National Low Carbon Fuel Standard
POLICY DESIGN RECOMMENDATIONS
19 JULY 2012
A Collaborative Study by
Institute of Transportation Studies
University of California, Davis
Department of Agricultural and Consumer
Economics and Energy Biosciences Institute
University of Illinois, Urbana-Champaign
Margaret Chase Smith Policy Center
and School of Economics
University of Maine
Environmental Sciences Division
Oak Ridge National Laboratory
International Food Policy Research Institute
Green Design Institute of
Carnegie Mellon University
L C
F S
2
!
$=8?6!6H@!3567895:!#%>1!16?AD!
!
The objectives of the National Low Carbon Fuel Standard (LCFS) Study were to (1) compare an
LCFS with other policy instruments, including the existing Renewable Fuel Standard (RFS2) and
a potential carbon tax, that have the potential to significantly reduce transportation greenhouse
gas (GHG) emissions from fuel use; and (2) propose a policy structure for an LCFS that would
be easy to implement, cost effective, and provide maximum economic gains to the consumers
and the society. The study is a collaboration between researchers from the following institutions:
Institute of Transportation Studies, University of California, Davis; Department of Agricultural
and Consumer Economics and Energy Biosciences Institute, University of Illinois, Urbana-
Champaign; Margaret Chase Smith Policy Center and School of Economics, University of
Maine; Environmental Sciences Division, Oak Ridge National Laboratory; International Food
Policy Research Institute; and Green Design Institute of Carnegie Mellon University.
This report builds on a series of papers and reports published over the past two years, including:
• Stacking low-carbon policies on the renewable fuels standard: Economic and greenhouse gas
implications
• Tradable credits system design and cost savings
• Energy security implications of a national LCFS
• Global land use change from US biofuels and finding effective mitigation strategies
• Policy options to address global land use change from biofuels
• Addressing uncertainty in life-cycle carbon intensity in a national LCFS
• Fuel electricity and plug-in electric vehicles in a national LCFS
Additional notes and discussion were also prepared on the following topics:
• Inclusion of marine bunker fuels in a national LCFS scheme
• Harmonizing low-carbon fuels policies
• Policy alternatives in reducing GHG emissions from transportation fuel uses
• Cost containment mechanism in the market-based credit markets
Individuals who contributed to the National Low Carbon Fuel Standard Study include the
following (names of the principal investigators are underlined):
• Institute of Transportation Studies, University of California, Davis: Sonia Yeh, Daniel
Sperling, Jamie Rhodes, Gouri Shanker Mishra, Nathan Parker, Julie Witcover, Christopher
Yang, Jeff Kessler, and David Ricardo Heres
• Department of Agricultural and Consumer Economics / Energy Biosciences Institute,
University of Illinois, Urbana-Champaign: Madhu Khanna, Hayri Onal, and Haixiao Hung
• Margaret Chase Smith Policy Center, and School of Economics, University of Maine:
Jonathan Rubin and Maxwell Brown
• Environmental Sciences Division, Oak Ridge National Laboratory: Paul Leiby
• International Food Policy Research Institute: Siwa Msangi and Miroslav Batka
• Green Design Institute of Carnegie Mellon University: Michael Griffin, Mathew Kocoloski,
Kimberly Mullins, and Aranya Venkatesh
In addition to research, the National LCFS Study also conducted extensive stakeholder outreach,
including the following activities:
3
!
• Presentation of seven webinars between April and June 2011 showing preliminary research
results to invited key stakeholders from industry groups, environmental NGOs, academic
scholars, and policy makers. Each webinar was attended by 40 to 70+ stakeholders and was
followed up with written comments from stakeholders and additional meetings between
researchers and stakeholders.
• Presentation of a one-day policy workshop in Washington DC in August 2011 where key
stakeholders discussed draft research results and preliminary policy recommendations.
• Co-hosting of a one-day workshop for policy makers in Washington DC in August 2011 with
the International Council on Clean Transportation (ICCT). The workshop was an update of
the progress of regional/state LCFS programs and a discussion forum for challenges and
future collaborations.
• Publication of seven research reports and two major reports summarizing key technical
analysis and policy recommendations.
• Presentation of research findings at conferences and workshops.
• Publication of journal articles and academic education on a national LCFS policy.
• Development of a National Low Carbon Fuel Standard website
(http://NationalLCFSProject.ucdavis.edu)!where we detail reports, journal articles,
stakeholder comments, relevant literature, and a collection of state/regional LCFS policies.
>?9A79F!
!
The study is funded by the Energy Foundation and the William and Flora Hewlett Foundation.
The views and opinions expressed in this paper are those of the authors alone and do not
necessarily represent those of any sponsoring organization.
%8965C6!
!
For project information, please contact Daniel Sperling ([email protected]) and Sonia Yeh
([email protected]), co-directors. Jamie Rhodes was the project managing director.
!
$CI98;:@AFG@96E!
The National LCFS Study has gone through an extensive internal and external peer-review
process participated in by more than a hundred stakeholders, including review of the seven
research reports, seven webinars, numerous face-to-face meetings and conference calls, regional
project meetings, and an one-day workshop in Washington DC discussing policy design
recommendations. We greatly appreciate all the comments and feedback provided to us. Though
their participation in no way represents an endorsement of the project conclusions nor proposed
policy design, we would like to acknowledge the following individuals/organizations (in no
particular order):
California Air Resources Board
American Petroleum Institute
Western States Petroleum Association
National Petrochemical & Refiners Association
4
!
Bob Epstein and Mary Solecki, E2
Christopher Hessler, AJW, Inc.
Ed Pike, Anil Baral, and Christopher Malins, International Council on Clean Transportation
Ralph Moran, Ruth Scotti, and Brian K Conroy, BP America, Inc.
Simon Mui, Natural Resources Defense Council
John Reese, Adam Nathan, Alex Nevill, Noor Yafai, and Angus Gillespie, Shell
Cathy Reheis-Boyd, Western States Petroleum Association
Dwight Stevenson, Tesoro Corporation
Harrison Sigworth, Jr., Jeffrey Jacobs, and Philip Heirigs, Chevron Corporation
Michelle Manion and Matt Solomon, Northeast States for Coordinated Air Use Management
Jason Mark and Patty Monahan, Energy Foundation
Lisa Holzman and Christopher Hessler, AKW, Inc.
Climate Change Secretariat, Environmental Assurance Division, Government of Alberta
W. David Montgomery, Sugandha D. Tuladhar, Robert A. Baron, and Paul D. Bernstein,
NERA Economic Consulting
Elisabeth Vrahopoulou and David L. Stern, ExxonMobil Refining and Supply
Cory-Ann Wind, Oregon Department of Environmental Quality
Douglas Scott, Illinois Environmental Protection Agency
Edward Jepsen, Wisconsin Department of Natural Resources
Ann McCabe, Ann McCabe & Associates, Inc.
Brendan Jordan and Amanda Bilek, Great Plains Institute
Ian Hodgson, Wojciech Winkler, and Ignacio Vazquez-Larruscain, Directorate General of
the Environment, European Commission
Paul Wieringa, Alternative Energy, Ministry of Energy and Mines, Government of British
Columbia
All the research reports have been submitted to the peer-reviewed journal Energy Policy to be
published in a special issue, “Low Carbon Fuel Policy.” We greatly appreciate the feedback and
comments provided by twenty-three anonymous academic reviewers. The special issue is
expected to be available online summer 2012. We also want to thank Lorraine Anderson for her
outstanding editing of the report.
!
1?FF@E6@A!C7656789!
Yeh, Sonia, Daniel Sperling, M. Griffin, Madhu Khanna, Paul Leiby, Siwa Msangi, James
Rhodes, and Jonathan Rubin. 2012. National Low Carbon Fuel Standard: Policy Design
Recommendations. Institute of Transportation Studies, University of California, Davis, Research
Report UCD-ITS-RR-12-10.
Also Available at SSRN: http://ssrn.com/abstract=2105897
! !
5
!
Executive)Summary)
The abundance and low cost of petroleum over the past 150 years has enabled rapid economic
growth and extraordinary mobility advancements. But dependence on petroleum fuels also has
large downsides, including dependence on insecure supplies, volatile prices causing high
economic costs, polluted and unhealthy air, climate change, and increasing threats to local
environments as production moves into more fragile areas.
The transition to low-carbon alternative transportation fuels is becoming more urgent. But their
introduction is inhibited by a long list of market conditions and failures. These include sunk
investments and technology lock-in by the automotive and energy industries, other forms of
technological and market inertia impeding investments in deployment and R&D, cartel pricing,
and the failure of markets to assign a price to greenhouse gas (GHG) emissions. Various policies
might be adopted to overcome these market conditions and barriers, ranging from pure market
instruments such as carbon taxes to prescriptive mandates and voluntary actions. Each has
different advantages and disadvantages. Some are easier to implement administratively, some are
more economically efficient, and some are more effective in accelerating investments. None is
perfect.
One of the most compelling, assuming some level of urgency, is a broad, performance-based
policy that targets greenhouse gas reduction—what we refer to as a low carbon fuel standard
(LCFS). In this report, we integrate scientific knowledge of alternative fuels—including an
assessment of economic, administrative, institutional, equity, political, and technological
considerations—to aid us in proposing a policy design for an LCFS for the United States. We
have aimed for a policy design that would be effective, economically efficient, and broadly
acceptable.
An LCFS is a policy designed to accelerate the transition to low-carbon alternative transportation
fuels by stimulating innovation and investment in new fuels and technologies. The goal is to
provide a durable policy framework that will stimulate innovation and technological
development. Since 2007, variations of an LCFS policy have been adopted by California,
1
the
European Union (Fuel Quality Directive, FQD),
2
and British Columbia (Renewable and Low-
Carbon Fuel Requirement Regulation, RLCFRR).
3
Other states in the United States have been
exploring the adoption of an LCFS policy, including states in the Midwest
4
and the
Northeast/Mid-Atlantic region,
5
and the states of Oregon
6
and Washington.
7
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
1
California Governor Executive Order S-01-07 (January 2007) http://www.arb.ca.gov/fuels/lcfs/eos0107.pdf
2
http://ec.europa.eu/environment/air/transport/fuel.htm
3
http://www.em.gov.bc.ca/RET/RLCFRR/Pages/default.aspx
4
http://shonic.net/LCFS/documents/LCFPagDoc.pdf
5
http://www.nescaum.org/topics/clean-fuels-standard
6
http://www.deq.state.or.us/aq/committees/lowcarbon.htm
7
http://www.ecy.wa.gov/climatechange/fuelstandards.htm
6
!
The design of an LCFS is premised on the use of technology-neutral performance targets and
credit trading, with the intent of harnessing market forces and providing industry with flexibility.
It is also premised on the use of life-cycle measurements of GHG emissions, to assure that
emissions are regulated effectively and scientifically. An LCFS is a hybrid of a regulatory and
market policy instrument. It does not include mandates for any particular fuel or technology and
as such does not attempt to pick winners or losers. Instead, it defines an average emissions
intensity standard—measured in grams CO
2
equivalent per mega-joule of fuel energy
(gCO
2
e/MJ)—that all energy providers must achieve across all fuels they provide. Many options
exist for meeting the standard. Regulated parties are free to employ any combination of strategies
that suits their particular circumstances and perspectives—including the purchase of credits from
other companies.
The breadth and reach of an LCFS, and the challenge of implementing an innovative policy,
means that adoption of a national LCFS will not be easy or straightforward and will require
careful analysis and design. It is necessary to address the cost-effectiveness of the policy
(compared with other similar GHG policies) and to analyze ease of administration, fairness,
equity, market flexibility, and impacts on energy security and sustainability. We have done so in
a companion report, National Low Carbon Fuel Standard: Technical Analysis Report (TAR).
This Policy Design Recommendations (PDR) report builds on insights and findings from the
TAR. Below we recommend key policy design principles that chart a path toward developing a
national LCFS policy.
!
1?GG5<D!*@C8GG@9A56789E!
!
Recommendation+1.+Adopt+complementary+policies+to+maximize+the+benefits+of+an+LCFS.+
An LCFS is designed to reduce GHG emissions and accelerate the introduction of nonpetroleum
transportation fuels. A wide variety of market conditions and market failures inhibit the
commercialization of nonpetroleum transportation fuels. These include the failure of markets to
assign a price to GHG emissions and other pollutants; sunk investments and technology lock-in
by the automotive and energy industries that make alternatives look disruptive and discourage
investments in new energy systems; network externalities where consumer decisions to purchase
electric and fuel cell vehicles are made separately from energy infrastructure decisions; the
market power of OPEC; high entry barriers in the automotive and fuels industries; R&D
underinvestment due to industry diffusion (especially in agriculture), R&D spillovers where
R&D findings cannot be fully captured, and learning-by-doing spillovers where societal savings
are not fully captured; conservative consumer behavior in buying new types of vehicles even
when they are economically superior; and volatile oil prices that create uncertainty that leads to
underinvestment in alternatives.
7
!
No single policy, including a carbon tax or an LCFS policy, can overcome all these market
conditions and failures. Additional, complementary policies are needed. Complementary policies
can be independently developed and targeted to address key underlying issues that are difficult to
address with broad policy solutions such as a carbon tax or an LCFS. Complementary policies to
an LCFS might include regulations that accelerate investments in new vehicle and fuel types,
basic energy and vehicle R&D, incentives for vehicles that use low-carbon fuels, policies to
decarbonize electricity generation, and sustainability requirements for fuel/feedstock production.
Many of these are already in place, adopted by local, state, or national governments—but many
are not. The success of an LCFS (or other policies to accelerate the use of low-carbon fuels)
would be aided by the adoption of complementary policies. More work is needed to carefully
evaluate complementary policies that could effectively maximize the full policy benefits of an
LCFS—and to identify existing policies that overlap or are not well aligned.!
!
Recommendation+2.+Modify+RFS2+to+incorporate+elements+of+an+LCFS,+or+replace+it+with+an+
LCFS.+
The most conspicuous example of an overlapping policy is the national Renewable Fuel
Standard, most recently updated in 2007 (RFS2). RFS2 requires specified volumes of several
types of biofuels, defined in terms of (life-cycle) carbon intensity thresholds. In contrast, an
LCFS would apply to all transport fuels, not just biofuels, and would base the requirements on
their life-cycle carbon intensity. This broader approach using a continuum of carbon intensities
would provide a stronger incentive for innovation for a broader range of fuels (including
electricity, natural gas, and hydrogen).
Our supporting studies conclude that implementing an LCFS alone or with RFS2 would be
superior to RFS2 alone in reducing GHG emissions, improving market incentives and flexibility,
and lowering domestic and international land use impacts. The impacts on energy security
relative to RSF2 would likely be beneficial. If an LCFS is to be adopted, two options are
possible: modify RFS2 to incorporate elements of an LCFS, or replace it with an LCFS. RFS2
and an LCFS could be complementary policies mutually reinforcing low-carbon fuel
development, or an LCFS could replace RFS2, acting as a new policy framework to drive low-
carbon and renewable fuel development.
!
Recommendation+3.+Initially+include+within+the+scope+of+the+LCFS+all+fuels+used+in+onDroad+
vehicles.+
In principle, it is desirable to include more types and uses of transport fuels in a national LCFS.
Including more fuels would result in greater GHG reductions and would enable more flexibility
in identifying low-cost mitigation options and increasing opportunities for regulated parties to
buy LCFS credits from a greater pool of options, thereby achieving LCFS targets in the most
cost-effective manner. Fuels used in on-road vehicles—cars, trucks, and buses—account for 80.3
8
!
percent of total transportation fuel use in the United States; we recommend including them in an
LCFS as their uses are easy to track and monitor. Electricity, hydrogen, and natural gas currently
account for less than 1 percent of total transportation fuel use in the United States, but will be
expanding as vehicle standards, the LCFS and other policies are implemented. These fuels
should be included in an LCFS, but because they tend to have lower life-cycle GHG emissions
than petroleum fuels, they would be used to generate credits for sale to petroleum fuel suppliers
(depending on verification that their carbon intensity is lower than that of gasoline and diesel).
Approximately 14.7 percent of transport fuels is used for ships and aviation. Including maritime
and aviation emissions within an LCFS would be challenging because ships and planes operate
across national boundaries. Ideally, they should be included so as to minimize emissions
leakage—whereby planes and ships evade LCFS regulations in the United States by purchasing
as much fuel as possible elsewhere. As other nations adopt LCFS rules—as EU nations have
already done—leakage will disappear, but the spread of carbon rules to ships and planes
traveling beyond US borders will likely be slow because they are regulated by international
agencies, which tend to act slowly. It may take a decade or more to establish a global policy
framework to regulate shipping and aviation GHG emissions. Nonetheless, just as the EU acted
unilaterally in capping aviation GHG emissions, regional and national policy initiatives could be
considered in the absence of international action.
Conventional transportation fuels used for off-road vehicles and outside the transportation sector
(for example, diesel fuel used for home heating) could be included in a national LCFS, but
implementation could be complex. We suggest not including these initially.
!
Recommendation+4.+Set+a+target+of+reducing+the+carbon+intensity+of+gasoline+and+diesel+by+10+
to+15+percent+by+2030.+
We recommend a target of reducing carbon intensity (CI) by 10 to 15 percent by 2030 based on
research findings of our national LCFS team. Carbon intensity is defined as life-cycle GHG
emissions (converted to carbon equivalence and expressed as gCO
2
e/MJ); the 10 to 15 percent
reduction is with respect to gasoline and diesel, the baseline fuels. The selection of a carbon
intensity reduction target calls for balancing a number of factors: the urgency of reducing GHG
emissions, expected costs of future energy supplies, expected economic impacts, variation in
costs and impacts across companies and regions, and the expected rate of induced innovation in
supplying low-CI fuels. Because there will be a lag between the time when an LCFS policy is
adopted and when investments and innovations occur, it is generally advisable to backload the
compliance schedule by starting with small annual CI reduction targets and steadily increasing
the size of the annual reduction percentages over time.
Recommendation+5.+Regulate+the+parties+responsible+for+producing,+importing,+or+supplying+
fuel.+
9
!
Regulated parties should generally be those parties responsible for producing or importing fuel
for consumption in the US transportation sector. For petroleum fuels used in transportation
(gasoline, diesel, jet fuel, bunker fuel), the regulated party should be oil refiners or importers,
along with blenders when biofuels are mixed with petroleum fuels.
For transportation fuels that are also used outside the transportation sector, the initial regulated
party should be the party responsible for supplying the fuel for transportation-sector applications.
These could be firms supplying fuel to vehicle fueling equipment, or firms owning the vehicle
fueling equipment, but not both. !
!
Recommendation+6.+Use+energy+efficiency+ratios+to+adjust+the+carbon+intensity+ratings+of+fuels+
for+diverse+propulsion+technologies.+
For an LCFS to account accurately for the full life-cycle impact of different fuels, the carbon
intensity (CI) ratings of fuels have to be adjusted by the differences in energy conversion
efficiency of vehicle engines. This adjustment is essential to correctly reflect the actual emission
reductions (in gCO
2
e per mile traveled) when replacing conventional fossil fuel with alternative
fuels that run on engines with much greater conversion efficiency, such as electric motors
compared to internal combustion engines. Adjustments are also required for fuel cell vehicles,
which are also more efficient than gasoline-powered vehicles.
These adjustment factors—energy efficiency ratios (EERs)—are best calculated by comparing
the fleet-average efficiencies of the alternative power train with the corresponding fleet-average
efficiencies of baseline fuel-vehicle technologies that the alternative fuel-vehicle technology will
displace. The values should be updated on a regular basis to ensure they adequately reflect the
evolving efficiency of vehicles on the road.
!
Recommendation+7.+Create+separate+fuel+pools+for+gasoline+and+diesel.+
We recommend that at least two separate fuel pools be established—for gasoline and diesel—
with the potential to establish additional fuel pools for jet and maritime fuels. A single fuel pool
could create incorrect incentives to increase diesel fuel sales because diesel would earn a more
favorable CI rating as a result of its higher EER compared to gasoline though actual
displacement may not have occurred. Without separate fuel pools, a refiner would have the
incentive to reduce the price of diesel fuel for sale to trucks or even foreign markets—with no
long-term GHG benefits.
To implement a fuel-pooling approach, CI reduction targets and EER values will need to be
established for each pool. Based on our research findings, we recommend unlimited LCFS credit
10
!
trading between pools in order to provide flexibility, lower compliance costs, and acknowledge
uncertainties in feedstock availability and technological progress.
Recommendation+8.+Regulate+fuels+according+to+their+lifeDcycle+GHG+emissions.+
An LCFS is premised on measuring all GHG emissions of a fuel from the source (oil well, coal
mine, farm field, and so forth) to the final point of consumption. This life-cycle approach is key
to comparing the emissions of different fuels. Calculation of life-cycle GHG emissions will
require modelers, and ultimately policy makers, to make decisions regarding modeling
approaches, system boundaries, and data sources. When multiple jurisdictions are involved, such
as nations, it will be important to harmonize the methodology used among different regulatory
agencies, creating a consistent approach for defining and measuring carbon intensity in fuels. We
recommend the following.
System boundaries. A national LCFS policy should adopt a standardized life-cycle assessment
(LCA) method for measuring fuel CI that reflects best practices and is transparent and consistent
across fuel types. Indirect emissions resulting from market-mediated effects should be evaluated
for potential inclusion when they (1) substantially impact fuel life-cycle carbon intensity (CI)
and (2) are closely linked to particular fuel supply chains (see Recommendation 9 for land use
emissions, which are the most significant indirect effect).
Spatial boundaries. Data inputs for LCA measures should be disaggregated enough spatially to
capture regional variability in supply chain emissions in ways that will incentivize greater use of
low-carbon feedstock/technology. As a convenient way of operationalizing boundary definitions,
we recommend using state boundaries for setting default CI values for biofuels, and load-
balancing area or higher levels of aggregation for electricity CI values.
Uncertainty and variability. LCA calculations, conducted for each step of the energy supply
chain, can be difficult to specify accurately. Differences in GHG emission estimates across
studies and models can be characterized as uncertainty. Uncertainty falls into three categories:
spatial and temporal variability, data limitations, and scientific uncertainty. Variability and data
limitations can be addressed through policy design and improved data collection and reporting.
Scientific uncertainty requires more research and is more difficult to accommodate but can also
be addressed through creative policy mechanisms (as indicated below for land use change
effects).
We recommend that the sources of uncertainty and variability be systematically identified and
carefully evaluated to determine default values (see next item) and to help design a more robust
GHG reduction target given uncertainties. We recommend that variability or uncertainty due to
data limitation be targeted with an opt-in reporting mechanism in the policy design to improve
11
!
data availability, reduce uncertainty, and incentivize innovation. We also recommend addressing
scientific uncertainty through adaptive management and targeted research.
Default values and opt-in mechanisms. Default CI values should be assigned to each energy path
to ease the reporting requirements of energy providers. If energy providers (the regulated
companies) can supply their fuel with lower emissions than the default values, they should be
allowed to opt in with their superior value. They would do so by documenting their lower
emissions. Allowing companies to opt in encourages innovation by rewarding producers for
reducing emissions.
