------------------------------------------------------------------------------- On 04/16/92 [L-S document 421027, 57 FR 13416, 9123 lines] ENVIRONMENTAL PROTECTION AGENCY 40 CFR Part 80 [AMS-FRL-4120-7] Regulation of Fuels and Fuel Additives; Standards for Reformulated and Conventional Gasoline AGENCY: Environmental Protection Agency. ACTION: Supplemental notice of proposed rulemaking. ----------------------------------------------------------------------------- SUMMARY: This supplemental notice of proposed rulemaking (SNPRM) describes the standards and enforcement scheme for both reformulated gasoline and for conventional gasoline sold in other areas. It also includes specific proposals for the emission models to be used in gasoline certification and enforcement. The SNPRM reflects a consensus that was reached through regulatory negotiation regarding certain provisions of the reformulated gasoline program. The preamble reflects the basis and purpose of this proposed rulemaking. A copy of the proposed regulatory language discussed herein may be obtained from Public Docket No. A-91-02 or from the contacts in the ADDRESSES section and is deemed to be part of this document. DATES: The comment period on this supplemental notice will extend through June 1, 1992. If no hearing is held on May 18, 1992, if a hearing is held to allow interested parties to comment on any specific provisions contained herein which were not in the NPRM for this rule, published July 9, 1991 (56 FR 31176). The comment period for the NPRM is also extended until such date. EPA will conduct a public hearing on this supplemental notice of proposed rulemaking on May 18, 1992, in Washington, DC, if anyone requests the hearing by May 1, 1992. The contact person listed below may be called regarding whether a public hearing will be held. EPA will conduct a public workshop on April 27 and 28, 1992, at the Best Western Domino's Farms Hotel, 3600 Plymouth Road, Ann Arbor, Michigan 48105; telephone (313) 769-9800. Discussion on the 27th will begin at 1 pm and be devoted to the issue of whether or not and, if so, how carbon monoxide (CO) should be included into the definition of VOC (as discussed in section II.A.1. of this proposal). Discussion on the 28th will begin at 9 am and be devoted to the complex model. Additional information concerning the agenda for the workshop and its location may be obtained from the contact person listed below, or from Michael Sklar at (313) 741-7817. ADDRESSES: Materials relevant to this SNPRM, including the regulatory language, are contained in Public Docket No. A-91-02, located at room M-1500, Waterside Mall (ground floor), U.S. Environmental Protection Agency, 401 M Street SW., Washington, DC 20460. 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. FOR FURTHER INFORMATION CONTACT: Joanne I. Goldhand, U.S. EPA (SDSB-12), Emission Control Technology Division, 2565 Plymouth Road, Ann Arbor, MI 48105, Telephone: (313) 668-4504. TO REQUEST COPIES OF THIS NOTICE CONTACT: Marie Tolonen, U.S. EPA (SDSB-12), Emission Control Technology Division, 2565 Plymouth Road, Ann Arbor, MI 48105, Telephone: (313) 668-4295. SUPPLEMENTARY INFORMATION: 1. Background This notice supplements the proposal for the reformulated gasoline program which was originally published July 9, 1991 (56 FR 31176) (hereafter, the NPRM). As did the NPRM, this notice describes the provisions of both a program to require the sale of gasoline which reduces emissions of toxics and ozone-forming volatile organic compounds (VOCs) in certain nonattainment areas and a program to prohibit the gasoline sold in the rest of the country from becoming more polluting. Since the NPRM was published, agreement has been reached through the regulatory negotiation process on an outline of these programs. This supplemental notice proposes adoption of the provisions of that agreement as well as detailed provisions not specifically covered by the agreement. This section will describe the history of EPA's efforts to develop a reformulated gasoline program and especially the events which have occurred since the NPRM was published. That notice contains a more detailed discussion of the early development of the program and further information regarding portions of the program described today which were first proposed at that time. The sections which follow discuss the methods for reformulated gasoline certification (sections II through IV) and enforcement (sections V through XI), anti-dumping requirements (sections XII and XIII), compliance audits (section XIV), federal preemption (section XV), the economic and environmental impacts of the program (section XVI) and finally certain statutory requirements (sections XVII through XXII). As described further in the NPRM, this rule has been developed through a process known as negotiated rulemaking as provided under the Negotiated Rulemaking Act of 1990, Public Law 101-648. That process involves creating an advisory committee under the Federal Advisory Committee Act 1 consisting of representatives of the groups which are likely to be substantially affected by the rule and the federal agency responsible for the rule. (See the NPRM for the members of the negotiating committee and a discussion of the process for selecting them.) In a negotiated rulemaking, such a committee meets to develop a proposed rule which will be acceptable to all parties. If consensus is reached on a proposed rule, it is published as an NPRM. The committee members and the entities they represent agree to support the proposal and not to seek judicial review of the final rule if it has the same substance and effect as the consensus proposal. NOTE 1 5 U.S.C. App. 1, et seq. In this case, EPA published an NPRM while the advisory committee was still conducting negotiations. The Agency believed that although consensus of the members on an acceptable rule was possible, an NPRM was required at that time if the rule was to be completed by the statutory deadline. The notice which was published described the outline of the reformulated gasoline program and options that committee members were considering. The negotiations continued after the NPRM was published and culminated in an Agreement in Principle which each of the regulatory negotiation committee members signed on August 16, 1991. The agreement stated the members' concurrence on an outline of the underlying principles of the reformulated gasoline and anti-dumping programs. The agreement, outline and several letters between EPA and the participants which further clarify the meaning of the outline are included in the docket for this rulemaking as items III-A-7 through 24. Generally, the agreed upon reformulated gasoline program would provide refiners with two modeling options and a testing option for determining whether fuels sold in 1995 and 1996 meet the reformulated gasoline requirements. The simpler of the modeling options (the simple model) is detailed in this SNPRM and allows certification based on a fuel's oxygen, benzene, heavy metal and aromatics content and Reid Vapor Pressure (RVP). Under the agreement, EPA would develop a more complex model (the complex model) through a rulemaking to be completed by March 1, 1993. The complex model is expected to provide a method of certification based on the above parameters plus sulfur, olefins and the temperature at which 90 percent of the fuel vaporizes (T90), as well as any other parameters for which sufficient data is available regarding their effects on ozone-forming volatile organic compounds (VOC), toxic air pollutants (toxics) or oxides of nitrogen (NOx) emissions. In the first two years of the program, testing would only be permitted to determine the NOx emission effects of oxygenates other than Methyl Tertiary Butyl Ether (MTBE). Testing would eventually be permitted to qualify for inclusion in the models the emission effects of such other parameters or the effects of fuel parameters beyond the range covered in the models. The agreed upon program would allow refiners to produce reformulated gasoline, either by meeting the applicable standards on a per gallon basis or by meeting the standards on average. The agreed upon averaging program ensures that averaging will not result in smaller overall reductions in pollutants than if averaging were not permitted. It does so through the use of adjusted emission and fuel composition standards for averaged fuels, caps on per gallon levels of the relevant parameters, and compliance surveys to be performed at retail stations. The outline contains two options for compliance with the requirement that conventional gasoline not cause greater emissions of certain pollutants than occurred in 1990. During 1995 and 1996 each refiner and importer may either use the complex model to show that its conventional fuel does not have greater toxics emissions than its fuel had in 1990 or meet certain exhaust benzene and fuel compositional caps. After 1997 each producer and importer must show using the complex model that its conventional fuel has no more emissions of exhaust toxics and NOx than its 1990 annual average. This supplemental notice proposes detailed reformulated gasoline and anti- dumping programs based on the regulatory negotiation consensus. The statutory provisions which form the basis for the agreement and this SNPRM were described in the NPRM, which may be consulted for further information regarding these provisions. II. Fuel Certification Requirements In accordance with section 211(k) of the Clean Air Act, EPA requires that in order for a gasoline to be certified as reformulated, it must contain at least 2.0 weight percent oxygen, no more than 1.0 volume percent benzene, and no heavy metals (unless a waiver is granted); result in no increase in NOx emissions; and achieve required toxics and VOC emission reductions. Toxics and VOC emission requirements and EPA's derivation of them are set forth below. Throughout the negotiation process, different procedures for certifying that a gasoline complies with the NOx, toxics, and VOC requirements were discussed. Pursuant to the consensus agreement, EPA proposes in this supplemental notice two modeling options and a testing method whereby the effects of fuel properties on emissions can be determined. Models offer several advantages over testing to determine emission effects. First, models can better reflect in-use emission effects since they can be based on the results of multiple test programs. Second, individual test programs may be biased, either intentionally or unintentionally, due to vehicle selection, test design, and analysis methods. Third, fuel compositions tend to vary due in part to factors beyond the control of fuel suppliers such as variations in crude oil compositions and the inherent variability of refining processes. As a result, without one or more modeling options, each batch of fuel would have to be tested to ascertain its emission performance. Such levels of testing are neither desirable (because of the potential for intentional bias in vehicle test programs) nor practical (because of the time and expense involved in vehicle testing). Fourth, models make more efficient use of scarce and expensive emission effects data than is possible otherwise. For these reasons, EPA believes that the modeling options outlined below are necessary for the reformulated gasoline program to achieve its environmental objectives and to minimize the costs of the program. The first modeling option being proposed is a simple emissions model (described below in section II.A). Enough is known about the emission effects of several parameters in a range of fuels to model these effects with confidence at this time. These fuel parameters are Reid vapor pressure, fuel oxygen, benzene, and aromatics; the sources of information used to develop the simple model are described in this section's discussion of the simple model. At the current time, not enough data is available on the fuel effects of other parameters to include them in the simple model without running the risk of under- or over-estimating the in-use emissions from reformulated gasolines. The available data, however, is sufficient to suggest that these other parameters (sulfur, T90, and olefins) have a directional effect on emissions. To prevent the emissions