appendix
C.
3000 12.0
2500
1
AddI
producfeo
du
*
20D0
omumption
(IA) 1500
a(hall
Lan reqLpkement
ODO
Ocrese in r
re5ie0)f du o1r~
0.8
1000
C
Offt
Addu
pnI
Figure4-3:a)aria
b
e
M
0,4
to
WO4UCOO4O
0n550L
0
a)
0
20
0
b)
11
120
0
so
30
tominimduie wtereqimnt.bLadeurmns
ranked
counties
production,
(109(har)
production
Currrra"uMD
90
50
MDprodrctn(109"er)
Cermdabiv
120
IS0
Figure 4-3: a) Marginal blue water consumption of rainfed and irrigatedl corn AF MD
production, counties ranked to minimize water requirements. b) Land requirements
of rainfed and irrigated corn AF MD production, counties ranked to minimize land
requirements.
The marginal resource cost curves are stacked to minimize either blue water consumption or land footprint, and they illustrate the trade-offs associated with blue water consumption and land use. In order to quantify these trade-offs, the average blue
water consumption footprint and areal productivity of the AF and HEFA pathways
were calculated for 10 billion liters of MD production under three different assumptions: rainfed cultivation; irrigated cultivation with blue water consumption footprint
minimized; and irrigated cultivation with areal productivity maximized. The results
are shown in Table 4.3. Blue water consumption footprint and areal productivity
range from 1.7 to 1660.3 lwater/lMD and 490 to 3710 lMD/ha/year, respectively.
4.2
Areal productivity benefits of irrigated biomass
cultivation
In practice, the decision to irrigate biomass cultivation would consider both blue
water consumption and areal productivity, not just one or the other as shown in the
marginal resource cost curves in Figure 4-3. Therefore, the areal productivity benefit
66
Table 4.3: Lifecycle blue water consumption footprint and areal productivity of 10
billion liters of MD production from each pathway, averaged over three different
assumptions.
Pathway
Rainfed
Irrigated,
Water use
minimized
Sugarcane AF
Corn AF
Switchgrass AF
Soybean HEFA
Rapeseed HEFA
Jatropha HEFA
Sugarcane AF
Corn AF
Switchgrass AF
Soybean HEFA
Rapeseed HEFA
Jatropha HEFA
11.1
6.33
5.06
2.00
1.72
1.98
3.35
1.22
1.91
0.490
0.894
2.06
262
18.8
180
53.4
1.93
20.6
3.15
1.64
2.19
0.710
1.01
2.17
Metric
Average
water
footprint
lwater/lMD
Average
areal
productivity
1000 lMD/ha
Irrigated,
Areal prod.
maximized
666
1170
1660
986
880
406
3.71
1.98
3.01
0.813
1.18
2.36
of blue water consumption footprint is calculated on a county basis for the AF and
HEFA pathways. This value, M, is expressed as A Y/A W, where:
10001
MD
M = Areal productivity benefit of irrigation
1
water
.
A Y = Irr. areal yield
1 0 0 0 1MD
[ha
" W = Irr. blue water consump.
ha
IMD
Rainfed areal yield
lwater
.
1000 1 MD
[ha]
Rainfed blue water consump. lwater
. IMD
AIMD W
M is calculated for each county in the contiguous US that GAEZ determines is
suitable for cultivation of the biomass feedstock of interest. A higher value indicates
that the areal productivity to be gained from irrigation comes with a relatively small
increase in blue water consumption footprint. Conversely, a low value indicates that
there is little areal productivity improvement to be realized, or that it comes at a
relatively large increase in blue water consumption footprint. An example of the
result of this calculation for the corn AF MD pathway is shown in Figure 4-4.
In order to compare the potential for areal productivity benefits from irrigation
between the AF and HEFA pathways, the results are ranked by the areal productivity
benefit of irrigation and plotted against cumulative MD production for all the AF and
HEFA MD production pathways. This is shown in Figure 4-5.
67
Figure 4-4: Areal productivity benefit of irrigation for corn AF MD production.
Figure 4-6 shows which fuel pathway will see the greatest areal productivity benefit
under an assumption of irrigated biomass cultivation for alternative MD production in
each county in the contiguous US. Relative to the other pathways, corn AF enjoys the
greatest benefit in the Plains States and the Mississippi Valley regions; switchgrass
AF in the Great Lakes, Northeast and West Coast regions; rapeseed HEFA in the
Rocky Mountain and Western States; and sugarcane AF and jatropha HEFA enjoy
the greatest benefit in the Gulf Coast regions of Texas and Louisiana. These results
are not prescriptive, as there are other factors that determine where certain crops
are grown. Rather, this analysis calculates the alternative MD pathway that could
realize the greatest areal productivity benefit from irrigation in each county.
It is important to note that the decision to irrigate in any particular location
depends on a number of additional factors, such as the impacts of irrigation practices
on local water stress and water quality. Further research is required to understand
these additional impacts of large-scale alternative MD fuel production.
68
100
10
-Sugarcane
1
--
Switchgrass AF
-Soybean
-Rapeseed
(1000 IMD/ha)
per
-Jatropha
0.1
AF
AF
-Com
HEFA
HEFA
HEFA
('.Wa.erMD)
0.01
0.001
0.0001
0
2
6
4
10
8
Cumulative MD production (109
12
14
I/year)
Figure 4-5: Areal productivity benefit of irrigation for alternative MD production.
Figure 4-6: MD production pathway with the greatest areal productivity benefit from
irrigation in each county.