Recommendation+9.+Address+GHG+emissions+from+land+use+change+(LUC)+through+shortDterm+
and+longDterm+policies.+
Most of a fuel’s life-cycle GHG emissions are directly measurable and within the energy supply
chain. But additional emissions can be caused when large amounts of land are diverted from
agriculture and other uses into energy production—which is the case with many biofuels and
some fossil fuels. The impacts of these land use changes (LUC) are complex and difficult to
quantify accurately—but accounting for them is important to assure that investments are directed
at those feedstocks with less impact. The effects can be large for land-intensive crops such as
corn but are much smaller for grass and tree feedstocks (if they are grown on marginal, degraded
land and/or if they avoid direct competition with food crops) and zero for biofuels made from
waste materials (crop and forestry residues and municipal solid waste). Oil sands production
induces small LUCs associated with soil and forest carbon emissions from peatland conversion.
We recommend adopting a flexible policy taxonomy that includes short-term and long-term
policies.
Short-term policies would induce or otherwise encourage immediate action to reduce use of
productive land for energy and other adverse impacts. They would encourage (1) using feedstock
that does not require additional land, such as wastes and agriculture residues, or feedstock that
requires less land, such as cellulosic feedstocks and algae; and (2) adopting measures that lower
LUC risk from land-using feedstock by (a) enhancing carbon sequestration and storage, (b)
encouraging the use of marginal, degraded, and abandoned land, and (c) prohibiting the
conversion of high-carbon, high-biodiversity, and environmentally sensitive areas. Despite
relatively large scientific uncertainty about LUC impacts, there are scientifically based and
increasingly well-developed estimates that should be used to ensure that only those fuels that
provide benefits are properly incentivized.
Long-term policy measures would combine short-term mitigation strategies with other incentive
mechanisms that offer the greatest potential for mitigating LUC over the long term. These
measures would encourage collaboration within and outside the biofuel supply chain to increase
investments in land use productivity, environmental protection, and carbon offset schemes. The
12
!
goal is to enhance economic productivity without compromising environmental or ecosystem
services. The regulatory process should establish rigorous and systematic evaluation frameworks,
coupled with intensified research, to assess options and implementation.
Recommendation+10.+Treat+all+crude+oils+as+part+of+the+overall+pool+of+transportation+fuels.+
Petroleum is not a uniform or homogenous liquid; it is a diverse mix of liquids comprised of
chains of hydrogen and carbon molecules. Initially California and the European Union (EU)
created a separate category of high-carbon-intensity crude oils within their LCFS and FQD. This
approach does not consider the reality that the CI of crude oils varies considerably, with some
conventional crudes, for instance, having higher CI values than some oil sands. It also runs the
risk of legal challenge from Canada, since targeting oil sands can be construed as discriminating
against a product of that country.
Instead of targeting specific high-carbon crudes, we recommend treating all crudes as part of the
overall pool of transportation fuels. We recommend adopting an approach that creates an
incentive to buy lower-CI crudes, invest in upstream improvements (such as carbon capture and
sequestration), and modify refinery designs to favor low-CI crudes. Each refinery (that is,
regulated party) would be assigned a benchmark value based on its CI in the baseline year. If it
exceeded this value, it would need to offset that increase by reducing GHG emissions in other
ways (or buying credits). If it reduced its crude oil CI, it could apply those reductions as credits
against its LCFS obligation. Some small additional shuffling of crude supply would occur—
whereby companies would send their lower-CI oil to US refineries and their higher-CI oil
elsewhere—but shuffling is a normal business practice for refineries in their effort to minimize
their costs. It is uncertain how much additional shuffling would occur. In any case, this shuffling
would diminish when other countries, starting with the EU, adopted a similar refinery-specific
approach. If the shuffling appeared to be significant, the extra transport energy consumed by
crude shuffling could perhaps be calculated and included (penalized) in the life-cycle
measurements for that crude.
!
Recommendation+11.+Harness+market+forces+using+LCFS+credits.+
As a general principle, it is desirable to harness market forces to achieve societal goals. An LCFS
does so by allowing companies to buy and sell credits. If a company prefers not to invest directly
in reducing GHG emissions to achieve its carbon-intensity target, it can buy credits from other
companies that can reduce emissions at less cost. The net effect is attainment of targets at less
overall cost.
Trading and banking. The efficiency and effectiveness of an LCFS credit market depends on the
design of the credit system, particularly the opportunities for trading and banking. Given the
uncertainties in feedstock costs and availability, their CI values, and the commercial success of
13
!
various biofuel refining technologies and fuel types (including “drop-in” bio-based gasoline and
diesel fuel), we recommend allowing unlimited trading of LCFS credits across the gasoline and
diesel fuel pools (and any others that might be created, such as jet fuels). Doing so can
significantly reduce compliance costs. For the same reason, banking of credits lowers the costs of
meeting the LCFS and stabilizes credit prices across compliance time periods.
Compliance and cost containment.!The design of an LCFS needs to address concerns about large
price swings that might result from unanticipated surges or crashes in economic growth, weather
and crop prices, and low-carbon fuel availability. While banking mechanisms reduce price
volatility, in extreme situations the number of banked credits available might not be sufficient to
avoid a credit price spike. To avoid the possibility of low-probability but high-impact price
spikes, we recommend the adoption of cost-containment mechanisms to protect regulated
companies and consumers.
Carbon credits from other programs/jurisdictions. Transportation energy is produced utilizing
many resources and technologies in many places across many political jurisdictions. GHG
emissions in some places and from some activities are, or will soon be, regulated by other (non-
LCFS) GHG programs (such as carbon caps on utilities and refiners or carbon taxes). These
energy activities are already incentivized to reduce emissions through other market instruments.
In these cases, when energy producers in other political jurisdictions are subject to other carbon
fees or taxes, including electricity producers subject to cap-and-trade fees, we recommend that
actual emission reductions along the supply chains being regulated by an LCFS be taken into
account through regular updates of default CI values, reflecting changes of emission intensity
aggregated over the industry, technology, or process over time.
Issues will arise, however, when obligated emission reductions in the other programs do not
actually occur but are met via credits, penalties, or fees, especially when there is a large disparity
between the actual or implicit carbon reduction costs or credit prices between the two programs.
We recommend recognizing these traded emission credits as actual emission reductions but
applying an adjustment factor to account for the price difference between programs, based on
published prices of credits traded in the same compliance period. For example, if a refiner pays
$15 per tonne CO
2
e of upstream emissions toward Alberta’s Specified Gas Emitters Regulation
(SGER) and LCFS credits are traded at $60 per tonne CO
2
e in the same compliance period, a
quarter of a carbon credit ($15/$60) can be counted as emission reductions. This is the same
approach we recommend for harmonizing a national LCFS with other LCFS jurisdictions (such
as British Columbia’s RLCFRR and the EU FQD), as discussed in Section 13.
+
Recommendation+12.+Implement+performanceDbased+sustainability+standards.+
Aside from GHGs, there are other important nonmarket impacts associated with energy
production. This group of sustainability concerns includes environmental sustainability
14
!
(conservation of air, water, soil, biodiversity, and land use) and social sustainability (human and
labor rights, local food security, rural development). The challenge is to determine the extent to
which an LCFS should include or be linked with rules to limit adverse impacts in these other
areas. Given the huge scale of energy production activities and their potentially large impacts,
and because an LCFS would play an instrumental role in stimulating large energy investments,
we believe that some sustainability safeguard mechanisms are needed. We recommend
formulating (1) minimum sustainability requirements, including conservation (not allowing
conversions of high-biodiversity and high-carbon-stock areas); and (2) reporting requirements
for specified impacts or voluntary certification.
A sustainability standard that includes key environmental and social impacts should be
performance based—it should not prescribe specific technology or practices but instead should
focus on measurable outcomes with clear expectations regarding performance, measurement,
verification, and enforcement. Effort should be made to identify incentive mechanisms that
motivate innovation beyond minimum compliance thresholds established by existing laws and
regulations.
Recommendation+13.+Harmonize+global+LCFS+policies.+
LCFS policies adopted in other countries and regions can vary significantly in policy design,
stringency levels, system boundaries, coverage of fuel types, and various other details. The goal
of harmonization is to create a consistent and acceptable approach for reducing the carbon
intensity of fuels to maximize the effectiveness and efficiency of the policies, while providing
individual countries and regions the freedom and flexibility to tailor the policies to their local
circumstances.
Harmonization can be achieved by adopting a globally consistent certification system, starting at
the feedstock level. Certificate harmonization allows for robust policy frameworks and thus more
room for policy and political differences, while still remaining effective. Achieving a
harmonized certification system will require an improved chain-of-custody tracking system in
order to provide transparent and reliable information about biofuel production across regions.
LCFS policies can be further harmonized between states and regions through credit
harmonization, which requires adopting unified methods (where possible) and using credit
multipliers to adjust non-unified aspects. These two methods allow credits or certificates to be
valued equivalently across regions and traded efficiently to comply with regional low carbon fuel
policies, even when they vary in stringency, system boundary, and fuel carbon ratings. Allowing
credits to be traded across countries or regions will increase policy effectiveness and efficiency
and lower the overall compliance costs. Fuel shuffling will also be reduced, hence strengthening
LCFS policy.
15
!
%89C:?E789E!
This report highlights thirteen key issues that must be addressed in the design of an LCFS for the
United States. The remainder of the report elaborates on these thirteen design issues, providing
more analysis and detail.
! )
16
!
)
Table)of)Contents)
)
!"!#$%&'!()$**+,-(....................................................................................................................(/(
%+01!(23(#24%!4%)(................................ ................................ ................................ ....................(56(
1&)%(23(%+01!)(.............................................................................................................................(57(
1&)%(23(3&8$,!)(................................................................................................ ........................... (57(
&4%,29$#%&24(............................................................................................................................(5:(
5( #2*;1!*!4%+,-(;21&#-(&4)%,$*!4%)(...........................................................................(<=(
<( ,3)<(+49(+(4+%&24+1(1#3)(................................................................................................(<<(
>( ;,28,+*(#2'!,+8!(+49()#2;!(......................................................................................(<?(
?( 0+)!1&4!(+49(%+,8!%)(.......................................................................................................(<:(
/( ;2&4%(23(,!8$1+%&24(+49(,!8$1+%!9(;+,%&!)(..............................................................(>5(
6( !4!,8-(!33&#&!4#-(,+%&2)(32,(9&'!,)!(;,2;$1)&24(%!#@42128&!)(...........................(>>(
7( )!;+,+%!(3$!1(;221)(32,(8+)21&4!(+49(9&!)!1(............................................................(>7(
:( 1&3!A#-#1!(#+,024(&4%!4)&%-(+)(,!8$1+%2,-(*!%,&#(...................................................(?5(
8.1( 1DE6@G!=8?9A5<7@E!JJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJ!KL(
MJN( 1O5675:!=8?9A5<7@E!JJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJ!KP(
MJL( Q9C@<65796D!59A!R5<75=7:76D!JJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJ!KM(
MJK( )@S5?:6!R5:?@E!59A!8O6T79!G@CH5 9 7EG E!JJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJ!PU(
B( 1+49($)!(#@+48!(8@8(!*&))&24)(....................................................................................(/<(
5=( 8@8(!*&))&24)(3,2*(#,$9!(2&1)(C&%@(@&8@!,(#+,024(&4%!4)&%-(.............................(/7(
55( 1#3)(#,!9&%)(........................................................................................................................(65(
VVJV( .<5A79F!59A!=59I79F!JJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJ!WL(
VVJN( %8GO:759C@!59A!C8E6!C896579G@96!JJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJ!WL(
VVJL( %5<=89!C<@A76E!S<8G!86H@<!O<8F<5GEXY?<7EA7C6789E!JJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJ!WW(
5<( ;!,32,*+4#!A0+)!9()$)%+&4+0&1&%-()%+49+,9)(.........................................................(6:(
5>( 8120+1(1#3)(;21&#&!)(.........................................................................................................(75(
VLJV( >@@AE68CI!C@<67S7C56@!H5<G897Z56789!JJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJ![N(
VLJN( %5<=89!C<@A76!H5<G897Z56789!JJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJ![L(
#24#1$)&24)(..............................................................................................................................(7?(
,!3!,!4#!)(.................................................................................................................................(7/(
!
! !
17
!
List)of)Tables)
Table 1. US transportation fuel consumption by fuel type and segment (trillions of Btu’s)!"""""""""""""""""""""""""""""""""""""""!#$(
%&'()!#"!*+,,&-.!/0,1&-2304!05!10(2/.!'&3)(24)3!&46!78!9&-:)93!""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""!;<(
%&'()!;"!==>!?&(+)3!+3)6!24!7&(250-42&@3!A7B*!""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""!;C(
%&'()!C"!8((+39-&92?)!6-2?)D9-&24!E )2:F 9)6 !&?)-& :) !==> 3!50-!)()/9-2/29. !&3!&!:& 30(24) !-)1(& /), )4 9 !"""""""""""""""""""""""""""""!;$(
%&'()!$"!G4!)H&,1()!05!/-)629!:)4)-&9204!-&9)3!50-!7IJ!+46)-!7&(250-42&@3!A7B*!"""""""""""""""""""""""""""""""" """""""""""""""""""""""!;K(
%&'()!L"!*2,1(252)6!2((+39-&9204!05!,+(92D100(!==>!/&(/+(&9204!50-!62)3)(!5+)(!"""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""!CM(
%&'()!N"!A25)D/./()!JOJ!),233204!/&9):0-2)3!&46!,06)(3!+3)6!50-!)()/9-2/29.P!50332(!5+)(P!&46!'205+)(!+46)-!9F)!Q*!
R=SGT!>B*#P!7&(250 -42&!R7 G >U T!A7 B*P!=Q !> =VP!& 46!U-2923F!7 0 ( + ,'2&!RU7T!>A7B>>!""""""""""""""""""""""""""""""""""""""""""!CC(
%&'()!W"!X?)-?2)E!0 5!1 -2/) !/0 49 &24 , ) 4 9!, ) /F & 423 , 3!24 !, & -Y )9D'&3)6!-):+(&92043!""""""""""""""""""""""""""""""""""""""""""""""""""!L$(
%&'()!K"!G(')-9&@3!9F-) )!1 -24 /21 &(!: -) )4 F 0+ 3) !: &3 !10 (2/2) 3!"""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""!LN(
!
List)of)Figures)
B2:+-)!M"!G/9+&(!R#<<<Z#<<KT!&46!1-0[)/9)6!R#<M<Z#<;<T!/F&4:)3!24!9F)!/&-'04!249)4329.!05!1)9-0()+,D'&3)6!5+)(3"
!""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""!#K(
Figure 2. Projected new car fuel consumption (miles per gallon) by technology type!"""""""""""""""""""""""""""""""""""""""""""""""!;L(
B2:+-)!;"!B+)(!)/040,.!2,1-0?),)49!&46!7X
#
!),233204!-)6+/9204!5-0,!F.'-26!&46!62)3)(!?)F2/()3P!/0,1&-)6!62-)/9(.!
E29F!9F)!,039!32,2(&-!/04?)49204&(!:&30(24)!,06)(3!,&6)!'.!9F)!3&,)!&+90,&Y)-!"""""""""""""""""""""""""""""""""""!;N(
B2:+-)!C.!V)(2?)-)6!'20,&33!.2)(63!24!9F)!Q429)6!*9&9)3!'.!-):204!R904!6-.!,&33\F&T!""""""""""""""""""""""""""""""""""""""""""""""""!CL(
Figure 5. Regional variability in mean GHG emissions for corn (top) and switchgrass (bottom) ethanol!""""""""""""""""!CN(
B2:+-)!L"!Cumulative distribution of life-cycle carbon intensity of electricity for different levels of regional
disaggregation!""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""!CW(
Figure 7. Opt-in CI values (average and ranges) and number of applications by feedstock pathway for California’s
LCFS as of November 2011!""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""!$M(
B2:+-)!W"!J)4)-&(!39-&9):2)3!50-!&66-)3324:!'205+)(!AQ7!"""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""!$C(
Figure 9. Comparison of life-cycle GHG emissions of gasoline produced from crude oil obtained from different
regions with a base scenario!""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""!$K(
B2:+-)!M<"!Q*!&?)-&:)!&46!-):204&(!(25)D/./()!JOJ!),2332043!5-0,!:&30(24)!R1-0'&'2(29.!6239-2'+92043T"!SGVV]!
S)9-0()+,!G6,24239-&9204!V)5)43)!V239-2/9!""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""!$K(
! !
18
!
Introduction)
A low carbon fuel standard (LCFS) is a technology-forcing policy designed specifically to target
reduction of greenhouse gas (GHG) emissions in the transportation sector. Since 2007, variations
of an LCFS policy have been adopted by California,
8
the European Union (Fuel Quality
Directive, FQD),
9
and British Columbia (Renewable and Low-Carbon Fuel Requirement
Regulation, RLCFRR).
10
Significant progress has also been made through state and regional
LCFS initiatives in the Midwest,
11
the Northeast/Mid-Atlantic states,
12
and the states of Oregon
13
and Washington.
14
The LCFS framework is specifically designed to advance broad climate policy objectives in the
unique context of the transportation sector given the need to overcome market barriers and
market inefficiency that economy-wide climate policy instruments, such as carbon taxes or cap-
and-trade schemes, fail to address. The primary objectives of an LCFS are to
• provide significant reductions in GHG emissions from transportation;
• stimulate innovation, technological development, and deployment of low-emission
alternatives for fueling the transport sector; and
• provide a durable framework for regulating GHG emissions in the transportation sector
within a broader portfolio of climate policies.
In advancing these objectives, an LCFS is notable for its structure as a technology-neutral
performance standard. It does not include mandates for any particular fuel, technology, or
compliance strategy and as such does not attempt to pick winners or losers. Instead, it defines an
average emissions intensity standard, measured in grams CO
2
equivalent per mega-joule of fuel
energy (gCO
2
e/MJ), which all regulated parties must achieve across all fuels they provide. Many
options exist for meeting the standard, and regulated parties are free to employ any combination
of strategies that suit their particular circumstances and perspectives.
Transportation-sector emissions abatement requires a multipronged approach, like a three-legged
stool comprising increased vehicle fuel efficiency, reduced carbon intensity of fuels, and reduced
vehicle miles traveled. Each of these “legs” offers potentially significant reductions in
transportation-sector emissions. An LCFS strategically targets only fuel carbon intensity; vehicle
fuel efficiency and vehicle use are better addressed using other policy approaches (for example,
CAFE standards and mass transit initiatives, respectively). This, along with other considerations
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
8
California Governor Executive Order S-01-07 (January 2007) http://www.arb.ca.gov/fuels/lcfs/eos0107.pdf
9
http://ec.europa.eu/environment/air/transport/fuel.htm
10
http://www.em.gov.bc.ca/RET/RLCFRR/Pages/default.aspx
11
http://shonic.net/LCFS/documents/LCFPagDoc.pdf
12
http://www.nescaum.org/topics/clean-fuels-standard
13
http://www.deq.state.or.us/aq/committees/lowcarbon.htm
14
http://www.ecy.wa.gov/climatechange/fuelstandards.htm
19
!
noted below, reflects the notion that an LCFS is best viewed as a strategically important
component within a portfolio of climate policies.
Significant progress is being made through a number of state and regional LCFS initiatives, but a
national LCFS could provide important benefits through policy uniformity at the national level.
Transportation fuel supply chains and the operational service areas of fuel suppliers often extend
far beyond the jurisdictional boundaries of state and regional initiatives. Imposing inconsistent
obligations under multiple state and regional LCFS policies could increase the complexity and
costs of compliance and precipitate multiple redundant regulatory systems. The potential for
differential treatment of individual fuels under multiple state and regional policies could also
increase incentives for shuffling fuels between jurisdictions on the basis of their treatment under
respective policies. A national LCFS policy that supersedes, integrates, or bridges emerging state
and regional policies could mitigate these risks, ensure credit transportability, and provide
additional opportunities for supplying low-carbon fuels through its broader coverage area,
thereby reducing compliance costs and increasing efficiency. Realizing these benefits requires a
national LCFS policy design that is consistent with existing and proposed state and regional
initiatives.
The principal objective of this report is to integrate available information and experience gained
from the earlier implementation of fuels policies and from a major set of collaborative studies
conducted over the past two years. In considering the various design alternatives and developing
our policy design recommendations, we followed the general principles that a policy should
• be scientifically grounded and defensible,
• be administratively easy to implement and broadly consistent with emerging state and
regional initiatives,
• be cost-effective in advancing the underlying policy objectives compared with other
alternatives,
• identify potential negative consequences and mitigate them as much as possible, and
• accommodate political dynamics likely to affect its implementation.
Integrating these principles was not always straightforward. Toward this end, we have attempted
to integrate the combined expert judgment embodied in the research team and in the thoughtful
comments and feedback provided by stakeholder participants.
As noted above, this report is released with a companion Technical Analysis Report (TAR),
which summarizes the technical and policy analyses of key issues informing and underlying the
recommendations presented here. In certain cases a single approach for addressing particular
policy features did not clearly emerge. In such cases alternate design decisions are presented and
expectations regarding their implications are summarized. This Policy Design Recommendations
(PDR) report provides concluding discussion of key policy design recommendations and
attempts to chart a path toward developing a national LCFS policy.
20
!
Each policy recommendation for a national LCFS policy is followed by a discussion
summarizing the issues addressed by the recommendation, possible design alternatives, and the
rationale for the recommended design.
1 Complementary)Policy)Instruments)
Recommendation+1.+Adopt+complementary+policies+to+maximize+the+benefits+of+an+LCFS.+
Key Issues
No single policy, including a carbon tax or an LCFS policy, can overcome all identifiable market
conditions and failures. Additional, complementary policies are needed. Complementary policies
can be independently developed and targeted to address key underlying issues that are difficult to
address with broad policy solutions. The key policy design questions are (1) Why are
complementary policies other than a carbon tax needed? and (2) What should the complementary
policies be?
Summary and Recommendations
A wide variety of market conditions and market failures inhibit the commercialization of
nonpetroleum transportation fuels. These include the failure of markets to assign a price to GHG
emissions; sunk investments and technology lock-in by the automotive and energy industries that
make alternatives look disruptive and discourage investments in new energy systems; network
externalities where consumer decisions to purchase electric and fuel cell vehicles are made
separately from energy infrastructure decisions; the market power of OPEC; high entry barriers
in the auto industry; R&D underinvestment due to industry diffusion (especially in agriculture),
R&D spillovers where R&D findings cannot be fully captured, and learning-by-doing spillovers
where societal savings are not fully captured; conservative consumer behavior in buying new
types of vehicles even when they are economically superior; and volatile oil prices that create
uncertainty, which leads to underinvestment in alternatives.
No single policy, including a carbon tax or an LCFS policy, can overcome all these market
conditions and failures. Additional, complementary policies are needed. Complementary policies
can be independently developed and targeted to address key underlying issues that are difficult to
address with broad policy solutions such as a carbon tax or an LCFS. Complementary policies to
an LCFS might include mandates that jump-start investments in new vehicle and fuel types,
basic energy and vehicle R&D, incentives for vehicles that use low-carbon fuels, policies to
decarbonize electricity generation, and sustainability requirements for fuel/feedstock production.
Many of these are already in place, adopted by local, state, or national governments—but many
21
!
are not. The success of an LCFS (or other policies to accelerate the use of low-carbon fuels)
would be aided by the adoption of complementary policies. More work is needed to carefully
evaluate complementary policies that could effectively maximize the full policy benefits of an
LCFS—and to identify existing policies that overlap or are not well aligned.!