benefits that would be obtained from the reformulated gasoline program from being undercut by changes in the values of these parameters, EPA is further proposing that each refiner's annual average levels of sulfur, T90 and olefins in reformulated gasoline not be allowed to exceed their 1990 annual averages for these parameters. EPA anticipates that as additional information becomes available through test programs in progress such as the Auto/Oil 2 program, it may be possible to include additional parameters in an emissions model. In particular, EPA anticipates that sufficient data will be available in 1992 or early 1993 from the Auto/Oil test program and other sources to quantify the emission effects of several additional parameters (including sulfur, T90, and olefins) for inclusion in an expanded model. NOTE 2 The Auto/Oil Air Quality Improvement Research Program is a cooperative research effort undertaken jointly by a number of major automobile and oil companies. This expanded model would be the second modeling option for fuel certification and is referred to here as the complex emissions model (described below in section II.B). Pursuant to the Agreement in Principle, EPA will issue a proposed rule by November 30, 1992 and a final rule by March 1, 1993 which will contain the specific details of the complex model. This complex model rulemaking would also address the "Phase II" reformulated gasoline VOC and toxics performance standards to take effect in the year 2000 as prescribed by section 211(k)(3). If EPA is unable to finalize the complex model rule by March 1, 1993, the required use of the complex model would be delayed one month for every month of delay in issuing the rule. This mechanism is intended to insure that the fuel producers continue to have sufficient lead time for refinery modifications prior to the effective date of the rule. EPA believes that gasoline suppliers should be required to use the most accurate and complete model available to certify their fuels in order to better ensure that the emission reductions that Congress intended reformulated gasoline to achieve actually occur in-use. However, the gasoline suppliers should also be provided with an adequate lead time in order to avoid fuel production shortfalls and economic inefficiencies brought about by changes to their refineries. EPA believes that four years is adequate lead time for fuel producers to make the necessary changes to their refineries to meet the reformulated gasoline requirements under the complex model. Some guidance on lead time is given by the Act's provision of over three years' lead time between promulgation of the rule and the start of the 1995 high ozone season. The Agency continues to believe, for the reasons expressed below, that less than a four-year lead time would be insufficient for a requirement to determine emission effects using the complex model. To prepare for implementation of the complex model, suppliers will have to determine which fuel formulations are most cost-effective for them based on the parameters included in the model and the size of such parameters' emission effects, develop the plans for refinery modifications and design any necessary refining equipment (such as desulfurization units) needed to produce such formulations, obtain the necessary permits and capital, construct the equipment, and complete start-up and equipment shakedown. Given the magnitude of the effort involved, EPA considers it reasonable to implement the complex model with four years' lead time. Therefore, EPA proposes that fuel suppliers be permitted to determine the emission effects of specific fuels by using either model (possibly augmented by testing as described below) for fuels produced before March 1, 1997 or four years after promulgation of the complex model, whichever is later. Until this date, fuel suppliers would have the option of using the complex model instead of the simple model to take advantage of the effects of parameters contained in the complex model but not contained in the simple model (as described in the following paragraphs). EPA further proposes that suppliers be required to use the complex model (appropriately augmented through testing) for fuels produced beginning March 1, 1997 or four years after promulgation of the model, whichever is later. EPA is further proposing that for fuel suppliers opting to use the simple model, each supplier's average annual levels of sulfur, T90 and olefins in reformulated gasoline not be allowed to exceed the refiner's 1990 annual average levels (as determined for the anti-dumping program described in sections XII and XIII). The available data strongly suggest that higher levels of sulfur, T90, and olefins result in higher emissions, although insufficient data exists at present to quantify these effects. These parameters therefore are not included in the simple model, and the effects of increases in the level of these parameters from their 1990 levels will not be reflected in predicted emissions using the simple model. Capping the levels of these parameters at their 1990 levels would help prevent in-use emissions from exceeding either the levels predicted by the simple model or the requirements of the Act. Further, EPA believes these levels will be achievable in 1995 since they were achieved in 1990. The Agency also believes that testing has a role in certification as a means of supplementing the models. Section III of this notice contains a detailed discussion of EPA's proposals regarding the conditions under which testing would be permitted, the manner in which test results would be used to supplement the models, and the minimum requirements for vehicle testing programs. Regardless of whether the emission effects of a gasoline are determined using the simple or complex model (with or without augmentation by vehicle testing results), each gasoline must comply with the requirements for reformulated gasoline individually, notwithstanding whether it is part of a slate of gasolines. On the other hand, credits earned from certain formulations of gasoline in a slate (including credits earned in part due to effects based on vehicle testing) may be used to show the compliance of other formulations in that slate. The credits provisions of the reformulated gasoline program are more fully discussed in section VIII. A. Simple Model As stated above, use of the simple model is a certification option for reformulated gasolines produced prior to March 1, 1997 (or until certification by the complex model is required). EPA proposes that a fuel be considered in compliance with the VOC, NOx, and toxics emission performance requirements under the simple model if it meets the compositional specifications described below. 1. VOC Emissions for Simple Model Fuels The Act requires reductions in emissions of ozone-forming VOCs. This interpretation is consistent with the focus of section 211(k) on the areas with the most extreme ozone pollution problem. Since the ozone-forming potential of methane is more than one order of magnitude lower than that of other types of volatile organic compounds commonly emitted from motor vehicles (including ethane), EPA proposes that VOC emissions be determined on a non-methane basis. EPA proposes to include ethane in VOC emissions since its ozone-forming potential is of the same order of magnitude as other straight-chain hydrocarbons and is much greater than that of methane.3 EPA currently includes ethane (but not methane) in its guidance regarding which VOC species should be included in airshed modeling used to support State Implementation Plans for ozone attainment.4 If EPA should change its guidance on which VOC species should be included in ozone modeling in the future, the definition of VOCs discussed above will be reconsidered. NOTE 3 Carter, William P.L., "Development of Ozone Reactivity Scales for Volatile Organic Compounds," presentation to EPA, 1991. The maximum ozone potential of methane is 0.0074 g ozone/g VOC; corresponding figures for ethane, propane, and n-pentadecane are 0.097, 0.23, and 0.101. NOTE 4 Draft Technical Memorandum entitled "Guidance for SIP Emissions for UAM Modeling," from William Laxton, Director, Technical Support Division, OAQPS, to all of EPA's regional offices. A final memorandum is expected by 12/31/91. The Agency solicits comment on the following concept: Carbon monoxide (CO) is not classified by EPA as a volatile organic compound. However, CO is a factor in ozone-forming photochemical reactions. A "mass-based carbon equivalent" could be assigned to CO emission reductions achieved by reformulated gasoline. This would provide a method by which the mass-based VOC increases attributable to increased volatility could be offset by a mass equivalent. Under this approach, EPA could assign the "mass carbon equivalent" by eliminating the oxygen mass from overall mass CO emissions, with adjustment made to account for the proportionately greater mass effect of carbon monoxide. It is suggested that EPA may have authority to limit such a provision to reformulated gasoline, given the requirement in section 211(k)(1) that EPA implement the program "taking into consideration * * * energy requirements." Under this approach, oxygen credits under section 211(k)(7) would not be applicable to reformulated gasoline to which the mass- based equivalent has been applied. EPA takes no position on this concept at this time, and invites comments on its technical and policy merits, as well as its legal basis. The Agency also requests interested parties to suggest other approaches which could enhance the role of oxygenates in reformulated gasoline including how atmospheric photochemistry can be accounted for in this regulatory framework. EPA intends to include a discussion of this concept in the agenda for the next complex model workshop to be held April 27 and 28. This portion of the workshop will begin at 1:00 pm on April 27. The remainder of the agenda will be devoted to the complex model. Participants in the regulatory negotiation process as well as any other interested parties are encouraged to participate in the workshop and provide comments on this concept as well as providing comments during any hearing on this proposed rulemaking, or in written comments on this proposal. In developing the final rule EPA will evaluate the record of comments and science with a view to allow the greatest flexibility for all oxygenates to lawfully compete in the marketplace. Under today's proposal, fuels sold at retail outlets must have an RVP during the high ozone season (June 1 through September 15) of no more than 7.2 psi in Class B areas and 8.1 psi in Class C areas.5 This period was chosen for the high ozone season because most of the ozone violations occur during this period. (See 56 FR 24242 for a discussion of the determination of this period.) Here Class B and Class C areas refer to those designated by the Phase II volatility control regulation (40 CFR part 80, 55 FR 23659, June 11, 1990) as requiring RVPs of 7.8 psi and 9.0 psi, respectively. (Class B areas correspond generally to the southern states and Class C areas to the northern states. The differences in climate between these two types of areas requires a corresponding difference in gasoline volatility to achieve the same emissions effect.) As discussed above, only the VOC emission effects of RVP and oxygen are included in the simple model. EPA projects that the VOC emission reduction in Class C areas from a fuel with an RVP of 8.1 psi and 2.0 weight percent oxygen will be sufficient to achieve the minimum 15% VOC emission reductions specified in section 211(k)(3) of the Act relative to the Clean Air Act baseline gasoline (which has an RVP of 8.7 psi). In Class B areas, an 8.1 psi RVP fuel with 2.0 percent oxygen (which would meet the 15% reduction requirement relative to the CAA baseline fuel) would actually have greater emissions than a fuel meeting EPA's Phase II RVP control standards for Class B areas (maximum RVP of 7.8 psi). EPA believes that when Congress designated Class B cities for inclusion in the reformulated gasoline program that it intended the reformulated gasoline program to provide emissions reductions in addition to those provided by the Phase II RVP requirements. If EPA merely required