------------------------------------------------------------------------------- on 02/26/93 [L-S document 449994, 58 FR 11722, 4406 lines] ENVIRONMENTAL PROTECTION AGENCY 40 CFR Part 80 [AMS-FRL-4556-7] RIN 2060-AD27 Regulation of Fuels and Fuel Additives: Standards for Reformulated Gasoline; Proposed Rule AGENCY: Environmental Protection Agency. ACTION: Notice of proposed rulemaking. ----------------------------------------------------------------------------- SUMMARY: This document proposes revisions to the simple model and the contents of a complex model to be used in the certification of reformulated gasoline, and associated procedures to assure compliance with the reformulated gasoline and antidumping programs. This notice additionally proposes "Phase II" reformulated gasoline emission performance standards which will take effect in the year 2000, as prescribed by section 211(k)(3) and authorized by section 211(c) of the Clean Air Act (CAA). The detailed proposed regulations described herein may be obtained from Public Docket No. A-92-12 or from the contacts listed in the ADDRESSES section. DATES: The comment period will close April 27, 1993. EPA will issue a subsequent notice indicating when and where the public hearing will be held. ADDRESSES: Comments may be submitted to U.S. EPA (RDSD-12) 2565 Plymouth Road, Ann Arbor, MI 48105. Materials relevant to this NPRM are contained in Public Docket No. A-92-12, located at room M-1500, Waterside Mall (ground floor), U.S. Environmental Protection Agency, 401 M Street SW., Washington, DC, 20460. Additional materials relevant to earlier related proposals are contained in Public Docket No. A-91-02 at the same location. The docket may be inspected from 8 a.m. until 12 noon and from 1:30 p.m. until 3 p.m. Monday through Friday. A reasonable fee may be charged by EPA for copying docket materials. EPA will provide a copy of the proposed regulations for this NPRM upon request. FOR FURTHER INFORMATION CONTACT: Michael Sklar (complex model), U.S. EPA (RDSD-12), 2565 Plymouth Road, Ann Arbor, MI 48105, Fuel Studies and Standards Branch, Telephone: (313) 741- 7817. Christine Brunner (Phase II performance standards), U.S. EPA (RDSD-12), 2565 Plymouth Road, Ann Arbor, MI 48105, Fuel Studies and Standards Branch, Telephone: (313) 668-4287. George Lawrence (enforcement and certification), U.S. EPA (6406J), 401 M Street, Washington, DC 20460, Eastern Field Office I, Telephone: (202) 233- 9307. To Request Copies of This Notice or the Proposed Regulations Contact: Marie Tolonen, U.S. EPA (RDSD-12), 2565 Plymouth Road, Ann Arbor, MI 48105, Fuel Studies and Standards Branch, Telephone: (313) 668-4295. SUPPLEMENTARY INFORMATION: I. Background The purpose of these proposed regulations is to improve air quality by requiring that gasoline be reformulated to reduce motor vehicle emissions of toxic and tropospheric ozone-forming compounds, as prescribed by section 211(k) of the Clean Air Act (CAA or the Act). This section of the Act mandates that reformulated gasoline be sold in the nine largest metropolitan areas with the most severe summertime ozone levels and other ozone nonattainment areas that opt into the program. It also prohibits conventional gasoline sold in the rest of the country from becoming any more polluting than it was in 1990, to ensure that refiners do not "dump" into conventional gasoline what they can no longer use in reformulated gasoline. Shortly after passage of the Clean Air Act Amendments of 1990, EPA entered into a regulatory negotiation with interested parties to develop specific proposals for implementing both the reformulated gasoline and related anti- dumping programs. These parties included representatives of the oil and automobile industries, vehicle owners, state air pollution control officials, oxygenate suppliers, gasoline retailers, environmental organizations, and citizens' groups. During the regulatory negotiation, EPA agreed to propose a two-step approach to reformulated gasoline. The first step would take effect in 1995 and utilize a "simple model" to certify that a gasoline meets applicable emission reduction standards. Under the second step, EPA would propose a "complex model" to supplant the simple model for certifying compliance with these standards, which would take effect on March 1, 1997 or 4 years after it is promulgated, whichever is later. EPA also agreed to propose the more stringent emission performance standards that the Act provides to take effect in the year 2000. Regulations to implement this first step were initially proposed on July 9, 1991 in a notice of proposed rulemaking (NPRM) (56 FR 13416). Today's notice includes approaches different from the previously proposed regulations related to the simple model as well as to the anti-dumping provisions. This proposal is supplemental to the previous proposals and is not intended to withdraw from consideration the various elements previously proposed, but provides additional proposals for public review and comment. (The reader may refer to the NPRM and SNPRM mentioned above, the Draft Regulatory Impact Analysis (DRIA), and Public Docket Numbers A-91-02 and A-92-12 for a thorough description of the goals and regulatory development of the reformulated and anti-dumping programs and discussions of a number of associated technical issues.) In this notice, the Agency is also proposing the complex model and associated revised enforcement provisions, as well as the year 2000 performance standards. In addition, EPA is proposing a range of NOX performance standards under section 211(c) of the Act as a complement to the VOC and toxics performance standards proposed under section 211(k). In a recent report on the chemistry of tropospheric ozone formation, the National Academy of Sciences emphasized the importance of NOX reductions in addressing this problem. The remainder of this preamble is organized into the following sections: II. Proposed Revisions to the Simple Model Proposal III. Complex Model IV. Vehicle Testing V. Phase I Performance Standards Using the Complex Model VI. Phase II Reformulated Gasoline Performance Standards VII. Enforcement Provisions VIII. Anti-Dumping Compliance and Enforcement Requirements for Conventional Gasoline IX. Anti-Dumping Requirements for Conventional Gasoline X. Environmental and Economic Impacts XI. Public Participation XII. Compliance With the Regulatory Flexibility Act XIII. Statutory Authority XIV. Administrative Designation and Regulatory Analysis XV. Reporting and Recordkeeping Requirements II. Proposed Revisions to the Simple Model Proposal A. Background 1. SNPRM As noted above, The Agency's SNPRM reflected the agreement reached in the regulatory negotiation that had been conducted to develop reformulated gasoline regulations under section 211(k) (as well as oxygenated fuels guidelines under section 211(m)). Added by the 1990 Clean Air Act Amendments, section 211(k)(1) directs EPA to issue regulations that, beginning in 1995, require the greatest reduction in emissions of ozone-forming [VOCs] and toxic pollutants achievable through the reformulation of conventional gasoline, taking into consideration the cost of achieving such emission reductions, any non air-quality and other air-quality related health and environmental impacts and energy requirements. Section 211(k)(3) specifies that the performance standards for 1995 to 2000 must require emission reductions, measured on a mass basis, equal to that achieved by a specified formula fuel or 15 percent of baseline emissions (the emissions of 1990 model year vehicles operated on a specified baseline gasoline), whichever is more stringent. For 2000 and beyond, the standard must be no less than the formula fuel or a 25 percent reduction from baseline emissions, whichever is more stringent. EPA can adjust this standard taking into account feasibility and cost, but in no case can it be less than 20 percent. Taken together, sections 211(k)(1) and 211(k)(3) call for the Agency to set standards that achieve the most stringent level of control, taking into account the specified factors, but no less stringent than those described by section 211(k)(3). For purposes of today's notice, it is useful to review the background of the SNPRM's proposed VOC and toxic standards. Given the short amount of time provided by the statute for promulgation of the reformulated gasoline standards, the regulatory negotiation committee agreed on the twostep approach noted above. The first step would utilize then-available information to construct a model that would determine a fuel's VOC and toxic emissions. Because the available information for quantifying the effects of fuel parameters on VOC emissions was limited in large part to the effects of Reid Vapor Pressure (RVP) and oxygen content, the so-called "simple model" took those fuel parameters into account in predicting a fuel's VOC emissions for purposes of compliance under the VOC performance standard. At the same time, it was understood that certain other fuel parameters affected a fuel's emissions, and that directionally higher levels of these other parameters were likely to increase VOC emissions. Until the effects of these parameters could be accurately quantified, the committee agreed that the average level of these parameters in each refiner's fuel should remain capped at 1990 levels. The second step was intended to take advantage of data then being developed to characterize the emissions effects of other fuel parameters. That data would be used to construct a "complex model" that would better predict a specific fuel's emissions based on these parameters. The complex model would thus also provide refiners with more options for reducing those emissions by changing the levels of these parameters in their gasoline formulations. The simple model would be available for at least the first two years of the program and would be replaced by the complex model four years after the complex model's promulgation. The Agency agreed to propose the four-year period to allow refiners lead-time in order to recoup investments made to comply with the simple model and to make any refinery changes necessary to comply with the standards under the complex model. However, under the agreement, the complex model could be used on a voluntary basis as soon as it was promulgated, so long as any fuel certified using the complex model achieved emission reductions as great as those it would have to achieve to be certified using the simple model. In developing the proposed VOC standards, it became clear that the formula fuel specified by section 211(k)(3) did not yield emissions reductions greater than the statutory minimum of 15 percent. To achieve the statutory minimum under the simple model, the regulatory negotiation committee members agreed that a fuel's RVP (measure of volatility) could not be greater than 8.1 pounds per square inch (psi) and its oxygen content had to comply with the 2.0 weight percent requirement prescribed by section 211(k)(2)(B). The more volatile a fuel, the more VOCs it emits, mostly through evaporation. The more oxygen in the fuel the lower the exhaust VOC emissions (up to a maximum of 2.7 weight percent oxygen per the simple model). Committee members considered the need for further emission reductions and the cost of obtaining them in deciding whether and to what extent VOC performance standards more stringent than the statutory minimum were achievable. For the reasons laid out in the SNPRM, EPA and the other committee members concluded that reformulated gasoline with 2.0 weight percent oxygen, 8.1 psi RVP in northern cities, and 7.2 psi RVP in southern cities represented the greatest emission reductions that could be achieved by reformulating gasoline, taking into account the specified factors. The committee sought the more stringent VOC reduction requirement for the South in order to assure that southern nonattainment areas obtained emission reductions beyond what those cities would obtain from the volatility reduction requirements separately required by section 211(h). (Due to higher temperatures in the South, emissions from a gasoline with a specified RVP are greater in the South than in the North. Under section 211(h), the volatility limits for southern areas were set lower than those applicable in northern areas in order to achieve comparable per vehicle emissions levels in the two regions). As explained in the SNPRM, however, the additional cost of achieving the more stringent standard in the South would be significant and made it difficult to impose a more stringent standard in the North, as well. In effect, the committee decided to allocate the greatest reductions from the 8.7 RVP baseline gasoline that reformulating gasoline could reasonably achieve in such a way as to roughly equalize the actual emission benefits of the program between northern and southern nonattainment areas. 