Discussion
The Kaya identity serves to define a relationship between GHG emissions and the factors that
influence them, where CO
2
is a function of population, GDP, and energy:
!!
!
= !"#$%&'(")!× !
!"#
!"#$%&'(")
×
!"#$%&
!"#
×
!!
!
!"#$%&
In the transportation sector, the Kaya equation can be rewritten as
!!
!
= !"#$%&'(")!× !
!"#$%!!"# $%&%'
!"#$%&'(")
×
!"#$%&
!"#$%!!"# $%&%'
×
!!
!
!"#!"#
Instead of GDP/population measuring economic growth, miles traveled /population measures
travel demand growth, or “mobility.” Energy/miles traveled measures the energy intensity of
travel, and CO2/energy measures the GHG intensity of energy use.
A broad policy approach will target the left-hand side of the Kaya identity—limiting total CO
2
emissions. Most of the economic models ignore externalities associated with climate mitigation,
as these costs are difficult to estimate and have rarely been quantified before. Yet these costs are
real, and without specific policies to address these market barriers, economic models
significantly overestimate the efficiency of the market and the potential to reduce GHG
emissions without complementary policies.
Typical market failures that have not been considered in economic studies include the following:
• Failure of markets to assign a price to GHG emissions
• Sunk investments and technology lock-in by the automotive and energy industries
• The principal agent problem where operators of rental cars, truck trailers, leased vehicles,
and cars for legislators/execs are not the buyers of the vehicles
• Network externalities where complementary products require
large nonrecoverable investments and investments that cannot be made by individual
consumers—such as when different vehicles or different infrastructures are needed (H2, bike
paths for biking, smart paratransit, and so on)
• The market power of OPEC (cartels, oligopolies, and the like)
• High entry barriers in the auto industry
• R&D underinvestment due to industry diffusion (especially in agriculture), R&D spillovers
where R&D findings cannot be fully captured, and learning-by-doing spillovers where
22
!
societal savings not fully captured
• Consumer behavior in buying new types of vehicles making decisions that underinvest in
efficiency (related to lack of information and loss aversion)
• Volatile oil prices that create uncertainty, which leads to underinvestment in alternatives
In view of these failures, it will be necessary to develop targeted policies other than policies that
rely on markets (including C tax and the cap=and-trade) to overcome these market barriers.
Similarly, complementary policies to an LCFS might include mandates that jump-start
investments in new vehicle and fuel types, basic energy and vehicle R&D, incentives for vehicles
that use low-carbon fuels, policies to decarbonize electricity generation, and sustainability
requirements for fuel/feedstock production. Many of these are already in place, adopted by local,
state, or national governments. For example, California has the Clean Fuels Outlet (CFO)
Regulation to ensure that an appropriate number of fueling stations will dispense a designated
fuel once a certain number of vehicles using that fuel are certified in California to the Low
Emission Vehicle (LEV) standard. California also has the Zero Emission Vehicle (ZEV)
mandates that support and accelerate the numbers of plug-in hybrids and zero emission vehicles
in California. More work is needed to carefully evaluate complementary policies that could
effectively maximize the full policy benefits of an LCFS. Existing policies that overlap or are not
well aligned should be deleted or modified. New policies that are complementary and synergistic
should be adopted, often at the local or state level.
2 RFS2)and)a)National)LCFS)
Recommendation+2.+Modify+RFS2+to+incorporate+elements+of+an+LCFS,+or+replace+it+with+an+
LCFS.+
Key Issues
The most conspicuous example of an overlapping policy is the national Renewable Fuel
Standard, most recently updated in 2007 (RFS2). An LCFS would work differently from RFS2 in
that it would incentivize fuel improvement, whereas RFS2 creates a tiered system with specific
GHG emission targets and volumetric requirements that fuels need to meet. Is a national LCFS
necessary, given RFS2? How would an LCFS improve upon RFS2? Should RFS2 be replaced by
an LCFS or should features of RFS2 be improved based on LCFS principles?
Summary and Recommendations
RFS2 requires specified volumes of several types of biofuels, defined in terms of (life-cycle)
carbon intensity thresholds. In contrast, an LCFS would apply to all transport fuels, not just
23
!
biofuels, and would base the requirements on their life-cycle carbon intensity. This broader
approach using a continuum of carbon intensities would provide a stronger incentive for
innovation for a broader range of fuels (including electricity, natural gas, and hydrogen).
Our studies conclude that implementing an LCFS alone or with RFS2 would generally be
superior to RFS2 alone in reducing GHG emissions, improving market incentives and flexibility,
and lowering domestic and international land use impacts. The impacts on energy security
relative to RSF2 would likely be positive or small, but not negative. If an LCFS is to be adopted,
two options are possible: modify the RFS to incorporate elements of an LCFS, or replace it with
an LCFS. RFS2 and an LCFS could be complementary policies mutually reinforcing low-carbon
fuel development, or an LCFS could replace RFS2, acting as a new policy framework to drive
low-carbon and renewable fuel development.
Discussion
An LCFS policy is a diverse, multifaceted policy that is geared specifically toward pushing low-
carbon fuel technology into the marketplace. While an LCFS is not a catch-all policy, it could
build upon the already existing RFS2 framework to drive low-carbon fuel deployment beyond
what RFS2 is already able to do.
RFS2 creates a volumetric fuel mandate and specifies carbon reduction tiers, or categories, to
lump renewable fuels into and to prequalify fuels. If a fuel meets a specific category
requirement, it is eligible to generate renewable identification number (RIN) credits for that fuel
type that can be applied to an obligated party’s regulated volume requirement.
An LCFS would define total carbon-reduction goals for regulated parties. Rather than requiring a
specific volume of fuel, it would require the regulated party to meet an aggregated baseline
carbon intensity (gCO
2
e/MJ) with its fuels. The regulated party could achieve this either by
increasing the proportion of low-carbon fuels it sells or by decreasing the carbon content of the
fuels it sells (making them lower in carbon intensity). If both RFS2 and an LCFS were in place,
RFS2 would provide volume-based incentives for fuels while the LCFS would provide carbon-
intensity-based incentives for fuels. If the volume of fuel required by RFS2 were greater than the
volume that would be achieved through low-carbon fuels required by the LCFS, RFS2 would act
as an additional deployment subsidy (Huang et al. 2012).
It is notable that RFS2 does not have provisions to allow novel fuels—many of which are
arguably needed to meet transportation emission reduction goals in the long term (IEA 2010;
McKinsey&Company 2009)—to count toward the standard. An LCFS would create provisions
for electricity, hydrogen, and novel fuels through use of an opt-in mechanism that would allow
for nonconventional-fuel providers to generate LCFS credits.
24
!
Our studies conclude that implementing an LCFS alone or with RFS2 would generally be
superior to RFS2 alone in reducing GHG emissions (see TAR Chapter 2), improving market
incentives and flexibility (see TAR Chapter 2), and lowering domestic and international land use
impacts (see TAR Chapter 5). The impacts on energy security relative to RSF2 would likely be
positive or small, but not negative (see TAR Chapter 6). These results are explained in detail in
the TAR.
RFS2 and an LCFS could be complementary policies mutually reinforcing low-carbon fuel
development, or an LCFS could replace RFS2, acting as a new policy framework to drive low-
carbon and renewable fuel development.
!
!
3 Program)Coverage)and)Scope)
Recommendation+3.+Initially+include+within+the+scope+of+the+LCFS+all+fuels+used+in+onDroad+
vehicles.+
Key Issues
As an LCFS aims to reduce the average carbon intensity of transportation fuels, the program
coverage and scope will affect the stringency of the standard as well as the degree of flexibility
for regulated parties to meet the standard. In principle, it is desirable to include more types and
uses of transport fuels in a national LCFS. Including more fuels would result in greater GHG
reductions and would enable broader and more flexibility in identifying low-cost mitigation
options, increase opportunities for regulated parties to buy LCFS credits from a greater pool of
options, thereby achieving LCFS targets in the most cost-effective manner. On the other hand, a
broadly defined coverage would impose more logistic and regulatory challenges and could
increase the risk of fuel shuffling.
Summary Recommendations
Fuels used in on-road vehicles—cars, trucks, buses, freight, and rail—account for 80.3 percent of
total transportation fuel use in the United States; we recommend including them in an LCFS as
their uses are easy to track and monitor. Small amounts of electricity, hydrogen, and natural gas
account for less than 1 percent of total transportation fuel use in the United States. These should
be included in an LCFS, but because they tend to have lower life-cycle GHG emissions than
petroleum fuels, they would be used to generate credits for sale to petroleum fuel suppliers
(depending on verification that their carbon intensity is lower than that of gasoline and diesel).
25
!
Approximately 14.7 percent of transport fuels is used for ships and aviation. Including maritime
and aviation emissions within an LCFS would be challenging because ships and planes operate
across national boundaries. Ideally, they should be included so as to minimize emissions
leakage—planes and ships evading LCFS regulations in the United States by purchasing as much
fuel as possible elsewhere. As other nations adopt LCFS rules, leakage will disappear, but the
spreading of carbon rules to ships and planes traveling beyond US borders will likely be slow
because they are regulated by international agencies, which tend to act slowly. It may take a
decade or more to establish a global policy/framework to regulate shipping and aviation GHG
emissions. Nonetheless, just as the EU acted unilaterally in capping aviation GHG emissions,
regional and national policy initiatives could be considered in the absence of international action.
Conventional transportation fuels used for off-road vehicles and outside the transportation sector
(for example, diesel fuel used for home heating) could be included in a national LCFS, but
implementation could be complex. We suggest not including these initially.
Discussion
A breakdown of current transportation fuel use by fuel type and segment is provided in Table 1,
based on data from the US Department of Energy’s Annual Energy Outlook 2011 (U.S. EIA
2011a). Reference case projections for 2035 (not shown here) indicate similar patterns of fuel
consumption but with increases in electricity, ethanol, and diesel fuel use by vehicles. These
figures indicate that liquid transportation fuels (particularly motor gasoline, diesel fuel, ethanol,
residual oil, and aviation fuels) account for more than 99 percent of current and expected future
transportation fuel use.
Table 1. US transportation fuel consumption by fuel type and segment (trillions of Btu’s)
26
!
Various other transportation fuels—including electricity, hydrogen, natural gas, liquefied
petroleum gases, and novel biofuels such as hydrocarbons produced from algae—currently make
only modest contributions to the transportation fuel mix; however, their use as transportation
fuels may increase considerably in certain scenarios. Such increases may be accelerated by an
effective national LCFS policy due to their potentially lower carbon intensity than petroleum.
However, recent evidence also suggests that not all alternative fuels are inherently low carbon,
and significant variations exist depending on their sources, production methods, and uses
(Burnham et al. 2011; Howarth, Santoro, and Ingraffea 2011; Samaras and Meisterling 2008).
Therefore, a consistent performance standard should also be applied to these alternative fuels that
are likely to play increasingly important roles in future transportation. Within this context,
parties supplying these fuels for transportation should be allowed to opt in to an LCFS policy for
the purpose of generating LCFS credits.
Notwithstanding these general points regarding fuel types covered by a national LCFS, several
segments of the transportation sector listed in Table 1 warrant special consideration.
Shipping. International maritime vessels can shift refueling patterns in response to even modest
price signals (Michaelis 1997; Mishra and Yeh 2011).
15
The shipping industry is likely to witness
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
15
The bunker fuel industry is highly cost competitive and ship operators choose their bunker source on the basis of small price
differentials. An often-quoted example is that of the introduction and subsequent repealing of a sales tax on bunkers sold at
the Port of Los Angeles / Long Beach in California (Michaelis 1997). Before the introduction of the tax in 1992, the LA/LB
bunker market had a monthly turnover of around 4.5 million barrels. After introduction of an 8.5-percent sales tax, bunker
sales dropped below 1 million barrels and shifted largely to Panama, which is en-route for many ships calling the ports of
LA/LB. As a result, the tax was rescinded within one year. None of the major bunker markets—Houston, Singapore,
Rotterdam, and LA/LB—impose any taxes on bunkers sold to international shipping.
Fuel
Freight
trucks
Freight
Rail
Rec.
Boating
Domestic
Shipping
International
Shipping
Air
Transport
Military
Bus
Total
%
Motor
Gasoline
354
212
25.3
16,881
63%
E85
2.63
0.0%
Distillate
Fuel Oil
(diesel)
4,029
528
51.8
154
59.3
161
232
5,691
21%
Compressed
Natural Gas
7.15
9.17
27.85
0.1%
Liquefied
Petroleum
Gases
16.41
0.17
19.6
0.1%
Electricity
23.3
0.1%
Residual Oil
54.7
732
17.5
804
3%
Jet Fuel &
Aviation
Gasoline
2,583
586
3,169
12%
Total
4,406
528
263
208
792
2,583
764
267
Percent
17%
2.0%
1.0%
0.8%
3.0%
9.7%
2.9%
1.0%
1
Includes light duty vehicles and commercial light trucks. Source: 2010 data from Table 37 in AEO 2011.
!
27
!
significant change in fuel use and increases in operating costs as a result of MARPOL Annex VI
regulation of sulphur emissions. The regulation will force the industry to transition from
inexpensive but dirty residual oil–dominated bunker fuels to cleaner, low sulphur distillate–
dominated bunker fuel by 2020/2025. Even in the absence of GHG regulation, fuel costs are
likely to increase sharply in the future and provide strong incentives to reduce fuel consumption
(Mishra and Yeh 2011). The structure of the bunker industry precludes the potential to regulate
the refinery or bunker supplier due to the high risk of leakage and thus adverse economic
impacts. Incentive-based policy, however, has the potential to work best for this particular
industry to take advantage of climate finance under market-based instruments (IMF 2011). Our
rough estimates suggest that if an LCFS were applied to the shipping industry and assuming no
leakage, which is very unlikely, it could contribute to a roughly 19-million-tonne GHG reduction
at US$33–330/tonne CO
2
e (total abatement costs will be US$0.7–4 billion to reduce the well-to-
wheel CO
2
e intensity of bunkers sold in the US by 10 percent) (Mishra and Yeh 2011).
Aviation. Including fuels used in international aviation, even on an opt-in basis, may be
complicated by commitments agreed upon under the 1944 Convention on International Civil
Aviation (CICA) that such fuel “be exempt from customs duty, inspection fees or similar
national or local duties and charges”
16
The CICA’s implications for a national LCFS are not
clear, however, because an LCFS is a performance standard, rather than a duty, fee, or charge.
Legal analysis is required before firm conclusions can be drawn in this respect. The aviation
industry has been particularly active in exploring potential low-carbon bio-jet fuels under the
renewable biofuel programs in the United States and the European Union.
17
Including aviation
fuels under a national LCFS could further strengthen these incentives (IMF 2011).
Military. Fuel used in tactical military applications should be exempt from a national LCFS on
the basis of national security interests. The US military has, however, taken a leadership role in
reducing the carbon intensity of its fuel supplies—reflected in key provisions in the Energy
Independence and Security Act of 2007 and the Great Green Fleet initiative.
Rail. Fuel used in rail transportation should also be allowed to opt in to an LCFS program to
generate credits. This would provide additional incentives for more efficient and low-carbon
transportation modes, such as freight hauling (Greene, Baker, and Plotkin 2011; U.S. DOT
2010), to compete with diesel-fueled trucks on a more level playing field. Careful analysis will
be needed in the future to determine the basis for comparison (for example, to arrive at
appropriate adjustment factors to account for the different efficiencies of the modes).
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
16
Convention on International Civil Aviation, Dec. 7, 1944, 61 Stat. 1180, 15 U.N.T.S. 295
17
The European Council recently adopted a directive (document 3657/08) that includes aviation activities in the EU
GHG emission allowance trading system (ETS). As of 1 January 2012, all flights arriving at or departing from an
EU airport will be included in the scheme. The directive is being challenged by industry trade groups in the
European Court of Justice (ECJ).
28
!
Nontransport. Certain nontransport applications of transportation fuels —such as home heating
with distillate (diesel) fuel and biofuels—have been considered for inclusion in regional LCFS
efforts, such as home heating oil (No. 2 distillate) in the Northeast/Mid-Atlantic LCFS. The
considerations are that (1) it will present significant challenges to separate out fuel oil used for
space heating vs. for transportation, with the proportion roughly 50/50 in the region; (2)
including nontransport applications will incentivize additional GHG emission reductions;
18
(3)
including nontransport applications will reduce the possibility of fuel shuffling, in which high-CI
fuels are diverted to stationary applications in exchange for equivalent fuels with lower CI. In
this particular case, the inclusion of heating oil could offer the additional benefit of simplifying
fuel accounting by focusing attention on transportation fuels supplied rather than on final use,
and allowing greater consistency with the RFS2 policy, which qualifies renewable fuels used for
home heating. Situations like this could be considered on a regional basis, as the goals of a
national LCFS should be broadly consistent with state or regional initiatives and should support
credit trading and potential policy integration (Kessler, Yeh, and Sperling 2012).
4 Baseline)and)Targets)
Recommendation+4.+Set+a+target+of+reducing+the+carbon+intensity+of+gasoline+and+diesel+by+10+
to+15+percent+by+2030.+
Key Issues
The policy baseline is the initial carbon intensity value from which reduction targets are
measured under an LCFS. Several factors influence the stringency of an LCFS: the baseline
carbon intensity, the target reduction, and the phasedown schedule (rate of reduction from the
baseline carbon intensity to the target). These factors affect not only the level of policy
stringency but also the cost-effectiveness and the economic impacts of an LCFS policy.
Summary Recommendations
We recommend a target of reducing carbon intensity (CI) by 10 to 15 percent by 2030 based on
research findings of the national LCFS team. Carbon intensity is defined as life-cycle GHG
emissions (converted to carbon equivalence and expressed as gCO
2
e/MJ); the 10-to-15-percent
reduction is with respect to gasoline and diesel, the baseline fuels. The selection of a carbon
intensity reduction target calls for balancing a number of factors: the urgency of reducing GHG
emissions, expected costs of future energy supplies, expected economic impacts, variation in
costs and impacts across companies and regions, and the expected rate of induced innovation in
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
18
http://www.ct.gov/dep/lib/dep/air/climatechange/lcfs_mou_govs_12-30-09.pdf
29
!
supplying low-CI fuels. Because there will be a lag between the time when an LCFS policy is
adopted and when investments and innovations occur, it is generally advisable to backload the
compliance schedule by starting with small annual CI reduction targets and steadily increasing
the size of the annual reduction percentages over time.
Discussion
The carbon intensity (CI) of petroleum fuels has been increasing over time because heavier and
more unconventional crude sources are being used (U.S. EIA 2010b), which require more energy
for extraction and processing (Brandt 2008, 2011). As the share of unconventional crude oil from
oil sands rises from 9.4 percent of US domestic petroleum consumption in 2010 to 18.1 percent
in 2020, the CIs of gasoline and diesel are expected to increase from 93.1 to 96.3 gCO
2
e/MJ for
gasoline and from 92.0 to 97.1 gCO
2
e/MJ for diesel over the period 2005–2035 (Figure!1)
(Rubin and Leiby 2012). On the other hand, the average CI of the transportation fuel mix will
remain relatively flat or decrease gradually, due to the mandated mix of biofuels under RFS2
(Huang et al. 2012; Rubin and Leiby 2012; Yeh and Sperling 2010).
As a result, the stringency of an LCFS policy will vary depending on the baseline selected,
ranging from the least stringent (that is, highest CI baseline if using the actual 2012 petro-
gasoline and petro-diesel fuel CIs) to the most stringent (that is, lowest CI baseline if using the
actual 2005 gasoline and diesel fuel CIs) (Rubin and Leiby 2012). We recommend using a
baseline CI of regulated transportation fuels for the most recent year for which there is data, such
as the CI of 2011 fuel mix for the gasoline and diesel fuel pool). This translates to a medium-
level stringency reflecting the actual baseline CI of the regulated fuel pool(s).
Figure 1. Actual (2000–2009) and projected (2010–2030) changes in the carbon intensity of
petroleum-based fuels. Source: Rubin and Leiby 2012.
!
30
!
LCFS targets define the CI reductions to be achieved within a certain time frame, such as a 10-
percent reduction by 2030. The national LCFS analysis examined a wide range of policy options
at varying CI baseline and target levels. These options used a range of assumptions, varying (1)
feedstock costs, availability, and CI values including the values of indirect land use change
(iLUC); and (2) policy options in treating high-carbon crude and the existence of other
complementary/contradictory policies as discussed in detail in our technical analysis papers
(Huang et al. 2012; Rubin and Leiby 2012).
As summarized in the TAR, our studies found that when an LCFS and RFS2 are combined, the
success of the LCFS policy depends on whether that policy can further incentivize low-carbon
feedstocks beyond those required by RFS2 at reasonable costs. The impacts of a national LCFS
are fundamentally determined by (1) the availability of low-carbon fuels, (2) the compliance
path, (3) the reference level CI of the fuel baseline, and (4) the degree of flexibility in the credit
system. Setting a 15-percent reduction target below a historical level such as that of 2005 can be
an achievable target provided there is a flexible credit trading and banking design (Huang et al.
2012; Rubin and Leiby 2012).
The phasedown schedule—that is, the amount of CI reduction annually within the policy time
frame—also affects the stringency of the policy. For example, LCFS policies in California and
British Columbia have both adopted a “technology-forcing” CI trajectory, in which modest
reductions in early years are followed by more substantial reductions later on (Table!2). This
backloading of CI reductions is intended to provide sufficient lag time to develop new low-
carbon fuel supplies, including R&D, construction of biorefineries, development of feedstock
supplies and infrastructure, and system integration (NRC 2011). Some stakeholders have
suggested that this approach may create additional challenges to financing low-carbon fuel
development, as modest initial reduction targets yield relatively low LCFS credit prices early in
the program. Our credit analysis study demonstrates that uncertainty in mitigation costs,
feedstock and technology availability, and credit prices can largely be mitigated via credit
trading and banking (Rubin and Leiby 2012).
Table 2. Summary comparison of policy baselines and CI targets
Policy design feature
CA LCFS
BC LCFS
OR
LCFS
1
WA LCFS
1
NEMA
LCFS
1
Gasoline baseline CI (gCO
2
e/MJ)
95.85
82.40
90.38
92.2
Baseline basis
2010
2
2010
2
2010
3
2007
2010
Reduction target
10%
10%
10%
10%
10%
Target year for achieving target
2020
2020
2022
2023
2023
Phasedown
Tech-
forcing
Tech-
forcing
Tech-
forcing
Tech-
forcing
Notes: 1. LCFS policy initiatives for Oregon, Washington, and the Northeast/Mid-Atlantic (NEMA) region are
in various stages of development. The information presented here reflects the authors’ current understandings of
31
!
likely policy designs based on publicly available information sources. This information may not reflect final
policy designs, if and when they are developed.
2. Baselines for California and British Columbia, and recommended baselines for WA were defined by
projecting a 2010 CI with 2007 data.