reformulated gasoline in Class B areas to meet the RVP requirement for Class C areas, then no additional reduction in VOC emissions would accrue to Class B areas from the first phase of the reformulated gasoline program beyond those mandated by Phase II RVP standards. EPA projects that relative to Phase II RVP control levels, a fuel with 7.2 psi RVP and 2.0 weight percent oxygen is necessary to provide VOC emission reductions to Class B areas similar to those obtained in Class C areas. NOTE 5 Lower RVP limits apply for fuels that comply under averaging. RVP controls also apply from May 1 to May 31 for facilities upstream of retail outlets. These issues are discussed elsewhere in this proposal. While requiring reformulated gasoline sold in Class B areas to have an RVP of no more than 7.2 psi goes beyond the minimum requirement stated in section 211(k)(3), section 211(k)(1) authorizes EPA to require emission reductions in Class B areas of this magnitude because they are achievable considering costs (see the draft regulatory impact analysis; docket identification number II-F- 7), other air quality, and non-air quality impacts and the energy implications of such a requirement. EPA cannot determine that greater reductions, by requiring even lower RVP levels, are warranted at this time for two reasons: (1) EPA's refinery modeling analyses have not examined the effects of RVP reductions on refinery operations at lower levels,6 and (2) EPA does not have sufficient test data to demonstrate emission benefits of lower volatility levels with confidence. Furthermore, extrapolating the results of these studies to lower levels may not be appropriate since the cost and emission effects of lower RVP levels are expected to respond non- linearly as RVP is decreased (because different chemical species and reformulation technologies would be affected than were considered in previous modeling efforts). Hence EPA is unable at this time to determine whether the cost and air quality effects of lower volatility standards warrant establishing lower RVP levels pursuant to section 211(k)(1). NOTE 6 The Bonner & Moore study ("Assessment of the Impacts on the Refining and Natural Gas Liquids Industries of Summer Gasoline Vapor Pressure Control," August 24, 1987, Bonner & Moore Management Science) examined the effects of reducing RVP outside of California to as low as 8.07 psi. Within California, the study examined the effects of RVP as low as 6.82 psi; however, the measurement of RVP for the study was subject to error on the order of 0.3 psi. The Turner Mason study (November 30, 1987) examined comparable RVP ranges. Furthermore, while greater reductions in RVP beyond 8.1 psi in Class C areas potentially may be cost effective, EPA believes that the 1995 implementation date provides insufficient leadtime for refiners to comply with a more stringent Class C standard in conjunction with a 7.2 standard in Class B areas and the toxics and NOx requirements. Given refiners' capacity to produce lower volatility gasoline with the available leadtime, requiring a greater reduction in RVP levels in Class C areas could be achieved only at the expense of relaxing the ability to produce 7.2 RVP gasoline for Class B areas. In addition, the lack of reliable refinery modeling data at this time, as discussed above, inhibits EPA's ability to determine whether further RVP reductions in Class C areas would be warranted. Therefore, EPA believes that to the extent the VOC reductions greater than section 211(k)(3) requires can be achieved, those greater reductions should be required in Class B areas, which otherwise would receive no benefit from the reformulated gasoline program. EPA believes that additional VOC reductions are obtainable if refiners are allowed to meet the RVP and oxygen standards through averaging. In the case of those refiners who can take advantage of averaging, EPA believes that average RVP for both Class B and Class C areas can be reduced by 0.1 psi to 7.1 and 8.0 psi, respectively, and that average oxygen concentration can be increased to 2.1 weight percent oxygen. These increments were determined as part of the regulatory negotiation consensus and would recapture the margin of safety that refiners could be expected to build into their compliance with per gallon requirements to reduce the risk of being found in violation. (See section VI.B.2 regarding compliance margins.) EPA believes the greater flexibility provided by averaging would offset the cost and difficulty of achieving these more stringent averaging requirements. EPA believes it appropriate under section 211(k)(1) to consider the potential of averaging to make greater reductions achievable, and where, as here, EPA finds averaging could make greater reductions achievable, to set more stringent averaged standards. Since refiners differ in the extent to which they can make use of averaging, EPA is proposing that refiners that want to average be required to meet RVP and oxygen standards that are more stringent than the non-averaged standards, as noted above. These tighter, averaged standards should have the potential to increase the environmental benefits of the reformulated gasoline program at no additional cost over the non-averaged standards. 2. NOx Emissions for Simple Model Fuels The Clean Air Act requires that there be no NOx emissions increase from reformulated fuels. 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 while no other simple model parameter appeared to have such an impact. Today's proposal was developed in the context of the negotiated agreement and the data then available. While the currently available data does not allow for quantifying relationships between oxygenate type and concentration and Nox emissions, it suggests that MTBE may contribute little or no NOx increase at concentrations of 2.0 to 2.7 weight percent oxygen, but that ethanol at a concentration of 3.5 weight percent oxygen may cause a NOx increase.7 EPA cannot definitively determine the effect of oxygenates on NOx emissions, due to a general lack of adequate data, a variety of concerns with the data that do exist (e.g., confounding fuel effects, limited vehicle types, testing variability, etc.), and a lack of understanding as to why different oxygenates may show different NOx effects. At the same time, EPA is aware of the benefits of oxygenates for reducing exhaust VOC, CO, and toxics emissions on a mass basis. NOTE 7 Data from EPA's Emission Factor Database and results from the Auto/Oil test programs. EPA proposes that during those months with ozone violations, 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, and thus be permitted for use in reformulated gasoline at any time and in any area. Because of the lack of data on the NOx effect of oxygenates, particularly at concentrations above 2.7 weight percent oxygen in the form of MTBE and above 2.1 percent in the form of other oxygenates, EPA cannot determine that all oxygen concentrations above the 2.1/2.7 limits will definitely increase NOx emissions. Given this, EPA proposes that each state have the discretion to waive the 2.1/2.7 weight percent oxygen limits during the months with ozone violations. In view of the uncertainty about oxygenate effects on NOx emissions and because of the known benefits of oxygenates for reducing exhaust VOC, CO, and toxics emissions on a mass basis, EPA proposes under the simple model that during those months without ozone violations any oxygenate up to 3.5 weight percent oxygen be presumed to result in no NOx emission increase unless a state requests that oxygenate levels be limited to those applicable during those months with ozone violations. A state may make such a request when it believes that the use of higher oxygenate levels would interfere with attainment or maintenance of another ambient air quality standard (other than ozone) or another air quality problem. This proposal parallels the Regulatory Negotiation Agreement of August 16, 1991 and the letter to the Renewable Fuels Association dated August 14, 1991. EPA requests comments on any implementation and other issues that might arise as a result of this provision, particularly how EPA should define months with ozone violations. EPA further proposes that parties wishing to market fuels with oxygen in excess of 2.1 weight percent in the form of oxygenates other than only MTBE (but subject to the oxygenate's waiver limit) during periods where they would be prohibited, as discussed above, may petition EPA to do so. Petitioners must demonstrate, through the use of data they generate, that use of the particular type and level of oxygenate will not adversely affect NOx emissions. EPA will expeditiously process such petitions. The detailed requirements for such test programs and the data required are described in section III. EPA requests comment on whether a less burdensome demonstration is warranted for approving oxygenate concentrations not up to 2.7 weight percent oxygen (as opposed to those above 2.7 weight percent) and if so, what such requirements should be. EPA believes that the proposed approach to NOx is consistent with the intent of section 211(k)(1) that the greatest reduction in ozone-forming volatile organic compounds be achieved during that portion of the year when ozone exceedences occur, taking into consideration cost and other factors. Allowing for increased use of a wide variety of oxygenates will increase the supply of oxygenate available for use in reformulated gasoline, thereby having a controlling effect on the cost for oxygenates, especially in the first years of the program. This increased supply of oxygenate may also allow for more nonattainment areas to opt-in (See NPRM Section II.F.2 regarding opt-in) and obtain the air quality benefits of the reformulated gasoline program earlier than would have otherwise been possible. Furthermore, allowing the States the right to limit the concentration of oxygenates in reformulated gasoline should prevent the occurrence of any negative nonair- quality or other air-quality impacts that the proposed approach might otherwise permit. EPA believes that this is an appropriate treatment of concerns related to NOx emissions effects of oxygenates given the current limitations of the data and of understanding of the possible effects. 3. Toxic Emissions Under the Simple Model Under section 211(k)(3), a reformulated gasoline's toxic emission performance must meet or exceed that of a specified formula fuel or a 15 percent reduction from that of baseline gasoline, whichever is greater. Under the simple model a fuel's toxic emissions are a function of its oxygen and benzene content, its VOC emission, and its level of benzene and non-benzene aromatics. If the fuel meets the requirements regarding oxygen and benzene content and VOC performance, its level of benzene and non-benzene aromatics must be sufficiently low such that the fuel meets or exceeds the toxic emissions requirements (described later in this section.). Since sufficient information either is not yet available or has not yet been fully analyzed to determine the proper coefficients for parameters that impact toxics emissions other than oxygenate type and oxygen, aromatics, and benzene concentration, the only variables which could be adjusted under the simple model to meet the toxic emission requirement are the benzene and non-benzene aromatic concentrations. The toxic emission equations proposed below would be used to determine a fuel's toxic emission reductions and could thereby determine the limits on aromatics content for fuels with various oxygenates, oxygen concentrations, benzene levels, and RVP levels. All five of the toxic air pollutants that section 211(k)(10) of the Act specifies for control through reformulated gasoline (benzene, 1,3-butadiene, polycyclic organic matter (POM), formaldehyde, and acetaldehyde) also fall under the category of VOCs. Under high ozone (summer) conditions, all five toxics are present in exhaust emissions, and only benzene is present in evaporative, running loss and refueling emissions (nonexhaust emissions). Benzene, an aromatic compound, is a natural component of gasoline and, as such, is present in gasoline vapor emissions. Exhaust emissions include unburned benzene and benzene formed from other aromatics during the