2. Ethanol Issues Following the release of the SNPRM, supporters of ethanol, an agriculturally derived oxygenate added to gasoline in some parts of the country, again raised the concern that the simple model's RVP requirement would effectively preclude ethanol from the RFG market during the summer months (VOC control season). Ethanol, when added to gasoline in the amount needed to satisfy the oxygen content requirement of the Act raises the volatility of the resulting blend by about 1 psi. Consequently, for the ethanol to be blended with the RFG under the simple model, a blendstock gasoline with an RVP low enough to offset the increase resulting from adding ethanol would have to be obtained. Ethanol representatives commented that obtaining such blendstocks would be both difficult and expensive, because "sub-RVP" blendstocks would be more costly to refine and because blendstock production would be controlled by petroleum refiners. Methyl tertiary butyl ether (MTBE), an oxygenate which does not boost a fuel's RVP, which is derived from natural gas and the petroleum product isobutylene and can readily be put through petroleum pipelines, was thought to be the oxygenate of choice for most refiners. Ethanol's representatives theorized that the oil industry would have a desire to use MTBE over ethanol and thus little incentive to make the sub-RVP blendstock necessary for ethanol blending. The ethanol industry contended that a reformulated gasoline program which they argued would effectively preclude ethanol was contrary to Congress' intent that ethanol have a role in the program. They argued that the oxygen content requirement of section 211(k)(2) was motivated in large part by a desire to expand markets for ethanol. They noted the strong support afforded the RFG legislative initiative by Members of Congress from agricultural states. They also cited statements in the legislative history indicating some members' expectation that the RFG program would provide an increasing market for ethanol. Ethanol representatives contended that the benefits of ethanol use justify its inclusion in the RFG program. Specifically, they explained that ethanol is currently made in the United States from domestically-grown grains, primarily corn, and thus represents an important domestic and renewable source of energy. They further explained that to the extent ethanol is used in place of imported petroleum products, it promotes the nation's energy independence and improves its balance of trade, and that ethanol use also strengthens the market for corn, consequently reducing the need for price supports. Moreover, as a biomass-based product, ethanol is potentially a renewable fuel to the extent the energy derived exceeds any fossil fuel energy consumed in producing the ethanol. In view of ethanol's importance to the nation's energy security and agricultural economy, ethanol representatives urged that the simple model proposal be revised to allow ethanol to effectively participate in the RFG market. They suggested several possible revisions. For example, they argued that the 1 psi waiver granted to certain ethanol blends by section 211(h) of the CAA be applied to ethanol-blended RFG under section 211(k). They reasoned that since Congress recognized in the provision requiring nationwide reductions in fuel RVP that ethanol required such a waiver, ethanol should receive a similar waiver if the VOC performance standard for RFG sold in the smoggiest cities were defined in terms of a required reduction in RVP. If the section 211(h) waiver were not available to RFG ethanol blends, the ethanol industry suggested that the VOC reduction requirement take into account that the variety of VOCs from gasoline differ in their ozone formation potential. While ethanol raises a fuel's volatility and thus its VOC emissions, they argued that the resulting VOCs are less ozone-forming than those that would otherwise occur. They urged that the 15 percent reduction requirement should thus be interpreted to require a 15 percent reduction in ozone-forming potential, not simply ozone-forming VOCs. Ethanol supporters suggested additional ways of encouraging or even requiring ethanol use in RFG. The Governors Ethanol Coalition, for instance, suggested that EPA require the RFG market to satisfy its oxygenate requirements through a minimum percentage of domestically produced renewable fuel. 3. October 1st Announcement EPA, the Department of Agriculture, the Department of Energy and other parts of federal government engaged in an extensive dialogue with all interested groups to address the ethanol industry's concerns with the RFG program. In response to the ethanol industry's suggestion, the applicability of the section 211(h) waiver to the RFG program under 211(k) was closely considered. For the reasons set forth in a memorandum available in the docket, the Acting General Counsel of EPA determined that it did not apply. Also explored was the ability to take reactivity into account in determining compliance with the requirement that "ozone-forming VOCs" be reduced by 15 percent. Section 211(k)(3), however, explicitly requires that reductions in ozone-forming VOCs be measured on a "mass basis." Although any reactivity benefits of emissions from ethanol blends, should they exist, could be taken into account in defining which VOCs are "ozone-forming," the statute does not appear to permit ozone-forming VOCs to be weighted on the basis of their ozone formation potential for purposes of determining compliance with the 15 percent reduction requirement. Based on ethanol's importance to the nation's energy and agricultural policy, President Bush on October 1, 1992 announced a plan to allow ethanol to effectively compete in the RFG program, with the expectation that, with barriers removed, ethanol use would grow. This plan is based upon provisions of section 211(k)(1) allowing the Administrator to take into consideration cost, energy requirements, and other specified factors in setting RFG performance standards. The most significant part of this plan called for EPA to "establish rules for reformulated gasoline in all northern cities that will have the effect of granting a one-pound waiver for the first 30 percent market share of ethanol blends, while achieving environmental benefits comparable to those provided for in EPA's proposed rule and regulatory negotiation." The environmental benefits of the proposed RFG program would be maintained by offsetting any increase in volatility of RFG containing ethanol with reductions in the volatility of the rest of the reformulated gasoline pool. Also pursuant of the plan, EPA is proposing to propose to establish rules allowing governors of southern states to extend the provisions made for ethanol in northern RFG-covered cities to covered southern cities as well. Other aspects of the plan include the establishment of an RVP trading program, the unrestricted optional early use of the complex model and oxygenate neutral NOX and toxics performance standards. Through today's notice the Administrator proposes these as well as a variety of other provisions relevant to his determination of section 211(k) factors. B. Reid Vapor Pressure Allowance 1. Overview EPA proposes that the RVP performance standard under the simple model provisions of the SNPRM and VOC performance requirements under the complex model be based upon the extent of use of renewable oxygenates such as ethanol. For northern areas under the simple model, EPA proposes that the RVP performance standard as proposed in the SNPRM of 8.1 psi (8.0 psi under the averaging provisions of the SNPRM) be reduced to 7.8 psi (7.7 psi under averaging) when no renewable oxygenates are used in reformulated gasoline, but be permitted to increase to 8.1 psi (8.0 psi under averaging) through the use of renewable oxygenates in 30% of the reformulated gasoline produced. For example, taking advantage of the proposal's averaging provisions, this would allow ethanol to be splash blended into 30% of reformulated gasoline and have a 1.0 psi greater RVP than other reformulated gasoline blends while still maintaining the in-use volatility and, thus, comparable environmental benefits proposed in the SNPRM. EPA is also proposing to permit a governor of a southern state to have applied in that state's ozone nonattainment areas provisions similar to those described above for northern areas. Specifically, under the simple model, EPA proposes that the 7.2 psi RVP standard proposed in the SNPRM for southern areas be reduced to 7.0 psi, but that refiners may meet an RVP standard of up to 7.2 psi if they use ethanol or other renewable oxygenates in up to 20 percent of the market covered in those areas. In addition, the Agency proposes a similar approach to standard setting under the complex model. It should be noted that this program in no way limits the ability of a refiner to use ethanol or any other renewable oxygenate in its gasoline to only 30 percent of its production (20 percent in the South). A refiner may blend as much ethanol into its gasoline as it desires, provided it meets its RVP (VOC) performance standards. It should also be noted that the ability of refiners to average RVP performance under the provisions of the SNPRM is not changed as a result of this program. RVP averaging can still be used to meet a refiner's RVP (VOC) performance requirements. In fact, as discussed below, EPA is proposing to expand the RVP averaging provisions of the SNPRM to include RVP trading as well. In order to provide greater flexibility for the use of renewable oxygenates, EPA proposes that refiners be permitted to either blend the renewable oxygenate themselves to raise their performance standard, or obtain commitments from other refiners to do so. This flexibility in connection with the RVP trading provisions will allow greater renewable oxygenate use by the refiners who can benefit from their use, and allow its use in those areas where it is most economical to do so. Today's action does not mandate the use of ethanol. Taking into account the statutorily specified factors, EPA is proposing these provisions for ethanol on the theory that allowing ethanol to effectively participate in the RFG market is important to the nation's energy security. While EPA is continuing to gather the relevant data and perform the requisite analyses, it appears that the use of renewable oxygenates such as ethanol as a motor vehicle fuel may reduce the nation's dependence on foreign oil, improve our balance of trade with other nations, and enhance our energy security./1/ At 2.7 wt% oxygen, use of ethanol (or ethanol in the form of ETBE) in 30 percent of RFG) would displace 2.3 percent of the volume of reformulated gasoline with ethanol. This is the equivalent of 360 million gallons of ethanol used (gasoline displaced) during the summer months alone in the areas currently covered by the reformulated gasoline program (including those areas that have already opted in), and could rise to as much as 510 million gallons if all eligible areas opted into the program. In addition to energy related benefits, the use of oxygenates made from renewable sources provides the potential for reducing emissions which contribute to global warming. NOTE /1/ The Final Report of the Interagency Commission on Alternative Fuels written for DOE and published in September, 1992 estimates that the displacement of crude oil with all oxygenate use under the reformulated gasoline and oxygenated fuels provisions of the CAA could reach roughly 200,000 barrels per day. In a private conversation on 1/8/92, Margaret Singh of Argonne National Laboratory, one of the authors of the Interagency Commission report, stated that roughly 10 percent of the crude oil displacement in that analysis was due to ethanol alone (this estimate does not take into account any changes which may result from today's proposal). As explained above, the statute prescribes for the near term a minimum 15 percent reduction in VOCs, and in the SNPRM EPA proposed RVP and oxygen requirements that it believed achieved this requirement. Under today's proposal, the simple model RVP standards for refiners or importers that use 30% renewable oxygenate is 8.1 psi, if compliance is met on a per gallon basis. Both the renewable and nonrenewable blends produced by those refiners or importers would have to meet this per gallon standard. If compliance is met on average, however, the RVP standard would be 8.0 psi. This could be met, for example, by two-thirds of a refiner's or importer's gasoline meeting 7.7 psi (nonrenewable blends) and one-third meeting 8.7 psi (renewable blend such as ethanol splash blended into a 7.7 psi blendstock). The effect of today's proposal on the cost of reformulated gasoline is expected to be small. The cost of reducing the RVP of gasoline by 0.3 psi in northern (VOC control region 2 as defined in the SNPRM) areas to offset the allowance for splash blended ethanol in 30% of the fuel is approximately 0.3 to 0.5 cents per gallon. However, the overall cost of the RVP performance allowance of 0.3 psi for the use of ethanol is expected to be less than 0.3 cents per gallon since ethanol is likely to be less costly than MTBE per gallon of reformulated gasoline produced./2/ NOTE /2/ "Evaluation of the Use of Ethanol and MTBE in Reformulated Gasoline," by Sobotka & Co., Inc., for USEPA, September 30, 1992. In addition, the cost of ethanol may decrease beginning in 1995 as a result of the proposed provisions. Under a program like that set forth in the SNPRM, ethanol would be expected to capture a sizeable portion of the RFG and oxygenated fuels markets during the winter months. (The oxygenated fuels program established by section 211(m) requires that gasoline sold in many carbon monoxide nonattainment areas include 2.7 percent oxygen during the winter months, beginning in 1992.) However, since the market during the summer was potentially small, ethanol production facilities could not operate year-round at a steady production level without storing vast quantities of ethanol during the summer for use during the winter. To the extent ethanol was also used during the summer months, more steady production could be achieved, with reduced capital expenditures for storage and peak production equipment. This would likely result in a reduction in the cost of producing ethanol. Year-round and increased use of ethanol may also increase the incentives to develop an established infrastructure for ethanol distribution and use, making it even more feasible and cost effective. In addition, expanded use of renewable oxygenates may encourage further research and development associated with their production and use, which could further reduce their cost and expand their market both as blends with gasoline and separately as alternative fuels. Expanded use of renewable oxygenates would further result in economic growth not only in the ethanol industry, but also in agriculture and related industries. Finally, since the consumption of grain to produce ethanol may raise the prices farmers receive for their crops, additional ethanol use may reduce the amount the government must pay in agricultural price supports. While the cost per gallon of reformulated gasoline of the RVP adjustment for the use of renewable oxygenates is small, the incentive per gallon of ethanol is substantial. An economic incentive for renewable oxygenate containing blends such as ethanol blends of 0.3 cents per gallon adds roughly 13.7 cents per gallon to the value of ethanol (assuming a base requirement of 2.7 wt% oxygen as described in section 4)./3/ NOTE /3/ The derivation of the value of the incentive is shown in Section VIII of the DRIA. In sum, section 211(k)(1) requires that gasoline be reformulated to achieve the greatest reductions in VOC emissions considering cost, energy, and the other specified factors, but not less than a 15% reduction during Phase I of the program and 25% (20%) during Phase II of the program. In determining how these statutory requirements can be met, EPA must balance the cost of RVP reductions in non-ethanol blended gasoline and the extent to which energy or other considerations warrant that the volatility associated with ethanol use be offset. Today, EPA is proposing that the volatility associated with a potential ethanol market share as high as 30 percent be offset through reductions for nonrenewable oxygenate blended gasoline such that the reformulated gasoline market achieve overall reductions meeting the minimum 15% reduction requirement. EPA requests comment on whether this proposal represents the proper balance. EPA believes that, in addition to the factors described previously concerning the energy and other benefits of renewable oxygenate use, the following are relevant to the issue: The cost to reduce the RVP of nonrenewable oxygenate blends below the previously proposed levels; the leadtime available for such reductions; and the market share ethanol and other renewable oxygenates could expect absent any such provisions. The expected market share absent these provisions is hard to predict, and at best 30% would allow ethanol and other renewable oxygenates to effectively participate in the reformulated gasoline program, including the opportunity to expand that participation beyond what would otherwise be possible. 