69
THIS PAGE INTENTIONALLY LEFT BLANK
70
Chapter 5
Conclusions and future work
RFS2 seeks to address a number of imperfections in the transportation fuels market,
including: imperfect competition leading to US economic vulnerability to volatile
prices; security and environmental externalities; and a lack of innovation and investment in alternative technologies. RFS2 targets dependence on expensive, volatile
sources of transportation fuels as a source of economic vulnerability for the US, however renewable fuels offer limited opportunities for the substitution of crude oil consumption, and are potentially more expensive. The policy also attempts to address
the collective action problems of security and environmental externalities by encouraging the consumption of alternatives to petroleum that have a lower GHG footprint
than conventional petroleum fuels. However, the regulatory tool employed is not the
most economically efficient mechanism, and RFS2 does not account for non-GHG
related environmental impacts. Finally, RFS2's ability to stimulate investment and
innovation in alternatives to petroleum fuels has been hampered by Stiglerian capture, and the regulatory uncertainty that has caused has slowed development of the
industry.
As US renewable fuels policy continues to evolve, the political and institutional
issues addressed in Chapter 2 of this thesis merit close attention. In particular production volume mandates, EPA compliance measures, and additional environmental
criteria will likely continue to be contentious issues, influenced by Stiglerian capture
by industrial interests. Fortunately, in the case of renewable fuels, undue influence
is partially offset by selective interest groups in opposition to each other. The challenge will be in ensuring that, in the face of this tension, the interests of the general
public are sufficiently represented to address the issues of economic vulnerability, and
security and environmental externalities. This is central to the future development
of socially optimal, equitable, renewable fuels policy in the US.
71
The water consumption and land resource requirements of alternative MD fuel
production are absent from the qualification requirements under RFS2 despite being
determinants of overall environmental sustainability. In addition, alternative MD fuel
production is expected to grow in the coming years. Therefore, this thesis contributes
to the literature by quantifying the water and land resource requirements of alternative MD fuel production. Chapters 3 and 4 demonstrate that the water consumption
footprint of alternative MD fuel is highly dependant upon the feedstock-to-fuel pathway considered, whether biomass feedstock is rainfed or irrigated, and the geographic
location of feedstock cultivation and fuel production. In addition, it is shown that
significant trade-offs exist between the water consumption footprint and areal land
requirements of alternative MD production.
The evolution of alternative fuels policy requires recognition of the non-GHG environmental impacts of production, and these findings could be used to develop water
consumption and areal productivity requirements for environmental certification of
fuels. In particular, these results could be integrated with a hydrological model to
determine the effects of alternative MD production on local fresh water availability
and water stress, and to determine the most effective allocation of arable land resources for alternative MD production. Furthermore, water consumption and areal
productivity characteristics could be used as inputs to an aggregate analysis of the
societal costs and benefits of alternative MD fuel production.
It is important to note that lifecycle GHG emissions contributing to climate
change, the only metric of environmental performance currently considered under
RFS2, is a global environmental impact, whereas water and land use are distinctly
local impacts. Water consumption and areal land footprints of alternative MD production are meaningful only insofar as the local scarcity of these resources are reflected
in decision making, and an area for future work could be the geo-spatial disaggregation of the localized value of fresh water and land resource use. The inclusion of
additional alternative MD pathways, such as FT biomass-to-liquids, algae, aqueousphase processing and pyrolysis, is also an area for further research.
72
Appendix A
Blue water consumption footprint
by lifecycle step
73
Table A. 1: Blue water consumption by lifecycle Step for conventional MD, FT and rainfed MD production pathways [lwater/lMD
Pathway
Conv. MD
Coal FT MD
NG FT MD
Sugarcane AF MD
Corn AF MD
Switchgrass AF MD
Soybean HEFA MD
Rapeseed HEFA MD
Jatropha HEFA MD
Soybean biodiesel
Rapeseed biodiesel
Jatropha biodiesel
Scenario
Feedstock
growth/extraction
Feedstock
transp.
Veg. oil
extraction
Veg. oil
transp.
Feedstock-to-fuel
conversion
MD fuel
transp.
Total
Low
Mid
High
Low
Mid
High
Low
Mid
High
Low
Mid
High
Low
Mid
High
2.33
3.16
5.71
10.31
11.20
12.20
5.69
5.90
6.22
-
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.10
0.10
0.11
0.09
0.12
0.16
-
-
1.72
1.72
1.72
3.87
4.85
5.87
0.00
0.00
0.00
4.69
10.92
14.86
4.59
8.80
16.88
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
4.08
4.91
7.46
14.21
16.08
18.11
5.73
5.94
6.26
4.81
11.04
15.01
4.71
8.95
17.07
Low
-
0.10
-
-
2.49
0.03
2.62
Mid
High
Low
Mid
High
Low
Mid
High
Low
Mid
High
Low
Mid
High
Low
Mid
High
Low
Mid
High
-
0.11
0.13
0.06
0.06
0.06
0.05
0.05
0.05
0.06
0.06
0.06
0.06
0.06
0.06
0.05
0.05
0.05
0.05
0.05
0.05
0.39
0.52
0.92
0.21
0.35
0.43
0.13
0.45
0.60
0.36
0.49
0.86
0.19
0.33
0.40
0.12
0.42
0.57
0.04
0.04
0.04
0.03
0.03
0.03
0.04
0.04
0.04
0.04
0.04
0.04
0.03
0.03
0.03
0.04
0.04
0.04
7.56
13.68
1.32
1.38
1.56
1.33
1.44
1.57
1.33
1.49
1.55
1.29
1.32
1.42
1.07
1.36
1.43
1.30
1.38
1.42
0.03
0.03
0.03
0.03
0.03
0.02
0.02
0.02
0.02
0.02
0.02
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
7.70
13.83
1.85
2.05
2.62
1.64
1.89
2.10
1.58
2.05
2.27
1.79
1.95
2.42
1.41
1.85
1.99
1.56
1.94
2.12
Table A.2: Blue water consumption by lifecycle step for irrigated MD production pathways [lwater/lMD
Pathway
Scenario
Feedstock
growth/extraction
Feedstock
transp.
Veg. oil
extraction
Veg. oil
transp.
