• Open Access

A legal analysis of the effects of the Renewable Fuel Standard (RFS2) and Clean Air Act on the commercialization of biobutanol as a transportation fuel in the United States

Authors


Correspondence: Timothy A. Slating, tel. + 217 333 6178, fax + 217 333 0623, e-mail: slating2@law.illinois.edu

Abstract

Biobutanol is currently a hot topic within discussions about second-generation biofuels. Its advocates point to the fact that it possesses a higher energy content than traditional bioethanol and, most importantly, that it is compatible with existing fuel distribution infrastructure. While traditional biobutanol production processes have long since suffered from an inability to produce it in an economically viable manner, several recent technological advances have spurred interest from the private sector and several companies are now actively pursuing the commercialization of biobutanol as a transportation fuel. As such, a legal analysis of the regulatory frameworks affecting this commercialization is highly relevant. In this study, we detail and analyze the two most import regulatory frameworks affecting the successful commercialization of biobutanol as a transportation fuel in the United States. First, we provide a thorough description of the US Renewable Fuel Standard (RFS2) and analyze its impact on biobutanol commercialization efforts. Next, we address the US Clean Air Act's so-called ‘substantially similar’ prohibition and detail the three distinct regulatory paths it creates for biobutanol commercialization. Finally, we conclude by exploring ways in which these regulatory frameworks could be altered to mitigate unjustified regulatory burdens. While our study focuses on the commercialization of biobutanol, its regulatory descriptions and analysis are equally informative in regards to the commercialization of other alcohol-based biofuels.

Introduction

As the US Renewable Fuel Standard (RFS2) increasingly mandates the commercialization of biofuels other than corn ethanol (US Congress, 2007a,b), second-generation biofuels must be developed to satisfy these mandates. Moreover, consensus is beginning to converge on the notion that optimal biofuels should be ‘drop-in’, or in other words, compatible with existing fuel-related infrastructure (Regalbuto, 2009; Babcock et al., 2011). As biobutanol has several unique infrastructure compatibilities and the potential to satisfy the RFS2's increasing mandates for second-generation biofuels, it is currently receiving much attention from both the research community (Ezeji et al., 2004, 2007; Atsumi et al., 2008; Dellomonaco et al., 2011; Higashide et al., 2011) and the private sector biofuels industry (Butamax Advanced Biofuels, 2011; Cobalt Technologies, 2011; Gevo, 2011).

The history of biobutanol production and commercialization

The term ‘biobutanol’ is generically used to refer to any form of butanol that is derived from naturally occurring biological feedstocks. While observations of biobutanol fermentation span back to the work of Louis Pasteur, the first major breakthrough in its efficient and economical production came with the development of the acetone-butanol-ethanol (ABE) process in 1916 (Arnold, 2008). Using the bacterium Clostridium acetobutylicum to ferment naturally occurring starches and sugars, the ABE process traditionally yields butanol, acetone, and ethanol in a ratio of 6 : 3 : 1 by weight, respectively (Gibbs, 1983). While the ABE process attained commercial deployment in the early to mid-20th century as a means of producing butanol for the chemicals industry, it fell out of favor in the late 1950s when petroleum became a cheaper feedstock for butanol production (Arnold, 2008).

In the early 1990s, research interest in the production of biobutanol was rekindled, and the yields from the ABE process were enhanced through the discovery and implementation of the mutant bacterium Clostridium beijerinckii BA101 (Wu et al., 2008). Using this bacterium and modern gas stripping techniques, the butanol, acetone, and ethanol yield ratio from the ABE fermentation process has been substantially increased to 44 : 23 : 1 by weight, respectively (Ezeji et al., 2004). Despite these advances, the ability of biobutanol produced via the ABE process to successfully enter the US fuel market remains questionable. Not only do complex issues surround the potential markets for the acetone and ethanol that are simultaneously produced by the ABE process (Wu et al., 2008), but as we will see below, the commercialization of this form of biobutanol as a transportation fuel could be hindered by its classification for purposes of the RFS2. As such, commercialization efforts for biobutanol produced through the ABE process are primarily focusing on chemical markets as opposed to transportation fuel markets (Cobalt Technologies, 2011; TetraVitae Bioscience, 2011).

Arguably the most important breakthrough in the production of biobutanol came in 2008 when the Escherichia coli bacterium was genetically modified to produce only isobutanol, the highest-octane isomer of butanol (Atsumi et al., 2008). With the advent of genetically modified microorganisms capable of producing biobutanol without significant production of co-products (i.e., acetone and ethanol), the private sector biofuels industry has taken notice, and efforts are underway to attempt to commercialize biobutanol as a transportation fuel (Butamax Advanced Biofuels, 2011; Gevo, 2011). The business model of these companies involves acquiring existing ethanol production facilities, retrofitting them to produce biobutanol, and also licensing their technologies as retrofit packages to other owners of ethanol production facilities. While first-generation biofuel feedstocks such as corn starch and Brazilian sugarcane are currently being considered the feedstocks of choice for biobutanol production, these companies claim that their proprietary biocatalysts are capable of producing biobutanol from any fermentable sugar.