3. The recommended baseline for Oregon was defined by projecting a 2010 CI using data from 2007 and 2009.
5 Point)of)Regulation)and)Regulated)Parties)
Recommendation+5.+Regulate+the+parties+responsible+for+producing,+importing,+or+supplying+
fuel.+
Key Issues
Identifying the point of regulation and the regulated parties determines who will be allowed to
generate and sell credits. How far upstream should the point of regulation be so as to maximize
administrative efficiency? Should alternative fuel providers other than oil companies, such as
biofuel producers, be regulated parties and allowed to generate credits? Should providers of
nonconventional transportation fuels such as electricity, natural gas, and hydrogen be regulated
parties, given that transportation application is only a small portion of the total energy these
energy companies provide?
Summary Recommendations
Regulated parties should generally be those parties responsible for producing or importing fuel
for consumption in the US transportation sector. For petroleum fuels used in transportation
(gasoline, diesel, jet fuel, bunker fuel), the regulated party should be oil refiners or importers,
along with blenders when biofuels are mixed with petroleum fuels.
For transportation fuels that are also used outside the transportation sector, the initial regulated
party should be the party responsible for supplying the fuel for transportation-sector applications.
These could be firms supplying fuel to vehicle fueling equipment, or firms owning the vehicle
fueling equipment, but not both. !
Discussion
An LCFS and relevant fuel policies (including RFS2 and the EU’s biofuel policy RED and
LCFS-like policy FQD) are technology-forcing policies (as opposed to demand-pull policies that
focus on creating demands directly). Thus fuel suppliers and importers are natural regulated
parties as they have sufficient control over fuels and/or feedstock sourcing and processing to
enable implementation of carbon-intensity-reduction strategies. They also have sufficient
32
!
knowledge of life-cycle emissions to fulfill compliance obligations and are sufficiently few in
number to enable effective administration and enforcement. In addition, fuel suppliers and
importers are in the position of making long-term commercial and R&D investments in
increasing the supply of low-carbon transportation fuels and have sufficient resources to manage
the trade of carbon credits.
For petroleum fuels used in transportation (gasoline, diesel, jet fuel, bunker fuel), the regulated
party should be oil refiners or importers, along with blenders when biofuels are mixed with
petroleum fuels.
Gasoline and diesel fuel producers can choose among five methods to meet LCFS targets:
1. Reduce the carbon intensity (CI) of gasoline and diesel.
2. Increase their use of alternative fuel blends in gasoline and diesel.
3. Substitute lower-CI for higher-CI biofuels in blends (for example, substitute low-
carbon ethanol for corn ethanol).
4. Sell more alternative fuels (for example, E85, B100, and CNG).
5. Purchase credits from other regulated parties or use credits banked in previous years.
The initial regulated party for those fuels currently used mostly for nontransportation purposes
(for example, natural gas, electricity, hydrogen) should be the party responsible for dedicating
the fuel to transportation-sector applications. This could be either the party supplying fuel to
vehicle fueling equipment or the owner of the vehicle fueling equipment, depending on the
circumstances. For fuels that are predominantly used outside the transportation sector, producers
and other upstream supply chain participants may have only limited influence over and interests
in the fuel used in transportation as it represents a very small market share of total use. In this
regard, incentives that focus on “demand-pull” may be more effective in incentivizing the
introduction of low-carbon fuels into the transport sector.
In the case of electricity used in transportation, the owners of vehicle-charging equipment could
potentially be allowed to act as regulated parties either directly or in some type of pooled
capacity without being regulated as an electric utility. This would provide greater flexibility in
increasing the supply of electricity to transportation, which is currently limited by the vehicle
fleet and charging infrastructure rather than by the supply of low-carbon electricity.
For example, Yang (2012) examined ways to allocate credits along the electric pathway in order
to maximize incentives to contribute electricity to lower transportation carbon intensity. From a
policy maker’s perspective, one approach to deciding who can claim LCFS credits is to consider
who is likely to use the proceeds from the sale of credits in a manner that will enhance the goals
of the LCFS—that is, lowering the fuel CI and increasing the amount of alternative fuel being
consumed. Increased infrastructure deployment can increase the amount of electricity used to
lower transportation CI by spurring additional plug-in electric vehicle (PEV) sales and increasing
33
!
charging opportunities for existing PEV drivers. Some have argued that smaller third-party
providers may influence the PEV market more than utilities would, because third-party charging
providers would receive LCFS credits only if they deployed charging infrastructure that is used
by PEV drivers, whereas utilities could obtain LCFS credits simply by virtue of having
customers who purchase PEVs. Utilities could also participate by installing charging equipment
and putting rules in place to prevent unfair competition. Allowing the infrastructure provider to
obtain the LCFS credit, rather than simply defaulting to the electric utility, especially for
infrastructure outside homes, would potentially increase the level of investment that results in
useful infrastructure or direct subsidy to PEV purchasers (Yang 2012).
6 Ene rgy)Efficiency)R atios)for)Div e r se ) Propulsion)Technologies)
Recommendation+6.+Use+energy+efficiency+ratios+to+adjust+the+carbon+intensity+ratings+of+fuels+
for+diverse+propulsion+technologies.
Key Issues
Some advanced fuel-engine combinations have superior efficiency and thus deliver more vehicle
miles traveled for the same amount of energy compared with gasoline internal combustion
engine (ICE) vehicles, resulting in lower carbon emissions on per mile basis (gCO
2
e/VMT). The
differences in efficiency are particularly large for all-electric drive trains and fuel cell vehicles.
To appropriately recognize actual emissions displaced by low-carbon fuels, fuel carbon intensity
(CI) should be adjusted to account for the superior efficiencies of advanced vehicular propulsion
systems. Proposed/adopted state, regional, and European LCFS policy programs have adjusted
the effective CI of fuels using energy efficiency ratios (EERs). However, the efficiencies of
gasoline/diesel vehicles are expected to increase substantially in response to increasingly
stringent efficiency standards (U.S. EPA 2011a, 2011b). As a result, the efficiency differences
between advanced vehicles and gasoline/diesel vehicles will change over time. Therefore the key
issues are how to calculate EERs (including issues of accounting for varying efficiencies across
fleet and time) and how often to update them.
Summary Recommendations
For an LCFS to account accurately for the full life-cycle impact of different fuels, the carbon
intensity (CI) ratings of fuels have to be adjusted by the differences in energy conversion
efficiency of vehicle engines. This adjustment is essential to correctly reflect the actual emission
reductions (in gCO
2
e per mile traveled) when replacing conventional fossil fuel with alternative
fuels that run on engines with much greater conversion efficiency such as electric motors
compared to internal combustion engines. Adjustments are also required for fuel cell vehicles,
which are also more efficient than gasoline-powered vehicles.
34
!
These adjustment factors—energy efficiency ratios (EERs)—are best calculated by comparing
the fleet-average efficiencies of the alternative power train with the corresponding fleet-average
efficiencies of baseline fuel-vehicle technologies that the alternative fuel-vehicle technology will
displace. The values should be updated on a regular basis to ensure they adequately reflect the
evolving efficiency of vehicles on the road.
Discussion
An LCFS targets the fuel side of the fuel-vehicle system to reduce GHG emissions. However, the
tight coupling of fuel-vehicle systems has important implications for GHG emissions when
vehicles switch to new propulsion technologies. These differences must be addressed to ensure
that LCFS incentives accurately reflect the actual emission reductions achieved by fuel
switching. This issue of propulsion technology shifts is relevant to many alternative fuels, but
electricity provides the most dramatic example. Electric motors use substantially less energy per
vehicle mile than combustion engines. As a result, substituting electricity for conventional
transportation fuels can provide substantial emissions benefits even if the CI of electricity is
higher per MJ of fuel delivered to the vehicle.
This issue of energy efficiency ratios was explored in recommendations developed for
California’s LCFS by Farrell and Sperling (2007), and the concept of EERs has been widely
adopted by state and regional LCFS initiatives, including the two policies that have been
implemented: EU’s FQD, and British Columbia’s RLCFRR. A couple of different approaches to
calculate EERs are possible, including the direct drive-train comparison and weighted average
approaches, both of which are discussed below. While the discussion and examples here focus on
electricity, EER values should be assigned to all fuels used in different drive trains with
significant differences in efficiency.
EER—direct drive-train comparison approach. In the direct drive-train comparison approach
to EERs, an EER is defined as the fuel consumption of a vehicle using an alternative fuel divided
by the fuel consumption of a vehicle using a conventional fuel (gasoline or diesel) for the same
amount of service delivered (miles traveled). Fuel consumption differences due to the fuel/drive-
train system are isolated from other factors by using fuel consumption data for vehicles with
different fuel/drive-train systems that are otherwise comparable in terms of vehicle class,
capacity, performance, and equipment. Table!3 shows the EER values used in California’s
LCFS, which were developed using this approach.
Table 3. EER values used in California’s LCFS
Light/medium-duty applications
(fuels used as gasoline replacement)
Heavy-duty/off-road applications
(fuels used as diesel replacement)
Fuel-vehicle combination
EER values relative to
Fuel-vehicle combination
EER values relative
35
!
gasoline
to diesel
Gasoline and gasoline-
ethanol blends
1.0
Diesel and diesel-renewable-/bio-
diesel blends
1.0
CNG–ICEV
1.0
CNG or LNG (spark-ignition
engines)
CNG or LNG (compression-
ignition engines)
0.9
1.0
Electricity–BEV or PHEV
3.4
Electricity–BEV or PHEV
H2–FCV
2.5
H2–FCV
Source: CARB 2011. Appendix A. Proposed regulation order. BEV: battery electric vehicle; PHEV: plug-in hybrid
electric vehicle; CNG: compressed natural gas; ICEV: international combustion electric vehicle; H2 FCV: hydrogen
fuel cell vehicle.
EER—weighted average approach. One way of adjusting the EER baseline for gasoline is to
compute fuel-specific EERs as averages across drive trains, weighted by the proportion of
vehicles with each drive train. This approach is illustrated for two time periods in Table!4 using
one particular set of estimates for drive-train efficiencies calculated for 2008 and projected for
2035 (Bandivadekar et al. 2008).
19
The same method is used for adjusting the average EER of
new drive-train technologies as they are commercialized more broadly. For instance, the mix of
different fuel cell and battery electric vehicles would be used to update EERs for those
technologies (relative to gasoline vehicles).
Table 4. Illustrative drive-train weighted average EERs for electricity as a gasoline replacement
Time
period
Subject fuel
Baseline fuel / drive
train
Proportion of
vehicles
Drive-train-specific
EER
Weighted average
EER
1
Electricity
Gasoline (ICE)
99.996%
4.00
4.00
Gasoline (HEV)
0.005%
2.81
2
Electricity
Gasoline (ICE)
85%
3.28
3.06
Gasoline (HEV)
15%
1.83
Notes: Subject fuel is the alternative fuel for which the EER is being calculated. Baseline fuel / drive train
reflects the baseline fuel / drive train system being displaced by the subject fuel. Proportion of vehicles reflects
the fraction of vehicles using each baseline fuel / drive-train system being displaced by the subject fuel. Time
period 1 estimates are provided for illustration purposes from values provided in Table 46 (U.S. EIA 2010a).
Drive-train-specific EER represents the EER computed as the ratio of fuel efficiencies for each combination of
displaced fuel / drive-train systems. Values based on tank-to-wheel energy use estimates from Table 7
(Bandivadekar et al. 2008), except Period 1 value for electricity as a substitute for gasoline ICE, which is based
on the Oregon LCFS analysis. Weighted average EER represents the mean of drive-train-specific EERs weighted
by the proportion of vehicles.
How often to update EER values. There are at least two options for updating EER values: (1)
EER values could be analyzed and revised on an ongoing basis; or (2) EER values could be
defined according to fleet fuel economy forecasts that cover relatively longer time periods. In the
latter case, EER values could define a schedule that mirrors expected changes in fleet fuel
economies. The first approach is used by California, the second by Oregon. Defining EER values
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
19
Computed as the ratios of fuel efficiency estimates provided in Table 7 of the referenced report: 1.77 / 0.54 = 1.8 for gasoline
hybrid electric vehicles (HEVs); and 0.99 / 0.54 = 3.3 for gasoline internal-combustion engines (ICEs).
36
!
for extended time horizons in principle provides greater certainty to fuel suppliers regarding
LCFS credit and deficit generation, though the changes in credits due to EER adjustments are
likely to be small and predictable. However, the uncertainty inherent in technology forecasts
suggests that periodic updating—perhaps every three or four years—may provide a more
technically sound basis for regulation.
EER values are expected to change over time because fuel-specific drive-train efficiencies are
expected to evolve at different rates. For example, gasoline engine systems are expected to
continue to improve over time as direct injection, continuously variable transmissions, and
hybrid-electric technologies (including stop-start) are improved and adopted. Likewise, battery
electric drivelines are also likely to improve significantly as battery management systems,
electric motors, and other control technologies are improved (Figure 2 and Figure 3). The impact
of new technologies will be moderated by the relatively slow turnover of the vehicle fleet;
however, the evolving nature of EERs needs to be captured in the policy design.
Figure 2. Projected new car fuel consumption (miles per gallon) by technology type. Source:
U.S. EIA 2012.
!"#
$"#
%"#
&"#
'""#
'!"#
'$"#
'%"#
'&"#
!""#
!"'"# !"'(# !"!"# !"!(# !")"# !")(#
!"#$%&'$()*"+$,"'$-&**./$$
0"&'$
*+,-#.,--#
/01234,5#
.67#
7893-:5,#/0;2:1#
<:,9,-#/0;2:1#
=-+4>:5#/0;2:1#$"#
=-+4>:5#/0;2:1#'"#
?@#A!""#B:-,C#
?@#A'""#B:-,C#
D<E#<:,9,-#E.?#
7893-:5,#E.?#
37
!
Figure 3. Fuel economy improvement and CO
2
emission reduction from hybrid and diesel
vehicles, compared directly with the most similar conventional gasoline models made by the
same automaker. Source: Lutsey 2010.
EER updating is especially important because of the increasing diversity of drive-train
technologies in the vehicle fleet. In particular, EER values for vehicles burning gasoline in
internal combustion engines will need to reflect the increasing penetration of hybrid-electric and
direct-injection technologies. As gasoline engine systems become more efficient, EER values
will be smaller for battery electric and fuel cell systems.
7 Separate)fuel)pools)for)gasoline)and)diesel)
Recommendation+7.+Create+separate+fuel+pools+for+gasoline+and+diesel.+
Key Issues
States and regions that have adopted LCFS policies generally divide the regulation into two
distinct fuel pools based on the petroleum product they displace—gasoline or diesel. The
intention is to minimize the potential for adverse incentives that encourage increased sale of
diesel fuel, which has a slightly lower CI value than gasoline, without actual investments in low-
carbon fuels, such as cellulosic biofuels, hydrogen, and so on. Should a national LCFS adopt the
same diesel vs. gasoline fuel pools approach? Should a single fuel pool be adopted to emphasize
regulatory simplicity and consistency across all fuels? Should more fuel pools be adopted for
each category of fuels and fuel substitutes, such as adding a jet fuel pool for jet fuel substitutes?
Summary Recommendations
38
!
We recommend that at least two separate fuel pools be established—for gasoline and diesel—
with the potential to establish additional fuel pools for jet and maritime fuels. A single fuel pool
could create incorrect incentives to increase diesel fuel sales if diesel earned a more favorable CI
rating as a result of its EER value against gasoline. Without separate fuel pools, a refiner would
have the incentive to reduce the price of diesel fuel for sale to trucks or even foreign markets—
with no long-term GHG benefits.
To implement a fuel-pooling approach, target CI reductions and EER values will need to be
established for each pool. Based on our research findings, we recommend unlimited LCFS credit
trading between pools in order to provide flexibility, lower compliance costs, and acknowledge
uncertainties in feedstock availability and technological progress.
Discussion
As discussed in Section 6, EER values are an important means of ensuring that LCFS credits and
deficits accurately reflect the emission intensities of fuels as they are used in vehicle propulsion
systems. EER values are only meaningful, however, when they are defined with respect to fuels
used as direct substitutes. Applying EER values for fuels that do not directly substitute for
baseline fuel use can result in inappropriate incentives and unintended consequences.
One fuel pool. The drawbacks of a single fuel pool have been discussed particularly in the
context of diesel fuel. In principle, EER values for diesel fuel used in vehicles should reflect the
efficiency advantage of shifting from spark-ignition gasoline engines to compression-ignition
diesel engines. Switching from gasoline to diesel fuel would appear to reduce emissions (per
mile traveled) by 11 to 18 percent based on the EER values adopted under California’s LCFS,
though this efficiency advantage exists only for diesel used as a substitute for gasoline in light-
and medium-duty vehicles. Diesel in light- and medium-duty vehicles currently accounts for less
than 3 percent of diesel consumption in transportation (AEO 2011, Table 37). As a result,
defining EER values for the whole diesel fuel pool based on the efficiency advantage of
compression-ignition engines would mischaracterize the emissions reduction of diesel
consumption and create inappropriate LCFS incentives for fuel suppliers.
There are two possible ways to more accurately implement this concept of displacement:
1. Regulate fuel pools separately (dual or multi-fuel pools), requiring regulated parties to track
fuel used in each fuel pool.
2. Regulate transportation fuels as a single pool using single EER values for each fuel, defined
as the average of pool-specific EER values weighted by the proportion of the fuel used in
each pool, similar to the example in Table!5.
39
!
A number of concerns have been raised in discussions regarding the appropriate EER value for
diesel fuel, including (1) the potential for an LCFS to subsidize diesel consumption in heavy-
duty vehicles at the expense of light- and medium-duty vehicles; (2) impacts of increased diesel
fuel usage on air quality and environmental justice; (3) technical and equity issues for refiners
related to increasing diesel supplies to US markets; and (4) the ability of diesel fuel to advance
the LCFS policy objective of driving innovation in low-carbon fuels. These concerns were
explored in the California LCFS policy design document (Farrell and Sperling 2007).
Two fuel pools. The California LCFS responded to these concerns in its LCFS policy by
defining and regulating CI reductions in separate fuel pools for diesel, gasoline, and their
respective substitutes. Other state and regional CFS policy initiatives appear to be following suit.
British Columbia’s RLCFRR took exception and first adopted a one-fuel-pool design but soon
found problems with refiners tying retail contracts to increase diesel sales. It is our understanding
that BC is working to revise its standard toward a two-fuel-pool system.
Under the two-fuel-pool system, fuel suppliers are required to meet the CI targets for each pool
separately. Selling more diesel fuels in the market by itself does not affect the carbon intensity of
diesel and gasoline pools unless there is actual emission displacement from more efficient diesel
vehicles in the gasoline fuel pool. This effectively prevents an LCFS from motivating increased
diesel fuel sales and alleviates related concerns. Concerns regarding uncertainties in technology
costs, maturity for commercialization, and feedstock availability of diesel substitutes can be
alleviated by unlimited LCFS credit trading between pools, which is expected to effectively
prevent costs from escalating (Huang et al. 2012; Rubin and Leiby 2012).
EER values for each fuel depend on its fuel pool—whether it is associated with a shift away
from gasoline or diesel engines. The impact of fuel pool designation on LCFS credit generation
can be significant. This is illustrated in Table!5. The example highlights the importance of
identifying regulated parties
20
and improved chain-of custody to accurately track fuel delivery to
end uses.
Table 5. An example of credit generation rates for CNG under California’s LCFS
Fuel pool
Gasoline
displacement
Diesel
displacement
CI value for CNG (gCO
2
e/MJ)
67.7
67.7
Pool-specific EER value
1.0
0.9
Target CI values (gCO
2
e/MJ)
2011
95.61
94.47
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
20
As discussed in Section 5, in order to incentivize infrastructure development and vehicle deployment, the initial
regulated party can be the party responsible for supplying the fuel for transportation-sector applications.
40
!
2020
86.27
85.24
Credit generation rates (gCO
2
e/MJ)
2011
27.9
19.2
2020
18.6
10.0
Notes: The CI value for CNG is the value associated with fuel Pathway Identifier CNG001, as specified in the
California LCFS Look-Up Tables. EER values and target CI values for each fuel pool are as specified in the
regulation order for the California LCFS. The numbers of credits generated are computed according to the formula
provided in the regulation order.
Multiple fuel pools. Fuels used in multiple transport modes and vehicle segments face similar
challenges regarding appropriate EER values. For example, electricity is used in both on-road
vehicles and rail transport; the efficiency advantages of electric motors may be substantially
different in these different modes. Moreover, EER values for jet fuel that reflect the relative
efficiencies of aircraft turbines and on-road vehicle engines would be completely meaningless.
Some means of addressing these issues is necessary to ensure that LCFS incentives are
meaningful, reflect the emissions intensity of fuels as they are used in vehicle propulsion
systems, and motivate changes in the fuel mix that efficiently advance underlying policy
objectives. To ensure technology neutrality, an LCFS policy should address these issues in a way
that is consistent across all fuels.
As a practical matter, the challenges associated with assigning alternative fuels to specific fuel
pools may not be overwhelming. This is due to their relatively modest contributions to
transportation fuel supply, the ownership structure of their distribution infrastructure, and the
composition of associated vehicle fleets. These challenges may increase over time, however,
particularly if increasing use of alternative fuels is associated with increasing diversification
across fuel distributors and vehicle segments.
An alternative approach—one fuel pool based on weighted average. If a dual-pool approach
is considered to be too complex and difficult to track and verify, there is an alternative: the
transportation sector could be viewed as an aggregation of multiple fuel pools but regulated as
one. Multiple fuel pools would be used to develop multiple EER values for each fuel. Pool-
specific EER values would be aggregated into a single average value for each fuel that would be
weighted by the proportion of fuel used in each fuel pool. This approach would ensure that EER
values accurately reflect the dynamics of fuel substitution over time, which motivated the dual-
pool approach, and would remove the need for each unit of fuel to be assigned to a particular fuel
pool.
This approach is illustrated in Table!6 for a simplified example of EER value for diesel fuel
under the single-pool approach. Two time periods are shown to illustrate the calculation of diesel
EER in the base year and future adjustments for changes in both fuel usage patterns (reflected in
the proportion of fuel used in each vehicle segment) and efficiency of vehicle propulsion systems
(reflected in EER values for each segment). Note that this approach could be applied to the dual-
41
!
pool case California has adopted for regulating fuels used in on-road vehicles, or it could be
generalized to accommodate many potential fuel pools, depending on the coverage and scope of
the policy.
Table 6. Simplified illustration of multi-pool EER calculation for diesel fuel
Year
2011
2020
Fuel pool
Light-duty
Heavy-duty
Light-duty
Heavy-duty
Proportion of fuel use
3%
97%
10%
90%
Pool-specific EER
1.2
1
1.16
1
Fuel average EER
1.006
1.016
Averaging formula
(3% * 1.2 + 97% * 1)
(10% * 1.16 + 90% * 1)
Notes: Pool-average EER is computed as the ratio of fuel consumption by vehicle propulsion systems in each fuel
pool. Fuel consumption data is from tank-to-wheel energy use estimates from the literature (Bandivadekar et al.
2008). Values for the proportion of fuel use within each pool are hypothetical and consider fuel use only in these
two vehicle segment fuel pools.
Implementing this approach would add some complexity for administering agencies, which
would be required to develop EER values for all fuels and monitor fuel allocation across all fuel
types over time. For example, each transportation mode (aviation, rail, on-road, and so on),
segment of the vehicle fleet (for instance, heavy-duty and on-road vehicles), and use of
transportation fuels (diesel use for home heating, for example) would need to be clearly defined
and tracked. Each fuel would then be evaluated to determine a fuel-specific EER value based on
technology mix and corresponding efficiency changes, as illustrated in Table!6. Fuel suppliers
would then adopt a single EER value for each fuel type and a single CI target each year defined
by the policy. This approach, however, has the downside of the incentives being too indirect and
too weak to motivate direct technology/vehicle substitution by fuel providers or third parties (see
Recommendation 5).