Feedstock-to-fuel
conversion
MD fuel
transp.
Total
Sugarcane AF MD
Low
Mid
0.20
512.33
0.10
0.10
-
-
5.93
10.92
0.03
0.03
6.26
523.38
High
2282.43
0.11
-
-
14.86
0.03
2297.43
Low
Mid
0.20
320.70
0.09
0.12
-
-
5.80
8.80
0.03
0.03
6.13
329.66
High
Low
2341.92
0.31
0.16
0.10
-
-
13.24
2.49
0.03
0.03
2355.35
2.93
Corn AF MD
Switchgrass AF MD
Soybean HEFA MD
Rapeseed HEFA MD
Jatropha HEFA MD
Soybean biodiesel
Rapeseed biodiesel
Jatropha biodiesel
Mid
503.33
0.11
-
-
7.56
0.03
511.03
High
5262.07
0.13
-
-
10.28
0.03
5275.51
Low
Mid
High
Low
Mid
High
Low
Mid
High
Low
Mid
High
Low
Mid
High
Low
Mid
High
0.58
1508.69
3044.73
0.63
1738.98
3476.49
0.91
277.10
1752.91
0.55
1432.76
2891.14
0.60
1651.34
3301.25
0.87
263.13
1664.58
0.06
0.06
0.06
0.05
0.05
0.05
0.06
0.06
0.06
0.06
0.06
0.06
0.05
0.05
0.05
0.05
0.05
0.05
0.79
0.52
0.64
0.36
0.35
0.38
0.38
0.45
0.40
0.75
0.49
0.60
0.44
0.33
0.36
0.36
0.42
0.38
0.04
0.04
0.04
0.03
0.03
0.03
0.04
0.04
0.04
0.04
0.04
0.04
0.03
0.03
0.03
0.04
0.04
0.04
1.50
1.38
1.43
1.51
1.44
1.52
1.47
1.49
1.46
1.39
1.32
1.35
1.25
1.36
1.40
1.38
1.38
1.37
0.03
0.03
0.03
0.02
0.02
0.02
0.02
0.02
0.02
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
3.02
1510.73
3046.94
2.60
1740.87
3478.49
2.88
279.14
1754.89
2.83
1434.71
2893.23
2.45
1686.89
3370.54
2.73
265.07
1666.46
THIS PAGE INTENTIONALLY LEFT BLANK
76
Appendix B
Maps of water consumption
footprint and areal productivity
77
Blue Water Consumption Footprint [lwater/'MD]
Areal productivity [Ilm/ha/year]
C
00
Not suitae
5000-100
1500-2000
2000-3000
rowth/no da
1000-1L500,
Figure B-1: Lifecycle blue water consumption footprint and areal productivity of rainfed and irrigated sugarcane AF MD
production.
Blue Water Consumption Footprint [Iwater/IMD]
Figure B-2: Lifecycle blue water consumption footprint and areal productivity of rainfed and irrigated corn AF MD production.
Blue Water Consumption Footprint [lwater/'MO
Areal productivity [ImD/ha/year]
-0
00
produt
10001500-200
2000 00
3000 -0
200-5000
50006200
io
20003000
n
50100d
100-10
S--000
1500
4M-W
400-500
Figure B-3: Lifecycle blue water consumption footprint and areal productivity of rainfed and irrigated switchgrass AF MD
production.
Blue Water Consumption Footprint ['water/IMD]
Notdstock rowth/no de
500-1
1000 1500
S1500-2000
2000 3000
3000 -4000
4000-5000
5000-6200
500 -1000C
1000 -1500
1500-2000
2000 -3000
3000 -4000
4000 -5000
Figure B-4: Lifecycle blue water consumption footprint and areal productivity of rainfed and irrigated soybean HEFA MD
production.
Areal productivity [Im/ha/year]
Blue Water Consumption Footprint [Iwater']MDI
4-4
*Not
suiwlt-
0-100
Not sut'ic
0 -2
2-4
4-6
6-8
8 -10
10 -12
Sro-tit/no d a300
-o
C{
s
estock owth/no da
100'-300
-5
500 - 1000
00- 1500
1500 -2000
2000-3000
3000-4000
400 - 500
00
300500- 1
1000-_ 1500
1500 - 2000
2000--3000
3000 -4000
4000 - 5000
500-6200
0 -100
100 - 306
300 -50
500 -1000-1000 -1500,
150 -2000
2000 - 3000
3000 - 4M0
4000-500
Figure B-5: Lifecycle blue water consumption footprint and areal productivity of rainfed and irrigated rapeseed HEFA MD
production.
Blue Water Consumption Footprint [lwater/MD]
Areal productivity [ImD/ha/year]
4-
-0
4-I
NotNo
fedtckoothn
suitable
stoc
10-00
5000- 6200
4000 -5000
Figure B-6: Lifecycle blue water consumption footprint and areal productivity of rainfed and irrigated jatropha HEFA MD
production.
THIS PAGE INTENTIONALLY LEFT BLANK
84
Appendix C
Marginal resource cost curves
85
2.0 -
3000 -2
2500
1.6
2000
1.2-
Blue water
consuptionLand
c" uMp)
requirement
(haIl
MD)
1500 -
0.8 1000 -
0.4 --.---.-
e ed
-
30
60
90
Cumulative MD production (109 /year)
120
fnigated
0.0
0\
0
I
-Rainfed
-Rainfed
150
/
0
30
60
90
Cumulative MD production (109 Vyear)
120
150
Figure C-1: a) Marginal blue water consumption of rainfed and irrigated sugarcane AF MD production, counties ranked to
minimize water requirements. b) Land requirements of rainfed and irrigated sugarcane AF MD production, counties ranked to
minimize land requirements.