The advantages of biobutanol

The recent focus on biobutanol production is substantiated by its many reported advantages as a transportation fuel. First, its biggest advantage is its relatively high energy content when compared to first-generation biofuels. Based on its low heating value, biobutanol derived from the fermentation of cornstarch has an energy content of 99 837 BTU per gallon (Wu et al., 2008). This equates to 86% of the energy content of gasoline, which possesses 116 090 BTU per gallon (Wu et al., 2008). By way of comparison, corn ethanol only possesses 76 330 BTU per gallon, which equates to an energy content roughly 66% that of gasoline (Wu et al., 2008). As such, if we put aside the effects of cost, simply account for the energy content of fuel (i.e., assume that distance traveled perfectly correlates with BTUs used regardless of fuel type), and compare a 16% biobutanol/84% gasoline blend (Bu16) with the 10% ethanol/90% gasoline blend (E10) widely available in the United States today, the widespread use of Bu16 would not only result in slightly greater (<2%) fuel economy for consumers but also would displace over twice as much hydrocarbon gasoline.

The second major advantage of biobutanol is that it reportedly has compatibilities with each stage of the existing ‘field-to-tank’ macro-level infrastructure needed to produce biofuels, bring them to market, and consume them. In regards to feedstock-related infrastructure, biobutanol can not only be produced from current first-generation biofuel feedstocks such as corn starch (Ezeji et al., 2007) and cane sugar (Butamax Advanced Biofuels, 2011), but its production from a cellulosic feedstock has also been demonstrated (Higashide et al., 2011). Additionally, the private companies currently working to commercialize biobutanol as a transportation fuel claim that it is compatible with existing biofuel production infrastructure in that their technologies allow existing corn ethanol plants to be minimally retrofitted to produce biobutanol in addition to ethanol (Butamax Advanced Biofuels, 2011; Gevo, 2011). Most importantly, it has been reported that biobutanol is compatible with current fuel distribution infrastructure since it does not produce the microbial-induced corrosion that is associated with ethanol (Wu et al., 2008). As such, it can potentially be blended with gasoline at the refinery, transported through existing fuel pipelines, and distributed at existing retail fueling stations. Furthermore, it has been reported that fuel blends containing as high as 100% biobutanol are compatible with existing spark-ignition internal combustion engines (Ramey, 2007; Szwaja & Naber, 2010).

While these reported advantages render biobutanol a prime candidate to act as a biofuel alternative to gasoline, two distinct regulatory frameworks will heavily impact its successful commercialization in the United States. First, the RFS2 will have a tremendous impact as it effectively mandates a captive market for biofuels (US Congress, 2007a,b). Second, the US Clean Air Act's (CAA) ‘substantially similar’ regulatory framework for new fuels and fuel additives effectively governs the lawful blending limits for biobutanol (US Congress, 1972; US Environmental Protection Agency, 2008). As such, this paper will first analyze the RFS2's treatment of biobutanol (both currently and prospectively) and then move on to analyze the three distinct options for the commercialization of biobutanol under the CAA, which each resulting in differing lawful blending limits. Finally, we will conclude by arguing that specific provisions of the CAA could create unjustified regulatory hurdles for the commercialization of biobutanol and suggesting unique regulatory reforms that could be implemented to rectify this situation.

Analysis of the US Renewable Fuel Standard (RFS2)

Overview of the RFS2

A basic idea underlying the enactment of any renewable fuel standard is that by legally mandating the commercialization of a specified volume of renewable fuels, a captive market will be created and therefore incentivize their development, further commercialization, and use. With this goal in mind, the US Congress enacted the first US Renewable Fuel Standard (RFS1) as a provision within the Energy Policy Act of 2005 (US Congress, 2005a,b). The RFS1 mandated that the US fuel supply contains 4.0 billion gallons (bg) of ‘renewable fuel’ in 2006, 4.7 bg in 2007, 5.4 bg in 2008, 6.1 bg in 2009, 6.8 bg in 2010, 7.4 bg in 2011, and 7.5 bg in 2012. For purposes of the RFS1, Congress broadly defined ‘renewable fuel’ to include all ‘motor vehicle fuel that[: (1)] is produced from grain, starch, oilseeds, vegetable, animal, or fish materials including fats, greases, and oils, sugarcane, sugar beets, sugar components, tobacco, potatoes, or other biomass; or [(2)] is natural gas produced from a biogas source, including a landfill, sewage waste treatment plant, feedlot, or other place where decaying organic material is found’. While specific provisions within the RFS1 attempted to incentivize the commercialization of biofuels other than corn ethanol, they proved to have little effect as a result of corn ethanol satisfying the definition of ‘renewable fuel’ and its existing consumption patterns tracking very closely with the RFS1's volumetric mandates (US Energy Information Administration, 2011). Essentially, the existing market for corn ethanol was already accomplishing what Congress had mandated.