combustion process. The four other toxic air pollutants subject to control by reformulated gasoline are not present in gasoline and hence are solely products of combustion. EPA proposes to regulate aggregate toxics emissions based on the sum of both exhaust and nonexhaust toxic emissions during the summer (April 1 through September 15). (The definition of summer and winter periods for toxics control is explained later in this section.) Under winter conditions, on the other hand, EPA is assuming that nonexhaust benzene (and in fact all nonexhaust VOC) emissions will be negligible relative to exhaust toxic emissions due to low ambient temperatures. EPA therefore proposes to regulate aggregate toxics emissions during the winter period (September 16 through March 31) based exclusively on total exhaust toxic emissions. As explained in the NPRM, since exhaust emission effects will likely vary between vehicles with varying emission performance levels, all data used to develop the exhaust emission correlations contained in the simple model are weighted by emitter subclass (based on available information) to reflect in- use fleet composition as per MOBILE4.1, consistent with the assumptions made concerning baseline exhaust emissions expressed in Section II.A.3.d below. Similarly, since nonexhaust emission effects vary between vehicles that pass and fail evaporative emission standards, all data used to develop the correlations contained in the simple model for nonexhaust emissions are weighted by evaporative emitter subclass (based on available information) to reflect the in-use fleet composition as per MOBILE4.1. a. Exhaust benzene emissions. Exhaust benzene emissions can be affected by fuel modifications in two basic ways. Some fuel effects will change the fraction of benzene in the exhaust, regardless of the total VOC mass that is emitted as exhaust. For instance, increasing or decreasing the level of benzene in a fuel will lead to a direct increase or decrease in the benzene fraction of exhaust emissions. Moreover, changes in the level of benzene precursors (primarily nonbenzene aromatics) will affect the amount of benzene that is produced during combustion, also changing the benzene fraction of exhaust VOC emissions. On the other hand, fuel modifications can affect the overall level of exhaust VOC emissions by affecting the efficiency of the engine or catalyst in burning hydrocarbons. In these cases, the benzene fraction of exhaust VOC emissions may stay relatively constant and benzene exhaust emissions will change proportionally with exhaust VOC emissions. Of course, some fuel modifications can produce a combination of these two effects. EPA proposes to analyze the effect of fuel modifications on exhaust toxic emissions by separating the two types of effects described above. This applies not only to benzene but to all five toxic air pollutants. With this approach, fuel modifications which change the level of exhaust VOC emissions are considered to change the levels of exhaust toxic emissions proportionally. Under the simple model, exhaust VOC emissions for both Class B and Class C areas are affected only by fuel oxygen content according to the following relationship: Exhaust VOC (g/mi)=Exhx[1 -(0.127xOx)/2.7] Ox refers to the fuel weight percent oxygen. Exh is the baseline level of nonmethane exhaust VOC emissions as determined from MOBILE4.1; for summer conditions Exh equals 0.46 g/mi, while for winter conditions Exh equals 0.68 g/mi. The term 0.127 represents the reduction in exhaust VOC emissions achieved when 2.7 weight percent oxygen is added to the fuel. This relationship is based on an analysis (contained in the docket to this rule) 8 of fuels containing MTBE in EPA's Emission Factor Database. NOTE 8 Christian E. Lindhjem, "Effect of Oxygenates on Emissions." With respect to the effects of fuel modifications on the benzene fraction of exhaust VOC emissions, fuel benzene and fuel aromatics appear to be the primary factors. EPA proposes that the correlation used to relate fuel benzene and aromatics to the weight fraction of benzene in exhaust VOC (nonmethane) emissions for both Class B and Class C areas be: {1.818+(0.9154xBz)+[0.109x(Arom -Bz)]}/100 where Bz is the volume percent of fuel benzene and Arom is the volume percent of fuel aromatics. This equation is based on a study by Chevron 9 and indicates that exhaust benzene emissions depend on benzene content and on non-benzene aromatics content. Combining exhaust VOC emissions with the effects of benzene and aromatics on the benzene fraction of VOC emissions, benzene emissions (grams per mile) would be: NOTE 9 Communication to EPA summarizing the following studies: "Study to Determine the Fate of Benzene Precursors in Gasoline", NIPER (Under CARB Agreement 150128-32), 1988; "Exhaust Benzene Emissions from Late-Model Vehicles", API Publication No. 841-44700, 10/88; "Vehicle Evaporative and Exhaust Emissions as Influenced by Benzene Content of Gasoline", NIPER (Under CRC CAPE-35-83 and U.S. DOE), 4/86. {1.818+[0.9154xBz]+[0.109x(Arom - Bz)]}/100xExhaust VOC where Exhaust VOC is the level of VOC nonmethane exhaust emissions in grams per mile as described above. This equation is assumed to be valid for both summer and winter conditions, based on EPA test results 10 showing benzene emissions to be proportional to exhaust VOC emissions at various test temperatures. NOTE 10 (Atmospheric Environment, vol. 23, no. 2, pp. 307-320, 1989; Atmospheric Environment, vol. 24A, no. 8, pp. 2105-2112, 1990). b. Nonexhaust benzene emissions. Benzene is the only toxic air pollutant that is emitted in measurable quantities from evaporative, running loss, and refueling vapors. Reductions in fuel benzene may be expected to result in proportional reductions in benzene emissions from all of these nonexhaust emission sources. The Agency proposes to include this proportional effect of fuel benzene on nonexhaust benzene emissions in the emissions model. In addition to fuel benzene content, two other fuel parameters--RVP and fuel oxygen content--can also affect nonexhaust benzene emissions. Both parameters affect both the total level and the benzene weight fraction of evaporative, running loss, and refueling VOC emissions. The effects of RVP on evaporative, running loss, and refueling VOC emissions are well characterized in MOBILE4.1 for Class C area summer conditions within a volatility range of 7.0 to 11.7 psi and for Class B summer conditions between 6.8 and 10.5 psi. The correlations used in the simple model are based on MOBILE4.1 and are valid for 6.6 to 9.0 psi for both Class B and Class C areas (the maximum RVP allowed under the simple model, however, is 8.4 psi). EPA further proposes the use of the formulae expressed below (based on the GM vapor model 11) to model the effects of RVP and fuel oxygen content on the benzene fraction of evaporative, running loss, and refueling VOC emissions. Due to differences in temperature conditions, slight differences in nonexhaust VOC emissions occur between Class B and Class C areas. As a result, separate standards for toxics emission performance are provided for Class B and Class C areas. NOTE 11 Communication to C.E. Lindhjem from S.R. Reddy, April 16, 1991. Evaporative benzene emissions from a given vehicle include hot soak emissions (evaporative emissions from a warm vehicle after it has been running) and diurnal emissions (evaporative emissions from a sitting vehicle as the daily ambient temperatures rise and fall). Hot soak emissions occur at higher temperatures than diurnal emissions and the relative volatility of benzene is slightly greater at higher temperatures. Therefore, the benzene fraction of hot-soak VOC emissions tends to be higher for a given fuel than that for diurnal VOC emissions. Running loss emissions occur at roughly the same fuel temperature as hot-soak emissions, and therefore have similar benzene fractions. Based on the emission factors contained in MOBILE4.1, evaporative and running loss benzene emissions tend to be dominated by emissions from vehicles with inoperative emission control systems (those vehicles likely to "fail" EPA's purge and pressure tests). The benzene fraction of evaporative and running loss emissions from vehicles with properly operating systems (those vehicles likely to "pass" EPA's purge and pressure tests) and from "fail" vehicles, however, are comparable. Hence EPA proposes that the benzene weight fraction of evaporative and running loss VOC emissions for a fuel be described by the following relationships, originally derived for "fail" vehicles. The hot soak and running loss benzene fraction of VOC equals: [Bz/100]x[1.4448 - (0.080274xRVP) - (0.0684xMTBE/2.0)] The diurnal benzene fraction of VOC equals: [Bz/100]x[1.3758 - (0.080274xRVP) - (0.0579xMTBE/2.0)] where Bz is the volume percent benzene, RVP is in psi, and MTBE is the weight fraction oxygen in the form of MTBE. The formulae for evaporative and running loss benzene emissions indicate that as oxygen in the form of MTBE increases, evaporative benzene emissions tend to decrease both in absolute terms and as a fraction of evaporative VOC emissions. Test data has shown that the presence of MTBE tends to reduce benzene's partial vapor pressure and, thus, evaporative and running loss benzene emissions.12 Test data with ethanol has not shown an effect on benzene emissions separate from its effect on overall evaporative VOC emissions. Data with other oxygenates is not yet available to determine whether an effect similar to that of MTBE exists. Therefore, the oxygenate term in the formulae expressed here applies only to MTBE. NOTE 12 Ibid. The formulae also indicate that as RVP decreases, evaporative and running loss benzene emissions also decrease but at a slower rate than total VOC emissions. Hence the benzene weight fraction of evaporative and running loss VOC emissions increases as RVP decreases. Applying these equations to CAA baseline gasoline results in a hot-soak and running loss benzene emission fraction of 1.14 percent of VOC and a diurnal benzene emission fraction of 1.04 percent. Evaporative and refueling benzene emissions (mg/mi) are then determined by the following formulae. Hot soak benzene emissions (mg/mi) equal: [Bz/100]xEvap VOCx0.679x[1.4448 - (0.080274xRVP) - (0.0684xMTBE/2.0)] Diurnal benzene emissions (mg/mi) equal: [Bz/100]xEvap VOCx0.321x[1.3758 - (0.080274xRVP) - (0.0579xMTBE)/2.0)] Running loss benzene emissions (mg/mi) equal: [1.4448 - (0.0684xMTBE)/2.0 - (0.080274xRVP)]x[Bz/100]xRunVOC Evap VOC is the evaporative VOC emissions in mg/mi, as determined below, 0.679 is the hot soak fraction of evaporative VOC emissions, 0.321 is the diurnal fraction of evaporative VOC emissions, and RunVOC is the running loss VOC emissions in mg/mi. These formulae are valid for fuel oxygen levels of up to 2.7 percent in the form of MTBE. Evaporative and running loss VOC emissions in mg/mi are determined by the following formulae. In Class B areas, Evap VOC (mg/mi) equals 1000x[0.7952 - (0.2461xRVP) + (0.02293xRVPxRVP)] In Class C areas, Evap VOC (mg/mi) equals 1000x[0.813 - (0.2393xRVP) + (0.021239xRVPxRVP)] In Class B areas, RunVOC (mg/mi) equals 1000x[(0.1096xRVP) - 0.734 + (0.002791xRVPxRVP)] In Class C areas, RunVOC (mg/mi) equals 1000x[0.2963-(0.1306xRVP)+(0.016255xRVPxRVP)] The relationship of fuel benzene levels to refueling benzene emissions (mg/ mi) using the General Motors model is given by: [1.3972-(0.0591xMTBE/2.0)-(0.081507xRVP)]x[Bz/100]xRefVOC where RefVOC is the total refueling VOC emissions in mg/mi, given by: 0.04x1000x[(0.1667xRVP)-0.45] The presence of MTBE tends to reduce benzene's vapor pressure and thus refueling benzene emissions; reductions in RVP tend to increase the benzene fraction of refueling VOC emissions while reducing refueling benzene emissions on a mass basis. Applying this equation to baseline gasoline results in a benzene fraction of refueling VOC emissions of 1.0 percent. c. Nonbenzene toxic emissions. As discussed above, the only regulated toxic pollutant present in unburned gasoline is benzene; hence non-benzene toxic emissions are present only in exhaust emissions. For summer fuels EPA proposes to use the results from the Auto/Oil study to determine the 1,3- butadiene, formaldehyde, and acetaldehyde fractions of exhaust VOC emissions. The Auto/Oil data as released, however, were modified slightly to exclude the acetaldehyde and formaldehyde results for ETBE and ethanol from one of the vehicles (car #5A) due to emission results which were confirmed as being in error. Furthermore, the effect of ETBE on the weight percent of acetaldehyde was based on the test results for ethanol due to the lack of adequate fuel comparability for ETBE-containing fuels. However, for this case, the ethanol results were adjusted based on a comparison of the ETBE and ethanol results on similar fuels for which data was available. The toxics emissions for summer fuels are determined by the following formulae. 