3000
2500
2.0
-
1.6
2000
Blue water
consumption
1.2
Land requirement
(lwat.AMo) 1500
(ha 1000 lm,,) 08
1
000
00
100
0.4
I
500
med
... ,--I
-
-Rainfed
a)
0
10.0
0
30
60
90
Cumulative MD production (109 I/year)
120
Irgated
-Rainfed
150
b)
0
30
60
90
120
150
Cumulative MD production (109I/year)
Figure C-2: a) Marginal blue water consumption of rainfed and irrigated corn AF MD production, counties ranked to minimize
water requirements. b) Land requirements of rainfed and irrigated corn AF MD production, counties ranked to minimize land
requirements.
. ...
...
....
..
.........
3000
2.0 -
2500
1.6
2000 -
1.2
Blue water
consumption 1500
Land requirement
(ha/OO Imo)
(ih10MD)
0.8 1000 -
00
00
0.4
500
-
-
Irrigated
-Irrgated
12aied
a)
0
30
60
90
Cumulative MD production (1091/year)
120
Rainfed
150
b)
0.0
0
30
60
90
120
150
Cumulative MD production (109 I/year)
Figure C-3: a) Marginal blue water consumption of rainfed and irrigated switchgrass AF MD production, counties ranked to
minimize water requirements. b) Land requirements of rainfed and irrigated switchgrass AF MD production, counties ranked
to minimize land requirements.
3000
2.0
2500
1.6
2000
1.2
Blue water
consumption
(le/iJlM)
Land requirement
(ha/1 000 I,,)
1500
0.8
1000
00
0.4
500
-Inigated
-
a)
-Irrigated
Rainfed
-Rainfed
0
0
30
60
90
Cumulative MD production (1091/year)
120
150
)0
0.0
30
60
90
Cumulative MD production (109 1/year)
120
150
Figure C-4: a) Marginal blue water consumption of rainfed and irrigated soybean HEFA MD production, counties ranked to
minimize water requirements. b) Land requirements of rainfed and irrigated soybean HEFA MD production, counties ranked
to minimize land requirements.
..
......
....
.... I......
1.
.... ..................
I..........................
...................
.-
.
2500
1.6-
2000
1.2
Blue water
consumption
(l1.,o)
Land requirement
(ha/1000 Iu,)
1500
0.81000
0.4-
500 -
-- Ingated
-rrigated
-Rainfed
-Rainfed
a)
D
0
0.0
30
6
90
90
Cumulative MD production (109 I/year)
120
1 3
0 60
12 0
150
Cumulative MD production (109 I/year)
Figure C-5: a) Marginal blue water consumption of rainfed and irrigated rapeseed HEFA MD production, counties ranked to
minimize water requirements. b) Land requirements of rainfed and irrigated rapeseed HEFA MD production, counties ranked
to minimize land requirements.
3000
2.0 -
2.
2500 -
1.6-
2000
1.2 -
Blue water
consumption
(L.AMD)
Land requirements
(ha/I 000 IMD)
1500
0.8
1000
0.4500 -
- igaed
a)
__
0
30
_
_
_
_
_
_
__
_
90
60
Cumulative MD production (109 Vyear)
_
_
120
_
Irrigated
-Rainfed
-Rinfed
_
_0.0
150
b)
0
30
90
60
Cumulative MD production (109 Vyear)
120
150
Figure C-6: a) Marginal blue water consumption of rainfed and irrigated jatropha HEFA MD production, counties ranked to
minimize water requirements. b) Land requirements of rainfed and irrigated jatropha HEFA MD production, counties ranked
to minimize land requirements.
.... ........
.. ......
THIS PAGE INTENTIONALLY LEFT BLANK
92
Appendix D
Maps of areal productivity benefits
of irrigated biomass cultivation
93
Figure D-1: Areal productivity benefit of irrigation for sugarcane AF MD production.
C.0
Figure D-2: Areal productivity benefit of irrigation for corn AF MD production.
I/I
-V
Figure D-3: Areal productivity benefit of irrigation for switchgrass AF MD production.
Figure D-4: Areal productivity benefit of irrigation for soybean HEFA MD production.
C-,
I
'-7
I
a'
F7~Q
00
Figure D-5: Areal productivity benefit of irrigation for rapeseed HEFA MD production.
-
Figure D-6: Areal productivity benefit of irrigation for jatropha HEFA MD production.
THIS PAGE INTENTIONALLY LEFT BLANK
100
Bibliography
[1] 96th Congress. Public Law 96-294, The Energy Security Act, 1980.
[2] D. Acemoglu. Diversity and Technological Progress. National Bureau of Economic Research, Chapters in: The Rate and Direction of Inventive Activity
Revisited:319-356, 2011.
[3] A. Aden, M. Ruth, K. Ibsen, J. Jechura, K. Neeves, J. Sheehan, B. Wallace,
L. Montague, A. Slayton, and J. Lukas. Lignocellulosic Biomass to Ethanol
Process Design and Economics Utilizing Co-current Dilute Acid Prehydrolysis
and Enzymatic Hydrolysis for Corn Stover. Technical Report June, National
Renewable Energy Laboratory, Golden, CO, 2002.
[4] US Energy Information Administration. Energy Timelines: Ethanol. http:
//www.eia.gov/kids/energy. cfm?page=tl ethanol, June 2008.
[5] US Energy Information Administration. Biofuels Issues and Trends. http:
//www.eia.gov/biofuels/issuestrends/, October 2012.
[6] US Energy Information Administration. Countries. http://www.eia.gov/
countries/index. cfm?view=production, November 2012.
Economic Efficiencies of
[7] US Energy Information Administration.
http://www.eia.gov/oiaf/aeo/otheranalysis/aeo_
High Oil Prices.
2006analysispapers/efhop. html, November 2012.
[8] US Energy Information Administration. Primary Energy Consumption By
Source and Sector. http: //www.eia. gov/totalenergy/data/annual/pecss_
diagram. cfm, September 2012.