As a result of the RFS1's inability to adequately incentivize the development of second-generation biofuels, Congress made substantial changes to the standard when it enacted the Energy Independence and Security Act of 2007 (US Congress, 2007a) and created what is now referred to as the RFS2 (US Congress, 2007b). Specifically, the RFS2 creates a system of nested mandates for four different categories of biofuels that are defined by the feedstocks they are produced from and their ability to reduce lifecycle greenhouse gas (GHG) emissions from a baseline defined as ‘the average lifecycle [GHG] emissions ··· for gasoline or diesel (whichever is being replaced by the renewable fuel) sold or distributed as a transportation fuel in 2005’ (Table 1). The broadest category is termed simply ‘renewable fuel’ and includes all transportation, heating, and jet fuels that are ‘produced from renewable biomass’ and have ‘lifecycle [GHG] emissions that are at least 20 percent less than’ the 2005 baseline. The second broadest category is called ‘advanced biofuel’ and is defined as ‘renewable fuel, other than ethanol derived from cornstarch, that has lifecycle [GHG] emissions that are at least 50 percent less than’ the 2005 baseline. The next category, “cellulosic biofuel”, is defined as “renewable fuel derived from any cellulose, hemi-cellulose, or lignin that has lifecycle [GHG] emissions that are at least 60 percent less than“ the 2005 baseline. Finally, the last category is called “biomass-based diesel” and is defined as ‘renewable fuel [that is either biodiesel or nonester renewable diesel and] has lifecycle [GHG] emissions that are at least 50 percent less than’ the 2005 baseline. As these definitions are nested, a given biofuel that satisfies the definition of a cellulosic biofuel or a biomass-based diesel also satisfies the definition for an advanced biofuel. Likewise, a fuel that satisfies the definition of an advanced biofuel also satisfies the definition for a renewable fuel.

Table 1. The Renewable Fuel Standard's (RFS2) fuel categories
CategoryGreenhouse gas reduction threshold (%)Permissible feedstockExamples of established fuel pathways
  1. a

    Percentage reductions are determined by comparing a given fuel pathway's greenhouse gas life-cycle analysis to that of a baseline, which is based on the greenhouse gas life-cycle analysis for either gasoline or diesel (whichever is being replaced by the renewable fuel) produced and consumed in 2005. Renewable fuels produced at facilities that commenced construction on or before December 19, 2007 are excluded from the requirement that they meet the 20% greenhouse gas reduction threshold (US Congress, 2007a,b).

  2. b

    Renewable biomass is defined as ‘(i) planted crops and crop residue harvested from agricultural land cleared or cultivated at any time prior to the enactment of this sentence that is either actively managed or fallow, and nonforested. (ii) Planted trees and tree residue from actively managed tree plantations on nonfederal land cleared at any time prior to enactment of this sentence, including land belonging to an Indian tribe or an Indian individual, that is held in trust by the United States or subject to a restriction against alienation imposed by the United States. (iii) Animal waste material and animal byproducts. (iv) Slash and precommercial thinnings that are from nonfederal forestlands, including forestlands belonging to an Indian tribe or an Indian individual, that are held in trust by the United States or subject to a restriction against alienation imposed by the United States, but not forests or forestlands that are ecological communities with a global or State ranking of critically imperiled, imperiled, or rare pursuant to a State Natural Heritage Program, old growth forest, or late successional forest. (v) Biomass obtained from the immediate vicinity of buildings and other areas regularly occupied by people, or of public infrastructure, at risk from wildfire. (vi) Algae. (vii) Separated yard waste or food waste, including recycled cooking and trap grease’. (US Congress, 2007a,b).

Renewable fuel20aRenewable biomassbEthanol derived from corn starch; butanol derived from corn starch
Advanced biofuel50aRenewable biomassb, with the exception of corn starch if the fuel is ethanolEthanol derived from sugarcane; ethanol derived from the noncellulosic portions of separated food waste
Biomass-based diesel50aRenewable biomass bBiodiesel derived from soy bean oil; biodiesel derived from canola oil
Cellulosic biofuel60aAny cellulose, hemi-cellulose, or lignin that is derived from renewable biomassbEthanol derived from cellulosic biomass; naphtha derived from cellulosic biomass

In order to determine which category a given biofuel qualifies for, the US Environmental Protection Agency (EPA) undertakes a GHG life-cycle analysis (LCA) for the biofuel. This analysis must account for the ‘quantity of [GHG] emissions (including direct emissions and significant indirect emissions such as significant emissions from land use changes) … related to the full fuel lifecycle, including all stages of fuel and feedstock production and distribution, from feedstock generation to extraction through the distribution and delivery and use of the finished fuel to the ultimate consumer’. The result of this GHG LCA is compared to the 2005 baseline for either gasoline or diesel in order to determine the biofuel's GHG reduction potential. Based on this and the feedstock the biofuel is produced from, the EPA determines which renewable fuel category the biofuel qualifies for and designates a fuel pathway for the biofuel. This fuel pathway sets out required production processes that must be complied with by manufacturers wishing to produce a biofuel that satisfies the fuel pathway's requirements and results in the same renewable fuel category designation. As of the writing of this article, the EPA has modeled and designated six fuel pathways that result in renewable fuels, four that result in advanced biofuels, two that result in cellulosic biofuels, two that result in biomass-based diesel, and one that qualifies as both a cellulosic biofuel and a biomass-based diesel. While the RFS2 provides a mechanism for biofuel producers to petition the EPA to designate new fuel pathways for new biofuel production processes, the EPA has yet to issue formal guidance on how to do so and, as such, one biofuels industry organization has complained that ‘the process for petitioning the [EPA] is somewhat unclear, as is the timeline and criteria for petition approval’ (Renewable Fuels Association, 2011a).