1,3-butadiene emissions in mg/mi equal: 0.00539x1000x(Exhaust VOC) where Exhaust VOC represents total exhaust VOC nonmethane emissions (including the effects of fuel oxygen) in grams per mile, and 0.00539 represents the weight fraction of 1,3-butadiene in baseline nonmethane VOC emissions (as determined by the Auto/Oil study 13. NOTE 13 Data received by EPA from the Auto/Oil Air Quality Improvement Research Program. Formaldehyde emissions in mg/mi equal: 0.01199xExhaust VOCx1000x(1+(0.42/2.7)x(MTBE)) for MTBE containing fuels, 0.01199xExhaust VOCx1000x(1+(0.358/3.55)x(ETOH)) for ethanol containing fuels, and 0.01199xExhaust VOCx1000x(1+(0.137/2.7)x(ETBE)) for ETBE containing fuels, where 0.01199 represents the weight fraction of formaldehyde in summer baseline nonmethane VOC emissions (as determined by the Auto/Oil study), Exhaust VOC represents total summer VOC (nonmethane) emissions in grams per mile, MTBE, ETOH, and ETBE refer to the weight fraction oxygen in the form of those oxygenates, and 0.42, 0.358, and 0.137 represent the increase in the weight fraction of formaldehyde emissions with the addition of 2.7 weight percent oxygen in the form of MTBE, 3.55 weight percent oxygen in the form of ethanol, and 2.7 weight percent oxygen in the form of ETBE, respectively. Acetaldehyde emissions in mg/mi equal: 0.00854xExhaust VOCx1000x(1+(0.078/2.7)x(MTBE)) for MTBE containing fuels, 0.00854xExhaust VOCx1000x(1+(0.865/3.55)x(ETOH)) for ethanol containing fuels, and 0.00854xExhaust VOCx1000x(1+(0.867/2.7)x(ETBE)) for ETBE containing fuels, where 0.00854 represents the weight fraction of acetaldehyde in summer baseline nonmethane VOC emissions (as determined by the Auto/Oil study), and 0.078, 0.865, and 0.867 represent the increase in the weight fraction of acetaldehyde emissions with the addition of 2.7 weight percent oxygen in the form of MTBE, 3.55 weight percent oxygen in the form of ethanol, and 2.7 weight percent oxygen in the form of ETBE, respectively. Emissions of polycyclic organic matter (POM) include a number of different, high molecular weight aromatics. There is no data quantifying the impacts of gasoline reformulations on POM emissions. At the present time, there are also no widely accepted test procedures for measuring POM in both the gaseous and particulate phases. In addition, POM emissions constitute a very small fraction of total toxic emissions (less than 2 percent). For these reasons, the Agency proposes that the emissions model consider POM emissions to be proportional to total exhaust nonmethane VOC emissions and not dependent on any particular fuel parameter. POM emissions in mg/mi equal: 0.00304x1000x(Exhaust VOC) where (based on EPA analyses 14), 0.00304 equals 0.0014 (the emissions of POM from the baseline fuel in grams per mile divided by 0.46 (the exhaust VOC emissions from summer baseline fuel in grams per mile). NOTE 14 "Analysis of the Economic and Environmental Effects of Methanol as an Automotive Fuel," U.S. Environmental Protection Agency, September 1989. Under winter conditions, EPA test results 15 indicate that the proportion of 1,3-butadiene in exhaust VOC emissions is the same as under summer conditions, while the mass of formaldehyde, acetaldehyde, and POM emissions are estimated to be the same as summer emissions. As a result, for all non- benzene toxics except 1,3-butadiene, the winter emissions are given by the equations expressed above with Exhaust VOC set equal to summer baseline exhaust VOC emissions (0.46 g/mi) rather than the winter value (0.68 g/mi). Winter exhaust 1,3-butadiene emissions (mg/mi), however, are to be determined by using the winter baseline exhaust VOC emissions of 0.68 g/mi. NOTE 15 Atmospheric Environment, op. cit. d. Baseline emissions. The derivation of baseline emissions used in the above formulae was described at length in section III.A. of the NPRM and the reader is referred to that document for discussion of that issue (56 FR 31179). Some changes and corrections have been made since the NPRM was published, and they are described below. i. Winter baseline gasoline. The winter baseline parameter values developed for the NPRM (56 FR 31180) have been recalculated to account for change in the length of the summer period from May 1-September 30 to April 1-September 15. This data is thus valid for use only during that period. Average values for additional parameter have been computed as shown in Table II-3. As part of the recalculation, the methodology was changed slightly from that described in the NPRM. In the final calculation of an average fuel parameter value, the contribution of each survey city's fuel consumption by month or bi-monthly period to the entire winter period was used. The final average fuel parameter value was then determined by a summation of all the cities' contributions over the entire winter period. Previously, a single value was obtained for each month or bi-monthly period which included the contribution of each city during that month or bi-monthly period. These values were then averaged to obtain the average winter value. Comments are requested on the determination of the winter values of the baseline parameters, particularly regarding the computation methodology used. Table II-3: Winter Baseline Fuel Composition API Gravity--60.4 Sulfur, ppm--338 Benzene, volume percent--1.62 RVP, psi--11.7 Octane, R+M/2--88.1 IBP, degrees F--87 T10, degrees F--111 T50, degrees F--199 T90, degrees F--332 End Point, degrees F--404 Aromatics, volume percent--26.4 Olefins, volume percent--11.9 Saturates, volume percent--61.7 ii. MOBILE4.1. The goal of EPA in developing the procedures for certifying fuel as meeting the reformulated gasoline requirements is to assure that a certified fuel will achieve the required emission reductions in-use. This goal necessitates the use of a fuel effects model which predicts in-use emissions. For the simple model, EPA has therefore used the MOBILE4.1 emissions model to determine the proposed baseline emission levels. For further discussion of the rationale behind this decision the reader is referred to the discussion in the NPRM. The final version of MOBILE4.1 was released on July 29, 1991, and is available from any regional office of EPA (August 26, 1991, 56 FR 42053) and the docket for this rulemaking. iii. Temperature conditions. MOBILE4.1 has been developed to predict motor vehicle emissions on an area-specific basis. In order to use MOBILE4.1, it therefore is necessary to specify a temperature range for the areas in which motor vehicle emissions are being evaluated. EPA proposes modeling baseline emissions under temperatures ranging from 71.6 to 91.6 degrees Fahrenheit in areas classified as Class C areas (9.0 psi RVP, classified as VOC Control Region 2 in section V.D.) and ranging from 69.4 to 94.0 degrees F in Class B areas (7.8 psi RVP, classified as VOC Control Region 1 in section V.D.). These temperatures represent the population-weighted average of minimum and maximum temperatures measured in each of 25 serious and worse ozone nonattainment areas during their ten worst ozone days in each of the months of July and August for the years 1986 to 1989 (in ten of the cities) and 1985 to 1987 (in the other fifteen cities).16 Refueling emissions were derived assuming an ambient temperature of 90 deg.F for both Class B and Class C areas. Distinguishing between the different areas did not appear justified given the similarity of Class B and Class C area temperatures, the relatively low magnitude of refueling emissions, and the wide range of times and temperatures at which refueling occurs during a day. 90 deg.F was considered to represent a severe case in order to account for average in-use refueling emissions on high ozone days. NOTE 16 Memorandum II-A-2 from Jeffrey A. Herzog and Stephen Mayotte to Public Docket No. A-91-02. For determination of winter baseline emissions, an average low temperature and an average high temperature of 39 deg.F and 57 deg.F, respectively, were utilized. These temperatures were estimated from the historical 30-year average low and high temperatures for the months of October through April for the 25 serious and worse ozone nonattainment areas.17 NOTE 17 Ibid. iv. Effects of Stage II refueling controls. As discussed in the NPRM, baseline emissions are assumed to include the benefits of a Stage II refueling vapor recovery program. The only change from the NPRM is that the efficiency of Stage II controls is now assumed to be 86 percent. EPA's regulatory impact analysis supporting refueling emission regulations estimated the efficiency of Stage II equipment to be 86 percent in areas such as California where the program is very strictly enforced. Because of the severity of ozone pollution in areas that will be covered by the reformulated gasoline program and because strong measures will be required to bring these areas into attainment, it is assumed that Stage II programs in these covered areas will be strictly enforced. v. Assumptions regarding enhanced inspection and maintenance programs. A large portion of motor vehicle emissions are attributable to a small fraction of vehicles whose emission levels are extremely high due to tampering or malmaintenance. Enhanced inspection and maintenance (I/M) programs, mandated by the Act for all serious, severe, and extreme ozone nonattainment areas, will address this category of emission sources by inspecting vehicles for proper maintenance of exhaust and evaporative emission control equipment. The Agency is in the process of developing the minimum criteria for enhanced I/M programs. In the NPRM, the Agency proposed to include the impacts of enhanced I/M programs on baseline emission projections since enhanced I/M programs will be in place when requirements for reformulated gasoline take effect. While the minimum criteria for enhanced I/M programs are still undefined, for the purposes of the simple model proposed in this notice, the program is assumed to include an anti-tampering gas cap check for evaporative and running loss emissions and a 2500 rpm idle test for exhaust hydrocarbons. These tests were chosen because EPA is confident that the definition of enhanced I/M will include tests at least this stringent. The in-use emission impacts of these potential I/M provisions were included in the MOBILE4.1 modeling to determine baseline emissions. The assumptions regarding enhanced I/M programs, for the purposes of the complex model, will be defined in the complex model rulemaking. e. Simple model performance of toxic emissions. Using the emissions effects proposed above and the assumptions described in section II.A.3.d. concerning baseline emissions, the following table lists EPA's estimated toxics emissions from Clean Air Act baseline summer gasoline and the formula fuel assuming the oxygenate type in the formula fuel is MTBE. The selection of MTBE for use in the formula fuel was based on the likelihood that MTBE will be the most heavily used oxygenate. In addition, MTBE yields slightly larger toxics emission reductions than other oxygenates tested to date due to its effect on nonexhaust benzene emissions. Since MTBE will be widely available for use in reformulated gasolines, EPA believes it is appropriate to base toxics emission standards on a formula fuel resulting in the greatest achievable reductions in toxic emissions. Table II-4.