2. Performance Standard Setting Process The RVP performance standards proposed today are intended to provide the same level of RVP and comparable environmental control as would be achieved by the SNPRM, while allowing ethanol and other renewable oxygenates to effectively participate in the reformulated gasoline program. This can be accomplished by adjusting the RVP performance standard with the use of renewable oxygenates while preventing the standard from going any higher than that proposed in the SNPRM. As such, EPA proposes to establish a performance standard for RFG not containing renewable oxygenates that is more stringent than that which was proposed in the SNPRM. EPA further proposes that if refiners commit to the use of renewable oxygenates in their fuel or purchase commitments from other refiners for the use of renewable oxygenates (if that proves more cost effective) their RVP performance standard would be relaxed accordingly (up to 8.1 psi). Finally, a refiner's performance standard could be relaxed even further through a process by which EPA provides RVP performance standard adjustments based on the refiner's commitment to use renewable oxygenate that other refiners might have used to take advantage of a less stringent RVP (VOC) performance standard, but chose not to. This RVP (and VOC) performance standard setting process is described in the following sections and shown in figure II.1. BILLING CODE 6560-50-M Figure II.1 RVP PERFORMANCE STANDARD SETTING PROCESS [INSERT: Flow diagram] BILLING CODE 6560-50-C a. Southern Performance Standards. The proposed approach taken in the North would not be required for southern areas covered by the RFG program. There are several reasons for this distinction. First, most of the RFG areas are in the North, so that removing barriers to ethanol use in the North will have a larger impact than removing these barriers in the South on the policies associated with ethanol use. Second, production of grain feedstocks and ethanol production are concentrated in the North. The higher tax benefits these northern areas offer ethanol together with lower transportation costs account for ethanol's greater use in the North. For these same reasons, EPA expects that the potential for growth in ethanol use is expected to be greater in the North. Third, the lower RVP standards applicable under the simple model in the South are likely to make additional RVP reductions to offset volatility increases associated with ethanol use more expensive for southern RFG areas. In sum, in striking the balance between achievable RVP reductions and the policy benefits associated with ethanol, EPA has reason to believe that for the South generally, the RVP costs may be higher and the resulting ethanol policy benefits may be lower. Thus, striking the balance somewhat differently for southern areas, President Bush's announcement did provide, nonetheless, that a modified version of the northern approach be available to governors of southern states to have applied at their discretion. A governor who believes that ethanol has a potentially strong market in his or her state could request that this modified approach, explained in more detail below, be applied in the state's RFG-covered areas. As summarized above while EPA believes that ethanol participation in southern RFG areas would advance the same policy concerns as ethanol participation in northern areas, EPA is not certain that ethanol would participate sufficiently even if the barriers are removed to justify the higher costs of offsetting the volatility increase in southern areas. However, southern governors may be in a good position to judge on a case-by- case basis the likely participation of ethanol in the market and the cost of removing the barriers. The modification of the northern approach for optional application in the South is as follows. EPA is proposing that the RVP standard for RFG not containing renewable oxygenates be reduced from 7.2 psi to 7.0 psi (6.9 psi under the averaging provisions of the SNPRM). The 0.2 psi reduction would permit 20 percent of the RFG market in an area to be splash-blended with ethanol without raising the average RVP of the RFG in the area. EPA further proposes as one option that the effective date for inclusion of a southern covered area be the beginning of the first VOC control season three years following receipt of the governor's letter requesting application of this approach in his or her state. EPA believes three years is the appropriate amount of time to provide adequate lead-time for refiners to prepare to make the necessary additional RVP reductions, and requests comment on the validity of this belief. While having the program automatically take effect three years after the request from the governor is a relatively straightforward approach, it does raise serious questions about whether section 211(k) authorizes a state to determine which federal standard applies in its area. At the same time, EPA is concerned whether receipt or non-receipt of a request from a governor is an adequate basis to support a different performance standard in that area. In light of these concerns, EPA is proposing a second option whereby the ethanol provisions described above would be implemented in a southern area through a rulemaking conducted subsequent to receipt of a request from the Governor. Upon receipt of any such request from a governor, EPA would promptly prepare a notice proposing extension of the ethanol-related provisions to the area or areas covered by the request. This would allow for full exploration of all the relevant issues, and would avoid the legal concerns mentioned above. This option is similar to the approach taken in the federal RVP regulations under section 211 (c) and (h), where RVP standards were set on a national basis with allowance for subsequent rulemaking to fine tune the regulations for specific localities. EPA requests comment on the relative merits of these two options. The energy related benefits of renewable oxygenates are not unique to the use of the simple model. Under the SNPRM, refiners will have the option of using the complex model in the first years of the program when the simple model is also available. After a specified period of time following the complex model's promulgation, only the complex model will be available for certification. In keeping with the proposal under the simple model, EPA proposes that the VOC performance standard stringency under the complex model be increased by the equivalent of 0.3 psi RVP. For Phase I of the reformulated gasoline program EPA proposes to determine the VOC equivalent of 0.3 psi RVP using as a baseline a fuel meeting the performance standards of the simple model (8.0 RVP, 2.1 wt% oxygen, 0.95 volume percent benzene, 26.3 percent aromatics, and CAA baseline values for all other fuel parameters), and under Phase II of the reformulated gasoline program one baseline fuel consistent with the VOC performance standard described in section VI of this notice (for purposes here, 2.0% oxygen, 105 ppm sulfur, 6.9 RVP, 202 deg.F T50, 316 deg.F T90, 23.0% aromatics, 6.5% olefins, and 1.0% benzene). As calculated from the complex model, EPA thus proposes that the Phase I and Phase II VOC performance standards be increased by 7.1 and 2.8 percentage points, respectively, in northern areas. If any southern areas are included in the program, the Phase I and Phase II VOC performance standards would be increased by 5.9, and 2.0 percentage points, respectively (for Phase I, the equivalent of 0.2 RVP determined using a baseline fuel of 7.1 RVP, 2.1 wt% oxygen, 0.95 volume percent benzene, 26.2 percent aromatics, and CAA baseline values for all other fuel parameters and, for Phase II, the equivalent of 0.2 RVP using as a baseline fuel a fuel with 2.0 wt% oxygen, 143 ppm sulfur, 6.5 RVP, 202 deg.F T50, 316 deg.F T90, 23.0% aromatics, 6.5% olefins, and 1.0% benzene). EPA requests comment on this methodology for setting the performance standards under the complex model, on the fuel assumptions used, on whether and to what extent the methodology unfairly disadvantages any refiner relative to another, and on whether there are any other methodologies or fuel assumptions which might be more appropriate. b. Performance Standard Adjustment Through Commitments for Renewable Oxygenate Use. EPA's proposed RVP and VOC standards are based on its estimate of the greatest reductions the market as a whole can achieve, taking into account the statutory factors. However, the ability and desire of refiners to take full advantage of today's proposed standards varies from refiner to refiner. To ensure that the policy objectives related to this proposal are fully served, it is important that the market be permitted to work as efficiently as possible. EPA is therefore proposing that a system of commitments be created to allow the greatest flexibility for the use of ethanol and other renewable oxygenates consistent with the applicable performance standards. In fulfilling their commitments, refiners could either blend renewable oxygenate themselves during the forthcoming VOC control season or, if meeting the VOC standards on average, also purchase commitments from other refiners. As long as refiners meet their RVP (VOC) performance standards in each area covered by the RFG program, the use of renewable oxygenate could easily be shifted from a refiner serving one area to a refiner serving another area where it is more economically advantageous to do so (based on state tax subsidies, distribution costs, etc.). Whether or not a refiner chose to fulfill its commitment through its own use of renewable oxygenate or through contracting for its use with another refiner, however, the refiner would nevertheless be liable for its commitment. Since various refiners produce vastly different amounts of gasoline, in order to maintain the same level of renewable oxygenate use under this program, trading of renewable oxygenates from one refiner to another must be performed on a volume basis. Under the simple model, refiners could elect to meet an RVP performance standard as high as 8.1 psi (8.0 under averaging) in the North and 7.2 psi (7.0 under averaging) in the South in proportion to their commitment to use renewable oxygenate in 30 percent and 20 percent of their fuel, respectively. Under the complex model refiners could lower the otherwise applicable VOC performance standard by as much as 7.1 percentage points in the North and 5.9 in the South under Phase I of the reformulated gasoline program, and 2.8 percentage points in the North and 2.0 percentage points in the South under Phase II, through the use of renewable oxygenates in 30 and 20 percent of their fuel, respectively. It is entirely up to each refiner to decide what amount of renewable oxygenate to use and, thus, which performance standard would be applicable to it. Refiners could utilize as much renewable oxygenates in producing their reformulated gasoline as they desire, but the ability to adjust the applicable performance standard in proportion to the amount used would be limited to 30 percent of their production in the North and 20 percent in the South. In order to provide refiners with the necessary flexibility to blend renewable oxygenates, EPA proposes to raise the maximum RVP allowed under the averaging provisions of the SNPRM to 8.7 psi for renewable oxygenate blends in the North and 7.9 in the South. Similarly, the minimum VOC performance standard under the complex model is proposed to be relaxed by 21.2 percentage points in the North and 24.6 percentage points in the South below the averaging standard under Phase I of the reformulated gasoline program. Under Phase II, EPA proposes that the minimum performance level be set at 8.5 and 8.3 percentage points in the North and South, respectively, below the performance standard described in section VI (taking into account a 1.5% increase in the standard under averaging). These levels are sufficient to provide a 1.0 psi margin for splash blending ethanol into 30 percent of reformulated gasoline downstream of the refiner. EPA proposes, however, that for reformulated gasoline not containing renewable oxygenates the maximum RVP remain at the already proposed level for the simple model of 8.3 RVP in the North and 7.4 in the South, and that the minimum VOC performance under the complex model remain at the levels proposed later in today's notice. The performance standard adjustment for blending renewable oxygenates has value only if the RVP or other parameters of gasoline are adjusted. Since these are determined by the refiner, EPA proposes that the refiners continue to be the party responsible for meeting the RVP performance standard, which is now a function of the amount of renewable oxygenate used. Since this is fully consistent with the provisions of the SNPRM, changes to the enforcement provisions proposed in the SNPRM are minimized. EPA considered various other options which would also place requirements on terminal operators (ethanol blenders), but considered such options as being too restrictive and burdensome. At the same time, unless refiners chose to comply with the per gallon performance standards, some blends (e.g., those splash blended with ethanol) would be likely to have an RVP that is higher than that of other blends by as much as 1.0 psi. (Under the per gallon standards, all of a refiner's fuel must meet its per gallon standard (e.g., 8.1 RVP if renewable oxygenate is used in 30 percent of its fuel) including any ethanol splash blends.) Therefore, in order to ensure that the RVP (and VOC) standard is met in each area covered by the reformulated gasoline program, it would be essential for the refiner to track and control the fraction of fuel with a higher RVP which is distributed to each area. Similarly, it would be essential for refiners to maintain a constant fraction of fuel with a higher RVP (lower VOC performance) over the course of the VOC control season. Otherwise, there could be temporal increases and decreases in the average emission performance of in-use fuels. As under the averaging provisions of the SNPRM, should temporal increases occur, the refiners run the risk of in-use compliance survey violations and thus, a more stringent performance standard the following year. c. Performance Standard Adjustment Through EPA Allocation of Uncommitted Renewable Oxygenate Use. Some refiners may decide not to use renewable oxygenates and forego the opportunity to meet a less stringent RVP (VOC) performance standard. So that other refiners might take advantage of that opportunity and thereby further advance the policy goals of today's notice, EPA proposes that a refiner's opportunity to elect a less stringent RVP (VOC) performance standard be forfeited for the following ozone control season if it does not commit to the use of renewable oxygenates (by themselves or through the purchase of commitments from other refiners). EPA would then adjust the RVP performance standard of other refiners in exchange for their commitment to use the amount of renewable oxygenate which had been left uncommitted. For example, a refiner which commits to use renewable oxygenate in 60 percent of its fuel and receives the associated RVP performance standard adjustment could raise its average RVP performance standard in the North under the simple model from 8.0 to 8.3 psi. This allocation system would also provide the means of effectively combining offsetting fuels in slates for the purpose of certifying them as RFG. Section 211(k)(4) permits gasoline to be certified as RFG as part of a "slate" of fuels that on average meet RFG requirements. While the proposed system of allocating RVP performance standard incentives would result in one refiner's fuel having an average RVP above the minimum 8.1 psi standard, by effectively combining in a slate that refiner's fuel with another refiner's fuel having an average RVP sufficiently below the standard, both refiner's fuels could still be certified as reformulated gasoline. In order to permit EPA to adjust the RVP (VOC) performance standards, information on reformulated gasoline and renewable oxygenate use must be made available on an area-specific basis. If RVP performance standard adjustments for renewable oxygenate use were forfeited by a refiner in one covered area and subsequently allocated to refiners serving another covered area, greater environmental control would be achieved in one area and less in another. To avoid such an outcome, EPA proposes that by November 1 prior to each ozone control season each refiner (and importer) report to EPA its commitment to use renewable oxygenates (and in what form), the amount of renewable oxygenate other refiners have committed to use for it, the amount of renewable oxygenate it has committed to use for other refiners, the amount of renewable oxygenate use eligible for making performance standard adjustments which is not committed to be used in each area covered by the reformulated gasoline program, its total reformulated gasoline production which was distributed for use in each covered area for the previous year, and the VOC performance characteristics (RVP under the simple model, all fuel parameters that affects VOC under the complex model) of the fuel which it distributes for use in each covered area. If and only if a refiner chooses to meet the VOC performance standards on a per gallon basis, commits to the use of a minimum of 30 percent of renewable oxygenate in the North and 20 percent in the South (if applicable), and is not committing to use renewable oxygenates for any other refiner, would its reporting requirements be simplified. In such a case the refiner would be required to report only its commitment to use at minimum 30 percent, and the volume of reformulated gasoline produced and renewable oxygenate used each year nationwide. By January 1 prior to each ozone control season, any uncommitted RVP (VOC) performance standard adjustments would be provided at no cost to other refiners for use in the same covered area for their commitment to use additional renewable oxygenate, and their performance standard would be adjusted accordingly. EPA proposes to allocate these adjustments among the refiners wishing to use them in the covered area where the adjustments are forfeited prorated based on the volume of reformulated gasoline produced by the various refiners requesting additional adjustments, and beginning with the category of refiners who fulfilled their commitment to use renewable oxygenates in previous years. The final RVP performance standard applicable for these refiners for the coming season would then be recalculated based upon the total amount of renewable oxygenate committed to be used. EPA's adjustment of RVP (VOC) performance standards, as proposed, would be specific to each area covered by the reformulated gasoline program in order to maintain each area's environmental benefits from the reformulated gasoline program. As such it would require that refiners track their fuel from the point of production to the point of its ultimate use and report to EPA the amount of reformulated gasoline and renewable oxygenates distributed for consumption in each area (except as described above under certain conditions for per gallon compliance). EPA recognizes that such a requirement changes current industry operating practices and as such may impose additional costs. However, EPA is unaware of any alternatives which would provide comparable environmental neutrality. EPA requests comment on any alternative means by which the environmental neutrality of the program can be maintained while minimizing the cost and burden on refiners. EPA also requests comment on the ability of the in-use compliance surveys to be revised to provide adequate certainty that the full environmental control is achieved in each area without requiring refiners to track their fuel. Despite the precautions described above, EPA is still seeking design improvements to fully assure that the program proposed today preserves the environmental control that would be achieved under the program proposed in the SNPRM. The program described above does not fully protect environmental neutrality if changes occur in the amount of fuel distributed to a given city by various refiners over time. Since the standards adjustment process is based on fuel volumes for the previous year, increases in the performance standard granted to some refiners following EPA adjustments would only be only offset if the refiners forfeiting RVP performance standard adjustments distribute the same or greater volume of reformulated gasoline to a given area as the year before. It is not until the following year that the adjustment would be made to offset this disparity. Despite this concern, EPA does not expect that there will be significant increases in RVP (decreases in VOC performance) in-use since decreases are just as likely as increases and not all refiners that opted to forfeit the performance standard adjustments would be expected to dramatically reduce the amount of fuel they distributed to a given area in the same year. Nevertheless, EPA requests comment on the this issue and on any means by which the environmental neutrality of the program can be more fully ensured. If during the course of the ensuing VOC control season a refiner uses less renewable oxygenates than it had committed to EPA, no enforcement action would be taken as long as that refiner achieves an RVP or VOC performance standard commensurate with its actual renewable oxygenate use. However, if a refiner uses less renewable oxygenates than it has committed to use, EPA requests comment on what remedy would be appropriate. One possible remedy would be that the refiner forfeit the ability to raise its performance standard through renewable oxygenate use in the future (to the extent the commitment was not reached). EPA would make these RVP (VOC) performance standard adjustments available to other refiners willing to commit to use the corresponding amount of renewable oxygenate as described above. If such a forfeiting refiner wanted to raise its performance standard through renewable oxygenate use in the future, it could do so only if unallocated RVP (VOC) performance standard adjustments were still available for the covered area(s) in question. Such refiners would have their RVP (VOC) performance standard adjusted only after all refiners who fulfilled their commitments the previous year have received any desired adjustments. EPA is concerned that the performance standard adjustment process described above does not grant refiners desiring to use additional renewable oxygenate adequate certainty of the availability of additional RVP (VOC) performance standard adjustments to fully take advantage of them. Furthermore, EPA is concerned that the refiners described in the previous paragraph who forfeited their opportunity to take advantage of RVP (VOC) performance standard adjustments for future years would be limited to receiving year to year adjustments from EPA if any are available and would never have any certainty of using any performance standard adjustments. As a result, EPA is also considering an approach whereby performance standard adjustments which are either given to EPA for subsequent distribution be permanent (instead of year to year) and whereby adjustments distributed by EPA would be permanent. Under such an approach, refiners would have the necessary certainty of being able to utilize the performance standard adjustments and be permitted to make refinery changes in accordance with such performance standard adjustments. Such an approach, however, prevents the year-to-year adjustments in reformulated gasoline volume necessary to ensure that any loss in environmental neutrality described above is limited to only one year. Another approach might be to maintain the year to year allocation but allow refiners who fell short of their commitments in previous years to regain their original opportunity to use renewable oxygenates for RVP (VOC) performance standard adjustments by making up their earlier shortfall without any RVP (VOC) adjustment. EPA requests comments on the above approaches, on the need for an allocation program given its potential limitations, and on any other approaches that would provide refiners with the necessary certainty while still preserving the environmental neutrality of the program. 3. RVP Trading Program Based on the performance standard adjustment process described above, if a refiner wishes to blend more than 30% of its production with renewable oxygenate, it would still be required to meet an 8.1 RVP standard. With regard to ethanol, this would require producing an ever lower RVP gasoline blendstock. In order to provide greater flexibility to refiners desiring to blend additional renewable oxygenates, EPA proposes to establish an RVP credit trading program (in addition to the averaging program proposed in the SNPRM) whereby refiners could obtain RVP credits from other refiners in lieu of reducing the RVP of their fuel. Under this program a refiner in Chicago could commit to use renewable oxygenate on behalf of a refiner in New York and then obtain RVP credits (from other refiners distributing fuel in Chicago) sufficient to compensate for any increase in RVP caused by increasing its use of renewable oxygenates. For the program to remain environmentally neutral the RVP credits could only be traded within a given area covered by the reformulated gasoline program. Otherwise, one area would have an average in-use RVP of less than 8.1 RVP, while the other area would have an average RVP of greater than 8.1 RVP. Such an RVP trading program could be conducted totally separate from any exchange in commitments for renewable oxygenate use or in conjunction with it. An example of the latter would be the sale of renewable oxygenate usage commitments and the purchase of RVP credits by a refiner wishing to use more than 30% renewable oxygenates and vice versa by a refiner wishing to use less than 30% renewable oxygenates. The former could potentially adjust its average RVP standard to as high as 8.7, while the average RVP standard of the latter refiner would be lower than 8.0. The fuels being marketed by both refiners would have to be in the same area, as mentioned above. Furthermore, if the program is to be environmentally neutral, RVP credits must be traded based on the relative volume of reformulated gasoline produced at any given RVP. Since RVP trading is a means by which refiners can choose to meet their RVP performance standard and is not used to determine the performance standard, EPA proposes that RVP trading be permitted to take place throughout the course of the control period and as such be based on the actual gallons of reformulated gasoline produced. EPA proposes that to the extent refiners take advantage of RVP trading, they be required to report to EPA at the end of the ozone control season the average RVP of the reformulated gasoline they produced and the number of RVP credits traded to and received from other refiners for each of the reformulated gasoline areas. EPA does not believe that the added flexibility resulting from a trading program allows for a more stringent performance standard, but requests comment on the Agency's legal authority it implement RVP trading absent a tighter performance standard for those refiners who trade. A more stringent performance standard was proposed in the SNPRM in connection with averaging, and this is merely an extension of that program. Furthermore, this trading program is intended to provide greater flexibility to refiners for the purpose of making it easier to blend renewable oxygenates. If the performance standard were made more stringent, it could eliminate any incentive for blending renewable oxygenates provided by a trading program. Since there is no RVP performance standard under the complex model, the RVP trading program described above cannot be used directly. However, the same motivation exists under the complex model as under the simple model to provide a trading program. Thus, under the complex model EPA proposes that refiners be permitted to trade VOC performance. EPA requests comment on this provision. As discussed above in regard to EPA's allocation process for performance standard adjustments with renewable oxygenate use, EPA is concerned that the requirement that refiners track and report to EPA the volume of reformulated gasoline and its RVP (VOC performance) on a covered area specific basis may represent a significant additional burden. In the interest of minimizing the cost and burden on refiners of the reformulated gasoline program, EPA would prefer to avoid, if possible, requiring refiners to track their fuel from the refinery to its ultimate use. However, to ensure the environmental neutrality of the RVP (VOC) trading program knowledge of the actual distribution to each area is needed. EPA requests comment on any methods other than tracking of fuel by which the environmental neutrality of the program can be ensured in all covered areas and on whether the cost of tracking would be greater than any economic incentive created by the flexibility of the trading program. 