8 Lif e K Cycle)Carbon)Intensity)as)Regulatory)Metric)
!
Recommendation+8.+Regulate+fuels+according+to+their+lifeDcycle+GHG+emissions.+
!
Key Issues
An LCFS is premised on measuring all GHG emissions of a fuel from the source (oil well, coal
mine, farm field, and so forth) to the final point of consumption. This life-cycle approach is key
to comparing the emissions of different fuels. GHGs are emitted from various sources within
transportation fuel supply chains. Life-cycle analysis (LCA) captures emissions sources from the
source to the end use of each fuel’s supply chain. But the results are subject to significant
uncertainty and variability—due to level of disaggregation, boundary definitions, scientific
uncertainty, and various assumptions and methods. How can regulations be designed to improve
42
!
precision and accuracy, taking into account the challenges of data availability, administration,
and enforcement, and the goal of stimulating innovation?
Summary Recommendations
Calculation of life-cycle GHG emissions will require modelers, and ultimately policy makers, to
make decisions regarding modeling approaches, system boundaries, and data sources. When
multiple jurisdictions are involved, such as nations, it will be important to harmonize the
methodology used among different regulatory agencies, creating a consistent approach for
defining and measuring carbon intensity in fuels. We recommend the following.
System boundaries. A national LCFS policy should adopt a standardized life-cycle assessment
(LCA) method for measuring fuel CI that reflects best practices and is transparent and consistent
across fuel types. Indirect emissions resulting from market-mediated effects should be evaluated
for potential inclusion when they (1) substantially impact fuel life-cycle carbon intensity (CI)
and (2) are closely linked to particular fuel supply chains (see Recommendation 9 for land use
emissions, which are the most significant indirect effect).
Spatial boundaries. Data inputs for LCA measures should be disaggregated enough spatially to
capture regional variability in supply chain emissions in ways that will incentivize greater use of
low-carbon feedstock/technology. As a convenient way of operationalizing boundary definitions,
we recommend using state boundaries for setting default CI values for biofuels, and load-
balancing area or higher levels of aggregation for electricity CI values.
Uncertainty and variability. LCA results throughout the supply chain, which an LCFS will rely
on to estimate emission benefits compared to baseline fossil fuels, are subject to significant
uncertainty and variability over space and time. We recommend that the sources of uncertainty
and variability be systematically identified and carefully evaluated to determine default values
(see next item) and to help design a more robust GHG reduction target given uncertainties. We
recommend that variability or uncertainty due to data limitation be targeted with an opt-in
reporting mechanism in the policy design to improve data availability, reduce uncertainty, and
incentivize innovation. We also recommend addressing scientific uncertainty through adaptive
management and targeted research. Uncertainty and variability distributions should be updated
regularly to reflect changes in science and technological progress.
Default values and opt-in mechanisms. Default values should be assigned to each energy path to
ease the reporting requirements of energy providers. If energy providers (the regulated
companies) can supply their fuel with lower emissions than the default values, they should be
allowed to opt in with their superior measurement value. They would do so by documenting their
lower emissions. Allowing companies to opt in encourages innovation by rewarding producers
for reducing emissions.
43
!
The use of default values leads to an “adverse selection” bias, which occurs when only fuels with
CI values lower than the default opt in with their lower values while fuels with CI values higher
than the default choose the default values. This results in systematic underestimation of actual
emission reductions by the LCFS (and less stimulation of innovation).
To minimize adverse selection bias, the downside of using default values and opt-in
mechanisms, we recommend (1) disaggregating fuels according to production method and other
parameters that have high impacts on GHG emissions; (2) minimizing the adverse selection bias
by periodically updating the distribution of fuels to eliminate fuels that have already been using
lower CI opt-in values; and (3) placing the default CI value at the high end of the distribution,
such as the 70th percentile and above, thereby incentivizing more reduction.
Discussion
GHGs are emitted from various sources within transportation fuel supply chains. The relative
contribution of sources varies considerably across different fuel types. For example, vehicle
tailpipe emissions account for roughly 80 percent of life-cycle GHG emissions from
conventional gasoline but effectively 0 percent of emissions from biofuels or fuel electricity.
21
As
a result, CI measurements based on tailpipe emissions alone would overstate the potential
benefits of biofuels and electricity. To ensure that GHG policies correctly incentivize fuels that
are truly low carbon, fuel CI measurements should capture emissions within the fuel’s entire life
cycle. A standardized and transparent life-cycle analysis model should be developed and adopted
for use within a national LCFS policy. Stakeholders should be engaged in reviewing and revising
the methodology, and the model should be regularly updated based on the best available science.
8.1 System)boundaries
To calculate fuel GHG emissions over a life cycle, it is important to define the boundaries
around each energy pathway and system. To estimate the CI of biofuels for RFS2, the US
Environmental Protection Agency (EPA) used a number of models and tools, including the
Argonne National Laboratory’s GREET model, Texas A&M’s Forestry and Agricultural Sector
Optimization Model (FASOM), and Iowa State University’s Food and Agricultural Policy
Research Institute’s (FAPRI) international agricultural models, as well as the Winrock
International database. The California Air Resources Board (CARB) adapted the GREET model
for California to calculate the life-cycle CI of all fuel pathways, and used the Global Trade
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
21
CO
2
emissions from the combustion of biofuels, particularly from crop-based feedstock, have been considered carbon neutral
since the same carbon has recently been removed from the atmosphere during biomass production and so is generally assumed
to have zero net global warming potential (GWP). Biofuels produced from wood-based feedstock cannot be considered as
carbon neutral due to the long lag time of resequestering the emitted carbon, and therefore they require separate consideration.
GHGs emitted from fuel electricity are emitted from the power plant, not the vehicle.
44
!
Analysis Project (GTAP) model coordinated by Purdue University for land use emissions for
biofuel pathways. The European Union developed rules and formulations for the calculation of
life-cycle CI in the 2009 EU RED. British Columbia uses GHGenius, an LCA model for Canada
that is analogous to the GREET model, to support its RLCFRR. These different major biofuel
programs rely on different models and input assumptions, and they draw different system
boundaries for the calculation of GHGs for different biofuel and fossil fuel pathways (Table!7).
Table 7. Life-cycle GHG emission categories and models used for electricity, fossil fuel, and
biofuel under the US (EPA) RFS2, California (CARB) LCFS, EU RED, and British Columbia
(BC) RLCFRR
Note: Table includes emissions from within the supply chain (solid box) and outside the supply chain (dash box);
the latter are often called indirect or market-mediated effects. Only highlighted areas have been considered in the
regulatory LCA analyses so far.
System boundaries—especially with respect to co-products, by-products, and indirect effects—
need to be carefully defined because different definitions of boundaries can result in quite
different emission calculations for some products (Wang, Huo, and Arora 2011). The production
process for corn ethanol, for example, not only produces ethanol but also large quantities of
valuable co-products used for animal feed, and the method used to allocate “credits” for the co-
products can result in very different GHG emission ratings (Wang, Huo, and Arora 2011). In
addition, the consideration of emission impacts outside of direct supply chains, often called
market-mediated response or indirect emissions (see Section 9), has been highly controversial in
the past few years, in the case of biofuels (Melillo et al. 2009; Searchinger et al. 2008b), fossil
fuel (Drabik and Gorter 2011), and electricity (McCarthy and Yang 2010).
Variability in system boundaries of fuel pathways should be minimized. Policy makers should
define the system boundary through a transparent process, propose a method to quantify
significant emissions that occur outside the supply chain due to market-mediated effects (often
referred to as indirect emissions), engage stakeholders in reviewing/revising the methods, and
!"#$%&'( !"#$ %#&'$ !($ '%$
)*+,-./$012+$34567-$148$
0,2.9:;,2$<=>$
0#?>@$
)*+,-./$A148$(-,$%B14C,$ 0#?>@$ DE#"$
)*+,-./$&:/,$@,7B14,$ 0#?>@$
)*+,-./$A:F,-7*/G$ 0#?>@$
347,241.*419$A148$(-,$%B14C,$ 0#"&3$ DE#"$
347,241.*419$012+$34567-$148$
0,2.9:;,2$<=>$
0#"&3$
347,241.*419$&:/,$@,7B14,$ 0#"&3$
347,241.*419$A:F,-7*/G$ 0#"&3$ DE#"$
06,9$148$0,,8-7*/G$E214-5*27$ D&!!E$ D&!!E$ &!)$ DHD,4:6-$
06,9$"2*86/.*4$ D&!!E$ D&!!E$ &!)$ DHD,4:6-$
E1:95:5,$ @>I!?$ D&!!E$ &!)$ DHD,4:6-$
)#**"'()%&'((
)*+,-./$A148$(-,$%B14C,$
347,241.*419$A148$(-,$%B14C,$
!"#$%&'()'%$*"&$+$*",-&
.%$/01()2&3$4546$7&
06,9$148$0,,8-7*/G$E214-5*27$
06,9$"2*86/.*4$
E1:95:5,$
+'&,-.","-/(
%B14C,-$:4$C,4,21.*4$+:J$
%B14C,-$:4$C2:8$/151/:7K$
,J514-:*4$
!"#$%&'()'%$*"&$+$*",-&
.%$/01()2&3$4546$7&
06,9$148$0,,8-7*/G$E214-5*27$
06,9$"2*86/.*4$
E1:95:5,$
45
!
regularly update the information based on the best available science. To avoid unintended
consequences and send correct policy signals regarding the true GHG impacts of fuels, we
recommend that indirect emissions resulting from market-mediated effects be carefully evaluated
for potential inclusion in fuel CI when they (1) substantially impact fuel CI and (2) are closely
linked to particular fuel supply chains as a result of fuel policy.
The existence of inconsistent system boundaries between policies creates the danger of ignoring
potentially significant emission sources, leading to confusion and to incorrect and conflicting
incentives for GHG emission reductions. It can also lead to leakage and shuffling of fuels and
credits. Minimizing shuffling and confusion and assuring compatible and consistent models,
assumptions, and outputs will require collaboration between responsible governments. Multilevel
stakeholder discussion will also be critical to ensure a collaborative and science-based approach
that leads to effective outcomes.
Overly expanded system boundaries can cause double counting, as the same emissions may
already be regulated in other regulations or carbon markets. If credits are allowed to be generated
from emission reductions within the expanded system boundary, double crediting can also occur.
This issue is further discussed in Section 11. The second challenge is that the characterization of
indirect effects often requires large-scale modeling and predictions, and therefore significantly
increases uncertainties as well as the resources and time required to update CI values associated
with indirect effects.
8.2 Spatial)boundaries)
Improving the spatial resolution of some parameters can significantly reduce known variations in
measurements. Perhaps the most blatant example is electricity production—where in some
regions much of the electricity is produced with hydropower, nuclear, solar, and wind, and in
others mostly with coal. But variations in biomass production emissions can also be large,
because of climate, soil conditions, irrigation, farming practices, carbon footprint of local
electricity, and electricity offsets generated from some refining facilities. Greater resolution and
disaggregation of spatial boundaries will increase the accuracy of fuel CI values and provide
incentives motivating efficient use of available resources (Elvidge et al. 2009). We recommend
that spatial boundaries reflect geographic units that capture meaningful variability and have
sufficient data available. The level of regional specificity may be different for different types of
data inputs and different fuel types.
Uncertainties in calculating biofuel CIs are dominated by uncertainties regarding land use
emission factors, N
2
O emissions from fertilizer application, land yields, and sources of
production energy. Biomass yields tend to be region specific, depending on local/regional
climate conditions such as temperature, rainfall, humidity, and soil type (Figure 4), resulting in
fuel CI values varying by a factor of 2 to 3 across the United States (Figure 5). Setting spatial
46
!
boundaries at the state level for calculating biofuel default CI values is desirable because it (1)
makes use of reliable public data that is collected annually, (2) effectively minimizes uncertainty,
and (3) incentivizes more efficient use of available resources.
One issue with regional disaggregation of fuel CI values stems from the commerce clause of the
US Constitution. A lawsuit against California’s LCFS that is currently under appeal questions
whether the assignment of CI values to midwestern corn ethanol represents an impermissible
interference with and discrimination against interstate commerce. The solution might be to use
more generic labels for GHG performance criteria, rather than labels based on geographic
regions.
!
Figure 4. Delivered biomass yields in the United States by region (ton dry mass/ha). Source:
Khanna, Onal, and Huang 2011.
47
!
Figure 5. Regional variability in mean GHG emissions for corn (top) and switchgrass (bottom)
ethanol. Source: Khanna, Onal, and Huang 2011.
Electricity—both for use as transportation fuel (powering electric vehicles) and as an input for
producing other fuels—presents several unique challenges with regard to establishing CI values.
These challenges stem from (1) the fact that electricity systems in the United States and
associated CI values are extremely heterogeneous and regionally specific; (2) the complex nature
of electricity supply that varies hourly and seasonally; (3) interconnections between regional
power grids; and (4) the lack of a system for tracking cross-boundary electricity flow.
Moving to larger regions of aggregation for determining electricity CI reduces the variability in
CI between regions. There is significantly greater spread in average CI when we look at the 112
eGRID power control areas (PCA) compared to the 10 NERC regions or the US average (a
single value) (Figure 6). The choice of spatial boundary for electricity allocation purposes can
have implications for the electricity CI value and the incentives that will result. Assigning every
electricity provider the same average CI value (for example, by defining one national region with
equal CI) would provide a uniform incentive to all electricity providers to provide electricity as a
transportation fuel, despite whether its actual CI values exceed the CI of gasoline. Using smaller
spatial boundaries would lead to greater variability in electricity CI and differential incentives
based on actual CI values. The distribution of electricity CI values becomes fairly robust at the
level of EGRID subregions and finer resolution, as shown in Figure 6.
48
!
Figure!6.!Cumulative distribution of life-cycle carbon intensity of electricity for different levels
of regional disaggregation. NERC:!North American Electricity Reliability Corporation; U.S.
Environmental Protection Agency’s Emissions & Generation Resource Integrated Database
(eGRID) has 26 subregions and 112 Power Control Area (PCA). Source: Yang 2012.
We recommend establishing CI values at the load-balancing area or higher level of aggregation
(encompassing dozens or more investor-owned and municipal utilities) based on the historical
average generating mix. This would give sufficient flexibility to choose the level of aggregation
depending on data availability and ease of administration. LCFS credits, electricity providers
should be required to provide detailed data on charging load, timing and location by a verifiable,
utility-grade meter. This information will be used for grid planning and CI calculations and also
ensure that PEV charging does not cause or exacerbate grid issues
8.3 Uncertainty)and)variability)
LCA calculations, conducted for each step of the energy supply chain, can be difficult to specify
accurately. Differences in GHG emission estimates across studies and models can be
characterized as uncertainty. Uncertainty falls into three categories: spatial and temporal
variability, data and model limitations, and scientific uncertainty. Variability and data limitations
can be addressed through policy design and improved data collection and reporting. Scientific
uncertainty requires more research and better models; it is more difficult to accommodate but can
also be addressed through creative policy mechanisms (as indicated below for land use change
effects).
49
!
One issue that cuts across virtually all the questions regarding life-cycle CI is the difference
between variation and uncertainty. In the case of uncertainty, parameters cannot be measured
because more scientific knowledge is needed or because data is difficult or impossible to collect.
In the case of variability, economic activities and technology usage vary dramatically—for
instance, in electricity generation—resulting in dramatically different emission characteristics
over space and across energy systems. In principle, virtually all variation can be eliminated by
increasing the resolution of regulations (for example, by specifying exactly where and how the
fuel is produced). But in practice, such fine resolution comes at a high cost and places a large
burden on regulators and the regulated parties. The challenge is to acknowledge and address
uncertainty, and for regulatory design to identify the optimal location on the spectrum from fine
disaggregation to high-level aggregation.
Different types of variability and uncertainty can be addressed and managed effectively
according to their sources. We recommend that variability or uncertainty due to data limitation
that can be reduced through data reporting be targeted with an opt-in reporting mechanism in
the policy design to improve data availability, reduce uncertainty, and incentivize innovation (see
discussion in Section 8.4).
Scientific uncertainty can be addressed through adaptive management and targeted research.
Scientific uncertainty that results from an incomplete understanding of particular factors (such as
nitrogen volatilization rate or the global warming potential of non-GHG gases) or relationships
between factors (potential impacts of factors outside of historical ranges) can be reduced
gradually as our understanding of the underlying factors and relationships evolves. Targeted
research initiatives should be established to accelerate reductions in scientific uncertainty
regarding factors that have large impacts on fuel CI. Adaptive management strategies should be
used to systematically incorporate new information into fuel CI measures. This can be
accomplished through periodic updating of fuel CI values to incorporate the most current
scientific information available, including information from targeted research initiatives.
Model uncertainty can be treated with a differentiated approach, including a transparent and
principled analysis of the effect of changes in market responses or the effect of changes in
market structure grounded in the policy objectives. Model uncertainty can be traced to several
root causes, including the idealized nature of mathematical models (for example, a computable
general equilibrium model vs. a partial equilibrium model), numerical approximation of
relationships, subjective or normative decisions inherent in LCA (for example, different methods
of allocating co-products), scientific uncertainty about the relations being modeled, and scenario
uncertainty. Model uncertainty can be differentiated according to these various causes and an
appropriate treatment can be developed to establish bounds for quantitative model results.
50
!
8.4 Default)values)and)optKin)mechanisms)
One means of increasing flexibility for regulated companies, reducing the regulatory burden, and
providing an incentive for technological innovation is to create default values for parameters and
an opt-in procedure for companies with innovative low-carbon products. Without default values,
the regulator would have to develop specific CI ratings for every activity in every energy
pathway, which would require substantial data collection, analysis, reporting, and collection for
each batch of fuel delivered. That is not feasible. Establishing default CI values representative of
fuel pathways, as has been done in most existing LCFS policies, can reduce data and analysis
requirements for the majority of fuel producers.
An opt-in mechanism allows fuel producers to generate CI values tailored to their fuels by
modeling and validating their own fuel production pathways. The opt-in mechanism
implemented in California’s LCFS allows fuel producers to (1) propose verifiable input values
within the existing pathways; or (2) request new, customized fuel pathways for their fuels.
Approved pathways are published with associated CI values and added to the list of default
values. Proprietary information can be redacted from fuel pathway descriptions to ensure
confidentiality. This approach appears to be effective: as many as 124 new CI values and
pathways had been approved as part of California’s LCFS as of today (Figure 7). Many of these
lower-than-default-CI fuels have enjoyed higher market prices than those using default values.
22
We recommend a similar approach for a national LCFS.
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
22
According to Pacific Ethanol’s Third Quarter 2011 Financial Report, the company’s corn ethanol CI rating is on average 18
gCO
2
e/MJ lower than the default value and has an approximately 2 to 4 cent per gallon premium compared with other corn
ethanol (http://www.pacificethanol.net/site/_documents/investors/PEIXQ311ConfCallPresentationFinal.pdf). Such a price
advantage translates to approximately $15–30/tonne CO
2
e avoided.
n=2
n=1
n=1
n=2
n=68
n=1
n=68
n=1
n=16
n=15
n=1
n=15
!"#$##%
!&#$##%
#$##%
&#$##%
"#$##%
'#$##%
(#$##%
)##$##%
)&#$##%
*+,-+./.0%12+340%546%
*+,-+./.0%742,04%8+0%
*+,-+./.0%7,92%8+0%
*+,-+./.0%:4/6.%8+0%
*+,3.6;42.%
<=>%
7,92%?6;42,0%
@,04//./%?6;42,0%
A,9B;C3%?6;42,0%
ACB49D42.%?6;42,0%
:4/6.%*.E.94B.%?6;42,0%
:;.46%A0C99F%?6;42,0%
!!"#$%&'()*%+,-./01+21)*1(3)4+56+,7
8
*9:;<+
1E.94B.%G40C./%
B4/,0+2.H-+./.0%942B.%
51
!
Figure 7. Opt-in CI values (average and minimum and maximum ranges) and number of
applications by feedstock pathway for California’s LCFS as of July 2012. The dash line
represents the 2020 gasoline and diesel 10% CI reduction targets.
The danger of adverse selection. A well-designed system of default values and opt-in reporting
mechanisms can effectively address the data and analytic requirements of LCA but can also
create a risk of adverse selection. Adverse selection occurs when fuel producers choose default
values only when they perform poorly (emissions above the default value) and propose new
values when they perform well (emissions lower than the default value). As a result, carbon
emission calculations for the entire fuel population will systematically underestimate the actual
emissions from fuel production. The biases created by adverse selection can be potentially large,
especially for fuel pathways with large variability or reducible uncertainty.
To reduce adverse selection, default values could be
• Established for a much expanded set of fuel pathways so as to address variation. Well-
differentiated fuel pathways can reduce the magnitude of CI variations among fuels
within each fuel pathway and thereby reduce the magnitude of potential adverse selection
impacts, as discussed in the TAR. An improved chain of custody compared with the
existing system is needed ensure the measurability and verifiability of these opt-in CI
values.
• Reviewed and updated periodically (every three to five years) to ensure that they
accurately reflect the CI of fuels using the default values. In particular, default CI values
may need to be adjusted upward to exclude opt-in reporting of fuels with relatively low
CI values.
• Set conservatively, such that 70 percent or more of fuels within each fuel pathway are
expected to have CI values below the default value. As discussed in the TAR and
underlying reports, this approach can effectively reduce the magnitude of adverse
selection impacts by more than 50 percent if using the average value.
Carbon capture and sequestration technology. Carbon sequestration technology may provide
important contributions to reducing fuel CI. It is the subject of considerable research and
development activity, which is expected to yield important innovations for advancing LCFS
policy objectives.
23
The diversity of technological approaches, the limited experience to date, and
the various concerns expressed regarding carbon sequestration suggest some special
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
23
The term carbon sequestration is used here to refer broadly to systems that secure carbon that would otherwise be emitted as
CO
2
away from the atmosphere for timescales relevant to climate policy objectives. This includes a diverse set of
technological approaches, including CO
2
capture from industrial waste streams (for example, power plant and refinery
emissions) or from the atmosphere for sequestration in geologic formations; CO
2
sequestration in mineral deposits, for
example via enhanced weathering processes; biological accumulation of soil carbon from changes in agricultural practices;
sequestration of bio-char in agricultural soils; and remote burial of biomass carbon in ocean basins or alluvial sediments.
52
!
consideration may be warranted. Guidelines should be established to ensure that LCFS credits
are granted only in cases where concerns are addressed regarding the measurability, verifiability,
additionality, permanence, and security of sequestered carbon.
9 Land)U se )C hange)GHG)Emissions)
Recommendation+9.+Address+GHG+emissions+from+land+use+change+(LUC)+through+shortDterm+
and+longDterm+policies.+
!