--Summer Toxic Emission Performance of Formula Fuel [Summer Toxic Air Pollutants (TAPs), mg/mi] Formula w/ Baseline MTBE (8.7 (8.7 RVP) RVP) Class Class Class Class B C B C Exhaust VOCs (g/mi) 0.46 0.46 0.42 0.42 Total VOCs (g/mi) 1.23 1.23 1.19 1.19 Exhaust Benzene (mg/mi) 30.1 30.1 22.3 22.3 Evaporative Benzene 4.3 3.8 2.6 2.2 Running Loss Benzene 4.9 4.5 2.9 2.6 Refueling Benzene 0.4 0.4 0.3 0.3 1,3-Butadiene 2.5 2.5 2.2 2.2 Formaldehyde 5.5 5.5 6.6 6.6 Acetaldehyde 3.9 3.9 3.8 3.8 POMs 1.4 1.4 1.3 1.3 Total TAPs 53.1 52.0 41.9 41.3 Using the emissions effects proposed above and the assumptions described in section II.A.5 concerning baseline emissions, the following table lists EPA's estimated toxics emissions from winter baseline gasoline and from the formula fuel assuming the oxygenate type in the formula fuel is MTBE. Table II-5.--Winter Toxics Performance of Formula Fuel [Winter Toxic Air Pollutants (TAPs), mg/mi] Winter Formula w/ baseline MTBE Class Class Class Class B C B C Exhaust VOCs (g/mi) 0.68 0.68 0.62 0.62 Total VOCs/1/ (g/mi) 0.68 0.68 0.62 0.62 Exhaust Benzene (mg/mi) 40.8 40.8 33.0 33.0 1,3-Butadiene 3.7 3.7 3.3 3.3 Formaldehyde 5.5 5.5 6.6 6.6 Acetaldehyde 3.9 3.9 3.8 3.8 POMs 1.4 1.4 1.3 1.3 Total TAPs 55.3 55.3 47.9 47.9 /1/ Assuming that wintertime evaporative emissions are negligible. (See earlier discussion.) Under section 211(k)(3), reformulated gasoline must meet the emissions performance of the formula fuel or the minimum performance standard specified in section 211(k)(3)(B), whichever is more stringent. For Class B areas, using the simple model with the formula fuel produces a 21.1 percent reduction in toxics emissions in the summer and a 13.5 percent reduction in the winter. For Class C areas, using the simple model and its modeling assumptions with the formula fuel produces a 20.7 percent reduction in toxics emissions in the summer and a 13.5 percent reduction in the winter. For purposes of toxics emission control, the winter period is assumed to be September 16 through March 31 since this period coincides with the time period during which winter gasoline will be produced. While summer gasoline would not be required at retail outlets until June 1, it would be required at terminals by May 1 and hence would be produced or imported by fuel suppliers at some earlier date. Fuel producers have indicated that production of summer gasoline could begin as early as March 1. In some cases, production of summer gasoline would not begin until after April 1, but in no case would it begin later than May 1. EPA believes that April 1 represents a reasonable average date for the beginning of summer gasoline production and proposes its use to determine summer and winter time periods for the purposes of the toxics compliance periods. When weighted according to fuel consumption (53.2 percent of gasoline is consumed during the winter period and 46.8 percent is consumed during summer), the annual average toxic emissions reduction is 17.1 percent from baseline levels in Class B areas and 16.9 percent in Class C areas. Based on the simple model correlations presented in this section, EPA believes that refiners are capable of achieving toxic emission reductions of this magnitude in conjunction with the VOC emission reductions discussed earlier. EPA believes that without the flexibility provided by an averaging program, requiring greater reductions in toxic emissions is not warranted at this time given refiners' need to produce gasoline at current and projected octane levels (more stringent toxics emissions standards would likely necessitate lowering aromatics levels, which would also reduce fuel octane levels) and the overall cost effectiveness of toxics emissions reductions relative to the corresponding health benefits, as discussed in the Regulatory Impact Analysis. As discussed in section VI.B.2, the Agency believes it appropriate for standards met on average to be more stringent than standards met on a per- gallon basis. Based on the discussion in section VI.B.2, EPA proposes that averaged toxic emission standards be 1.5 percentage points more stringent than the relevant per-gallon standards. Adding a 1.5 percentage point margin to the Class B and Class C results above would result in an 18.6 percent reduction requirement in Class B areas and an 18.4 percent reduction requirement in Class C areas; given the uncertainties in measuring toxic emission levels and the levels of fuel parameters that affect toxic emissions, and given the additional compliance and regulatory burden involved in establishing and enforcing separate Class B and Class C area standards, EPA believes that a single year-round standard is appropriate. EPA proposes that this standard be set at a level 18.5 percent lower than emissions from the annual average baseline emission level. Under the authority of section 211(k)(1) to set tighter standards, EPA believes that the greater flexibility and reduced cost afforded to gasoline refiners and importers by an averaging program allow EPA to require a greater reduction in toxics emissions than is required under section 211(k)(3). In addition, EPA estimates that the approximate 1.5 percentage point margin will be more than sufficient to recoup any compliance margin refiners would have otherwise had to maintain to ensure achievement of the toxics requirements in the absence of an averaging program. In sum, the tighter averaged standard should have the potential to increase the environmental benefits of the reformulated gasoline program while not increasing the cost of obtaining those benefits. For suppliers who opt to certify their gasolines on a per-gallon basis, EPA proposes that separate summer and winter toxics performance standards be based on the performance of the formula fuel under summer and winter conditions, respectively. Using the simple model, the summer performance standard would be a 21.1 percent reduction in toxic emissions in Class B areas and a 20.7 percent reduction in toxic emissions in Class C areas, relative to summer baseline gasoline. The winter performance standard would be a 13.5 percent reduction in toxic emissions relative to winter baseline gasoline. EPA believes that applying the annual averaged emission reductions (17.1 in Class B areas and 16.9 in Class C areas) as separate summer and winter per gallon standards would not be appropriate, since such standards would essentially require a greater and less cost effective reduction in toxics emissions in the winter months than is achieved by the winter formula fuel but would not reduce total toxic emissions. B. Complex Model As stated in the introduction to this section, EPA will issue a proposed rule no later than November 30, 1992 and a final rule by March 1, 1993 which will contain the specifics of a complex model to evaluate the emissions effects of a larger number of fuel parameters than are included in the proposed simple model. The complex model will be developed in a fashion similar to the simple model. However, the specific relationships used to relate simple model parameters to emissions may change as additional data becomes available and as the Agency's projections of the effectiveness of enhanced I/M programs and Stage II refueling controls develop. These relationships will be defined as part of the development of the complex model. While EPA believes that it is important to use the most accurate and complete model available for fuel certification, EPA also believes that fuel suppliers need adequate lead time to modify and adjust their production processes. Therefore, use of the complex model is not required prior to March 1, 1997. Beginning on March 1, 1997 (or four years after promulgation of the complex model, whichever is later), however, all reformulated gasoline must be certified by the complex model (augmented as appropriate by vehicle testing results). This timing was developed as part of the regulatory negotiation and, as discussed earlier, it provides the time required to develop the additional capacity needed to supply sufficient quantities of reformulated gasoline and provides adequate lead time for refiners to make any necessary refinery changes. Until March 1, 1997, refiners who produce reformulated gasoline would have a choice of certifying their fuel by using the simple model, the complex model, or by augmenting the models with vehicle testing (section III). EPA has developed two options for application of the complex model during the first two years of the program. Under the first option, if a refiner opts to utilize the complex model before March 1, 1997, the reformulated gasoline can have no worse VOC, NOx, or toxic emissions performance than would be predicted by the complex model for a simple-model fuel (minimum 2.0 percent oxygen, maximum 1.0 percent benzene, and maximum RVP of 8.1 psi in Class C areas and 7.2 psi in Class B areas) having that refiner's average 1990 levels of sulfur, olefins, and T90. This requirement would prevent fuel suppliers from supplying higher-emitting fuels than would be required under the simple model by electing to use the complex model to evaluate emissions performance. Since the complex model may contain parameters capped under the simple model and may also attribute larger emission effects for one or more simple model parameters, emission reductions for a fuel evaluated under the complex model may be larger than for the same fuel evaluated under the simple model. For example, under the simple model, a fuel producer with sulfur levels below the CAA baseline fuel level achieves a certain emission reduction due only to the parameters contained in the simple model. Under the complex model, however, that fuel producer would likely be able to claim an emissions benefit for its low sulfur level and relax the requirements on simple model parameters. The resulting fuel would meet the performance standards according to the complex model but would fall short of the standards according to the simple model. Because this option requires such producers to produce fuels that meet the required performance according to the simple model, such producers would be required to produce fuels that would achieve lower in-use emissions than required according to the complex model. However, once the complex model is required beginning March 1, 1997, such producers would be able to reduce the extent of reformulation needed to meet the requirements of the Act. As a result, this option may require capital expenditures during the first two years of the program from such producers that would not be required after March 1, 1997. For example, under this option, a fuel supplier with low levels of sulfur, T90, or olefins would be required to reduce RVP to a greater extent than would be required in 1997, when the complex model is required and they would be able to take credit for their low sulfur, T90, or olefin levels (assuming the complex model includes these parameters). To some extent, however, this added cost might be offset to the extent these expenditures would be required to meet the Phase II standards which take effect in 2000. This option would preserve the environmental benefits that would be realized using the simple model. However, it may also provide greater flexibility to fuel suppliers with higher 1990 baseline levels of sulfur, T90, and olefins, thereby effectively "rewarding" fuel suppliers with higher-emitting 1990 baseline fuels. The second option EPA is considering would allow refiners to certify fuels using only the complex model during the initial years of the program without any reference to simple model fuel performance. This option is not included in the negotiated agreement and, as noted above, this option may result in higher emissions prior to 1997 than would the first option. However, this option would be more cost effective than the first option since it would allow refiners to make one refinery change which would be effective both before and after 1997. Additionally, the Clean Air Act sets absolute emission standards for reformulated gasoline, and the complex model will reflect the best available model of emissions by incorporating these parameters; hence, it arguably should be allowed to supplant the simple model as soon as possible. Finally, the many parameters of the complex model give refiners more methods of reformulating gasoline than does the simple model, thereby allowing refiners to choose the method which is most cost effective for them. On the other hand, the emission reduction requirements for reformulated