4. Definition of Renewable Oxygenate Use While the performance standards described above are adjustable based upon the amount of renewable oxygenates used in the reformulated gasoline produced, some constraints need to be placed on what constitutes the use of renewable oxygenates if both the environmental and other policy benefits of today's proposal are to be achieved. EPA proposes that renewable oxygenate be defined for the purposes of this program as those oxygenates that are derived from renewable resources such as agricultural products and waste products such as sewage sludge or waste cellulose and does not include oxygenates produced from non-renewable resources such as petroleum, natural gas, coal, or peat. Furthermore, since some oxygenates such as ethers are currently made from both renewable and non-renewable resources, EPA proposes that only the portion of oxygenates derived from renewable resources be considered as renewable oxygenate. At the present time, the only oxygenate EPA is aware of that is produced in significant quantities that meets the definition of renewable oxygenate is ethanol. However, the definition is not meant to be restricted to ethanol, and is intended to encompass numerous other oxygenates, particularly the ethanol portion of ethers such as ETBE and ETAE which in the future may be produced in significant quantities. EPA further proposes that the use of renewable oxygenates to meet the 30 percent market share described above for northern areas and 20 percent for southern areas be based upon an assumed base oxygen content in the reformulated gasoline of 2.7 wt%. A 30 percent market share would, thus, represent 0.81 wt% oxygen in the form of renewable oxygenates in a refiner's reformulated gasoline production in the North, and 0.54 wt% oxygen in the South. In order to allow the greatest degree of flexibility for refiners, EPA proposes that the use of any renewable oxygenate in any concentration be prorated on the basis of weight percent oxygen for credit toward the 0.81 wt% oxygen base. The reformulated gasoline provisions require only 2.0 wt% oxygen in reformulated gasoline, but a base oxygen content of 2.7 wt% would better advance the energy related policies of today's proposal, and is still consistent with the maximum oxygen content for which no NOX emission increase is assumed under the simple model in section D below (as well as under both the negotiated agreement and SNPRM). Going beyond 2.7 wt% to as high as 3.5 wt% would provide a disincentive for the use of renewable oxygenates other than ethanol since they are currently limited to a maximum of 2.7 wt% and, thus, would be required to be used in more than 30 percent of the market to achieve the same level of renewable oxygenate use. Furthermore, setting a base oxygen content greater than 2.7 wt% oxygen would require the use of the complex model pursuant to the provisions of section D. This would represent a disincentive for certain refiners to use renewable oxygenates and, thus could potentially result in less renewable oxygenate use than setting the baseline at 2.7 wt%. EPA invites comment on setting the baseline oxygenate content at 2.7 wt% and on what level would best ensure that the energy related goals of today's proposal are achieved in a manner consistent with section 211(k) of the CAA. If oxygenates are used that are derived from both renewable and nonrenewable resources, EPA proposes as one option that such oxygenates be prorated in direct proportion to the amount of their oxygen content derived from renewable resources (e.g., ETBE derived from ethanol which is produced from biomass would achieve the same credit as the amount of ethanol used to make it). Not all renewable oxygenates, however, are identical with respect to their environmental benefits. One significant difference between oxygenates is their potential to cause vapor pressure increases when mixed with other fuels in the distribution system and vehicle fuel tanks, a phenomenon referred to as commingling. ETBE does not exhibit a commingling effect, nor does it produce the dramatic rise in blend vapor pressure above 100 deg.F that ethanol does. As a result, from an environmental standpoint, the use of ethanol in the form of ETBE is preferable to direct ethanol blending. The value added per gallon of ethanol from the RVP incentive proposed in this section would be roughly 13.7 cents, assuming an RVP control cost of one cent per RVP-gallon. This economic incentive would be constant regardless of the concentration of ethanol added or the form (splash blended ethanol or ETBE) since it is based on RVP. However, independent from today's proposal there is an inherent added incentive for blending ethanol at 3.5 wt% instead of 2.7 wt% oxygen since the former causes less of an RVP increase per gallon of ethanol used (ethanol blended to both 2.7 wt% and 3.5 wt% oxygen cause roughly a 1.0 psi RVP increase). Because of this inherent incentive EPA expects most ethanol blends would be at 3.5 wt% oxygen through the use of the complex model and oxygen credit trading. As a result, ETBE would always be forced to compete with ethanol at 3.5 wt% oxygen instead of ethanol at 2.7 wt% oxygen where ETBE would be at less of a disadvantage. In order to offset this disadvantage, EPA is also requesting comment on an option that renewable oxygenates such as ETBE that do not exhibit commingling effects be given the same credit at 2.0 wt% oxygen that is given to ethanol (or other renewable oxygenates which exhibit commingling effects) at 2.7 wt% oxygen. ETBE use in 30 percent of the market would, under this option, represent 0.60 wt% oxygen in the form of renewable oxygenates in a refiner's reformulated gasoline production. This approach would provide ETBE blends with an additional incentive of roughly the same magnitude as exists for blending ethanol at 3.5 wt% oxygen relative to 2.7 wt% oxygen. Commingling itself is an issue which is not unique to this proposal. It was not addressed in the context of the SNPRM but, nevertheless, represents an environmental impact whenever oxygenates such as ethanol which exhibit commingling effects are blended into gasoline. Today's proposal is not expected to dramatically change the magnitude of any commingling effects inherent under the SNPRM. While today's proposal may increase the market share of ethanol, it also contains incentives for ETBE and other ethers which do not exhibit commingling effects. As such there could be no increase or even a decrease in the amount of ethanol used to meet the reformulated gasoline program requirements resulting from today's proposal. Furthermore, even if there were an increase in the use of ethanol blends, today's proposal allows for its use to be shifted to those areas where it is most economical. As a result, it is possible that some areas would have virtually 100 percent ethanol blended reformulated gasoline while others may have virtually no ethanol blended reformulated gasoline. Such a scenario would reduce concerns over commingling. Finally, even if both ethanol use and commingling do rise, the increase may be small. At present EPA has not been able to adequately quantify the commingling effects resulting from the use of ethanol blends, but requests comment on how significant such effects may be both under the provisions of the SNPRM and under the provisions of today's proposal. C. NOX Neutrality A second element of President Bush's announcement was for EPA to establish a NOX performance standard which was neutral with respect to the type of oxygenate being used. Under the simple model proposed in the SNPRM, and in keeping with the agreement reached through regulatory negotiation, EPA proposed that "MTBE in concentrations up to 2.7 weight percent oxygen and other oxygenates in concentrations up to 2.1 weight percent oxygen be assumed not to increase NOX emissions." While "based on data available during the regulatory negotiations, it appeared that fuel oxygen content and the type of oxygenate used may have an impact on NOX emissions," the data did "not allow for quantifying relationships between oxygenate type and concentration and NOX emissions." Thus, based on the available information, caps were placed on oxygen content in order to permit the sale of reformulated gasoline and limit any potential NOX emission increase until the complex model could be developed and implemented. Since the time the SNPRM was developed a great deal of additional data has become available which has gone into the development of the complex model (to be discussed later). Under the complex model, all oxygenates are treated the same, i.e., based on oxygen content. There does not appear to be a significant difference between various oxygenates in their NOX forming tendency. Also, under the complex model, other fuel changes that accompany the blending of oxygenates into the fuel (reductions in the concentrations of aromatics, olefins, sulfur, etc.) have been found to offset the NOX increases caused by increasing the oxygen concentration. The results, which are described in the DRIA, show that on average 4-5 percent reductions in NOX emissions are expected. Thus, the complex model is directionally consistent with the data used for the simple model. However, refiners would not be prohibited from adjusting their base gasoline in anticipation of dilution so that a NOX increase could occur. EPA does not anticipate that refiners will have an economic incentive to do this. At most, as described in section V.C. below, a one percent increase in NOX emissions from certain fuels might occur. Given the much larger reductions projected on average, it is highly unlikely that a net NOX increase could result under the simple model with oxygenate up to 2.7 wt% oxygen. As a result, EPA proposes that in keeping with the provisions of the SNPRM and the more recent data, all oxygenates be assumed to yield a no NOX emission increase up to 2.7 wt% oxygen under the simple model. EPA requests comment on this view and the resulting proposal, and on the legal adequacy of relying on this expectation that no NOX increase will occur. EPA also requests comment on whether any NOX increase associated with oxygenates at or below 2.7 weight percent oxygen (should one be found to occur) may be considered de minimis and on whether EPA has legal authority to disregard such a NOX increase as de minimis. The one option available to prevent a NOX increase from every batch of fuel would be to require refiners to use the complex model in 1995 (under the early use provisions described in section E. below) solely to certify that their fuels did not increase NOX emissions. Given the effect of oxygenate on the other parameters of gasoline and the lack of economic incentive to worsen fuel quality via other means, this option should not pose a leadtime concern. EPA requests comment on this additional option as well. D. Toxics Neutrality In addition to NOX neutrality, President Bush also requested that the toxics standards be set in such a manner as to be neutral with respect to the type of oxygenate used. EPA based the toxics performance standard in the SNPRM on the use of MTBE as the oxygenate in the fuel on the assumption that this would be the predominant oxygenate used to produce reformulated gasoline. Due to a characteristic of MTBE unique among oxygenates, evaporative benzene emissions are suppressed, and MTBE achieves a greater toxics performance on a weight percent oxygen basis than the other oxygenates. As a result, the toxics performance standard was proposed in the SNPRM in a manner which disadvantaged the other oxygenates by making it more costly for them to be utilized in reformulated gasoline. As is discussed in section VI, control of toxics emissions in reformulated gasoline is a very costly means of reducing cancer incidences. For this reason, basing the toxics standard on the sole oxygenate which achieves the greatest toxics control is not appropriate. Thus, EPA proposes that the toxics emission performance standard under the simple model be based upon the non-MTBE oxygenates. Applying this approach, the toxics performance standard required by the act is the greater of a 15 percent reduction or the performance of the formula fuel containing 2.0 wt% oxygen in the form of a non-MTBE oxygenate. Since a formula fuel containing non-MTBE oxygenates does not achieve the 15 percent minimum, EPA proposes that the air toxics performance standard under the simple model be 15 percent. In accordance with the provisions proposed in the SNPRM, EPA also proposes that the year-round averaging standard be 16.5 percent. E. Unrestricted Early Use of the Complex Model The CAA requires all refiners to reduce the VOC and air toxics emission forming tendency of their fuels by a minimum 15 percent below that of the 1990 average in-use fuel as defined in section 211(k)(10)(B) (referred to here as the CAA baseline). Due to the limited data available at the time of the regulatory negotiation (and SNPRM development), however, few of the effects of fuel parameters on emissions could be quantified and placed in the simple model even though they were known qualitatively to exist. As a result, EPA proposed in the SNPRM that all refiners hold such fuel parameters (sulfur, olefins, and T90) at or below their 1990 baseline levels. Because different refiners have significantly different 1990 baselines of sulfur, olefins, and T90, this proposal required some refiners to produce reformulated gasoline which was significantly less polluting than that produced by other refiners. In order to maintain the same overall level of environmental control if refiners opted to use the complex model prior to its required use in 1997, EPA further proposed in the SNPRM that refiners who opted to use the complex model early would be required to maintain whatever emission performance they would otherwise have achieved using the simple model through the use of their own 1990 baselines. This was viewed by some refiners, however, as an impediment to early use of the complex model since they could be required to achieve greater environmental control in 1995-96 than in 1997. In response, EPA also proposed that since in the southern areas covered by a reformulated gasoline program, refiners would be required under the proposal to achieve greater than the minimum VOC control required by the Act, they be permitted the option to use the CAA baseline instead of their own individual baselines. In northern areas covered by the reformulated gasoline program refiners would still be required to use their own 1990 baselines. In President Bush's announcement, he requested EPA to provide refiners with the broadest range of options possible in "meeting the pollution reduction requirements of the Act," and called for the "unrestricted" early use of the complex model. EPA today proposes that all refiners be permitted to utilize the complex model before its required use in 1997 to certify reformulated gasoline to be sold in all parts of the country with their performance measured against either their own 1990 baseline