Key Issues
Most of a fuel’s life-cycle GHG emissions are directly measurable and within the energy supply
chain. But additional emissions can be caused when large amounts of land are diverted from
other uses (such as agriculture) into energy production—which is the case with many biofuels
and some fossil fuels. Accounting for these land use changes (LUC) is necessary to develop
more accurate life-cycle CI values and to assure that the LCFS sends correct signals to fuel
suppliers. Given evolving scientific knowledge about land use change effects, what is the best
policy mechanism for addressing these land use changes?
Summary Recommendations
The impacts of these land use changes (LUC) are complex and difficult to quantify accurately—
but accounting for them is important to assure that investments are directed at those feedstocks
with less impact. The effects can be large for land-intensive crops such as corn but are much
smaller for grass and tree feedstocks (if they are grown on marginal, degraded land and/or if they
avoid direct competition with food crops) and zero for biofuels made from waste materials (crop
and forestry residues and municipal solid waste). Oil sands production induces small LUCs
associated with soil and forest carbon emissions from peatland conversion. We recommend
adopting a flexible policy taxonomy that includes short-term and long-term policies.
Short-term policies would induce or otherwise encourage immediate action to reduce use of
productive land and other adverse impacts. They would encourage (1) using feedstock that does
not require additional land, such as wastes and agriculture residues, or feedstock that requires
less land, such as cellulosic feedstocks and algae; and (2) adopting measures that lower LUC risk
from land-using feedstock by (a) enhancing carbon sequestration and storage, (b) encouraging
the use of marginal, degraded, and abandoned land, and (c) prohibiting the conversion of high-
carbon, high-biodiversity, and environmentally sensitive areas. Despite relatively large scientific
uncertainty about LUC impacts, we recommend using iLUC factors selected from science-based
53
!
ranges so that LUC policy has a transparent basis in emissions and integrates easily with existing
policies.
Long-term policy measures would combine short-term mitigation strategies with other incentive
mechanisms that offer the greatest potential for mitigating LUC over the long term. These
measures would encourage collaboration within and outside the biofuel supply chain to increase
investments in land use productivity, environmental protection, and carbon offset schemes. The
goal is to enhance economic productivity without compromising environmental or ecosystem
services. The regulatory process should establish rigorous and systematic evaluation frameworks,
coupled with intensified research, to assess options and implementation.
Discussion
The term land use change (LUC) refers to changes in the way land is used to support human
activities. The most dramatic examples are deforestation and land clearing to support agricultural
production. LUC contributes to approximately 17 percent of global GHG emissions (Metz et al.
2007). When land is cleared, carbon stored in the natural vegetation and soils is released to the
atmosphere, primarily as CO
2
.
Biofuel production can cause LUC GHG impacts through three distinct mechanisms: (1) clearing
land to produce biofuel feedstock can cause emissions from above- and below-ground carbon —
called direct land use change emissions; (2) changes in the amount of carbon stored in
agricultural soils caused by changes in farming practices—generally included in direct land use
change emission accounting; and (3) clearing land to meet the demand for conventional
agricultural products expansion where agricultural lands are displaced by biofuel feedstock
production domestically or in other countries—called indirect land use change (iLUC) emissions.
LUC is inherent to all farming activities, including production of biofuel feedstock. Emissions
from clearing land for biofuel production, especially in high-carbon areas such as forests and
peatlands, have been shown to offset the carbon savings of displacing fossil fuels with bioenergy,
and the payback period
24
can be longer than decades (Fargione et al. 2008; Gibbs et al. 2008).
LUC raises a number of concerns beyond increased GHG emissions, including higher or more
volatile food prices, which affect the poor in greater proportion and magnitude (FAO 2008a;
FAO et al. 2011); conversion of high-biodiversity areas; overuse or degradation of local water or
land resources; damage to important ecosystem services (Donner and Kucharik 2008; Koh and
Wilcove 2008; Welch et al. 2010); and disruption of local land ownership or other social patterns
(Toulmin 2009). Depending on location and management practices, however, LUC from biofuel
production can also improve resource productivity, sequester carbon, or provide an additional
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
24
Payback period is defined in Fargione et al. (2008) as “how many years it takes for the biofuel carbon savings
from avoided fossil fuel combustion to offset the losses in ecosystem carbon from clearing land to grow new
feedstocks.”
54
!
income source for rural populations (Berndes, Bird, and Cowie 2010; Lapola et al. 2010; Tilman
et al. 2009).
RFS2 requires “renewable biomass” crops and crop residues to come from agricultural land
(cropland, pastureland, Conservation Reserve Program land, and possibly rangeland) cleared or
cultivated before the law was enacted (U.S. EPA 2010). Thus, theoretically direct LUC should
not occur within the direct biofuel production system, though this ironically increases the
likelihood of iLUC. Due to the complexities of LUC, no single policy is likely to adequately
address biofuel LUC.
Mitigation strategies to address unwanted biofuel LUC effects fall into three categories: (1) use
feedstocks that require less land, (2) adopt measures that lower LUC risk for land-using
feedstocks, and (3) invest in productivity gains, environmental protection, and carbon accounting
methods that reduce the scope for biofuel (and other) LUC. In general, moving down the list,
policy targets broaden—from within to beyond the biofuel supply chain—and involve more
investment and coordination and a longer time to come to fruition. These three strategies are
summarized in Figure 8.
Figure 8. General strategies for addressing biofuel LUC. Source: Witcover, Yeh, and Sperling,
2012.
Theoretically, if GHG emissions from all LUC activities were covered globally and priced
accordingly, iLUC would still exist but iLUC emissions would be accounted for within the
global policy. However, recognizing (1) the difficulty of reaching an international consensus to
address LUC within a reasonable time frame, (2) the need to mitigate the potential unintended
consequences of increased biofuel production before the problem becomes irreversible, and (3)
55
!
the need to encourage the production of feedstock that has low risk of LUC, we make the
following short- and long-term policy recommendations, broken down by the three categories
listed above.
Short-term policy recommendations:
1. Use feedstocks that require less land. Encourage feedstocks that come from wastes or
agricultural residues, or that require less land, such as cellulosic feedstock and algae. Current US
policies already contain some of these incentives such as setting volumetric requirements and
providing tax incentives and subsidies for cellulosic ethanol.
2. Adopt measures that lower LUC risk for land-using feedstocks.
a. Soil carbon accumulation from switching to energy crops or adopting low-till farming
practices can reduce life-cycle GHG emissions and therefore should be included in the
life-cycle CI rating through an opt-in mechanism. Carbon accumulated in soils due to
changes in crops or farming practices can be re-emitted to the atmosphere if farmers
revert to previous crops and practices. The treatment of soil carbon accumulation in
biofuel CI measures must balance this risk of re-emission against the potential
importance of soil carbon emissions effects. Uncertainty about the magnitude of soil
carbon effects should be treated consistently with other sources of LCA uncertainty (see
the discussion in 8.3, Uncertainty and Variability).
b. Although RFS2 disallows crops and crop residues to come from agricultural land,
broadly defined, cleared or cultivated before the law was enacted, the conversion of
marginal/degraded/abandoned land should be allowed and encouraged, provided that
adequate definitions are adopted through transparent policy discussion.
c. The conversion of high-carbon, high-biodiversity, and environmentally sensitive areas
and wetlands should be prohibited.
d. Employing an iLUC factor provides a relatively straightforward, plausible approach that
can contribute to addressing biofuel LUC in a way that explicitly recognizes emission
effects. Choosing a single number from a science-based range for use as an iLUC factor
is defensible for policy purposes if the assigned factor sends unambiguous signals to
investors encouraging use of less land-intensive feedstocks in a way that reflects
(transparent) policy decisions regarding acceptable risk. Issues associated with the use of
an iLUC factor, including uncertainty about its magnitude, are not unique to LUC
emissions. Any policy that considers feedstock LUC emissions will face this same
challenge. To address the issue of uncertainty, we recommend selecting a “risk-based”
value from a distribution of LUC emissions, such as the methodology demonstrated in
Griffin et al. (2012). The report uses probability distributions to match acceptable risk
for meeting a policy target (for example, 75 percent certainty of success in achieving a
56
!
10-percent threshold of GHG savings) to a particular iLUC factor value within the
distribution. The probability distribution of the iLUC factor can be updated regularly to
reflect scientific improvements in capturing the true uncertainty distribution.
Long-term policy recommendations:
3. Invest in productivity gains, environmental protection, and carbon accounting methods that
reduce the scope for biofuel (and other) LUC. The largest potentials for reducing global LUC lie
outside of the biofuel sector. Incentivizing producers to take concrete steps to lower the LUC risk
for land-using feedstocks, or to invest in technology/management that lowers LUC, within and
outside of the biofuel supply chain, provides the greatest potential for reducing LUC. However,
stringent evaluation criteria must be developed to ensure that these policies are implementable
(that is, effective, efficient, robust, and fair to all players) and feasible (available, practical,
integratable with other policies, and transparent) (Witcover, Yeh and Sperling 2012). The
challenge of guaranteeing additionality (reductions would not have occurred anyway), no
leakage (reductions in the project area will not shift elsewhere), and permanence (conserved
carbon will not be released in the future) are shared across all project-level certification schemes
discussed below. These processes will take time, thus we recommend that long-term biofuel
LUC policies consider the following:
a. Encourage project-level yield increase, efficient use of co-products, and/or system
integration to reduce losses within biofuel supply chains. Measuring how much
production is truly additional, however, is complicated given many field-level sources of
output variation such as weather, prices, and local policies. Output from similar local
production systems or from project land in the past can help set benchmarks for normal
production levels.
b. Allow carbon offsets, up to the value of the default LUC factor, through established
international programs dealing with LUC such as the initiative on Reducing Emissions
from Deforestation and Degradation (REDD), which offers compensation for preventing
emissions from tropical forest conversion; or the Kyoto Protocol’s clean development
mechanism (CDM), limited to projects related to LUC, which allows developing
countries to earn carbon credits that can be sold to industrialized nations, then used to
meet their treaty obligations.
c. Engage in broad-based LUC policy initiatives that encourage investments in productivity,
environmental protection, and carbon accounting or sequestration measures involving all
sectors beyond bioenergy. Reaching international agreements about land use policy could
take decades, as coordinated action involves extraordinary technical and political
challenges. Msangi et al. (2012) demonstrate that modest productivity gains for staple
crops in sub-Saharan Africa, or environmental protection around the Amazon, could
reduce or even completely offset a biofuel iLUC factor determined by the EPA. Forging
57
!
an alliance between bioenergy and other land-using sectors, and moving toward common
land-use policies for all sectors, is critical for meeting twenty-first-century needs without
compromising environmental or ecosystem services. A balance must be struck between
achieving a scale for coordinated efforts to have meaningful effects and avoiding new
layers of bureaucracy that introduce unnecessary inefficiencies.
One last policy issue: the inclusion of market-mediated LUC effects in biofuel LCA raises
additional questions such as how biofuel policy should treat other potentially significant market-
mediated emissions like those from agricultural management changes due to LUC (included in
RFS2 but not in California’s LCFS) or rebound effects in fossil fuel markets. Examples of
attempts by other market-based carbon policy proposals to incorporate indirect effects include
adjusting for leakage at borders for carbon tax and cap-and-trade policies, or within carbon
accounting programs like REDD (Murray 2008). LCFS might need amending to include
previously ignored indirect effects.
10 GHG)Emissions)from)Crude)Oils)with)Higher)Carbon)Intensity)
Recommendation+10.+Treat+all+crude+oils+as+part+of+the+overall+pool+of+transportation+fuels.+
Key Issues
Given the depletion of conventional fossil resources, there is increasing reliance on domestic and
imported unconventional oil resources that take more energy to extract, process, and transport.
Under an LCFS, any significant GHG benefits and debits associated with the use of
transportation fuels should be accounted for and treated equally, as discussed in Sections 8 and 9.
Fuels from certain petroleum resources—including Canadian oil sands, oil shales, and other
heavy crudes, together identified as high-carbon-intensity crude oils—can generate substantially
greater GHG emissions than most, but not all, conventional crude oils, potentially negating any
benefits achieved through the introduction of low-carbon fuels.
Some argue that in restricting the carbon content of fuels, a national LCFS would adversely
affect energy security by preventing the use of reliable high-carbon unconventional oils. This, it
is said, would encourage US reliance on less secure oil imports. It would then either lead to
export of those oil sands to other countries, resulting in little net reduction!in!global!CO
2
!(crude!
shuffling!CO2!leakage);!or it would lead to reduced global use of unconventional oils from
stable, competitive sources, hence greater global reliance on insecure or cartelized conventional
oil sources.
Summary Recommendations
58
!
Petroleum is not a uniform or homogenous liquid; it is a diverse mix of liquids comprised of
chains of hydrogen and carbon molecules. Initially California and the European Union (EU)
created a separate category of high-carbon-intensity crude oils within their LCFS and FQD. The
EU has persisted with a unique category for oil sands with a distinct set of regulations and
targets. This approach does not consider the reality that the CI of crude oils varies considerably,
with some conventional crudes, for instance, having higher CI values than some oil sands. It also
runs the risk of legal challenge from Canada, since targeting oil sands can be construed as
discriminating against a product of that country.
Instead of targeting specific high-carbon crudes, we recommend treating all crudes as part of the
overall pool of transportation fuels. We recommend adopting an approach that creates an
incentive to buy lower-CI crudes, invest in upstream improvements (such as carbon capture and
sequestration), and modify refinery designs to favor low-CI crudes. Each refinery (that is,
regulated party) would be assigned a benchmark value based on its CI in the baseline year. If it
exceeded this value, it would need to offset that increase by reducing GHG emissions in other
ways (or buying credits). If it reduced its crude oil CI, it could apply those reductions as credits
against its LCFS obligation. Some small additional shuffling of crude supply would occur—
whereby companies would send their lower-CI oil to US refineries and their higher-CI oil
elsewhere—but shuffling is a normal business practice for refineries in their effort to minimize
their costs. It is uncertain how much additional shuffling would occur. In any case, this shuffling
would diminish when other countries, starting with the EU, adopted a similar refinery-specific
approach. If the shuffling appeared to be significant, the extra transport energy consumed by
crude shuffling could perhaps be calculated and included (penalized) in the life-cycle
measurements for that crude (though constructing a counterfactual baseline might be onerous and
even impossible).
Discussion
Given the depletion of conventional fossil resources, there is increasing reliance on domestic and
imported unconventional oil resources that take more energy to extract, process, and transport.
25
These additional emissions should be captured and reduced. However, the fuel CIs of crudes
from different sources tend to overlap, as do those of conventional crudes and unconventional
sources such as oil sands from Canada (Figure 9). Flaring practices in particular, which emit
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
25
The U.S. EIA estimates production of unconventional crude oils (primary from Canada’s oil sands and
Venezuela’s Orinoco belt) will reach 4 million barrels per day higher in 2035 than in 2008 and will represent 5.6
percent of the global liquid fuels supply in 2035. US production of oil shale is projected to reach 0.4 million
barrels per day in 2035 in the reference case. The EIA also projects that relatively high prices will encourage
growth in global coal-to-liquid (CTL), gas-to-liquid (GTL), and biofuel production, from a combined total of 1.8
million barrels per day in 2008 to 8.4 million barrels per day in 2035, or 8 percent of total liquids supplied (U.S.
EIA 2010a).
59
!
large quantities of methane gas into the atmosphere from countries including Russia, Nigeria,
Iran, Iraq, and Algeria (Elvidge et al. 2009), also result in conventional crude with high carbon
intensity. As a result, the crude mix by region (Griffin et al. 2012) may not be significantly
differentiated (Figure 10).
!
Figure 9. Comparison of probability distributions of life-cycle GHG emissions of gasoline
produced from crude oil obtained from different regions (90-percent confidence interval for the
U.S. average represented by dashed lines). Source: Griffin et al. 2012.
!
!
Figure 10. The probability distributions of US average and regional life-cycle GHG emissions of
gasoline. PADD: Petroleum Administration for Defense District (PADD 1 = East Coast, PADD
2 = Midwest, PADD 3 = Gulf Coast, PADD 4 = Rocky Mountain, PADD 5 = West Coast).
Source: Griffin et al. 2012.
!
Individual refineries, however, are typically designed to take a particular slate of crudes, and
their emissions vary depending on the level of complexity and type of refined products produced.
We recommend treating all crudes as part of the overall pool of transportation fuels instead of
60
!
targeting specific high-carbon crudes. Each refinery would be assigned a benchmark based on its
CI rating in the baseline year. Each year, records of the volumes and crude marketing names
(also known as “marketable crude oil name”) for all crudes delivered to the US market by
regulated parties would be used to calculate refinery-specific crude CI values. If a refinery
exceeded this value, it would need to offset that increase by reducing GHG emissions in other
ways (or buying credits). If it reduced its crude oil CI, it could apply those reductions as credits
against its LCFS obligation. This approach might result in small amounts of additional shuffling
of crude supply, if companies sent their lower-CI oil to US refineries and their higher-CI oil
elsewhere. This shuffling would be reduced if other countries, starting with the EU, adopted a
similar refinery-specific approach. If the shuffling appeared to be significant, the extra transport
energy consumed by crude shuffling could be calculated and included (penalized) in the life-
cycle measurements for that crude (though constructing a counterfactual baseline might be
onerous and even impossible). By setting individual refinery-specific targets, this approach
provides the greatest incentive to innovate and to reduce the carbon intensity of crudes.
This recommended approach for handling high-carbon-intensity crude oils has the following
benefits:
• Preserves program benefits of accounting for all GHG emissions and debits associated
with the use of transportation fuel uses of each regulated party equally across all fuel
types
• Ensures consistent treatment of all crudes regardless of their origins
• Improves accounting of GHG emissions from production and transport of crude oil
• Promotes innovation by allowing companies to earn credits using innovative methods to
reduce crude carbon intensity or to shift to low-carbon fuel
Producers could reduce the CI of upstream operations—that is, of extraction, upgrading, and
refining. This could be done by improvements in the energy efficiency of technology and
systems or by carbon capture and sequestration. There is limited published information on the
costs to reduce the CI of high-carbon crude, specifically Canadian oil sands. The technologies
and costs are uncertain and not proven at commercial scale. A review of limited data in the
literature suggests that 0 to 7 percent of oil sands’ life-cycle CI (up to 35 percent of production
GHG emissions) could be reduced at less than $0.25 per barrel of oil (/BBL), and a maximum
reduction of 8 percent of life-cycle CI (40 percent of production GHG emissions) could be
achieved for a cost of $9/BBL. These figures translate into a carbon cost of between $28 and $87
per metric ton (MT) of CO
2
(Leiby and Rubin 2012).
There may be other innovative CI reduction strategies that can reduce a sizable percentage of
crude production emissions economically, such as carbon capture, utilization, and storage
(injecting captured CO
2
into depleted oil wells to recover untapped oil; CO
2
-enhanced oil
recovery), and flaring and venting reduction. As with other emission reduction credits discussed
61
!
in Sections 7 and 8, evaluation criteria need to be developed to ensure measurability,
verifiability, additionality, permanence, and security of these carbon reduction strategies.
Some raise concerns that a refinery-specific approach has the potential downside of shuffling
high-carbon crudes to countries that do not have climate policies and disproportionally affecting
the import of Canadian oil sands, which may have significant energy security implications. The
concern about carbon leakage and impacts on domestic economy is shared across all
climate/environmental policies, though the energy security concern deserves additional attention
and analysis. A study by Leiby and Rubin (2012) on the energy security impacts of an LCFS
concludes that policies that discourage high-carbon-intensity crude oil (HCICO) will lower the!
average!CI!of!crude!consumption in the United States, resulting in lower compliance obligation,
lower credit prices, and overall positive net energy security benefits. This occurs due to either (1)
shuffling (sending Canadian oil sands elsewhere instead of exporting them to the United States,
which is likely to occur regardless of LCFS policy (U.S. EIA 2011b), and/or changing the crude
mix that U.S. companies purchase), (2) reduced oil sands production, (3) investments in
technologies that reduce upstream GHG emissions during extraction and production of HCICO,
or (4) companies purchasing national LCFS credits and continuing to import Canadian oil sands
to the United States. Detailed discussion of these scenarios and the costs and benefits calculation
can be found in Leiby and Rubin (2012).
11 LCFS)Credits)
Recommendation+11.+Harness+market+forces+using+LCFS+credits.+
Key Issues
As a general principle, it is desirable to harness market forces. An LCFS does so by allowing
companies to buy and sell credits. If a company prefers not to invest directly in reducing GHG
emissions to achieve its carbon-intensity target, it can buy credits from other companies that can
reduce emissions at less cost. The net effect is attainment of targets at less overall cost. Key
questions regarding structuring and supporting credit trading in a national LCFS include: Should
regulated parties be allowed to trade credits across fuel pools? Should banking be allowed? Is a
cost containment mechanism such as a safety-valve price for credits necessary to insure against
the outcome that future compliance is more expensive than anticipated or otherwise
technologically infeasible? Should regulated parties be able to use LCFS credits in other carbon
programs and vice versa?
Summary Recommendations
62
!
Trading and banking. The efficiency and effectiveness of an LCFS credit market depends on
the design of the credit system, particularly the opportunities for trading and banking. Given the
uncertainties in feedstock costs and availability, their CI values, and the commercial success of
various biofuel refining technologies and fuel types (including “drop-in” bio-based gasoline and
diesel fuel), we recommend allowing unlimited trading of LCFS credits across the gasoline and
diesel fuel pools (and any others that might be created, such as jet fuels). Doing so can
significantly reduce compliance costs. For the same reason, banking of credits lowers the costs of
meeting the LCFS and stabilizes credit prices across compliance time periods.
Compliance and cost containment.!The design of an LCFS needs to address concerns about
large price swings that might result from unanticipated surges or crashes in economic growth,
weather and crop prices, and low-carbon fuel availability. While banking mechanisms reduce
price volatility, in extreme situations the number of banked credits available might not be
sufficient to avoid a credit price spike. To avoid the possibility of low-probability but high-
impact price spikes, we recommend the adoption of cost-containment mechanisms to protect
regulated companies. Such mechanisms would reduce uncertainty and accelerate capital
formation for low-carbon fuel production and deployment.
Carbon credits from other programs/jurisdictions. Transportation energy is produced
utilizing many resources and technologies in many places across many political jurisdictions.
GHG emissions in some places and from some activities are, or will soon be, regulated by other
(non-LCFS) GHG programs (such as carbon caps on utilities or carbon taxes). These energy
activities are already incentivized to reduce emissions through other market instruments. In these
cases, when energy producers in other political jurisdictions are subject to other carbon fees or
taxes, including electricity producers subject to cap-and-trade fees, we recommend that actual
emission reductions along the supply chains being regulated by an LCFS be taken into account
through regular updates of default CI values reflecting changes of emission intensity aggregated
over the industry, technology, or process over time.
Issues will arise, however, when obligated emission reductions in the other programs do not
actually occur but are met via credits, penalties, or fees, especially when there is a large disparity
between the actual or implicit carbon reduction costs or credit prices between the two programs.
In principle, policy should try to avoid imposing “double penalties” on regulated parties for the
same unit of emissions. We recommend recognizing these traded emission credits as actual
emission reductions but applying an adjustment factor accounting for the price difference
between programs based on published prices of credits traded in the same compliance period.
For example, if a refiner pays $15 per tonne CO
2
e of upstream emissions toward Alberta’s
Specified Gas Emitters Regulation (SGER) and LCFS credits are traded at $60 per tonne CO
2
e in
the same compliance period, a quarter of a carbon credit ($15/$60) can be counted as emission
reductions. This is the same approach we recommend for harmonizing a national LCFS with
63
!
other LCFS jurisdictions (such as British Columbia’s RLCFRR and the EU FQD), as discussed
in Section 13.