gasoline under the simple model are considered to be achievable and cost effective; therefore, equivalent emission reductions under this option also would be achievable and cost effective, considering suppliers' freedom to choose either model and the additional flexibility the complex model offers refiners. Since this option would allow suppliers with low 1990 baseline levels of sulfur, T90, or olefins to claim these benefits of their fuels, this option effectively rewards suppliers of lower-emitting 1990 baseline fuels. However, under this option a supplier with very low levels of sulfur, T90, and olefins might be able to meet the standards using the complex model with RVP levels that exceed the per-gallon RVP caps established as part of the simple model and might thereby affect the ability of the Agency to enforce compliance with the requirements of the Act while the simple model is in use. Therefore, EPA proposes that this option include the caps on RVP included in the simple model averaging program. Finally, this option could result in smaller emission reductions during the first two years than the 15 percent emission reduction goal implied by the Act. Fuel suppliers with high levels of sulfur, T90, and olefins would meet (and in some cases exceed) the 15 percent reduction requirement relative to their 1990 fuels, but not necessarily relative to the Clean Air Act baseline fuel. At the same time, fuel suppliers with low levels of these parameters would be able to meet the requirements of the Act, but with smaller reductions in emissions relative to their 1990 fuels. To resolve these problems, EPA proposes that the second option be restricted to Class A and Class B areas only. The VOC performance standard in such areas would be set equal to the projected emissions of a simple model fuel (7.2 psi RVP, 2 percent oxygen, 1 percent benzene, and other parameters set equal to 1990 industry average levels) using the complex model. The Agency believes that this performance standard is appropriate since it would require the same emissions performance for all fuel suppliers while still providing suppliers greater flexibility in meeting the requirements of the Act. EPA believes that providing this additional flexibility is of greater significance for Class B area fuels than for Class C area fuels because of the greater capital and operating expenditures needed to achieve the much lower Class B RVP levels. In addition, EPA believes that the second option would have a significantly smaller effect on enforcement in Class A and B areas than in Class C areas. The enforcement problems associated with this proposal are considerably simpler to overcome for areas that are geographically distinct and are served by distinct fuel distribution networks. Class C areas that are mandated for inclusion in this program often overlap, and many of the areas that either already have opted into the reformulated gasoline program or are eligible to do so are adjoining. Class A and B areas, by contrast, tend to be served by distinct fuel distribution systems. Finally, the Agency believes the potential diminution of emission reductions in Class A and B areas would be offset by the increase in the number of areas that would be able to opt into the reformulated gasoline program. The potential increase in opt-in would result from the enlarged supply of usable oxygenates resulting from this option: since the complex model is anticipated to include parameters with significant reduction potential for NOx as well as VOC and toxics emissions, this option would allow fuel suppliers the flexibility to utilize a wider range of oxygenates in Class B areas. While EPA is not aware of any interactive effects (as defined in section III) among the parameters contained in the simple model, EPA anticipates that fuel parameters with dilution and interactive effects will be identified in the future, and that fuel suppliers may wish to have such parameters incorporated in the complex model to simplify certification of fuels with such parameters. If fuel parameters have negative dilution or interactive effects, then mixing of fuels containing these parameters in the fungible fuel supply could result in degradation of the emission performance of all fuel in the fungible fuel supply. Therefore inclusion of such fuel parameters in the complex model may not be appropriate. The issue of how to include fuel parameters with dilution or interactive effects in the complex model will be dealt with in more detail in the subsequent rulemaking that will define the complex model. EPA requests comment on this issue at this time. III. Vehicle Testing to Augment the Emission Models The negotiated agreement is largely silent on the use of vehicle testing to augment 18 the emission models. The agreement does state that "vehicle testing will be permitted to qualify new parameters but not to modify the coefficients of existing model parameters" and further states that as new parameters are added to the complex model, the model shall be used to quantify the effect of the new parameters. To the extent that the proposals in this section go beyond those discussed in the NPRM or outlined in the agreement, EPA believes that they are consistent with the intent of the agreement and the provisions of the Act. Comments are requested on the specific proposals presented in this section. NOTE 18 The distinction between "augmenting" the complex model through vehicle testing and "revising" the model itself is discussed more fully in Section III.A.1. A. Purposes, Objectives, and Limitations of Vehicle Testing 1. Purpose of Vehicle Testing Vehicle testing is the primary way that the effects of various gasoline formulations on motor vehicle emissions can be determined. As described above and in the NPRM, data from vehicle testing programs forms the bulk of the basis for the simple model described above. This will also be the case for the complex model when it is developed. At the same time, when the subject of fuel certification by vehicle testing is discussed, most people envision a single test program of two or three fuels with the decision to certify being derived solely from the results of this single test program. EPA believes that fuel certification through such a single test program is inherently less reliable than certification through a testing-based model due to the strong likelihood that a far greater amount of testing was used to develop the model than that involved in any single test program and due to the fact that the potentially varying and conflicting results of numerous test programs can be considered together in a model. Even when no other data on the emissions effect of a fuel parameter exists, the statistical variance associated with any limited testing program raises the concern that a fuel will show emission effects during testing that would not occur in-use. Therefore, EPA proposes that testing only be permitted in conjunction with the models to augment them where fuel effects on emissions are not covered in the models. A distinction is drawn between "augmenting" a model and "updating" or "revising" a model. Augmentation involves modifying a model's predicted emission effects based on the results of vehicle testing submitted to EPA by industry that quantify the emission effects of new parameters or the extension of emission effects from already-included parameters, as discussed at length in this section. Augmentations to a model would be valid for a limited period of time and would apply only to those fuel suppliers requesting the use of the augmentation or claiming emission effects from the fuel parameter for which the augmentation was developed. Augmentations would be permitted on a temporary basis only as discussed below in Section III.A.5. Updates or revisions to a model would involve changes to the base model (to which further augmentations would be applied), and would affect all fuel suppliers. Revisions to the model would be developed by EPA and are expected to involve a rulemaking process. Revisions may involve new parameters, extension of the effects of already-included parameters, or changes to the coefficients of already-included parameters. EPA generally envisions that augmentations that are valid at the time the model is being revised would be proposed as revisions to the model. Although it is likely that any such augmentations would be proposed and accepted as a revision to the model given the extent of the data required for the augmentation, whether such augmentations would be proposed and finalized as revisions to the model would depend on the level of statistical confidence in the test result, various factors such as the existence of valid concerns with the original data since the time of the augmentation, and test results or other data obtained by EPA or other parties that dispute the conclusions drawn from the testing performed for the augmentation. The most likely time for concerns with the original data to come to light would be in comments provided on a proposal. As a result, in most instances, EPA would anticipate that augmentations would be proposed as revisions to the model. As discussed in Section II, data with which to develop an emission model is limited for many fuel parameters. The simple model includes only some of the fuel parameters that are known to have an effect on emissions. EPA has chosen to include in the simple model only those parameters for which the emission effects have been quantified with sufficient assurance to justify their inclusion. The complex model required for use in 1997 and to be released in 1993 is intended to include a number of additional parameters whose effects on emissions are now being substantiated and quantified through ongoing Agency and industry test programs. These parameters include sulfur, T90, and olefins. Additional parameters which affect emissions will periodically be incorporated in the complex model as they are discovered and quantified over time. In order to encourage fuel suppliers to identify and develop innovative and cost effective fuel reformulations that reduce emissions and to permit their use prior to such time as they could be incorporated into the complex model, EPA considers the use of vehicle testing to augment the models to be an important alternative to fuel certification by modeling alone. 2. Objectives of the Vehicle Testing Process Under the Simple Model EPA believes that fuels certified by vehicle testing should be evaluated in conjunction with the most complete emission model available to more accurately determine the emission benefits of the fuels being tested. Therefore, EPA proposes that with the following exception, vehicle testing not be permitted to augment the simple model. Approval to use oxygenates at concentrations greater than 2.7 weight percent oxygen in the form of MTBE 19 or 2.1 weight percent oxygen in the form of other oxygenates, up to the waivered limit for the oxygenate in question, would require the submittal to EPA of data that demonstrates that the oxygenate in question does not increase NOx emissions. EPA would evaluate such data, along with data already available and obtained from other sources, and process such petitions expeditiously. For such fuels, VOC and toxics emissions would still be determined using the simple model. States would be permitted to prohibit specific oxygenates in non-VOC controlled reformulated gasolines at levels in excess of 2.1 weight percent oxygen (2.7 weight percent oxygen in the form of MTBE) as per section II.A.2 unless the Administrator finds that specific oxygenates do not increase NOx at higher levels. If the Administrator were to make such a finding, the oxygenate in question would be permitted in reformulated gasolines up to the level specified in the finding without further restriction. NOTE 19 Note that the waivered limit for MTBE corresponds to an oxygen concentration of 2.7 weight percent. Hence a fuel supplier wishing to use MTBE at greater concentrations would have to complete the waiver process as well as the vehicle testing process outlined in this section. EPA further proposes that to obtain approval to use an oxygenate at such elevated levels, a formal vehicle testing program to augment the simple model be required as outlined in