[9] US Energy Information Administration. Short-term Energy Outlook. http:
//www. eia. gov/forecasts/steo/report/global-oil. cfm, November 2012.
[10] US Energy Information Administration. US Energy Information Administration. http: //www. eia. gov/electricity/monthly/epm-table-grapher. cfm?
t=epmt_5_6_a, August 2012.
[11] US Energy Information Administration. Petroleum & Other Liquids. http:
//www.eia.gov/dnav/pet/pet-cons
101
wpsup-k_4.htm, March 2013.
[12] US Federal Aviation Administration. Destination 2025. Technical report, US
Federal Aviation Administration, Washington, DC, 2011.
[13] US Central Intelligence Agency. The World Factbook. https: //www. cia.
gov/library/publications/the-world-factbook/geos/us.html,
November 2012.
[14] US Environmental Protection Agency. Renewable Fuel Standard: Obligated
Parties and Exporters of Renewable Fuel. http: //www. epa. gov/oms/regs/
fuels/obligatedparties.htm, December 2011.
[15] US Environmental Protection Agency. 2011 EMTS.
otaq/fuels/rfsdata/201lemts.htm, October 2012.
http://www.epa.gov/
[16] US Environmental Protection Agency. Renewable Fuel Standard. http: //www.
epa.gov/otaq/fuels/renewablefuels/index.htm, October 2012.
[17] US Environmental Protection Agency. Renewable Fuels Standard. http: //
www. epa. gov/otaq/fuels/renewablefuels/index. htm, March 2013.
[18] A. F. Alhajji and D. Huettner. OPEC and World Crude Oil Markets from 1973
to 1994: Cartel, Oligopoly, or Competitive? The Energy Journal, pages 31-60,
2000.
[19] D. T. Allen, C. Allport, K. Atkins, J. S. Cooper, R. M. Dilmore, L. C. Draucker,
K. E. Eickmann, J. C. Gillen, W. Gillette, W. M. Griffin, W. E. Harrison, J. I.
Hileman, J. R. Ingham, F. A. Kimler, A. Levy, C. F. Murphy, M. J. O'Donnell,
D. Pamplin, G. Schivley, T. J. Skone, S. M. Strank, R. W. Stratton, P. H.
Taylor, V. M. Thomas, M. Q. Wang, and T. Zidow. Framework and Guidance
for Estimating Greenhouse Gas Footprints of Aviation Fuels. Technical report,
Air Force Research Laboratory, Wright-Patterson Air Force Base, 2009.
[20] P. A. Almoguera and A. M. Herrera. Testing for the cartel in OPEC: noncooperative collusion or just noncooperative? https://www.msu.edu/-herrer20/
documents/OPEC.pdf, August 2007.
[21] A. Andrews, R. Pirog, and M. F. Sherlock. The US Oil Refining Industry: Background in Changing Markets and Fuel Policies. Technical report, Congressional
Research Service, Washington, DC, 2010.
[22] R. E. Bailis and J. E. Baka. Greenhouse Gas Emissions and Land Use Change
from Jatropha Curcas-based Jet Fuel in Brazil. Environmental Science & Technology, 44(22):8684-91, November 2010.
[23] B. Bao, M. M. El-Halwagi, and N. 0. Elbashir. Simulation, Integration, and
Economic Analysis of Gas-to-Liquid Processes. Fuel Processing Technology,
91(7):703-713, July 2010.
102
[24] N. A. Carter. Environmental and Economic Assessment of Microalgae-derived
Jet Fuel. Master of science, Massachusetts Institute of Technology, 2012.
[25] N. A. Carter, R. W. Stratton, M. K. Bredehoeft, and J. I. Hileman. Energy
and Environmental Viability of Select Alternative Jet Fuel Pathways. In 47th
AIAA/ASME/SAE/ASEE Joint Propulsion Conference & Exhibit, number August, San Diego, 2011.
[26] United Nations Environmental Programme World Conservation Monitoring
Centre. Protected Areas Inventory. http: //www. unepwcmc .org, 2007.
[27] Y. W. Chiu and M. Wu. Assessing County-Level Water Footprints of Different
Cellulosic-biofuel Feedstock Pathways. Environmental Science & Technology,
46(16):9155-62, August 2012.
[28] S. Combs. Percentage of US Corn Used to Produce Ethanol and Price per
Bushel, 1980-2007. http://www.window. state.tx.us/specialrpt/energy/
renewable/exhibits/exhibit13-10.php, 2013.
[29] European Commission Joint Research Commission. Global Land Cover 2000.
http: //bioval. jrc. ec. europa. eu/products/glc2000/products. php, 2006.
[30] Chevron Corporation. Diesel Fuels Technical Review. Technical report, Chevron
Corporation, 2007.
[31] US Grain Council. United States Grain Council. http: //www. grains. org/
images/DDGSX20Weekly%2OReports/USGC%20-%20DDGS%2OWeekly%2OMarket%
20Report%20August%203.2012.pdf, August 2012.
[32] K. Crane, A. Goldthau, M. Toman, T. Light, S. E. Johnson, A. Nader,
A. Rabasa, and H. Dogo. Imported Oil and US National Security. Techni-
cal report, RAND Corporation, 2009.
[33] M. Crocker and R. Andrews. The Rationale for Biofuels. In Thermochemical
Conversion of Biomass to Liquid Fuels and Chemicals, number 1, chapter 1,
pages 1-25. Royal Society of Chemistry, Lexington, KY, 2010.
[34] A. Dancs, M. Orisich, and S. Smith. The Military Cost of Securing Energy.
Technical report, National Priorities Project, 2008.
[35] J. Davis. Feasability Study for Insuring Dedicated Energy Crops. Technical
Report 817, AgriLogic Consulting, LLC, Arlington, TX, 2011.