or the CAA baseline. Not only does this proposal grant refiners greater flexibility, but it also makes it easier for refiners to use renewable oxygenates such as ethanol in producing their fuel. III. Complex Model A. Baseline Emissions Under the Complex Model In the SNPRM, EPA proposed that the emission baseline for the simple model would be based on light-duty vehicle emissions using a July 11, 1991 version of MOBILE4.1 with an assumption for enhanced I/M that is now more representative of EPA's requirements for basic I/M (57 FR 52950). This proposal was based on the best information available at the time. Since that time, a new version of the MOBILE model, MOBILE5.0, has been developed and both the basic and enhanced I/M requirements have been promulgated. These changes affect the baseline calculations and raise the issue as to whether and when to require the revised baseline. The choice of MOBILE model and I/M program determines the level of baseline emissions established for baseline vehicles. The primary impact of this choice on baseline emissions is through changes in the ratio of baseline exhaust to non-exhaust VOC emissions. Since various reformulations will reduce exhaust and non-exhaust VOC emissions to varying degrees, changing the ratio of baseline exhaust and non-exhaust VOC emissions can affect the overall VOC emission reduction expected from a specific fuel reformulation. NOX emission performance is unaffected since NOX is only emitted via the exhaust. Toxics performance is only slightly affected (much less than VOC), since non-exhaust toxics comprise only a small fraction of total toxics emissions. As described in section V.C., the proposed complex model projects that typical simple model fuels will substantially reduce both exhaust and nonexhaust VOC emissions. The result is that the overall effectiveness of various Phase I reformulations is expected to be relatively insensitive to the baseline exhaust to non-exhaust VOC ratio. Also, given the no NOX increase requirement and the effect of oxygenate on other fuel parameters, EPA expects fuels produced under the complex model to produce even greater reductions in exhaust VOC than those shown in section V.C., making their performance even less sensitive to the choice of MOBILE model. Thus, EPA is confident that fuels certified using a MOBILE4.1 baseline would meet at least the minimum 15 percent VOC reduction requirement even with a MOBILE5.0 baseline. Thus, there would appear to be little or no air quality or other impact of delaying implementation of a revised baseline based on MOBILE5.0. The assumptions made in the SNPRM with respect to the appropriate enhanced I/M program to assume for Phase I may also be reasonable through the end Phase I. While enhanced I/M is required in many ozone nonattainment areas starting January 1, 1995, its full implementation may be later than this in several important ways including a phase-in of emission cut points and a phase-in of the percentage of the fleet being tested. It is therefore possible that areas will not realize the full in use benefits of those programs until near the end of the Phase I period. The complex model, however, will be available as an option as early as 1995, and as proposed would not be mandated for use in certifying VOC controlled fuel until the 1997 ozone season at the earliest. At the same time, changing the baseline in 1997 could potentially change the overall stringency of the VOC performance standard for some refiners and require them to invest additional capital to further control RVP or other fuel parameters. The statutory structure established by Congress creates a two-phase program (beginning in 1995 and 2000, respectively), and maintaining a consistent level of emission performance through the first five years of the program is consistent with this structure. Therefore, EPA proposes to retain the baseline proposed in the SNPRM for use with the complex model prior to 2000. This is based on the lack of adverse air quality impact, EPA's confidence that the required statutory minimum reductions are not at risk, and the potential disruption in the latter years of Phase I for at least some refiners. EPA requests comment on all of these points, including comment on utilizing an emissions baseline based on either the official version of MOBILE4.1 or MOBILE5.0 beginning in 1997. EPA also requests comment on the appropriateness of the I/M program assumptions discussed above and the appropriateness of other I/M program assumptions. The Agency specifically requests comment from the States on the type of enhanced I/M program they plan on implementing (e.g., annual, biennial, emission cut-points, etc.), their projected schedules for implementation, and the timetable for any expected phase-ins. EPA invites comment on the degree of change in fuel formulation which might occur and their environmental, energy, cost and leadtime consequences. Changes in the baseline modeling assumptions, if appropriate, would also lead to corresponding changes in the Phase I complex model standards. With the onset of the Phase II performance standards in 2000, the overall stringency of the standards is already changing and a new baseline based on MOBILE5.0 will not, by itself, be the cause of new investment by refiners. Also, by this time, enhanced I/M programs should be fully operational in nearly all reformulated gasoline areas. Therefore, for Phase II EPA proposes that the baseline for the complex model be based on light-duty vehicle and light-duty truck emissions using MOBILE5.0 with enhanced I/M. B. Complex Model for Exhaust Emissions 1. Data Sources The relationship between fuel properties and exhaust emissions is complex and the theory behind such relationships continues to be developed. As a result, EPA has asked industry, state regulatory agencies, and other organizations with relevant test data to make their data available to the Agency to ensure that this rule is based on as much relevant information as possible. The complex model described in the following section is based on data generated from a number of exhaust emissions testing programs. These programs, their design intent, and their limitations are discussed in the Draft RIA. Data from these programs were excluded from EPA's analysis if the data were not based on a valid FTP measurement cycle, if the vehicle in question did not employ 1990-equivalent emission control technology, or if nonmethane hydrocarbon measurements were not taken. The Agency believes its analysis considered all available, valid, and relevant data on the exhaust emissions effect of fuel modifications when used in 1990 model year and equivalent vehicles. Since these test programs generally involved different vehicles, different fuels, and in some cases different test procedures, the analysis required to determine the relationship between fuel properties and emissions is complex. However, EPA believes that the model proposed below appropriately considers and addresses these complexities. 2. Analysis Method Exhaust emissions are affected by both vehicle and fuel characteristics, which makes identification of fuel effects on emissions difficult. EPA has considered and in some cases tested a number of analysis methods, many of which were presented during a series of public workshops held by EPA over the past year. Information regarding these methods can be found in Docket A-92- 12. The approach chosen by EPA to analyze the data described in section III.B.1 is summarized below and is discussed more fully in the Draft RIA. EPA utilized statistical analysis techniques to determine the effects of fuel modifications on exhaust emissions of VOC, NOX, and toxics. At a series of five public workshops held over the past year, the Agency has presented its data sources, proposed analysis methods, and preliminary emissions models for public review and comment. The Agency has also requested other organizations to share their data, analysis expertise, and emissions models at these workshops. The Agency believes that today's proposal appropriately incorporates the comments and suggestions regarding the analysis process received at the workshops and during the course of the past year from industry, state and federal government authorities, and other interested parties. The resulting exhaust emission models have been validated through confirmatory testing and analysis. EPA's analysis separated exhaust emissions into fuel components and vehicle components. In all test programs analyzed by EPA, the single most significant determinant of the level of emissions from a given vehicle on a given fuel was the vehicle itself. Fuel properties exert a much smaller influence on exhaust emissions than do vehicle characteristics such as emission control system technology, vehicle mileage, catalyst efficiency, oxygen sensor efficiency, engine size, engine design, vehicle size, fuel efficiency, and vehicle maintenance. To identify the effects of fuel property modifications on emissions, EPA found it necessary to identify the effect of each vehicle on emissions and separate this effect from the fuel effects. The fuel components of exhaust emissions were separated into two main categories. The first category consisted of the effects of individual fuel parameters. For example, the effect of sulfur on hydrocarbon emissions was best modeled by a relationship containing a linear sulfur term and a second- order sulfur term (a term of the form c1S /2/, where c1 is a constant and S is the sulfur level). The second category of fuel terms consisted of interactive effects between two fuel parameters. EPA's analysis of NOX emissions, for example, found that oxygen's effect on NOX is related to the aromatics level of the fuel. This effect cannot be represented as an oxygen or aromatics effect alone but must be represented as an interactive term of the form c2XA, where c2 is a constant, X is the oxygen level, and A is the aromatics level. The effect of each fuel term was determined separately for each technology group within each emitter class for NOX and VOC emissions. Technology groups and emitter classes, their importance in determining the effects of fuel modifications on emissions, and their definitions are discussed more fully in the Draft RIA and simple model SNPRM. The linear terms for all fuel parameters were retained, while only those second-order and interactive terms that were statistically significant at the 85 percent level were retained. Retention of non-significant linear terms reflects the general consensus of public workshop participants regarding appropriate analysis and modeling methods and a desire to keep this potentially very complex model as simple, yet as accurate, as possible. EPA requests comment on the appropriateness of the significance level used to exclude terms from the model and the inclusion of non-significant linear terms. Because vehicles can have different emission control systems, the Agency anticipated that fuel modifications would have different emission effects on different types of cars. EPA has found it necessary to develop terms unique to technology groups to assure the predictive power of the complex model. EPA also has found that vehicles classified as "high emitters" respond very differently to fuel modifications than do normal emitting vehicles. Hence the terms described above were defined and determined separately for each technology group within each emitter class for VOC and NOX emissions. EPA chose a different approach when modeling emissions of exhaust toxics. Many of the studies available to EPA did not measure exhaust toxics, so the data on exhaust toxics emissions is less complete than the data on exhaust VOC and NOX emissions. This problem is particularly acute for high emitters. In addition, baseline estimates of toxics emissions are not available from the MOBILE models, so a different source of baseline toxics estimates had to be found. To overcome the limited amount of data on high emitters, the Agency chose to combine normal and high emitters into a single emitter class while retaining the technology group distinction. To overcome the absence of baseline toxics data from the MOBILE models, the resulting toxics model was then adjusted to compensate for differences between the average VOC exhaust emission levels for those vehicles for which toxics data were available and the 1990 model year, in-use emission levels projected by the MOBILE models. This process is discussed in more detail in the DRIA. Since the proposed models for normal and high emitter VOC emissions differ substantially, and since the compounds classified as toxic in section 211(k) are VOCs, the Agency is concerned that this approach based on absolute mass emissions may not accurately represent the effects of fuel changes on toxics emissions from high emitters. The Agency has considered modeling toxics emissions as mass fractions of total VOC emissions (as was proposed in the simple model NPRM and SNPRM) but has found models of toxics in terms of absolute mass to have very high predictive power. However, this conclusion could be the result of limited high emitter toxics data. In addition, the mass fraction approach can be more readily extrapolated to winter conditions than can the absolute mass emissions approach. EPA requests comment on the appropriateness of the mass fraction and absolute mass emissions approaches. EPA focused its modeling efforts on those fuel parameters thought to affect vehicle emissions based on available test results. Oxygen, RVP, aromatics, olefins, T90, sulfur, and benzene were identified as relevant fuel parameters in the negotiated agreement. T50 was identified as a potentially significant parameter by a number of testing programs and at several public workshops. EPA anticipates that further refinements to the model will be made prior to finalization of this rule. First, the Agency expects to convert its fuel distillation curve data to a percent evaporated basis, thereby replacing T50 and T90 with terms such as E200 and E300 (the volume percentage of the fuel evaporated at 200 deg.F and 300 deg.F, respectively). The percent evaporated terms are thought to be more directly relevant to the generation of exhaust emissions than the distillation parameters. In addition, the percent evaporated values for a mixture of fuels can be calculated directly from the percent evaporated values of each component of the mixture, while the distillation values for a mixture must be measured separately. This characteristic has important implications for certification and enforcement. Second, the Agency will attempt to determine the influence of the type of aromatic compounds in fuels, specifically heavy aromatics, on exhaust emissions, since preliminary test data suggests that both the level and type of aromatic compounds may affect emissions. Third, EPA expects to include additional emission