For each calendar year through 2022, the RFS2 mandates specific volumes of each category of biofuels (with the exception of biomass-based diesel, which only has specific volumes detailed through 2012 and will have future mandates set at a later date) that must be introduced into US commerce by fuel refiners, blenders, and importers (Fig. 1). To use 2022 as an example, 36 bg of renewable fuels must be introduced into US commerce and 21 bg of this must be advanced biofuels. Of the 21 bg advanced biofuel mandate, 16 bg must be cellulosic biofuel and an as-yet to be determined volume must be biomass-based diesel. While many commentators assert that a 15 bg mandate exists for corn ethanol, they are actually referring to the difference between the renewable fuel mandate and the advanced biofuel mandate. While it is highly likely that this 15 bg difference will be satisfied by the commercialization of corn ethanol, the entire 36 bg renewable fuel mandate could, in theory, be satisfied by the use of advanced biofuels. For example, if 5 bg of biomass-based diesel end up being mandated for 2022, the RFS2's mandates could hypothetically be satisfied by the use of: (1) 5 bg of biomass-based diesel and 31 bg of cellulosic biofuels; (2) 20 bg of biomass-based diesel and 16 bg of cellulosic biofules; (3) 5 bg of biomass-based diesel, 16 bg of cellulosic biofuels, and 15 bg of advanced biofuel; or (4) 5 bg of biomass-based diesel, 16 bg of cellulosic biofuels, and 15 bg of renewable fuel.

Figure 1.

The Renewable Fuel Standard's (RFS2) yearly volumetric mandates for its four categories of biofuels. *While the RFS2 only specifies a yearly mandate for biomass-based diesel through 2012 and requires future yearly mandates to be determined at a later date, the EPA has indicated that these future mandates will be at least 1 billion gallon yr−1.

The RFS2's current treatment of biobutanol

Only one RFS2 fuel pathway has currently been modeled and designated for biobutanol (US Congress, 2007b). In order for a form of biobutanol to qualify for this pathway, its production must involve the fermentation of corn starch via a dry mill process that utilizes either natural gas, biogas, or biomass for process energy. Based on the GHG LCA modeled for this pathway, which relies on data from the traditional ABE fermentation process as a result of publicly released data not being available for newer production processes, the EPA has found that compliant biobutanol meets the 20% GHG reduction threshold necessary to be classified as a renewable fuel. More specifically, the EPA concluded that biobutanol produced in accordance with this pathway results in a 31% GHG emissions reduction from the 2005 baseline for gasoline and ‘based on the uncertainty in the land use change assumptions’, this 31% reduction is actually the midpoint of a range with the 95% confidence interval around this midpoint ranging from 20% to 40% reduction (US Environmental Protection Agency, 2010b).

Additionally, the EPA has assigned biobutanol an Equivalence Value of 1.3 for purposes of complying with the RFS2's volumetric mandates (US Environmental Protection Agency, 2010b). Because different biofuels have different energy contents, their use displaces different amounts of petroleum and as such, the EPA assigns them Equivalence Values based on their energy content in order to ensure that the RFS2's mandates result in roughly the same amount of petroleum displacement regardless of which biofuels are used to satisfy its mandates. Ethanol is used as the baseline with an Equivalence Value of 1.0 and since biobutanol has a roughly 30% greater energy content than ethanol, biobutanol is assigned an Equivalence Value of 1.3. Thus, if we use 2022 as an example, the RFS2's 16 bg cellulosic biofuel mandate could be satisfied with the use of: (1) 16 bg of cellulosic ethanol; (2) 12.3 bg of cellulosic biobutanol; or (3) some other combination of cellulosic biofuels totaling 16 bg based on their respective Equivalence Values. Even if biobutanol manufacturers attempted to demand a higher per-gallon price from RFS2 obligated parties (e.g., fuel refiners, blenders, and importers) than ethanol producers currently do, the fact that these obligated parties would need to purchase, blend, and distribute fewer physical gallons of biobutanol in order to satisfy their RFS2 mandates could provide a significant value proposition.

The RFS2's prospective treatment of biobutanol

Since the US production of corm ethanol is on pace to satisfy the 15 bg portion of the RFS2's mandates for 2022 that can only be satisfied by ‘renewable fuels’ (Renewable Fuels Association, 2011b), the RFS2 will likely do little to incentivize the production and use of biobutanol so long as it remains only classified as a renewable fuel. On the other hand, if new biobutanol production processes emerge that result in the establishment of new RFS2 fuel pathways that are designated as resulting in either advanced biofuels or cellulosic biofuels, then the RFS2's captive markets for these categories of biofuels would likely incentivize the production of biobutanol. While the RFS2's captive market for fuels classified only as renewable fuels is effectively saturated, those for advanced biofuels and cellulosic biofuels are anything but.