Discussion
11.1 Trading)and)banking)
LCFS credits are generated when regulated parties supply fuel with CI values lower than the
regulatory standard. Credits can be sold to regulated parties for whom credit purchasing
represents the least costly compliance option. In this way LCFS credit trading enables least-cost
reductions across all transportation fuel suppliers.
Credit trading between fuel pools allows regulated parties to apply LCFS credits generated in
one fuel pool to compliance obligations in any other fuel pool. It is applicable only in the context
of a policy design that disaggregates transportation fuels across multiple fuel pools (for example,
gasoline, diesel, and their respective substitutes), as discussed in Section 7. Because significant
uncertainty exists regarding projected technological advancements in producing cellulosic
biofuels as diesel or gasoline substitutes (Huang et al. 2012; Rubin and Leiby 2012), it could be
significantly more expensive to meet national LCFS requirements for equally stringent
phasedown paths if each market were required to meet its target separately. The ability to trade
credits across gasoline and diesel markets provides a straightforward mechanism to solve this
fundamental underlying uncertainty about future advances in biofuel technology.
Credit banking serves a similar function. It allows regulated parties to reserve or “bank” LCFS
credits generated in one period to satisfy compliance obligations in a future period. It represents
a type of intertemporal credit trading within firm, which can be used to hedge against risks of
higher future compliance costs. Our study results indicate that allowing banking of credits would
lower the costs of meeting an LCFS and stabilize credit prices (Rubin and Leiby 2012). Banking
provides additional temporal flexibility for regulated parties to meet increasingly stringent CI
standards (per the phasedown schedule). This is because regulators do not know the most cost-
effective time path for reducing fuel CI. Were they clairvoyant, an optimal phasedown path
could be specified and banking would be redundant. Trading across time, as banking allows, is
particularly suited to a carbon-mitigation system like an LCFS. This is because the
environmental impacts of CO
2
emissions result from cumulative emissions across time, and
earlier reductions can only reduce the total effect.
11.2 Compliance)and)cost)containment)
The trading price for LCFS credits represents an important signal to fuel suppliers and
prospective low-carbon fuel developers. In particular, the credit price provides a signal to market
participants regarding the relative cost-effectiveness of available compliance options. Parties
capable of reducing fuel CI at costs lower than the credit price will generally invest in achieving
64
!
those reductions, while parties that cannot generate CI reductions at such costs will generally
comply by purchasing credits. The policy design should provide a clear and coherent price signal
to the market. Absent a clear signal, investment decisions will be based on more abstract and
uncertain notions of LCFS credit value, which will generally yield inefficient investment
decisions and increase compliance costs.
In addition to enabling efficient investment decisions, the credit price signal can facilitate
efficient access to capital markets. The high capital requirements of developing fuel supplies in
general, and low-carbon fuel supplies in particular, have the potential to create a bottleneck for
scaling up production capacity. Efficient access to capital markets can therefore accelerate
deployment to some extent. A clear LCFS credit price signal should help low-carbon fuel
developers demonstrate conformance with financing criteria by establishing the value of
supplying low-CI fuels.
The design of an LCFS, however, also needs to address concerns about large price swings that
could potentially result from unanticipated developments in availability of low-carbon fuels,
economic growth, and weather and crop prices. While a banking mechanism does reduce price
volatility, the number of banked allowances available at any given time may limit their efficacy.
To improve the structure and integrity of an LCFS program, mechanisms need to be developed to
address potentially severe spikes in prices due to low-probability but high-impact scenarios.
Such mechanisms will reduce perceived uncertainty and accelerate capital formation for low-
carbon fuel production and deployment, and improve support for the program overall.
The need for allowance price containment mechanisms is highlighted by the experience with
southern California’s REgional CLean Air Incentives Market (RECLAIM), a cap-and-trade
program aimed at reducing NOx and SOx emissions from industry and electricity utilities. In
2000, unanticipated regulatory-driven disruptions in the electricity sector and a weather-driven
fall in hydroelectricity generation led to a spike in demand for electricity from fossil-fuel
generators and consequently a spike in allowance prices. In the absence of any price containment
mechanisms, the price of NOx allowances traded in 2000 exceeded $45,000 per ton NOx,
compared to the average price of $4,284 per ton traded in 1999 (Burtraw et al. 2006).
One potential price containment mechanism is a price ceiling (also called a safety valve), where
support is triggered when allowance prices reach a predefined level. At this stage, participants
can purchase unlimited allowances from the regulator at the ceiling price. Trigger prices usually
increase over time—for example, the ceiling price in New Zealand’s Emissions Trading Scheme
(ETS) will increase by NZ$5 per annum (http://www.climatechange.govt.nz/emissions-trading-
scheme/) while the ceiling prices in California’s cap-and-trade program will increase annually by
5 percent plus inflation measured by the consumer price index (CPI) (Enion 2012).
65
!
When unlimited allowances can be introduced at the trigger price, quantity uncertainty replaces
price uncertainty and risks the entire environmental integrity of an LCFS program. This is the
reason cited for not allowing a price ceiling in the EU ETS program (Moslener and Sturm 2008).
For example, participants may undermine emission goals by buying allowances at current low
safety-valve prices, and banking and using them sometime in the future when caps as well as
ceilings are more stringent (Murray, Newell, and Pizer 2008). This can be addressed by what is
known as a soft price ceiling or allowance reserve (Fankhauser and Hepburn 2010; Murray,
Newell, and Pizer 2008). With a soft price ceiling, a small percentage of allowances are saved in
the reserve and can be used at a trigger price level (Fell et al. 2011). Most recent legislative
proposals in the United States for GHG reduction programs—such as the Waxman-Markey and
Kerry-Boxer bills, as well as California’s cap-and-trade program, which went live in 2012—have
adopted the allowance reserve option to reduce price volatility.
Careful selection of a ceiling price is necessary to ensure the overall integrity of an LCFS
program. A low price ceiling will mean the mechanism is triggered more frequently and, in
extreme cases, may convert an LCFS program into a carbon tax regime, as was observed in the
Danish ETS program between 2000 and 2003 (Jacoby and Ellerman 2004). A very high trigger
price, on the other hand, will reduce price volatility to a smaller extent and only safeguard
participants against more extreme spikes. The trigger price also depends upon the size of the
reserve—a smaller reserve will necessarily warrant a higher price ceiling.
Since the reserve is created by taking away allowances allocated in any given period and thus
reducing the allowance budget, this increases the stringency of the program and hence the
potential for higher allowance prices. In the California cap-and-trade program, this situation is
addressed by increasing the allowable number of offsets by an equal amount. More careful study
is needed to implement a price containment mechanism for a national LCFS. Table! reviews
price containment mechanisms in market-based regulations.
Table 8. Overview of price containment mechanisms in market-based regulations
Cap-and-trade
program
Price containment
mechanism
Details
CA RECLAIM
(year 2000
version)
No price floors or ceilings
EU ETS
No price ceiling. Penalties of
€40/ton and €100/ton are
imposed in Phases I and II
respectively.
Paying penalties does not release participants from the obligation
to reduce emissions—excess emissions must also be offset in the
following compliance period (Fankhauser and Hepburn 2010).
Danish ETS
(2000–2003)
Hard price ceiling. Ceiling
fixed at a level sufficiently
below the marginal cost.
Due to the low ceiling and stringent nature of the cap (emission
caps were 30% below average annual emissions in 1994–1998),
the system operated as a “tax with tradable exemptions” (Jacoby
and Ellerman 2004).
CA cap-and-trade
Soft price ceiling. Around
Three levels of price supports. For 2013, those price triggers will
66
!
4% of the allowances will be
stowed away into a reserve
every quarter.
be $40, $45, and $50. Correspondingly, the reserve is divided
equally into three tiers.
NZ ETS
Hard price ceiling initially at
low levels but rising rapidly
Hard price ceiling at NZ$12.50 (US$10) per metric ton currently.
After December 2012, the ceiling will double to NZ$25 (US$20)
per metric ton of GHG. Subsequently, the NZ ETS Review Panel
(2011) recommends increasing the ceiling by NZ$5 per annum to
reach NZ$50 in 2017.
Regional
Greenhouse Gas
Initiative
No price ceiling
However, reaching some predefined CO
2
allowance price
thresholds will trigger expanded use of offsets—from a default
3.3% of a power plant’s total compliance obligation to 5% and
even 10%.
The RFS2 program allows the EPA to sell cellulosic biofuel credits when it determines that
quantities of available cellulosic biofuels are below levels required in the Energy Independence
and Security Act of 2007. For 2011 the EPA has set the price at $1.13 per credit (75 FR 76790)
(Rubin and Leiby 2012). Rubin and Leiby (2012) set the safety-valve credit price in their study
at $300/MtCO
2
e, which is equivalent to a $30/MtCO
2
e carbon tax, or about a $0.38/gallon
ceiling on the cost of reducing gasoline CI by 10 percent (Rubin and Leiby 2012). We
recommend that a national LCFS include provision for substantial flexibility and trading of
credits, with a credit safety-valve price and at least limited banking. While these mechanisms can
entail trade-offs between managing the uncertain cost of compliance and assuring regulatory
effectiveness, they seem essential for the workability of a national LCFS in a global motor fuel
market. Further examination of the implications of alternative safety-valve levels and alternative
banking systems is merited.
11.3 Carbon)credits)from)other)programs/jurisdictions)
The production, transport and delivery of transportation energy utilize a wide range of resources
and technologies across the entire life-cycles, which take place in many places across many
political jurisdictions. Energy producers in other political jurisdictions can be, or maybe soon,
subject to other carbon caps or taxes. If upstream energy activities subject to an LCFS are
already required to reduce emissions elsewhere, we recommend that actual emission reductions
along the supply chain be recognized through regular updates of default CI values using industry
averages.
When emission reductions, trading, and fees are covered under separate market mechanisms, it is
appropriate to recognize these emission reductions and regularly update the average emissions
factor to reflect changes of emission intensity aggregated over the industry, technology, or
process over time. As any additional emission reductions (or debts) by individual companies
beyond policy requirements are appropriately rewarded in the other regulated markets, adopting
an industry, technology, or process average also avoids any potential issue associated with
67
!
double counting/crediting/penalizing. This approach also appropriately recognizes the trade and
jurisdictional authority of emission sources regulated under other jurisdictions.
Issues will arise, however, when obligated emission reductions in the other programs do not
actually occur but are met via credits, penalties, or fees, especially when there is a large disparity
between the actual or implicit carbon reduction costs or credit prices between the two programs.
In principle, policy should try to avoid imposing “double penalties” on regulated parties for the
same unit of emissions. We recommend recognizing these traded emission credits as actual
emission reductions but applying an adjustment factor accounting for the price difference
between programs based on published prices of credits traded in the same compliance period.
For example, if a refiner pays $15 per tonne CO
2
e of upstream emissions toward Alberta’s
Specified Gas Emitters Regulation (SGER)(Table 9) and LCFS credits are traded at $60 per
tonne CO
2
e in the same compliance period, a quarter of a carbon credit ($15/$60) can be counted
as emission reductions. This is the same approach we recommend for harmonizing a national
LCFS with other LCFS jurisdictions (such as British Columbia’s RLCFRR and the EU FQD), as
discussed in Section 13.
Table 9. Alberta’s three principal greenhouse gas policies
Name of policy
Main sectors covered / Likely
emission reduction
Description
Specified Gas
Emitters
Regulation
(SGER)
Industrial facilities
(electricity, oil and gas,
other)
Likely reduction in annual
emissions in 2020 relative to
no policy: 1.5–5 Mt CO
2
e
This regulation, which took effect in July 2007, sets GHG
intensity (emissions per unit of production) targets for all
facilities emitting more than 0.1 Mt CO
2
e per year. The target
for a facility beginning operation before 1999 is 12% below the
average intensity for 2003–05. Newer facilities are exempt for
their first three years of operation and then face targets that
gradually increase to reach, in the ninth year of operation, 12%
below the intensity measured in the third year. Facilities with
emissions higher than their targets can comply by making
payments of $15 per tonne CO
2
e into the Climate Change and
Emissions Management Fund (see below) and by purchasing
offset credits from projects in Alberta.
CCS Major
Initiatives
Industrial facilities
(electricity, oil and gas,
other)
Likely reduction in annual
emissions in 2020 relative to
no policy: 1.5–5 Mt CO
2
e
In 2009 the Alberta government selected four large-scale CCS
projects to receive grants totaling $2 billion over 15 years. The
projects—a coal-fired power plant retrofit, an oil sands
upgrader, an underground coal gasification project, and a CO2
pipeline—are expected to start up by 2015. However, it is not
yet certain that all four projects will be constructed.
Climate Change
and Emissions
Management
Fund (CCEMF)
Industrial facilities
(electricity, oil and gas,
other)
Likely reduction in annual
emissions in 2020 relative to
no policy: 0.5–1.5 Mt CO
2
e
The $15 per tonne payments into the CCEMF, made under the
SGER, are reinvested in a wide range of emission reduction
projects. In 2007–10, $256 million were paid into the CCEMF;
to date $126 million has been committed to approved projects.
68
!
Source: Bramley et al. 2011
12 PerformanceKBased)Sustainability)Standards)
Recommendation+12.+Implement+performanceDbased+sustainability+standards.+
Key Issues
Aside from GHGs, there are other important nonmarket impacts associated with energy
production. This group of sustainability concerns includes environmental sustainability
(conservation of air, water, soil, biodiversity, and land use) and social sustainability (human
and labor rights, local food security, rural development). The challenge is to determine the
extent to which an LCFS should include or be linked with rules to limit adverse impacts in
these other areas.
Summary Recommendations
Given the huge scale of energy production activities and their potentially large impacts, and
because an LCFS would play an instrumental role in stimulating large energy investments, we
believe that some sustainability safeguard mechanisms are needed. We recommend formulating
(1) minimum sustainability requirements, including conservation (not allowing conversions of
high-biodiversity and high-carbon-stock areas); and (2) reporting requirements for specified
impacts or voluntary certification.
A sustainability standard that includes key environmental and social impacts should be
performance based—it should not prescribe specific technology or practices but instead should
focus on measurable outcomes with clear expectations regarding performance, measurement,
verification, and enforcement. Effort should be made to identify incentive mechanisms that
motive innovation beyond minimum compliance thresholds established by existing laws and
regulations.
Discussion
As biofuel production increased in the early 2000s, new studies began to link this increased
production to increased risk of adverse environmental impacts (Donner and Kucharik 2008;
Miller, Landis, and Theis 2007; Robertson et al. 2008) and of social and economic impacts (FAO
2008b; Rajagopal et al. 2007; Tenenbaum 2008), casting doubt on the real GHG benefits of some
biofuels (Fargione et al. 2008; Gibbs et al. 2008; Searchinger et al. 2008a). Even though there are
69
!
vastly different views on the nature and magnitude of causal relationships between biofuel
policies and diverse environmental impacts, land use changes, and global food prices, there have
been increasing efforts to adopt sustainability requirements to minimize potential social and
environmental damage and unintended consequences. These efforts are intended to provide
environmental and social safeguards for biofuels directly or indirectly via biofuel subsidies, tax
credits, demand mandates, and other mechanisms.
In the past few years, sustainability requirements for biofuel production have been
adopted/implemented by the Netherlands (Cramer et al. 2006, 2007; NEN 2009), the United
Kingdom (RFA 2009), Germany (German government 2007; WWF 2006), the European Union
(EC 2008), and California (CEC 2008). International organizations, including the United
Nations Food and Agriculture Organization (FAO), the UN Environment Programme (UNEP),
and the G8’s Global Bioenergy Partnership (GBEP), have encouraged and supported the
research, modeling, and negotiation efforts among stakeholders at the country level. There are
also more private and public efforts in promoting certifications, facilitating information
sharing, and developing guidelines for sustainability best management practices (BMP). Many
new, especially commodity-based, biofuel-targeted certifications have recently been or are
being established, such as the Roundtable on Sustainable Palm Oil (RSPO), the Roundtable on
Responsible Soy (RTRS), the Better Sugarcane Initiative (BSI), the Council on Sustainable
Biomass Production (CSBP, focusing on second-generation feedstock), and the Roundtable on
Sustainable Biofuels (RSB, focusing on creating internationally consistent sustainability criteria
and certification schemes). A more detailed review of these recent activities can be found
elsewhere (Endres 2010; Lewandowski and Faaij 2006; van Dam et al. 2008; Winrock
International 2009).
Due to the importance and complexity of sustainability issues, many of which are irreversible,
we recommend a hybrid of a minimum standard plus self-reporting approach: (1) require
independent verification or voluntary certification of meeting two mandatory minimum
requirements: no newly converted land in high-biodiversity areas, and no newly converted land
in high-carbon-stock areas; and (2) encourage/require self-evaluation and reporting against
sustainability principles and criteria,
26
plus verification; or alternatively, voluntary
certification.
27
Self-reporting and verification. The self-reporting requirement would encourage regulated
parties to develop self-evaluations against sustainability principles and criteria agreed to via a
stakeholder process and to report these evaluations. The self-evaluation would not be required to
demonstrate meeting all of the sustainability principles and criteria, but its data accuracy would
need to be verified by an independent auditor. Independent verification would also be required to
ensure that the two mandatory minimum requirements were fully met. The intention would be to
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
26
The list of principles and criteria should ideally cover environmental, social, and economic criteria developed by
environmental agencies in consultation with the public.
27
Establishing mechanisms for dispute will be a critical component of certification schemes.
70
!
take a first step toward creating an easy means of reporting sustainability performance,
encouraging innovation by regulated parties in developing metrics for evaluating sustainability
performance, and collecting verifiable data. Annual evaluation and periodic review of the
program would need to be conducted to determine whether further adjustments or an alternative
system were needed in the future.
Demonstration of compliance. To demonstrate compliance, the following requirements should
be met:
• submit information on biofuels’ compliance with the two mandatory minimum requirements
as specified as (1) above and self-evaluation and reporting or voluntary certification against
other principles and criteria as specified as (2) above, and
• arrange for an adequate standard of independent auditing.
Challenging issues that require significant progress toward the implementation of sustainability
standard include improvement in chain-of-custody (CoC) from fuel-to-field level; and potential
WTO challenges to some of the sustainability standards.
CoC in the current biofuel carbon accounting schemes, including RFS2 and California’s LCFS,
does not track feedstock to the field level, only to refinery gates. Establishing sustainability
requirements would require tracking CoC to the field level. Extending CoC tracking to the field
level would incentivize fuel providers to report agricultural practices that have large carbon-
reduction benefits such as no-till, reduced fertilizer use, and yield improvement that cannot be
acknowledged under the current carbon accounting CoC system.
The CoC tracking system that’s best for the United States need not to be the same as the one
required in the European Union’s RED sustainability requirement: the mass balance CoC. The
mass balance CoC requires that certified/verified feedstock not be separated from the
feedstock/biofuel and that it stays with the finished products along the supply chain. A mass
balance CoC system would be challenging to establish in the United States for two reasons: (1)
in an LCFS, carbon credits may be separated from fuels in the future, as in RFS2 where RIN
credits can be sold separately from fuels, allowing for more flexible trading of credits; (2) food
commodities such as soybeans have very complex supply chains; in this case, soybean oil is
only a by-product and thus its CoC is hard to establish. For these reasons, a more flexible CoC
system such as book-and-claim might be more desirable, especially in the early years of
compliance. In a book-and-claim system, end users (in this case, the fuel suppliers/importers)
submit certificates that guarantee the production of a certain quantity of sustainable biomass,
but the certified products can be delivered anywhere.
The WTO might challenge mandatory sustainability standards and even carbon intensity
standards (for example, the recent WTO dispute in the European Union regarding the
calculation of carbon intensity for biofuels and Canadian oil sands). There is a chance that the
recommendations proposed here would largely avoid WTO challenges since (1) the mandatory
minimum requirements propose here regarding no newly converted land in high-biodiversity
71
!
and high-carbon-stock areas are consistent with the EU-RED requirement and have the least
likelihood of provoking WTO challenges, and (2) mandatory reporting and voluntary
compliance with the broader sustainability principles and criteria was considered to have the
least likelihood of violating WTO rules.
28
!
!
13 G lo b a l)LC FS)Policies)
Recommendation+13.+Harmonize+global+LCFS+policies.+
Key Issues
LCFS policies adopted in other countries and regions can vary significantly in policy design,
stringency levels, system boundaries, coverage of fuel types, and various other details. The goal
of harmonization is to create a consistent and acceptable approach to reducing the carbon
intensity of fuels to maximize the effectiveness and efficiency of the policies, while providing
individual countries and regions the freedom and flexibility to tailor the policies to their local
circumstances.
Summary Recommendations
Harmonization can be achieved by adopting a globally consistent certification system, starting at
the feedstock level. Certificate harmonization allows for robust policy frameworks and thus more
room for policy/political differences while still remaining effective. Achieving a harmonized
certification system will require an improved chain-of-custody tracking system in order to
provide transparent and reliable information about biofuel production across regions.
LCFS policies can be further harmonized between states and regions through credit
harmonization, which requires adopting unified methods (where possible) and using credit
multipliers to adjust non-unified aspects. These two methods allow credits or certificates to be
valued equivalently across regions and traded efficiently to comply with regional low carbon fuel
policies, even when they vary in stringency, system boundary, and fuel carbon ratings. Allowing
credits to be traded across countries or regions will increase policy effectiveness and efficiency
and lower the overall compliance costs. Fuel shuffling will also be reduced, hence strengthening
LCFS policy.
Discussion
!
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
28
See review and more discussion in Lendle and Schaus 2010; Yeh et al. 2009.
72
!
The goal of harmonization is to create a consistent and acceptable approach to reducing fuel CI
while allowing individual regions the freedom to incorporate a policy that makes the most
sense in that region. Presently, each governing body has its own interpretation of how an LCFS
should be structured and implemented. Because LCFS policies are aimed at reducing CO
2
emissions from the transportation sector within regional boundaries as part of global reduction
goals, harmonization is necessary if LCFS programs are to be effective at the aggregate level at
reducing global GHG emissions. As with any policy, two key areas determine policy success:
policy effectiveness and policy cost. Any harmonization strategy thus must serve to improve
policy effectiveness (aid in incentivizing behaviors that reduce GHG emissions) while lowering
implementation costs.
Improving policy effectiveness. Currently, a gallon of low-carbon fuel or a unit of energy (such
as an MJ) of low-carbon fuel (for example, sugarcane ethanol) may receive different CI ratings
under different low-carbon fuel policies due to methodological difference. This can create
incentives for shuffling of fuels from regions that assign higher CI values to regions that assign
lower CI values to the same fuels as long as the costs of shuffling are lower than the costs of
compliance (Kessler, Yeh, and Sperling 2012; Leiby and Rubin 2012). Shuffling will reduce the
effectiveness of low-carbon fuel policies by appearing to achieve GHG emission reductions on
paper even though no net GHG emission reduction takes place in reality. In the worst case, net
emissions could actually increase due to the extra transport distance required to shuffle fuels
and/or feedstock. The incentives for shuffling fuel between LCFS regions will disappear if fuels
and feedstock are treated equally across policies.