testing data as they become available, and such data may affect the results of the modeling process discussed above. Specific test programs expected to provide additional data include the second half of Auto/ Oil's T50/T90 study and EPA's Phase II reformulated gasoline test program. EPA's Phase II test program data is available and has been used to confirm the validity of the model presented in Sec. 80.45. After using either the EPA Phase II or Auto/Oil T50/T90 studies to confirm the revised model, EPA expects to include the data used in this confirmation in its database to develop a final model. Fourth, EPA anticipates changing the confidence level required to allow additional terms into the model to 90 percent, in keeping with past EPA practice. Finally, EPA expects to complete its analysis of the importance of technology group and emitter class distinctions prior to finalization of this rule. Preliminary analyses confirm that the original classification scheme does capture significant differences in the emissions performance of fuels, as can be seen by examining the differences in the exhaust emission equations for specific technology group/emitter class categories as presented in the Draft RIA. These distinctions may be changed if further analysis indicates that changes are necessary to adequately estimate the effects of fuel modifications on vehicle emissions. At the present time, EPA is considering reducing the number of technology group categories for normal emitters by eliminating technology groups which contribute a very small proportion of total mobile source emissions or combining such groups with other, similar technology groups. EPA is also considering subdividing high emitters by their emission characteristics, particularly their exhaust hydrocarbon to NOX ratio, rather than by vehicle technology. This modification reflects EPA's belief, supported by preliminary field information, that one or more emission control components on high emitters tend to be malfunctioning, which renders a classification scheme based on vehicle equipment potentially suspect. EPA requests comment on these potential changes in its technology group definitions and on the other potential changes in its analysis approach discussed above. EPA will place its revised version of the complex model for exhaust emissions (referred to as "Iteration 4" of the model), which incorporates the changes discussed above, and its supporting materials in the docket for review and comment as soon as it is completed. The American Petroleum Institute (API) has also developed its own version of a complex model, which has been placed in the docket for this rulemaking. EPA requests comment on the various aspects of API's statistical and engineering approach and on its overall suitability for use in EPA's complex exhaust emission model. 3. Complex Exhaust Emissions Model Equations As was discussed in the SNPRM, EPA believes that the exhaust emission effects of fuel modifications are different when applied to vehicles with different emission control and fuel distribution technologies. In addition, EPA believes that the emissions effects of fuel modifications also depends on the level of emissions from the vehicle. During development of the complex model, preliminary investigation of these assumptions have supported the approach described above. The complex model incorporates these effects on an emissions-weighted basis, i.e., fuel effects on each technology group/emitter class category were weighted according to the contribution to in-use emissions of 1990 vehicles from that category. To determine the contribution of each technology group to in-use emissions, EPA calculated the fraction of 1990 model year sales of light duty vehicles and trucks from each technology group. EPA used non-California sales figures for this analysis since vehicles certified as "California-only" vehicles are required to meet a different set of emission standards than other cars sold in the U.S., and since California will have established its own reformulated gasoline program by 1996 that is expected to exceed the stringency of the federal program. EPA requests comment on the exclusion of California vehicles from its calculation of 1990 sales fractions. EPA then adjusted this fraction to reflect the differences in average emissions among technology groups where such information was available. This adjusted fraction is represented by the "W" constants in the model proposed in Sec. 80.45. Because the equations are in the form of exponential equations, it is necessary to normalize each equation to a known emission value on a known fuel for each technology group. The resulting proportionality constant is represented by the "k" constants in the model proposed in Sec. 80.45. In effect, the "k" constant transforms a percentage increase or decrease into a gram per mile figure corresponding to that percentage change from baseline emissions. The rationale for and derivation of the "k" and "W" constants are discussed in more detail in the DRIA. Since the effects of fuel modifications on emissions differ across technology groups and emitter classes, EPA has developed separate models for each technology group-emitter class combination for which sufficient data were available. The emissions performance of a given fuel reformulation for a specific pollutant can be determined by taking the following steps. First, each technology group-emitter class model for the pollutant in question would be evaluated for the specific fuel formulation being evaluated. Second, the resulting predicted emissions for each technology group-emitter class combination would be weighted by the estimated contribution of that combination to total miles traveled and these weighted emissions would be added together. Third, the total predicted emissions would be compared to the baseline emissions level for the pollutant in question. The emissions performance of the fuel reformulation would be defined as the percentage change in total emissions from baseline emission levels, which are discussed below in section B. The specific equations that comprise the complex model can be found in Sec. 80.45 of the regulations proposed for this rule. For the most part, their derivation is straightforward and can be found in the Draft RIA. In some cases, available data were insufficient to develop models for every technology group-emitter class category. In such cases, the weighting factors for the categories which could be modeled were renormalized to adjust for the absence of data. This process is discussed more fully in the Draft RIA. The Draft RIA and the proposed regulations also discusses the range of parameter values for which these equations are valid. As discussed in the SNPRM, refiners are required to submit data to augment the model if they wish to certify fuels with properties that fall outside this range. EPA requests comment on deviations between the model's predictions and the data used to develop the model, on the model equations themselves, on their derivation, and on the ranges over which these equations are considered valid. The model presented in Sec. 80.45 is based on emissions data generated under summer conditions. While the VOC performance standard for reformulated fuels applies only in the summer, the toxics and NOX requirements apply year- round. Hence, the complex model must cover typical winter, as well as summer, fuels. However, the model presented in Sec. 80.45 is based on data generated under summer conditions and cannot be used directly for fuels with typical winter RVP levels. EPA proposes to use a modified version of the complex model as presented in Sec. 80.45 to evaluate winter fuels. This modified version would assign an RVP value equal to that of summer baseline gasoline, i.e., 8.7 psi, for all non-VOC controlled fuels when evaluating the toxics and NOX emissions performance of such fuels (including winter baseline fuels). RVP's impact on canister loading and subsequent purging is thought to be the primary cause of its effects on exhaust emissions. Since data do not exist on the effects of winter fuels on canister loading under winter conditions, the Agency believes it is not appropriate to attempt to model the effects of winter RVP levels on emissions. Fixing the RVP value of winter fuels at baseline levels for modeling purposes effectively removes the influence of RVP on winter exhaust toxics and NOX emissions. EPA requests comment on this approach, particularly on the ability of the non-RVP factors in the complex model to accurately predict winter toxic and NOX emissions. C. Complex Model for Nonexhaust Emissions The model for nonexhaust emissions is based on data generated by EPA to develop its MOBILE emission inventory models. The Agency believes that the relationship between fuel properties and nonexhaust emissions is less complex and better understood than for exhaust emissions. Data from EPA's ongoing evaporative emissions testing has been used to develop MOBILE4.1 and MOBILE5.0. EPA believes this data to be sufficient to model the relationship between fuel properties and nonexhaust emissions. Additional information about MOBILE4.1 and MOBILE5.0 can be found in Dockets A-91-02 and A-92-12. Non-exhaust emissions are less affected by vehicle design and are influenced by fewer fuel characteristics than are exhaust emissions. In addition, the theoretical principles involved in evaporative emissions are better understood, and evaporative emission control technologies are more consistent across vehicles than are exhaust emission controls. EPA is in the process of developing an evaporative emissions model based on fundamental theoretical principles and actual test data that are expected to be more accurate and more widely applicable to oxygenated fuels than MOBILE4.1. A preliminary version of this model was discussed at a public workshop held on August 25, 1992, and materials related to this model have been placed in the docket for this rulemaking. At this time, however, the enhanced evaporative emissions model is not complete and is not expected to be completed prior to publication of the final rule. As a result, the Agency is proposing correlations relating RVP to nonexhaust emissions derived from the MOBILE4.1 emissions model issued November 1991. To develop the correlations shown below, MOBILE4.1 was used with temperatures of 69 to 94 degrees Fahrenheit for Class B areas and 72 to 92 degrees Fahrenheit for Class C areas, assuming the existence of an enhanced inspection and maintenance program. In addition, the presence of Stage II evaporative emissions recovery systems with an overall vapor recovery efficiency of 86 percent was assumed (as discussed in the SNPRM and NPRM). Should the enhanced evaporative emissions model not be developed in time for inclusion in the final rulemaking, EPA proposes to utilize the July 11, 1991 version of MOBILE4.1 to determine the effect of RVP on nonexhaust VOC emissions under the Phase I performance standards (i.e., 1995-1999) and to utilize MOBILE5.0 to do the same under the Phase II performance standards (i.e., 2000 and beyond). This would be in keeping with the proposed baseline emissions described in section A above and, again, minimize the changes refiners have to undergo in the transition from the simple to the complex model. The Agency requests comment on this proposed basis for determining non-exhaust VOC emissions, including comment on the appropriate MOBILE model and I/M assumption. As discussed in the SNPRM, the Agency's correlation between fuel benzene content and non-exhaust benzene emissions is based on General Motors' proprietary model of tank vapors, and its independent confirmation by data generated by EPA using a number of fuels. Both the derivation and verification of the non-exhaust benzene emissions model is discussed more fully in the Draft RIA. EPA requests comment on the use of the correlation between fuel benzene content and non-exhaust benzene emissions derived from the General Motors model and the sufficiency of EPA's independent confirmation of its results. As was also discussed in the SNPRM, EPA believes that nonexhaust VOC and toxics emissions (including running loss, hot soak, diurnal, and refueling emissions) depend primarily on the RVP of the fuel. The Agency proposes that the nonexhaust VOC and toxics equations found in Sec. 80.45 of the proposed regulations be used to determine nonexhaust VOC and toxics emissions, respectively. These equations relate VOC emissions to the RVP of the fuel. Nonexhaust benzene emissions are determined by the level of benzene in the fuel and the level of nonexhaust VOC emissions. EPA requests comment on this approach and on the equations proposed in Sec. 80.45, the derivation of which is discussed more fully in the Draft RIA. IV. Vehicle Testing As was discussed in the simple model SNPRM, the complex model may be augmented based on the results of vehicle testing. EPA considers the testing program discussed in the SNPRM to be appropriate in terms of the fuels, number of vehicles, and test procedures specified. However, the Agency now believes the analysis process described in the SNPRM to be inconsistent with the process used to develop the complex model and may be overly complex. As a result, EPA proposes that the data generated during vehicle testing be analyzed statistically as described in Sec. 80.48. This analysis would fit a stepwise regression model to the data that includes the following terms: A vehicle term, a linear term for the parameter being tested, and squared and interactive terms involving the parameter being tested if significant at the 90 percent confidence level. The analysis would treat normal emitters and high emitters as separate populations but would not subdivide the emitter categories into technology groups for analysis purposes, since the test fleet compositions specified in Sec. 80.60 assure that the normal and high emitter test fleets include technology groups that reflect their contribution to in- use emissions. In addition, the statistical power of the test program would be compromised by subdividing each emitter class test fleet into technology groups since the number of vehicles in each technology group would be small. EPA believes that this approach is more consistent with the process used to develop the complex model, is less susceptible to manipulation and gaming by testing organizations, and is capable of extracting the maximum amount of information from the test data. In today's proposal, EPA has deleted Sec. 80.57 and Sec. 80.58 from its proposed regulations. These sections specified procedures for determining the presence and significance of nonlinear and interactive emission effects of fuel parameters being tested. EPA believes the regression-based analysis approach described in Sec. 80.48 to be more accurate and less complex than the procedures described in Sec. 80.57 and Sec. 80.58 of the April 16, 1992 proposal. Since the regression-based approach explicitly determines nonlinear and interactive effects, the Agency considers the deleted sections to be unnecessary and burdensome. EPA requests comment on these proposed deletions and on the proposed modification to the testing proposal outlined in the SNPRM. V. Phase I Performance Standards Using the Complex Model A. Baseline Emission Estimates The following table summarizes the Agency's proposed estimates of VOC and NOX emissions from in-use, 1990 technology vehicles when fueled with Clean Air Act baseline gasoline. All VOC emission estimates are given in terms of non-methane, non-ethane hydrocarbon emissions. In today's notice, EPA further proposes to redefine "ozone forming VOCs" as used in section 211(k) to exclude ethane. This redefinition is consistent with the Agency's most recent definition of VOC as "any compound of carbon * * * which participates in atmospheric photochemical reactions * * * other than the following, which have been determined to have negligible photochemical reactivity: Methane, ethane, * * *." (See 40 CFR 51.100(s) February 3, 1992), which excludes both ethane and methane from its definition of ozone forming VOCs (57 FR 3941). Ethane comprises approximately three percent of total hydrocarbon emissions but, like methane, is far less reactive in the atmosphere than other hydrocarbons. Table V-1.