The first form of biobutanol that could potentially hit the US fuel market is a corn-based isobutanol that is currently projected to be commercialized in the first half of 2012 (Gevo, 2011). While this form of biobutanol could never qualify as a cellulosic biofuel (since it is not derived from a cellulosic feedstock), it could qualify as an advanced biofuel if its production process could be shown to result in the requisite 50% GHG reduction. Although this seems like a formidable task based on the 31% GHG reduction that the EPA has determined for the existing biobutanol pathway, advanced biofuel classification seems more attainable if you consider the fact that the EPA's LCA actually showed between 20% and 40% GHG reduction based on uncertainties in its indirect land use change values (US Environmental Protection Agency, 2010b). Bearing on this issue is the fact that despite predictions about increased biofuel production causing significant indirect land use change (Searchinger et al., 2000; Tyner et al., 2010), more recent research is beginning to show that US biofuels production has resulted in little (if any) indirect land use change (Kim & Dale, 2011). Moreover, the current RFS2 pathway for biobutanol is based on the outdated ABE fermentation process and the LCAs for new processes will likely benefit from the fact they produce higher biobutanol yields and do not require energy intensive processes to separate the acetone and ethanol co-products. Additionally, the energy efficiency of producing biofuels from corn is beginning to be enhanced through the implementation of advanced production technologies such as corn oil extraction and raw starch hydrolysis. As current business models involve licensing biobutanol production technologies as retrofit packages to existing corn ethanol facilities, it is not difficult to imagine producers looking to implement these emerging types of advanced production technologies at the same time they undergo the requisite retrofit process to produce biobutanol. Not surprisingly, as of the writing of this paper, one private company is in the process of working with the EPA to establish an advanced biofuel pathway for its corn-based isobutanol (Voegele, 2011).

In addition to commercialization strategies that involve licensing biobutanol production technologies to existing corn ethanol facilities, another private company is also looking to produce biobutanol from Brazilian sugarcane (Butamax Advanced Biofuels, 2011). Ethanol derived from the fermentation of Brazilian sugarcane is currently classified as an advanced biofuel and was found by the EPA to result in a 61% GHG reduction from the 2005 baseline for gasoline (US Environmental Protection Agency, 2010b). While the EPA has yet to model and designate a fuel pathway for sugarcane biobutanol, when it modeled and designated the RFS2 corn biobutanol pathway, the EPA merely imported its feedstock values for corn ethanol and made adjustments to account for biobutanol production processes and its higher energy content. If the EPA follows this same methodology in evaluating the requisite fuel pathway application that would need to be submitted for biobutanol produced from Brazilian sugarcane, it is highly likely that the new pathway would be designated as resulting in an advanced biofuel.

Finally, the production of biobutanol from cellulosic feedstocks has also been reported. While one public company has announced the ability of its microbes to produce isobutanol from fermentable sugars derived from cellulose (Gevo, 2011), researchers at UCLA have demonstrated the synthesis of isobutanol directly from cellulose (Higashide et al., 2011). If these production processes develop to commercial scale, the resulting cellulosic isobutanol will potentially qualify as a cellulosic biofuel under the RFS2 so long as it satisfies the requisite 60% GHG reduction threshold.

Analysis of commercialization paths under the Clean Air Act (CAA)

Overview

As biobutanol will likely enter the transportation fuel market as an additive to be blended with gasoline, the US Clean Air Act's (CAA) ‘substantially similar’ regulatory framework, which regulates the amount of biobutanol that can lawfully be blended in fuel, makes up the second most important regulatory framework affecting biobutanol's commercialization (US Congress, 1972; US Environmental Protection Agency, 2008). As opposed to the RFS2, which could operate to incentivize the increased use of biobutanol as a transportation fuel, this framework could have the affect of limiting the US fuel market size for biobutanol. To use a vastly over-simplified example, if the yearly US consumption of gasoline-based fuel blends were 100 bg and biobutanol could only be lawfully blended in volumes up to 10% under the CAA's regulatory framework, then this framework would effectively work to limit the US fuel market for biobutanol to 10 bg per year. In the context of ethanol, the maximum volume that can lawfully be blended nationwide pursuant to the CAA is commonly referred to as the ‘blend wall’ and as we will see, the contours of this wall are much less concrete in regards to biobutanol.

The CAA regulates the amount of biobutanol (or any other fuel additive) that can lawfully be blended with gasoline through the operation of its so-called ‘substantially similar’ prohibition (US Congress, 1972). Specifically, the CAA prohibits any fuel or fuel additive manufacturer from commercializing any fuel or fuel additive that is not ‘substantially similar’ to any fuel or fuel additive used by the EPA in its testing process for the certification of vehicle and engine emissions. If a manufacturer desires to commercialize a fuel or fuel additive that is not ‘substantially similar’ to certification fuel, then it must either receive a new fuel waiver from the EPA or try to commercialize its fuel or fuel additive pursuant to a previously granted fuel waiver. As such, this framework produces three distinct regulatory paths for the commercialization of biobutanol as a transportation fuel additive.

Commercialization pursuant to the Substantially Similar Rule

While the CAA explicitly prohibits the commercialization of fuels and fuel additives that are not ‘substantially similar’ to the fuel and fuel additives authorized by the EPA to be used in the certification of vehicle and engine emissions, the statute does not define what constitutes a ‘substantially similar’ fuel or fuel additive (US Congress, 1972). Instead, Congress opted to leave the task to the discretion of the EPA, which issues its so-called Substantially Similar Rule to set out the specific requirements that fuels must meet in order to be considered ‘substantially similar’ and therefore fall outside the CAA's prohibition against commercialization (US Environmental Protection Agency, 2008). Likewise, a fuel additive is considered to be ‘substantially similar’ if the finished fuel that results from it being blended in its recommended concentration complies with the requirements of the Substantially Similar Rule. For example, if a finished fuel can only comply with the requirements of the Substantially Similar Rule so long as it contains a maximum concentration of 10% of a given fuel additive, then the CAA prohibits manufacturers of that fuel additive from commercializing it for use in a concentration greater than 10%.