[36] M. 0. S. Dias, A. V. Ensinas, S. A. Nebra, R. Maciel Filho, C. E. V. Rossell,
and M. R. W. Maciel. Production of Bioethanol and Other Bio-based Materials
from Sugarcane Bagasse: Integration to Conventional Bioethanol Production
Process. Chemical Engineering Research and Design, 87(9):1206-1216, September 2009.
103
[37] R. Dominguez-Faus, S. E. Powers, J. G. Burken, and P. J. Alvarez. The Water
Footprint of Biofuels: A Drink or Drive Issue?
Environmental Science &
Technology, 43:3005-3010, 2009.
[38] B. Donnis, R. G. Egeberg, P. Blom, and K. G. Knudsen. Hydroprocessing of
Bio-Oils and Oxygenates to Hydrocarbons. Understanding the Reaction Routes.
Topics in Catalysis, 52(3):229-240, January 2009.
[39] T. Edwards. Alternative Aviation Fuels Evaluation to Support Certification,
2011.
[40] EIA. Definitions of Petroleum Products and Other Terms. Technical Report
May, U.S. Energy Information Administration, 2010.
[41] A. V. Ensinas, S. A. Nebra, M. A. Lozano, and L. M. Serra. Analysis of
Process Steam Demand Reduction and Electricity Generation in Sugar and
Ethanol Production from Sugarcane. Energy Conversion and Management,
48(11):2978-2987, November 2007.
[42] M. Falkenmark and J. Rockstroem. The New Blue and Green Water Paradigm:
Breaking New Ground for Water Resources Planning and Management. Journal
of Water Resources Planning and Management, 132:129-132, 2006.
[43] J. E. Fargione, T. R. Cooper, D. J. Flaspohler, J. Hill, C. Lehman, D. Tilman,
T. McCoy, S. McLeod, E. J. Nelson, and K. S. Oberhauser. Bioenergy and
Wildlife: Threats and Opportunities for Grassland Conservation. BioScience,
59(9):767-777, October 2009.
[44] G. Fischer, F. 0. Nachtergaele, S. Prieler, E. Teixeira, G. Toth, H. van
Velthuizen, L. Verelst, and D. Wiberg. Global Agro-ecological Zones (GAEZ
v3.0) - Model documentation. Technical report, International Institution for
Applied Systems Analysis, 2009.
[45] Scientific Food & Agricultural Organization United Nations Educational
and Cultural Organization. Soil Map of the World. http: //www. fao. org/
fileadmin/templates/nr/images/resources/images/SoilMap-hires. pdf,
1974.
[46] International Institute for Applied Systems Analysis. Global Agro-ecological
Zones (GAEZ v3.0). http://www.gaez.iiasa. ac.at/w, 2012.
[47] Center for Responsive Politics. American Petroleum Institute. http: //www.
opense cret s. org/lobby/f irmsum. php?id=D000031493&ye ar=2009, October
2012.
[48] Center for Strategic & International Studies.
"NOPEC" Legislation and U.S. Energy Security.
http://csis.org/publication/
nopec-legislation-and-us-energy-security, June 2011.
104
[49]
P. Gleick. Water and Energy. Annual Review of Energy and the Environment,
19(1):267-299, January 1994.
[50] J. M. Griffin. OPEC and World Oil Prices: Is the Genie Back in the Bottle?
Energy Studies Review, 4(1), 1992.
[51] J. I. Hileman, H. M. Wong, I. A. Waitz, D. S. Ortiz, J. T. Bartis, M. A. Weiss,
and P. E. Donohoo. Near-Term Feasibility of Alternative Jet Fuels. Technical
report, Massachusetts Institute of Technology, Cambridge, MA, 2009.
[52] T. A. Howell. Irrigation Efficiency. In Encyclopedia of Water Science, pages
467-472. Marcel Dekker, Inc., New York City, 2003.
[53] D. Humbird, R. Davis, L. Tao, C. Kinchin, D. Hsu, and A. Aden. Process
Design and Economics for Biochemical Conversion of Lignocellulosic Biomass
to Ethanol. Technical Report May, National Renewable Energy Laboratory,
Golden, CO, 2011.
Lifecycle
[54] Hong Huo, Michael Wang, Cary Bloyd, and Vicky Putsche.
Soybeanof
Assessment of Energy Use and Greenhouse Gas Emissions
derived Biodiesel and Renewable Fuels. Environmental Science & Technology,
43(3):750-6, February 2009.
[55] IATA. A Global Approach to Reducing Aviation Emissions. Technical report,
International Air Transport Association, Switzerland, 2009.
[56] International Food Policy Research Institute. Global Agricultural Extent
http://www.asb. cgiar.org/BNPP/phase2/bnpp-phase2_datasets.
v2.0.
htm, 2002.
[57] S. M. Jordaan, D. W. Keith, and B. Stelfox. Quantifying Land Use of Oil
Sands Production: A Lifecycle Perspective. Environmental Research Letters,
4:15, 2009.
[58] V. Karplus. Climate and Energy Policy for U.S. Passenger Vehicles: A
Technology-Rich Economic and Policy Analysis. PhD thesis, MIT, 2011.
[59] C. W. King and M. E. Webber. Water Intensity of Transportation. Environmental Science & Technology, 42(21):7866-72, November 2008.
[60] D. Kumar and G. S. Murthy. Impact of Pretreatment and Downstream Processing Technologies on Economics and Energy in Cellulosic Ethanol Production.
Biotechnology for Biofuels, 4(1):27, January 2011.
[61] J. R. Kwiatkowski, A. J. McAloon, F. Taylor, and D. B. Johnston. Modeling the
Process and Costs of Fuel Ethanol Production by the Corn Dry-grind Process.
Industrial Crops and Products, 23(3):288-296, May 2006.
105
[62] Argonne National Laboratory. Greenhouse Gases, Regulated Emissions, and
Energy Use in Transportation (GREET) Model. http: //greet. es. anl. gov/,
2011.