--Baseline VOC and NOX Emissions (MOBILE4.1 with SNPRM I/M), g/mi Class Class B C Summer: Exhaust VOC emissions 0.446 0.446 Hot soak VOC emissions .265 .230 Diurnal VOC emissions .125 .109 Running loss VOC emissions .431 .390 Refueling VOC emissions .040 .040 Total summer VOC emissions 1.307 1.215 Winter exhaust VOC emissions .660 .660 NOX emissions .66 .66 The following table summarizes the Agency's estimates of toxics emissions from in-use, 1990 technology vehicles when fueled with Clean Air Act baseline gasoline. These estimates were derived from the complex emission models described in section III and were based on the July 11, 1991 pre-release version of MOBILE4.1 emission inventory model with the I/M program assumptions described in the SNPRM. As discussed in the SNPRM, separate estimates were made for nonexhaust toxics emissions from Class B and C areas and for summer and winter emissions. Table V-2.--Baseline Toxics Emissions (MOBILE4.1 with SNPRM I/M), mg/mi Winter Summer /1/ Class Class Class B C B & C Exhaust benzene emissions 25.37 25.37 22.04 Formaldehyde emissions 3.80 3.80 3.79 Acetaldehyde emissions 2.06 2.06 1.81 1,3-butadiene emissions 3.28 3.28 3.00 POM emissions 1.50 1.50 1.45 Diurnal benzene emissions 1.30 1.13 Hot soak benzene emissions 3.02 2.63 Running loss benzene emissions 4.92 4.46 Refueling benzene emissions 0.42 0.42 Total baseline toxics emissions 45.67 44.64 32.05 /1/ Based on EPA's proposed complex model evaluated for winter baseline gasoline and assuming that wintertime evaporative emissions are negligible, as is discussed in the SNPRM (57 FR 13416). B. Complex Model Performance of the "Formula Fuel" The proposed complex model equations predict that the formula fuel would produce a reduction in exhaust VOC emissions of 12.3 percent and no reduction in nonexhaust VOC emissions when measured relative to the Clean Air Act baseline fuel. The resulting overall VOC emissions reduction of the formula fuel is 6.1 percent in Class B areas and 6.5 percent in Class C areas. The proposed complex model further indicates that the annual average toxics performance of an MTBE-based formula fuel would be an 11.7 percent reduction in Class B areas and an 11.2 percent reduction in Class C areas. The annual average toxics performance of the ethanol-based formula fuel would be an 11.5 percent reduction in Class B areas and an 11.0 percent reduction in Class C areas. These are lower reductions than those discussed in the SNPRM. The current estimate are believed to be more accurate given they are based on the proposed complex model rather than on the simple model. The Agency estimates that formula fuels using other oxygenates, such as ETBE or TAME, would result in VOC and toxics performance similar to that of ethanol-based formula fuels. C. Complex Model Performance of Simple Model Fuels The simple model specifies a number of fuel parameters to assure satisfactory emissions performance under the simple model. Many of the complex model relationships between these fuel parameters and emissions performance are different than the corresponding relationships in the simple model. As a result, the complex model predicts different emission performance levels than does the simple model for a given fuel formulation. The following tables summarize the predicted simple model and complex model performance levels for four different fuels relative to two specified baseline fuels. The first baseline fuel is the CAA baseline fuel as specified in the Act. The second baseline fuel is identical to CAA baseline fuel except for its RVP, which is set at the level required by Phase II volatility controls in Class B areas. This second baseline fuel is referred to as "Class B baseline fuel." The four summer fuels described in Table V-3 would meet the proposed VOC requirements for Class C and Class B areas, respectively, under averaging standards, and assuming that winter fuels are reformulated to a similar extent, would meet the toxics requirements. Under the simple model, the toxic emission effects of different oxygenates vary, so two different fuels with aromatics levels sufficient to assure compliance with the simple model toxic emission performance standards are shown for each area. Tables V-4 and V-5 summarize the summer performance of the fuels described in Table V-3 relative to the CAA baseline fuel according to both the simple and complex emission models. Table V-6 summarizes the emissions performance of Fuels 3 and 4 relative the Class B baseline fuel using both models. Tables V-4, V-5, and V-6 use baseline emissions derived using the July 11, 1991 pre- release version of MOBILE4.1 with essentially a basic I/M program. Table V-3.--Sample Fuel Compositions Class C Class B Fuel Fuel Fuel Fuel 1 2 3 4 Fuel characteristics: Oxygenate MTBE EtOH MTBE EtOH Oxygen, weight percent 2.1 2.1 2.1 2.1 Aromatics, volume percent 26.3 24.3 26.2 23.9 Olefins, volume percent 9.2 9.2 9.2 9.2 Benzene, volume percent 0.95 0.95 0.95 0.95 Sulfur, parts per million weight 339 339 339 339 T50, degrees Fahrenheit 218 218 218 218 T90, degrees Fahrenheit 330 330 330 330 RVP, pounds per square inch 8.0 8.0 7.1 7.1 Table V-4.--Simple and Phase I Complex Model Summer Emissions Reductions of Fuels 1 and 2 in Class C Areas Relative to CAA Baseline Gasoline [In percent] Simple Complex model model reductions reductions Fuel Fuel Fuel Fuel 1 2 1 2 Exhaust VOC 9.9 9.9 13.0 15.4 Nonexhaust VOC 23.9 23.9 23.9 23.9 Total VOC 18.8 18.8 19.9 20.8 NOX 0.0 0.0 -0.8 0.1 Toxics 22.4 23.4 22.4 23.1 Table V-5.--Simple and Phase I Complex Model Summer Emissions Reductions of Fuels 3 and 4 in Class B Areas Relative to CAA Baseline Gasoline [In percent] Simple Complex model model reductions reductions Fuel Fuel Fuel Fuel 3 4 3 4 Exhaust VOC 9.9 9.9 14.3 17.0 Nonexhaust VOC 51.4 51.4 51.4 51.4 Total VOC 37.2 37.2 38.7 39.6 NOX 0.0 0.0 -2.5 -2.0 Toxics 25.7 27.1 26.7 27.9 Table V-6.--Simple and Phase I Complex Model Summer Emissions Reductions of Fuels 3 and 4 in Class B Areas Relative to Class B Baseline Gasoline [In percent] Simple Complex model model reductions reductions Fuel Fuel Fuel Fuel 3 4 3 4 Exhaust VOC 9.9 9.9 13.2 15.9 Nonexhaust VOC 29.8 29.8 29.8 29.8 Total VOC 21.3 21.3 22.7 23.8 NOX 0.0 0.0 -2.5 -2.0 Toxics 22.2 23.7 17.3 18.7 It should be noted that the above fuels show slight NOX emission increases when using the complex model. This occurs because Fuels 1-4 assume no effect of oxygenate on fuel sulfur, T50, and T90, and a minimal effect on fuel aromatics. As discussed in the Draft RIA, when oxygenates typical effect on these fuel parameters are taken into account, which EPA expects will happen on average under the simple model, these NOX emission increases disappear and NOX emissions should actually decrease slightly at both 2.1 and 2.7 weight percent oxygen, regardless of the oxygenate used. The same holds true for the results shown in Tables V-7 through V-10 below. For comparison, Tables V-7 through V-10 summarize the summer performance of the fuels described in Table V-3 relative to the CAA baseline fuel according to two alternative Phase I complex models. The complex model used to derive the results in Tables V-7 and V-8 utilized baseline emissions from MOBILE5.0, while the complex model used to derive the results in Tables V-9 and V-10 utilized baseline emissions from the official version of MOBILE4.1. Both alternative Phase I complex models assumed basic I/M. A discussion of both of these alternative models can be found in the Draft RIA. Table V-7.--Summer Emissions Reductions of Fuels 1 and 2 in Class C Areas Relative to CAA Baseline Gasoline Using a Complex Model Based on the Official Version of MOBILE5.0 [In percent] Complex model reductions Fuel Fuel 1 2 Exhaust VOC 13.1 15.4 Nonexhaust VOC 24.8 24.8 Total VOC 17.6 19.1 NOX -0.8 0.1 Toxics 23.2 23.8 Table V-8.--Summer Emissions Reductions of Fuels 3 and 4 in Class B Areas Relative to CAA Baseline Gasoline Using a Complex Model Based on the Official Version of MOBILE5.0 [In percent] Complex model reductions Fuel Fuel 3 4 Exhaust VOC 14.3 17.0 Nonexhaust VOC 49.2 49.2 Total VOC 28.9 30.5 NOX -2.5 -2.0 Toxics 27.4 28.5 Table V-9.--Summer Emissions Reductions of Fuels 1 and 2 in Class C Areas Relative to CAA Baseline Gasoline Using a Complex Model Based on the Official Version of MOBILE4.1 [In percent] Complex model reductions Fuel Fuel 1 2 Exhaust VOC 13.0 15.4 Nonexhaust VOC 25.8 25.8 Total VOC 21.3 22.1 NOX -0.8 0.1 Toxics 22.9 23.6 Table V-10.--Summer Emissions Reductions of Fuels 3 and 4 in Class B Areas Relative to CAA Baseline Gasoline Using a Complex Model Based on the Official Version of MOBILE4.1 [In percent] Complex model reductions Fuel Fuel 3 4 Exhaust VOC 14.3 17.0 Nonexhaust VOC 51.3 51.3 Total VOC 39.1 40.0 NOX -2.5 -2.0 Toxics 27.0 28.3 D. Phase I Complex Model Performance Standards 1. Phase I Complex Model Toxics Standards Under CAA section 211(k)(3), during 1995 through 1999, a reformulated gasoline's toxic emissions performance must meet or exceed that of a specified formula fuel or a 15 percent reduction from that of baseline gasoline, whichever is greater. As discussed in the SNPRM, the simple model predicts that the formula fuel achieves an average annual toxic emissions reduction of 16.5 percent, thus setting the toxic emissions standard for refiners certifying their reformulated gasolines under the simple model. The complex model for toxics emissions proposed above, however, predicts a different toxics emissions performance for the formula fuel. The complex model indicates that the annual average toxics performance of an MTBE-based formula fuel would be an 11.7 percent reduction in Class B areas and an 11.2 percent reduction in Class C areas. The annual average toxics performance of the ethanol-based formula fuel would be an 11.5 percent reduction in Class B areas and an 11.0 percent reduction in Class C areas. The Agency estimates that formula fuels using other oxygenates, such as ETBE or TAME, would result in VOC and toxics performance similar to that of ethanol-based formula fuels. Since these emission reductions are smaller than the 15 percent minimum reduction specified in the Act, EPA proposes that the toxics performance standard under the complex model be set 15 percent below the toxics emission level of baseline gasoline. EPA has considered basing its toxics performance standard on the performance of simple model fuels rather than the statutory requirements. The Agency is not proposing such an approach in today's notice for the reasons outlined in the SNPRM. In particular, the Agency is concerned that such standards would have an adverse impact on refiners whose 1990 baseline levels of sulfur, aromatics, T50, and T90 are higher than CAA baseline levels of these parameters. Since the simple model does not include these parameters while the complex model does, such refiners may face additional expenditures in 1997 to comply with the statutory requirements. A more stringent set of standards in 1997 would raise the cost of the program substantially. As discussed in the Draft RIA, EPA does not believe that more stringent controls are cost effective in the case of toxics. EPA requests comment on the proposed toxics standards. 2. Phase I Complex Model VOC Standards Both the simple model and the complex model predict that the VOC emission performance of the formula fuel would be less stringent than the 15 percent standard specified in the Act. As a result, Phase I VOC performance standards will not be determined by the performance of the formula fuel. As discussed in the NPRM and SNPRM, the Agency believes that Fuels 3 and 4 are producible, with Fuel 3 considered more representative of typical Class B simple model fuels. Since such fuels (which will be required in Class B under the simple model) would achieve VOC emissions reductions of 38.7 to 39.6 percent relative to the CAA baseline fuel under the complex model, EPA proposes that the Phase I performance standard for VOC emissions in Class B areas be set by the performance of Fuel 3, as predicted by the complex model. This performance standard amounts to a 38.7 percent reduction in VOC emissions relative to CAA baseline levels and a 22.7 percent reduction from the VOC emission level of Class B baseline gasoline. The above VOC reductions were based on fuels just complying with the 0.95 volume percent benzene standard (with averaging) and aromatic content adjusted accordingly. Simple model fuels with lower benzene levels could have higher levels of aromatics. Due to beneficial effect of reduced aromatics on VOC emissions, Fuels 3 and 4 described above produce slightly greater VOC reductions than simple model fuels with lower benzene and higher aromatic levels. For example, with benzene levels at 0.80 volume percent, the VOC reduction decreases to 37.5-38.4 percent relative to CAA baseline gasoline and 21.1-22.3 percent relative to Class B baseline gasoline (see DRIA). EPA requests comment on setting the Class B VOC performance standard based on the performance of simple model fuels. EPA also requests comment on utilizing Fuel 3 to set this performance and, if Fuel 3 was not used, which particular simple model fuel should be used to set the performance standard. EPA is also considering setting the Class C VOC standard on the performance of simple model fuels. As shown above, Fuels 1 and 2 achieve greater than the 15 percent minimum statutory requirement. However, with a lower benzene level of 0.8 volume percent is