The current iteration of the Substantially Similar Rule allows biobutanol to be lawfully blended with gasoline in volumetric concentrations of roughly 11.5–12.5% (i.e., Bu11.5–Bu12.5), depending on the density of the finished fuel (US Environmental Protection Agency, 2008). Specifically, the Rule provides that in order for a fuel containing alcohol (excluding methanol) to be considered ‘substantially similar’ to certification fuel, it ‘must contain no more than 2.7 percent oxygen by weight’. While the oxygen content of biobutanol is known to be 21.6% by weight and gasoline traditionally contains no oxygen, the density of gasoline fluctuates and, as such, it is impossible to calculate a precise volumetric blending limit for biobutanol that results from the Substantially Similar Rule's limit of 2.7% oxygen by weight. Nonetheless, conservative estimates report that the Substantially Similar Rule permits the lawful commercialization of Bu11.5 (US Department of Energy, 2011), while biobutanol manufacturers prefer to rely on the liberal estimate of the Rule permitting Bu12.5 (Gevo, 2011).

Finally, it is important to note that commercializing a biobutanol blend pursuant to the limits of the Substantially Similar Rule is currently the regulatory path of least resistance (Fig. 2). So long as a biobutanol manufacturer has annual sales that total less than $50 million, all it needs to do prior to lawful commercialization is submit a fuel additive registration form to the EPA that, in addition to other basic information, specifies a recommended volumetric blending concentration for its biobutanol that results in a finished fuel compliant with the Substantially Similar Rule. The EPA will only reject the application if the recommended concentration specified would clearly result in a noncompliant fuel and so long as it is plausible that the recommended concentration could result in a complaint finished fuel, the EPA will register the additive knowing that the blenders of finished fuels remain bound by the CAA's substantially similar prohibition. Based on this reasoning, the EPA has registered two forms of biobutanol for use in a recommended volumetric blending concentration of 12.5%.

Figure 2.

Commercialization path for biobutanol fuel blend pursuant to the current Substantially Similar Rule.

Commercialization pursuant to an existing fuel waiver

The second regulatory path for the commercialization of biobutanol as a transportation fuel additive involves attempting to register it for use in a concentration that complies with a previously granted fuel waiver (Fig. 3). Once the EPA grants a waiver of the substantially similar prohibition for a particular fuel blend, that fuel waiver is applicable to all similarly situated manufacturers and the EPA is without authority to explicitly revoke the waiver. As such, one biobutanol manufacturer has indicated an intention to commercialize higher-level blends in accordance with a series of fuel waivers that were previously granted by the EPA (Butamax Advanced Biofuels, 2011).

Figure 3.

Possible commercialization path for a Bu16–Bu17 blend pursuant to either the DuPont Waiver or the Octamix Waiver.

In the 1980s, the EPA granted the so-called ‘DuPont Waiver’ and the ‘Octamix Waiver’. While both of these waivers were originally sought for the commercialization of methanol/gasoline fuel blends, the explicit conditions that the EPA opted to include in granting theses waivers arguably allows biobutanol to be blended in concentrations of roughly 16–17% by volume (i.e., Bu16–Bu17). Specifically, the DuPont Waiver allows for the lawful commercialization of finished fuel blends that ‘consist[] of a maximum of 5.0 volume percent methanol, a minimum of 2.5 volume percent co-solvent (ethanol, propanols, GTBA, and/or butanols) in unleaded gasoline[,]’ and possess a ‘maximum concentration of up to 3.7 weight percent oxygen in the final fuel’ (US Environmental Protection Agency, 1985). Likewise, the Octamix Waiver also allows for the lawful commercialization of fuel blends consisting of a ‘maximum of 5 percent by volume methanol’ and a ‘minimum of 2.5 percent by volume cosolvent’ (including ‘butanols’) so long as ‘a maximum concentration of up to 3.7 percent by weight oxygen in the final fuel is observed’ (US Environmental Protection Agency, 1988). As both of these waivers merely specify the ‘maximum’ amount of methanol allowed and the ‘minimum’ amount of butanol allowed, it is arguable that they permit the lawful commercialization of fuel blends containing 0% methanol and as high a percentage of biobutanol as can be achieved within the confines of the finished fuel containing no more than 3.7% oxygen by weight. Again, as it is impossible to calculate a precise volumetric blending limit for biobutanol that results in a finished fuel containing on oxygen weight of 3.7% without knowing the density of the finished fuel, it is commonly estimated that this limit would permit Bu16–Bu17 blends.

Regardless of the language in these two waivers seeming to permit the lawful commercialization of Bu16–Bu17 blends, it is highly uncertain how the EPA would respond to an attempt to register these blends (Fig. 3). The simple fact is that when the EPA originally granted these waivers, it never conducted vehicle and engine emissions tests using Bu16–Bu17 blends and, as such, it is highly unlikely that the EPA would allow their commercialization pursuant to these waivers. While a biobutanol manufacturer attempting to pursue this path could always legally challenge the EPA's registration decision in court, it is highly uncertain how a court would rule because there is no precedent for a fuel additive manufacturer attempting to commercialize a completely new form of biofuel in accordance with a previously granted fuel waiver that was originally directed at a different type of additive. Finally, even if the EPA opted to allow the registration of a Bu16–Bu17 blend pursuant to either the DuPont Waiver or the Octamix Waiver, the biobutanol manufacturer would also face the uncertainty associated with the very high probability that some third-party will try to legally challenge the EPA's decision.