Reducing compliance costs. Setting of different stringency levels in different LCFS policies
leads to different compliance costs and a mix of compliance options. If the compliance options
are limited to local/regional resources, the compliance costs will be higher than if mitigation
options can be shared across regions. Similarly, if the compliance goals of regional LCFS
policies can be harmonized (compliance in one region counted toward compliance in another
region) while ensuring that the overall stringency level can be achieved, significant cost savings
and less fuel/feedstock shuffling will occur (Eggert and Greaker 2012).
13.1 Feedstock)certificate)harmonization)
The European Union has developed a harmonized certification scheme at the EU level. If similar
certification schemes are adopted across all low-carbon fuel policies, certificates will be able to
be traded across regions, yielding harmonized low-carbon feedstock for fuel production. Other
low-carbon fuel policies should look at the possibility of utilizing current certification schemes
to assess compliance in some capacity. For instance, if a certified product could be used to
support pathway approval or proof of pathway compliance under the California LCFS,
harmonization of feedstock certificate would occur. Once feedstock certificate is harmonized,
each LCFS implementation could apply their rules for calculating CI based on the same pathway
information that is consistent across LCFS regions.
73
!
13.2 Carbon)credit)harmonization)
Unlike a feedstock or fuel certificate that carries information that qualifies a feedstock or a fuel,
an LCFS credit is effectively an allowance to emit 1 tonne of CO
2
(where allowances are limited
by the baseline of each standard). There is considerable appeal in generating credits under low-
carbon fuel policies, as they allow for flexibility in assessing compliance and could conceivably
be transferable to other GHG emission reduction programs such as cap-and-trade.
We have identified two different approaches to achieving credit harmonization: (1) unifying the
credit generation methodology, or (2) using credit multipliers to treat further regional policy
differences, such as the inclusion of iLUC (indirect land use change) and treatment of HCICO
(high-carbon-intensity crude oil) in calculating fuel CI values.
Methodology unification. Standardizing the method of doing LCA calculations is essential for
the harmonization of LCFS/fuels programs. As discussed, the first step toward this
harmonization will be to develop a credible and consistent Chain-of-custody (CoC) tracking and
reporting mechanism; this will allow the same information to be accurately conveyed to each
region. Using the life-cycle approach, it can be expected that each state or region will have its
own lookup table for determining CI values. This is due to the fact that average input materials,
production processes, and emission factors differ from region to region, as each region will have
different resource and technology mixes for fuel production (Griffin et al. 2012; Yang 2012).
If standard methodologies and procedures are established, including underlying assumptions, for
determining how CI is calculated, there will be no need for a standardized baseline, a
standardized year target, or a single CI lookup table to allow for credit trading across regions. A
credit represents a tonne of CO
2
reduction across the entire harmonized region calculated using a
consistent methodology (Kessler, Yeh, and Sperling 2012).
Credit multiplier. For one region to accept credits from another region, it will be necessary for
the credits to be valued similarly. If the underlying methodology for generating a credit in one
region is not identical to how another region generates credits, there will clearly be a disconnect
between the credits, and their values will not be identical, thus hindering credit trading.
Credit multipliers could be effectively utilized if a credit market were established that
normalized all credits generated in different policy regions to a reference LCFS region, such as a
national LCFS in the United States, or a theoretical LCFS region, to generate “equivalent
credits” (if a reference region is chosen) or “theoretical credits” (if a theoretical LCFS region is
used) to allow for credit trading. While this approach could work, it creates unnecessary policy
redundancy and confusion, which could substantially increase implementation costs.
In summary, the first step toward harmonization is to adopt a globally consistent certification
system, starting at the feedstock level. Certificate harmonization allows for robust policy
frameworks and thus more room for policy/political differences while still remaining effective.
Achieving a harmonized certification system will require a full chain of custody in order to
74
!
provide transparent and detailed information for biofuel production across regions. Overall the
long-term, methodology unification will be the most effective and easy to implement toward
credit harmonization, allowing credits to be traded across countries or regions to increase policy
effectiveness and efficiency and lower the overall compliance costs.
Conclusions)
This report discusses thirteen key issues that must be addressed in the design of an LCFS for the
United States. The underlying technical analyses supporting various recommendations are
summarized in the Technical Analysis Report (TAR).
! )
75
!
References)
Bandivadekar, A., K. Bodek, L. Cheah, C. Evans, T. Groode, J. Heywood, E. Kasseris, M. Kromer, and
M. Weiss. 2008. On the road in 2035: Reducing transportation’s petroleum consumption and
GHG emissions. Cambridge, Massachusetts: MIT Laboratory for Energy and the Environment.
Berndes, G., N. D. Bird, and A. Cowie. 2010. Bioenergy, land use change and climate change
mitigation. IEA Bioenergy:ExCo:2011:04.
Bramley, Matthew, Marc Huot, Simon Dyer, and Matt Horne. 2011. Responsible action? An
assessment of Alberta’s greenhouse gas policies. The Pembina Institute.
Brandt, Adam R. 2008. Converting oil shale to liquid fuels: Energy inputs and greenhouse gas
emissions of the Shell in situ conversion process. Environmental Science and Technology
42:7489–95. doi: 10.1021/es800531f.
———. 2011. Upstream greenhouse gas (GHG) emissions from Canadian oil sands as a
feedstock for European refineries. A report prepared for the European Commission.
Burnham, Andrew, Jeongwoo Han, Corrie E. Clark, Michael Wang, Jennifer B. Dunn, and Ignasi
Palou-Rivera. 2011. Life-cycle greenhouse gas emissions of shale gas, natural gas, coal,
and petroleum. Environmental Science and Technology 46:619–27. doi:
10.1021/es201942m.
Burtraw, D., A. E. Farrell, L. H. Goulder, and C. Peterman. 2006. Lessons for a cap-and-trade
program. In Managing Greenhouse Gas Emissions in California.
CARB (California Air Resources Board). 2011. Staff report: Initial statement of reasons for
proposed rulemaking. Proposed amendments to the Low Carbon Fuel Standard.
CEC (California Energy Commission). 2008. Assembly Bill 118 (Núñez, Statutes of 2007,
Chapter 750): Implementation of the Alternative and Renewable Fuel and Vehicle
Technology Program.
Charpentier, Alex D., Oyeshola Kofoworola, Joule A. Bergerson, and Heather L. MacLean.
2011. Life cycle greenhouse gas emissions of current oil sands technologies: GHOST
model development and illustrative application. Environmental Science and Technology
45:9393–9404. doi: dx.doi.org/10.1021/es103912m.
Cramer, J., Erik Wissema, Ella Lammers, Daan Dijk, Hans Jager, Sander van Bennekom, Ewald
Breunesse, Robert Horster, Caroline van Leenders, Wim Wolters, Helma Kip, Hugo
Stam, Andre Faaij, and Kees Kwant. 2006. Criteria for sustainable biomass production.
Sustainable Production of Biomass project group, the Netherlands.
Cramer, J., Erik Wissema, Mariska de Bruijne, Ella Lammers, Daan Dijk, Hans Jager, Sander
van Bennekom, Ewald Breunesse, Robert Horster, Caroline van Leenders, Steven
Wonink, Wim Wolters, Helma Kip, Hugo Stam, Andre Faaij, and Kees Kwant. 2007.
Testing framework for sustainable biomass. Sustainable Production of Biomass project
group, the Netherlands.
Donner, Simon D., and Christopher J. Kucharik. 2008. Corn-based ethanol production
compromises goal of reducing nitrogen export by the Mississippi River. Proceedings of
the National Academy of Sciences 105:4513–18. doi: 10.1073/pnas.0708300105.
76
!
Drabik, Dusan, and Harry de Gorter. 2011. Biofuel policies and carbon leakage. AgBioForum
14:104–10.
EC (European Council). 2008. European Parliament legislative resolution of 17 December 2008
on the proposal for a directive of the European Parliament and of the Council amending
Directive 98/70/EC as regards the specification of petrol, diesel and gas-oil and
introducing a mechanism to monitor and reduce greenhouse gas emissions from the use
of road transport fuels and amending Council Directive 1999/32/EC, as regards the
specification of fuel used by inland waterway vessels and repealing Directive 93/12/EEC
(COM(2007)0018 – C6-0061/2007 – 2007/0019(COD)).
Eggert, Håkan, and Mads Greaker. 2012. Trade policies for biofuels. The Journal of
Environment and Development 21:281–306. doi: 10.1177/1070496512443541.
Elvidge, Christopher D., Daniel Ziskin, Kimberly E. Baugh, Benjamin T. Tuttle, Tilottama
Ghosh, Dee W. Pack, Edward H. Erwin, and Mikhail Zhizhin. 2009. A fifteen year record
of global natural gas flaring derived from satellite data. Energies 2:595–622. doi:
10.3390/en20300595.
Endres, Jody M. 2010. Clearing the air: The meta-standard approach to ensuring biofuels
environmental and scoial sustainability. Virginia Environmental Law Journal 28 (1).
Enion, Rhead. 2012. Auction prices in California’s cap-and-trade program.
Ensys Energy. 2010. Keystone XL assessment, final report.
Fankhauser, S., and C. Hepburn. 2010. Designing carbon markets. Part I: Carbon markets in
time. Energy Policy 38:4363–70.
FAO (Food and Agriculture Organization). 2008a. High level conference on world food security:
The challenges of climate change and bioenergy. Soaring food prices: Facts, perspectives,
impacts and actions required.
———. 2008b. The state of food and agriculture. Biofuels: Prospects, risks and opportunities.
FAO, IFAD, IMF, OECD, UNCTAD, WFP, World Bank, WTO, and IFPRI. 2011. Price
volatility in food and agricultural markets: Policy responses. Interagency report to the
G20, UN Food and Agriculture Organization.
Fargione, Joseph, Jason Hill, David Tilman, Stephen Polasky, and Peter Hawthorne. 2008. Land
clearing and the biofuel carbon debt. Science 319:1235–38. doi:
10.1126/science.1152747.
Farrell, Alexander E., and Daniel Sperling. 2007. A low-carbon fuel standard for California, part
2: Policy analysis. Institute of Transportation Studies, University of California, Davis,
Research Report UCD-ITS-RR-07-08.
Fell, H., D. Burtraw, R. D. Morgenstern, and K. L. Palmer. 2011. Soft and hard price collars in a
cap-and-trade system: A comparative analysis. Journal of Environmental Economics and
Management 61:107–18.
German government. 2007. Regulation regarding requirements for the sustainable production of
biomass for use as biofuels (Biomass Sustainability Regulation—BioNachV).
77
!
Gibbs, Holly K., Matt Johnston, Jonathan A. Foley, Tracey Holloway, Chad Monfreda, Navin
Ramankutty, and David Zaks. 2008. Carbon payback times for crop-based biofuel
expansion in the tropics: The effects of changing yield and technology. Environmental
Research Letters 3:034001.
Greene, David L., Howard H. Baker Jr., and Steven E. Plotkin. 2011. Reducing greenhouse gas
emissions from U.S. transportation. Prepared for the Pew Center on Global Climate
Change.
Griffin, W. Michael, Matthew Kocoloski, Kimberley A. Mullins, and Aranya Venkatesh. 2012.
Addressing uncertainty in life-cycle carbon intensity in a national low carbon fuel
standard. Energy Policy special issue on Low Carbon Fuel Policy: submitted.
Howarth, Robert W., Renee Santoro, and Anthony Ingraffea. 2011. Methane and the greenhouse-
gas footprint of natural gas from shale formations. Climate Change. doi: 10.1007/s10584-
011-0061-5.
Huang, Haixiao, Madhu Khanna, Hayri Onal, and Xiaoguang Chen. 2012. Stacking low carbon
policies on the renewable fuels standard: Economic and greenhouse gas implications.
Energy Policy special issue on Low Carbon Fuel Policy: accepted.
IEA (International Energy Agency). 2010. Energy technology perspectives 2010.
IMF (International Monetary Fund). 2011. Market-based instruments for international aviation
and shipping as a source of climate finance. Background paper for the report to the G20
on mobilizing sources of climate finance.
Jacoby, H. D., and A. D. Ellerman. 2004. The safety valve and climate policy. Energy Policy
32:481–91.
Kessler, Jeff, Sonia Yeh, and Daniel Sperling. 2012. Moving towards a harmonized state,
regional, and global low-carbon fuel policy. Manuscript draft. Institute of Transportation
Studies, University of California, Davis.
Khanna, Madhu, Hayri Onal, and Haixiao Huang. 2011. Economic and greenhouse gas reduction
implications of a national low carbon fuel standard. A report submitted to the National
Low Carbon Fuel Standard Study. University of Illinois, Urbana-Champaign.
Koh, L., and D. Wilcove. 2008. Is oil palm agriculture really destroying tropical agriculture?
Conservation Letters 1:60–64.
Lapola, David M., Ruediger Schaldach, Joseph Alcamo, Alberte Bondeau, Jennifer Koch,
Christina Koelking, and Joerg A. Priess. 2010. Indirect land-use changes can overcome
carbon savings from biofuels in Brazil. Proceedings of the National Academy of Sciences
107:3388–93. doi: 10.1073/pnas.0907318107.
Leiby, Paul N., and Jonathan Rubin. 2012. Energy security implications of a national LCFS.
Energy Policy special issue on Low Carbon Fuel Policy,
http://dx.doi.org/10.1016/j.enpol.2012.06.058.
Lendle, Andreas, and Malorie Schaus. 2010. Sustainability criteria in the EU Renewable Energy
Directive—Consistent with WTO rules? ICTSD Information Note No. 2.
78
!
Lewandowski, I, and A.. Faaij. 2006. Steps towards the development of a certification system for
sustainable bio-energy trade. Biomass and Bioenergy 30:83–104. doi: 10.3141/2252-06.
Lutsey, Nicholas P. 2010. Comparison of emissions, energy, and cost impacts of diesel and
hybrid models in the U.S. in 2010. Transportation Research Record: Journal of the
Transportation Research Board 2252: 40–48.
McCarthy, Ryan, and Christopher Yang. 2010. Determining marginal electricity for near-term
plug-in and fuel cell vehicle demands in California: Impacts on vehicle greenhouse gas
emissions. Journal of Power Sources 195:2099–2109. doi:
10.1016/j.jpowsour.2009.10.024.
McKinsey and Company. 2009. Roads toward a low-carbon future: Reducing CO
2
emissions
from passenger vehicles in the global road transportation system.
Melillo, Jerry M., John M. Reilly, David W. Kicklighter, Angelo C. Gurgel, Timothy W. Cronin,
Sergey Paltsev, Benjamin S. Felzer, Xiaodong Wang, Andrei P. Sokolov, and C. Adam
Schlosser. 2009. Indirect emissions from biofuels: How important? Science 326:1397–99.
doi: 10.1126/science.1180251.
Metz, B., O. R. Davidson, P. R. Bosch, R. Dave, and L. A. Meyer. 2007. Climate change 2007:
Mitigation of climate change. Contribution of Working Group III to the Fourth
Assessment Report of the Intergovernmental Panel on Climate Change. New York:
Cambridge University Press.
Michaelis, L. 1997. Special issues in carbon / energy taxation: Marine bunker fuel charges. Paris:
OECD.
Miller, Shelie A., Amy E. Landis, and Thomas L. Theis. 2007. Environmental trade-offs of
biobased production. Environmental Science and Technology 41:5176–82.
Mishra, Gouri Shankar, and Sonia Yeh. 2011. Inclusion of marine bunker fuels in a national
LCFS scheme. A draft report prepared for the National Low Carbon Fuel Standard
Project. Institute of Transportation Studies, University of California, Davis.
Moslener, U., and B. Sturm. 2008. A European perspective on recent trends in US climate
policy. European Environment 18:257–75.
Msangi, Siwa, Miroslav Batka, Julie Witcover, and Sonia Yeh. 2012. Analysis of iLUC impacts
under an LCFS policy: Exploring impact pathways and mitigation. Energy Policy special
issue on Low Carbon Fuel Policy: submitted.
Murray, B. C. 2008. Leakage from an avoided deforestation compensation policy: Concepts,
empirical evidence, and corrective policy options. Durham, NC: Nicholas Institute for
Environmental Policy Solutions, Duke University.
Murray, Brian C., Richard G. Newell, and William A. Pizer. 2008. Balancing cost and emissions
certainty: An allowance reserve for cap-and-trade. National Bureau of Economic
Research, Working Paper No. 14258.
NEN (Dutch Standardization Institute). 2009. Dutch Technical Agreement (NTA 8080):
Sustainability criteria for biomass for energy purposes.
79
!
NRC (National Research Council of the National Academies). 2011. Renewable fuel standard:
Potential economic and environmental effects of U.S. biofuel policy. Washington DC:
National Academies Press.
NZ Emissions Trading Scheme Review Panel. 2011. Doing New Zealand’s fair share.
emissions trading scheme review 2011: final report. Wellington: Ministry for the
Environment.
Parker, Nathan C., Quinn Hart, Peter Tittman, M. Murphy, Richard Nelson, Ken Skog, E. Gray,
A. Schmidt, and Bryan Jenkins. 2010. Development of a biorefinery optimized biofuel
supply curve for the western United States. Biomass and Bioenergy 34:1597–1607.
Rajagopal, D., S. E. Sexton, D. Roland-Holst, and D. Zilberman. 2007. Challenge of biofuel:
Filling the tank without emptying the stomach? Environmental Research Letters
4:044004.
Rajagopal, Deepak, and Richard Plevin. 2012. Price effects in global fuel markets may reduce or
reverse the climate benefits of renewable fuel policies. Energy Policy special issue on
Low Carbon Fuel Policy: submitted.
RFA (Renewable Fuels Agency). 2009. Carbon and sustainability reporting within the
Renewable Transport Fuel Obligation, Technical guidance part one.
Robertson, G. Philip, Virginia H. Dale, Otto C. Doering, Steven P. Hamburg, Jerry M. Melillo,
Michele M. Wander, William J. Parton, Paul R. Adler, Jacob N. Barney, Richard M.
Cruse, Clifford S. Duke, Philip M. Fearnside, Ronald F. Follett, Holly K. Gibbs, Jose
Goldemberg, David J. Mladenoff, Dennis Ojima, Michael W. Palmer, Andrew Sharpley,
Linda Wallace, Kathleen C. Weathers, John A. Wiens, and Wallace W. Wilhelm. 2008.
Sustainable biofuels redux. Science 322:49–50. doi: 10.1126/science.1161525.
Rubin, Jonathan, and Paul N. Leiby. 2012. Tradable credits system design and cost savings for a
national low carbon fuel standard for road transport. Energy Policy special issue on Low
Carbon Fuel Policy, doi: 10.1016/j.enpol.2012.05.031.
Samaras, Constantine, and Kyle Meisterling. 2008. Life cycle assessment of greenhouse gas
emissions from plug-in hybrid vehicles: Implications for policy. Environmental Science
and Technology 42:3170–76. doi: 10.1021/es702178s.
Searchinger, Timothy, Ralph Heimlich, R. A. Houghton, Fengxia Dong, Amani Elobeid, Jacinto
Fabiosa, Simla Tokgoz, Dermot Hayes, and Tun-Hsiang Yu. 2008a. Use of U.S.
croplands for biofuels increases greenhouse gases through emissions from land use
change. Science 319:1238–40. doi: 10.1126/science.1151861.
———. 2008b. Use of U.S. croplands for biofuels increases greenhouse gases through emissions
from land use change. Supporting online material. Science 319:1238–40. doi:
10.1126/science.1151861.
Stern, N. 2007. The Economics of Climate Change: The Stern Review, Cambridge University
Press, Cambridge.
80
!
Tenenbaum, David J. 2008. Food vs. fuel: Diversion of crops could cause more hunger.
Environmental Health Perspectives 116:A254–A257.
Tilman, David, Robert Socolow, Jonathan A. Foley, Jason Hill, Eric Larson, Lee Lynd, Stephen
Pacala, John Reilly, Tim Searchinger, Chris Somerville, and Robert Williams. 2009.
Beneficial biofuels—The food, energy, and environment trilemma. Science 325:270–71.
doi: 10.1126/science.1177970.
Toulmin, C. 2009. Securing land and property rights in sub-Saharan Africa: The role of local
institutions. Land Use Policy 26:10–19. doi: 10.1016/j.landusepol.2008.07.006.
U.S. DOT (Department of Transportation). 2010. Transportation’s role in reducing U.S.
greenhouse gas emissions. Volume 1: Synthesis report.
U.S. EIA (Energy Information Administration). 2010a. Annual energy outlook 2010.
———. 2010b. Assumptions to the Annual energy outlook 2010: Oil and gas supply module.
———. 2011a. Annual Energy Outlook 2011.
———. 2011b. Canada energy data, statistics and analysis—Oil, gas, electricity, coal.
http://www.eia.doe.gov/emeu/cabs/Canada/Background.html.
———. 2012. Annual energy outlook 2012.
U.S. EPA (Environmental Protection Agency). 2010. Renewable Fuel Standard Program (RFS2)
regulatory impact analysis (EPA-420-R-10-006).
———. 2011a. 2017 and later model year light-duty vehicle greenhouse gas emissions and
corporate average fuel economy standards.
———. 2011b. Draft regulatory impact analysis: Proposed rulemaking for 2017–2025 light-duty
vehicle greenhouse gas emission standards and corporate average fuel economy
standards.
van Dam, Jinke, Martin Junginger, André Faaij, Ingmar Jurgens, Gustavo Best, and Uwe
Fritsche. 2008. Overview of recent developments in sustainable biomass certification.
Biomass and Bioenergy 32:749–80.
Wang, Michael, Hong Huo, and Salil Arora. 2011. Methods of dealing with co-products of
biofuels in life-cycle analysis and consequent results within the U.S. context. Energy
Policy 39:5726–36. doi: 10.1016/j.enpol.2010.03.052.
Welch, Heather L., Christopher T. Green, Richard A. Rebich, Jeannie R. B. Barlow, and
Matthew B. Hicks. 2010. Unintended consequences of biofuels production: The effects of
large-scale crop conversion on water quality and quantity. U.S. Geological Survey Open-
File Report 2010–1229.
Winrock International. 2009. Building capacity to monitor compliance with biofuel sustainability
standards.
Witcover, Julie, Sonia Yeh, and Daniel Sperling. 2012. Policy options to address global land use
change from biofuels. Energy Policy special issue on Low Carbon Fuel Policy:
submitted.
WWF (World Wildlife Fund). 2006. Sustainability standards for bioenergy. WWF Germany and
Öko Institut.
81
!
Yang, Christopher. 2012. Fuel electricity and plug-in electric vehicles in a low carbon fuel
standard. Energy Policy special issue on Low Carbon Fuel Policy. doi:
10.1016/j.bbr.2011.03.031.
Yeh, Sonia, and Daniel Sperling. 2010. Low carbon fuel standards: Implementation scenarios
and challenges. Energy Policy 38:6955–65. doi: 10.1016/j.enpol.2010.07.012.
Yeh, Sonia, Daniel A. Sumner, Stephen R. Kaffka, Joan M. Ogden, and Bryan M. Jenkins. 2009.
Implementing performance-based sustainability requirements for the Low Carbon Fuel
Standard—Key design elements and policy considerations. Institute of Transportation
Studies, University of California, Davis, Research Report UCD-ITS-RR-09-42.