[63] Oak Ridge National Laboratory. LandscanTM Global Population Database.
http://www.orni.gov/landscan/, 2003.
[64] R. K. Lattanzio. Canadian Oil Sands: Lifecycle Assessments of Greenhouse
Gas Emissions. Technical report, Congressional Research Service, Washington,
DC, 2012.
[65] P. Lobo, D. E. Hagen, and P. D. Whitefield. Comparison of PM Emissions from
a Commercial Jet Engine Burning Conventional, Biomass, and Fischer-Tropsch
Fuels. Environmental Science & Technology, 45(24):10744-9, December 2011.
[66] P. C. Lobo, E. F. Jaguaribe, J. Rodrigues, and F. A. A. da Rocha. Economics of Alternative Sugar Cane Milling Options. Applied Thermal Engineering, 27:1405-1413, 2007.
[67] I. d. C. Macedo. Sugar Cane's Energy. Uniio da Indlstria de Cana-de-Agulcar
(Brazilian Sugarcane Industry Association), 2nd edition, 2007.
[68] J. Mandel. Court Rejects Refiners' Challenge to EPA Biodiesel Blending Rules.
New York Times, 2010.
[69] H. C. Mantripragada. Techno-economic Evaluation of Coal-to-Liquids (CTL)
Plants and Their Effects on Environment and Resources. PhD thesis, Carnegie
Mellon University, 2010.
[70] Y. Masutomi, K. Takahashi, H. Harasawa, and Y. Matsuoka. Impact Assessment of Climate Change on Rice Production in Asia in Comprehensive Consideration of Process/Parameter Uncertainty in General Circulation Models.
Agriculture, Ecosystems & Environment, 131(3-4):281-291, June 2009.
[71] D. A. T. McConnachie. Climate Policy and the Airline Industry: Emissions
Trading and Renewable Jet Fuel. PhD thesis, MIT, 2012.
[72] L. McPhail, P. Westcott, and H. Lutman. The Renewable Identification Number System and U.S. Biofuel Mandates. Technical report, US Department of
Agriculture Economic Research Service, 2011.
[73] F. Mei. Mass and Energy Balance for a Corn-to-Ethanol Plant. PhD thesis,
Washington University, 2006.
[74] G. E. Metcalf. Federal Tax Policy towards Energy. In National Bureau of
Economic Research, volume 21. MIT Press, 2007.
106
[75] E. Mielke, L. D. Anadon, and V. Narayanamurti. Water Consumption of Energy
Resource Extraction, Processing, and Conversion. Technical Report October,
Belfer Center for Science and International Affairs, Harvard Kennedy School,
Cambridge, MA, 2010.
[76]
Minnesota Technical Assistance Program. Ethanol Benchmarking and Best
Practices. Technical report, Minnesota Technical Assistance Program, 2008.
[77] T. D. Mitchell and P. D. Jones. An improved method of constructing a database
of monthly climate observations and associated high-resolution grids. International Journal of Climatology, 25(6):693-712, 2005.
[78] D. Molden, editor. Water for Food, Water for Life. International Water Management Institute, 2007.
[79] D. Morris. Ethanol: A 150 Year Struggle Toward a Renewable Future. Institue
for Local Self-Reliance, Peoria, Il, 1993.
[80] S. Mueller. Detailed Report: 2008 National Dry Mill Corn Ethanol Survey.
Technical report, University of Illinois at Chicago, Chicago, 2010.
[81] G. D. Najafpour. Biochemical Engineering and Biotechnology. Elsevier B.V.,
2007.
[82] Earth Resources Observation and Science Data Center. Global Land Survey. http://eros.usgs.gov/#/FindData/Products_andData_Available/
GLS, 2000.
[83] US Department of Agriculture. National Agricultural Statistics Service. http:
//quickstats.nass .usda.gov/, February 2013.
[84] US Department of Agriculture Forest Service. Water Supply Stress Index Model.
http: //www.wassiweb. sgcp.ncsu.edu/, March 2013.
[85] US Department of Defense. Office of the Assistant Secretary of Defense for
Operational Energy Plans and Programs. http: //energy. defense . gov/.
[86] US Department of Energy. Reduce Oil Dependance Costs.
fueleconomy.gov/feg/oildep. shtml, November 2012.
http: //www.
[87] National Research Council of the National Academies. Renewable Fuel Standard: Potential Economic and Environmental Effects of US Renewable Fuel
Policy. Technical report, The National Academies Press, 2011.
[88] M. Olson. The Rise and Decline of Nations. Yale University Press, 1984.
[89] Intergovernmental Panel on Climate Change. IPCC Fourth Assessment Report:
Climate Change 2007. Technical Report WG3.2.6.4, UNFCCC, 2007.
107
[90] Consultative Group on International Agricultural Research Consortium for
Spatial Information. Shuttle Radar Topography Mission. http: //srtm. csi.
cgiar.org/, 2006.
[91] Food & Agricultural Organization. Global Forest Resources Assessment. http:
//www.fao.org/forestry/32203/en/, 2000.
[92] Food & Agricultural Organization. Global Forest Resources Assessment. http:
//www.fao.org/forestry/fra/fra2005/en/, 2005.
[93] Food & Agricultural Organization. GAEZ Frequently Asked Questions. http:
//www.fao.org/nr/gaez/faqs/en/, 2013.
[94] M. Pearlson, C. Wollersheim, and J. I. Hileman. A Techno-economic Review
of Hydroprocessed Renewable Esters and Fatty Acids for Jet Fuel Production.
Biofuels, Bioproducts and Biorefining, 7:89-96, 2013.
[95] S. Phillips, A. Aden, J. Jechura, and D. Dayton. Thermochemical Ethanol via
Indirect Gasification and Mixed Alcohol Synthesis of Lignocellulosic Biomass.