Commercialization pursuant to a new fuel waiver

The third and most burdensome regulatory path for the commercialization of biobutanol as a transportation fuel additive involves seeking a new fuel waiver from the EPA (Fig. 4). Specifically, the CAA permits the EPA to waive the substantially similar prohibition, through the grant of a fuel waiver, if a ‘manufacturer of any fuel or fuel additive’ submits an application that establishes ‘that such fuel or fuel additive or a specified concentration thereof, and the emission products of such fuel or fuel additive or specified concentration thereof, will not cause or contribute to a failure of any emission control device or system (over the useful life of the motor vehicle, motor vehicle engine, nonroad engine or nonroad vehicle in which such device or system is used) to achieve compliance by the vehicle or engine with the emission standards with respect to which it has been certified’. While the grant of a new fuel waiver would definitely be required for the commercialization of biobutanol blends greater than Bu16–Bu17, it would also be required for blends greater than Bu11.5–Bu12.5 if the EPA refuses to allow higher-level blends pursuant to either the DuPont Waiver or the Octamix Waiver.

Figure 4.

Commercialization path for a higher-level biobutanol blend pursuant to a new fuel waiver.

If we unpack the CAA's fuel waiver language, some important and possibly misleading ideas emerge. First of all, the language makes it perfectly clear that the fuel waiver applicant bears the burden of establishing that the waiver fuel will not negatively affect vehicle and engine emissions control systems. Nonetheless, when the EPA conditionally granted its most recent fuel waiver, which involves a fuel blend containing 15% ethanol and 85% gasoline (i.e., E15), it relied almost exclusively on data supplied by the US Department of Energy as opposed to the data and analysis submitted by the waiver applicants (US Environmental Protection Agency, 2011). Second, while the language appears to indicate that a waiver will only be granted if the waiver fuel has no adverse affects on the emissions control systems in a variety of vehicle and engine types, when the EPA conditionally granted the E15 waiver, it segregated these types of vehicles and engines and granted the fuel waiver for use in some while prohibiting the use of E15 in others. Specifically, the E15 waiver permits the use of E15 in model year 2001 and newer on-road vehicles and prohibits its use in all other vehicles and engines. Finally, while the language seems to require that the applicant prove a negative proposition (i.e., that the waiver fuel ‘will not cause or contribute to a failure of any emission control device or system’), the EPA has developed a workable process for applicants to establish their burden of proof. Applicants are allowed to use ‘reliable statistical sampling and fleet testing protocols’ to establish their burden or they can develop a ‘reasonable theory regarding emissions effects ··· , based on good engineering judgment’ (US Environmental Protection Agency, 2010a). Nonetheless, if the applicant relies on the latter approach, it must conduct enough testing to produce adequate data for a statistical analysis that confirms the theory.

When the applicant submits its fuel waiver application, it must include data and analysis that addresses four distinct areas. First, data must be submitted that demonstrates the waiver fuel's immediate and long-term effects on exhaust emissions. If the fuel is predicted to only have an immediate effect, then the applicant is permitted to conduct ‘back-to-back’ emissions testing, which ‘involves measuring, sequentially, the emissions from a particular vehicle, first operated on a base fuel not containing the waiver request fuel or fuel additive and then on a base fuel containing the additive or the waiver request fuel’ (US Environmental Protection Agency, 1979). If the fuel is not predicted to only have immediate emissions effects, then the applicant must submit data that covers the entire useful lives (120 000 miles) of the engines in the test fleet. Second, the applicant must also submit data that demonstrates the immediate and long-term effects that the waiver fuel has on evaporative emissions and is again allowed to rely on back-to-back testing if only an immediate affect is predicted. Third, data must be submitted that demonstrates the compatibility of the waiver fuel with the material components of vehicles and engines because exhaust and evaporative emissions can be substantially increased as a result of material compatibility issues. Finally, the applicant must also submit data that addresses how the waiver fuel might affect the driveability and operability of vehicles and engines. The idea here is that if a fuel negatively affects driveability or operability, there is a likelihood that drivers or operators might attempt to modify their vehicles or engines into a configuration that differs from the one certified by the EPA.

Taken as a whole, these requirements render the fuel waiver process very costly, burdensome, and highly uncertain (Fig. 4). Once an applicant compiles this data and submits its application to the EPA, the application is released to the general public for comment. Additionally, the CAA provides that the EPA has only 270 days from receipt of the application to either grant or deny the fuel waiver. Nonetheless, the fuel waiver process, which used to not involve mandatory public comment and was subject to a provision that treated applications as granted if the EPA failed to act within 180 days, has turned into a rather lengthy ordeal. For example, while the original E10 waiver from 1979 was granted by operation of law with very little public comment after the EPA failed to act within 180 days of receiving the waiver application, the EPA's conditional grant of the recent E15 waiver was not finalized until nearly 2 years after the EPA received the original application and involved approximately 78 000 public comments officially submitted to the EPA. While the EPA obviously exceeded its 270 day time limitation, the waiver applicants consented to multiple extensions.

Discussion

To those unfamiliar with the US regulatory system, it might seem odd that a single federal agency could oversee one regulatory framework that incentivizes the development of a given industry while, at the same time, it also oversees another regulatory framework that potentially hinders that same industry. It is arguable that the EPA-administered regulatory frameworks affecting the successful commercialization of biobutanol as a transportation fuel do just this. On the one hand, the RFS2 incentivizes the commercialization of biobutanol, while on the other, the CAA's ‘substantially similar’ framework regulates the amount of biobutanol that can be blended with gasoline and therefore limits its potential market.

As detailed above, a biobutanol manufacturer is forced to choose between three different regulatory paths when commercializing its product as an additive to be blended with gasoline. First, it can commercialize its biobutanol for use in a Bu11.5–Bu12.5 blend pursuant to the current Substantially Similar Rule (Fig. 2). Second, it could attempt to take the very uncertain path of attempting to commercialize a Bu16–Bu17 blend pursuant to either the DuPont Waiver or the Octamix Waiver (Fig. 3). Finally, it could attempt to take the extremely costly and even more uncertain path of seeking a new fuel waiver from the EPA to allow for the lawful commercialization of higher-level biobutanol blends (Fig. 4).

Before articulating a few ways in which this regulatory framework could be reformed to better facilitate the commercialization of biobutanol, it is important to set out two theoretical principles that help justify our suggestions. First, it is well established that when regulated activities result in the creation of social value, then regulatory reforms can be justified in order to mitigate unreasonable regulatory burdens that impede our ability to capture that social value (Slating & Kesan, 2011). Here, the idea is that since the current CAA ‘substantially similar’ regulatory framework creates substantial burdens for the commercialization of higher-level biobutanol blends and increased use of biobutanol as a transportation fuel would produce social value through its ability to reduce GHG emissions and enhance US energy security, then regulatory reforms can be justified. Second, it is a firmly established notion among law and economics scholars that the compliance-related burdens associated with a given regulatory framework should not outweigh the societal harms that framework is intended to mitigate (Bardach & Kagan, 1982). As it applies to biobutanol, the burdens created by the CAA's ‘substantially similar’ regulatory framework outweigh the harms it is intended to mitigate since biobutanol fuel blends have been shown to actually decrease regulated emissions (Wallner et al., 2009). Specifically, research has shown that when a Bu10 blend is compared to pure gasoline in an engine that has been calibrated for gasoline use, the two blends produce similar amounts of carbon monoxide emissions and the Bu10 blend produces lower oxides of nitrogen emissions (Wallner et al., 2009). As such, regulatory reforms are justified in order to mitigate these regulatory burdens and allow us to more efficiently capture the social value that could be created by the commercialization and use of higher-level biobutanol fuel blends.

Our most basic suggestions for reforming the CAA's regulatory framework involve actions that the EPA could take in the absence of Congressional action or formal notice and comment rulemaking. First of all, the EPA could simply issue a new Substantially Similar Rule that sets out a precise volumetric blending limit for biobutanol. This would provide a degree of certainty that is lacking under the current Rule, which simply sets an oxygen weight percentage limitation that cannot be equated to a precise volumetric blending limit. As the current Substantially Similar Rule does this for methanol, our suggestion is far from radical. Furthermore, we suggest that the EPA should set this volumetric blending limit for biobutanol at 17% or, in the alternative, increase the allowable oxygen weight percentage to 3.7 for all fuels. E10 fuel blends, which contain roughly 3.7% oxygen by weight, have been lawfully commercialized in the United States for over 30 years with improved engine emissions compared to pure gasoline. As such, it is not surprising that a 3.7% oxygen weight limitation is the most common condition contained in alcohol-based fuel waivers granted by the EPA. In order to efficiently capture the social value of emerging alcohol-based biofuels such as biobutanol, there is no reason that the EPA should not update its outdated Substantially Similar Rule to allow all fuels to lawfully contain up to 3.7% oxygen by weight.

Our final suggestion for regulatory reform involves a Congressional change to the CAA to provide a ‘fast-track’ review process for new fuel waivers involving biofuels that are capable of satisfying the RFS2's mandates. We propose that once the EPA has designated a fuel pathway for a given fuel under the RFS2, then manufacturers of that fuel should be allowed to seek higher blending limits for that fuel by submitting a new fuel waiver application that is not subject to formal public notice and comment and must be acted upon by the EPA within 180 days. Furthermore, we suggest that these fuel waiver applicants be allowed to designate which types of vehicles and engines they are seeking a fuel waiver for and not have to submit data addressing the fuel's effects on other types of vehicles and engines. If the EPA is willing to conditionally grant a fuel waiver for use in some vehicles and engines and not others, as it did with the recent E15 waiver, then why should applicants not be allowed to designate which vehicles and engines they are seeking a waiver for and avoid the costs associated with gathering data that addresses the waiver fuel's effects on other types of vehicles and engines?

As the biobutanol industry remains in its infancy, these forward-looking regulatory reforms could have profound impacts. Increasing lawful blending limits and mitigating the burdens associated with seeking new fuel waivers provides the type of certainty necessary to incentivize continued investment in this promising fuel. In order for society to fully capture the social value of biobutanol and other emerging biofuels, the burdens associated with the regulatory frameworks effecting their successful commercialization must be fully justified. Where these regulatory burdens are not justified, we must actively seek out regulatory reforms in order to facilitate the commercialization of socially beneficial biofuels.

Acknowledgements

This research has been funded by the Energy Biosciences Institute at the University of Illinois at Urbana-Champaign.

Ancillary