Technical Report April, National Renewable Energy Laboratory, 2007.
[96] D. Pimentel and T. W. Patzek. Ethanol Production Using Corn, Switchgrass,
and Wood; Biodiesel Production Using Soybean and Sunflower. Natural Resources Research, 14(1):65-76, March 2005.
[97] R. Schnepf and B. D. Yacobucci. Renewable Fuel Standard (RFS): Overview
and Issues. Technical report, Congressional Research Service, Washington, DC,
2012.
[98] J. L. Schnoor, 0. C. III Doering, D. Entekhabi, E. A. Hiler, T. L. Hullar, G. D.
Tilman, W. S. Logan, N. Huddleston, and M. J. Stoever. Water Implications of
Biofuels Production in the United States. Technical report, National Reseach
Council, Washington, DC, 2008.
[99] C. D. Scown, A. Horvath, and T. E. McKone. Water Footprint of U.S. Transportation Fuels. Environmental Science & Technology, 45(7):2541-53, April
2011.
[100] Independant Chemical Information Service. Independant Chemical Information
Service. http: //www. icispricing. com/, August 2012.
[101] H. Shapouri, J. A. Duffield, and M. Wang. The Energy Balance of Corn Ethanol:
An Update. Technical report, United States Department of Agriculture, 2002.
[102] J. Sheehan, V. Camobreco, J. Duffield, M. Graboski, and H. Shapouri. Life
Cycle Inventory of Biodiesel and Petroleum Diesel for Use in an Urban Bus.
Technical Report May, National Renewable Energy Laboratory, Golden, CO,
1998.
108
[103] J. J. Sheehan. Biofuels and the Conundrum of Sustainability. Current Opinion
in Biotechnology, 20(3):318-24, June 2009.
[104] S. Siebert, P. Doell, S. Feick, J. Hoogeveen, and K. Frenken. Global Map of
Irrigation Areas. http: //www. f ao. org/nr/water/aquastat/irrigationmap/
index10.stm, 2007.
[105] J. L. Smith. Inscrutable OPEC: Behavioral Tests of the Cartel Hypothesis.
Technical report, MIT, 2003.
[106] J. G. Speight. The Chemistry and Technology of Petroleum. CRC Press, 4th
edition, 2007.
[107] M. D. Staples. Personal communication with Glenn Johnston, VP Gevo. Email,
February 2012.
[108] M. D. Staples. Personal communication with Hussain Abidi, Post-doc at MIT.
Email, January 2012.
[109] M. D. Staples. Personal communication with Kevin Weiss, CEO Byogy. Email,
March 2012.
[110] M. D. Staples. Personal communication with Wei Huang, VP LS9.
January 2012.
Email,
[111] G. Stigler. The Theory of Economic Regulation. The Bell Journal of Economics
and Management Science, 2(1):3-21, 1971.
[112] R. W. Stratton, H. M. Wong, and J. I. Hileman. Quantifying Variability in
Lifecycle Greenhouse Gas Inventories of Alternative Middle Distillate Transportation Fuels. Environmental Science & Technology, 45(10):4637-44, May
2011.
[113] K. Strzepek, J. Baker, W. Farmer, and C. A. Schlosser. Modeling Water Withdrawal and Consumption for Electricity Generation in the United States. Technical Report 221, Joint Program on the Science and Policy of Global Change,
MIT, Cambridge, 2012.
[114] US Geological Survey. Global Land Cover Characteristics Database version 2.0.
http://edcdaac.usgs.gov/glcc/glcc.html, 2001.
[115] V. Tidwell, A. C. Sun, and L. Malczynski. Biofuel Impacts on Water. Technical
Report January, Sandia National Laboratories, Albuquerque, NM, 2011.
[116] Honeywell UOP. Controlling Production of Transportation Fuels from Renewable Feedstocks. Patent: WO 2009/151692 A2, 2009.
[117] S. Vaswani. Biodiesel from Algae. Technical Report November, SRI Consulting,
Menlo Park, CA, 2009.
109
[118] K. Viscusi, J. E. Harrington, and J. M. Vernon. Economics of Regulation and
Antitrust. Massachusetts Institute of Technology, 2005.
[119] M. Wackernagel, N. B. Schulz, D. Deumling, A. C. Linares, M. Jenkins, V. Kapos, C. Monfreda, J. Loh, N. Myers, R. Norgaard, and J. Randers. Tracking
the Ecological Overshoot of the Human Economy. Proceedings of the National
Academy of Sciences of the United States of America, 99(14):9266-71, July
2002.
[120] M. Wang, H. Huo, and S. Arora. Methods of Dealing with Co-products of
Biofuels in Lifecycle Analysis and Consequent Results within the US Context.
Energy Policy, 39(10):5726-5736, October 2011.
[121] M. Wang, Ha. Lee, and J. Molburg. Allocation of Energy Use in Petroleum
Refineries to Petroleum Products: Implications for Life-Cycle Energy Use and
Emission Inventory of Petroleum Transportation Fuels. InternationalJournal
of Life cycle Assessment, 9(1):34-44, 2004.
[122] M. Wang, M. Wu, and H. Huo. Lifecycle Energy and Greenhouse Gas Emission Impacts of Different Corn Ethanol Plant Types. Environmental Research
Letters, 2(2):024001, April 2007.
[123] M. Wu, M. Mintz, M. Wang, and S. Arora. Consumptive Water Use in the
Production of Ethanol and Petroleum Gasoline. Technical report, Argonne
National Laboratory, 2009.
[124] H. Yang, Y. Zhou, and J. Liu. Land and Water Requirements of Biofuel and
Implications for Food Supply and the Environment in China. Energy Policy,
37(5):1876-1885, May 2009.
[125] F. S. Zazueta. Evaporation Loss During Sprinkler Irrigation. Technical report,
University of Florida, Gainsville, 2011.
110