this section. Based on results from the testing program, petitioners would have to demonstrate that the oxygenate at such concentrations does not increase NOx emissions. For such fuels, VOC and toxics emissions still would be determined using the simple model. EPA requests comment on whether less burdensome requirements (relative to those outlined in the remainder of section III.) are appropriate for oxygenate concentrations between 2.1 and 2.7 weight percent oxygen, and if so, what such requirements should be. 3. Objectives of the Vehicle Testing Process Under the Complex Model EPA believes that the objective of testing under the complex model should be to evaluate the emission effects of fuels whose emission effects cannot be adequately represented by the model. Such fuels would include fuels claiming emission effects from parameters not included in the complex model as well as fuels containing complex model parameters at levels beyond the range covered by the model. Without this constraint, it may be possible for a fuel producer to use the statistical variance associated with testing to demonstrate emission effects through the testing option which would not be demonstrated in-use, when tested to a greater degree, or when modeled. For example, a fuel that would fail to meet the VOC requirement by a small margin when evaluated under the complex model could be tested and potentially pass due to the testing error associated with any vehicle testing program. In addition, allowing testing of existing modeled parameters would essentially make the complex model, and thus, the emission performance standards, a fluid target. Fuel producers would lose the certainty associated with a fixed model and the confidence that their capital investments will be useful for at least a fixed amount of time. Therefore, EPA proposes that vehicle testing be used only to determine the emission effects of the parameter(s) not adequately represented by the complex model. The complex model would be used to determine the emission effects of fuel parameters covered by the model since the model would likely be based on more data than would be generated by any individual test program. The emission effects of the fuel in question would be determined by combining the emission effects determined through vehicle testing with the emission effects predicted by the complex model. Furthermore, EPA proposes that each testing program be used to identify the effects of only one new fuel parameter, unless the changes in other fuel parameters are a natural and inherent consequence of the primary fuel modification. These proposals, taken together, would alleviate the concerns raised above. In addition, EPA proposes that fuel suppliers opting to augment the complex model through vehicle testing must examine the extent to which emissions are affected when fuels containing the fuel parameter(s) being tested are mixed with other fuels. The Agency is concerned with two potential problems when different fuels are combined. First, the emission effects of a parameter, as determined from vehicle testing, may not behave linearly as fuels with one level of the parameter are mixed with fuels with different levels of the same parameter. The degree to which this process occurs is referred to in this notice as the parameter's dilution effect. Second, the emission effects of various fuel parameters may be affected by the presence or level of other fuel parameters. The degree to which this process occurs is referred to in this notice as the interactive effect. If such effects were to be present, actual emission performance of the fuel mixture in-use could be worse than the emission performance predicted from the complex model augmented by vehicle testing results. Therefore, EPA proposes that the testing process be structured so as to identify dilution and interactive effects. Since the presence of adverse dilution and interactive effects could seriously undermine the in-use effectiveness of this program, EPA believes that the only alternative to testing for such effects would be to segregate the fuel in question throughout the distribution system. Even this alternative may not be fully satisfactory, since such fuels would still be mixed with other fuels in vehicle fuel tanks. 4. Limitations on Vehicle Testing In addition to the limitations on testing described in the previous two sections, EPA proposes that petitioners be required to obtain advance approval from the Agency for proposed vehicle testing programs. EPA would only consider petitions to augment the model based on the results of approved testing programs. EPA would further retain the discretion to evaluate other data when evaluating petitions to augment the complex model and when determining the nature, extent, and limitations of the augmentation. Petitioners would be required to include the following information when submitting a test program plan for approval: the fuel parameter to be evaluated for emission effects; the number and description of vehicles to be used in the test, including model year, model name, VIN number, mileage, emission performance, and technology type; the fuels to be used in the testing program, characterized as defined in section B.4.; the pollutants and emission categories intended to be evaluated; the methods and precautions to be used to ensure that the effects of the parameter in question are independent of the effects of other parameters already included in the complex model; a description of the quality assurance procedures to be used during the test program, and the identity and location of the organization performing the testing. For test programs that focus only on exhaust emissions, petitioners would have to include a justification as to why nonexhaust emissions should be assumed to be unaffected by the fuel parameter in question. EPA fully anticipates, and would encourage petitioners to submit the information listed above in stages beginning with the most general and ending with the most specific in order to streamline the approval process and eliminate wasted effort. EPA would provide petitioners with a justification for rejection of a proposed testing program that fails to provide adequate information and assurances as described above. Rejected programs could be modified to address Agency concerns and re-submitted for approval. These provisions would provide the Agency with greater assurance that petitioners would not selectively report test results to the Agency that support their petitions. Petitioners would still be able to "game" the testing process by pre-screening vehicles to obtain a test fleet with the desired sensitivity to the proposed parameter. However, such a test fleet would have to be re-tested as part of the formal test program and hence would be subject to the variability inherent in vehicle testing, which would tend to reduce the gaming benefits from pre-screening. EPA believes that the risks and costs associated with re-testing would dissuade petitioners from attempting to manipulate the testing process in this manner. EPA further proposes that the results of all approved testing programs be submitted to the Agency, even if the parameter in question proves not to provide an emission benefit. The Agency believes this requirement is necessary to ensure that all available data is at the Agency's disposal when evaluating proposed augmentations to the complex model and when updating the model itself. EPA does not intend to use this provision to limit legitimate, innovative test programs. Rather, EPA is only interested in preventing the creation of artificial fuel parameters that claim to be the source of emission effects which are in reality only normal statistical variability. For example, a fuel's 10 percent distillation point (T10) is closely related to its RVP. A testing program to identify the effects of T10 may indicate that an emission effect from T10 exists when the effect is actually due to differences in the fuels' RVPs or to statistical variability. At the same time, some measure of a fuel's volatility above 100 deg. F (the RVP test is conducted at 100 deg. F) could be very relevant to running losses, where tank temperatures can reach 120-135 deg. F. A proposed test program to identify the effects of T10 would require the petitioner to identify specific measures to be taken to isolate the emission effects of T10 from those of RVP, which is anticipated to be included in the complex model. In this example, EPA might require that the candidate and candidate-baseline fuels contain identical RVP levels. This provision would eliminate one potential means by which petitioners would be able to "game" the testing process and produce fuels that meet requirements under the model but do not meet requirements in-use. 5. Duration of Acceptance of Emission Effects Determined by Vehicle Testing The Agency is concerned that fuel suppliers not be allowed to claim emission effects in perpetuity based on the testing program described in this section due to the lower statistical confidence in the effects compared to those included in an updated complex model. The Agency also recognizes the need for fuel suppliers to recoup investments made to reformulate gasoline, including investments to utilize the emission effects identified through vehicle testing. EPA therefore proposes that petitioners be permitted to use emission effects determined through vehicle testing only for a limited period of time. In general, this period of time would extend until an updated version of the complex model takes effect. As discussed in section 1, EPA anticipates that most currently valid augmentations to the complex model would be proposed for inclusion in the updated model. Assuming that no serious, valid comments were received arguing against inclusion, such augmentations would be included in the updated model. Updates to the complex model will be proposed no more than five years apart. Since some augmentations may be in place for a relatively short period of time before the model is next updated, the Agency may not be able to adequately assess the augmentation. However, if a proposed update to the complex model is issued within three years of the time at which the augmentation takes effect, then in certain circumstances, fuel suppliers would be permitted to continue using the augmentation to determine the emission effects of reformulated gasolines. Specifically, if the Agency does not formally accept, reject, or modify the augmentation in question for inclusion in the updated complex model, then the augmentation would remain available until the next update to the model takes effect. If the Agency reviews the augmentation and either excludes the augmentation entirely or includes the augmentation in a modified form, then the augmentation would remain available for five years from the date the augmentation took effect or for three years of fuel production, whichever is shorter. This provision, however, would apply only to those refiners that either contributed 50 percent or more of the costs directly attributable to testing in support of the augmentation, or that have already begun producing a fuel utilizing the augmentation at the time of the proposal. In the latter case, the refiner would be able to continue producing fuel utilizing the augmentation up to the maximum fraction of fuel production which had previously utilized the augmentation and only to the extent (on average) that the augmentation had been used (e.g., up to or down to the average concentration or level of a new parameter or the extension of an existing parameter). Fuel suppliers not meeting either of these two criteria would be able to use the augmentation until the date the update to the model is promulgated. The minimum allowable period of five years from augmentation approval or three years of production of a certified fuel, whichever is shorter, is intended to provide fuel suppliers which invested substantially in the augmentation through either vehicle testing or refinery modifications with essentially the same period of time to recoup the costs regardless of when EPA grants them the augmentation. By restricting the continued use of the augmentation only to those fuel suppliers who would otherwise be most economically disadvantaged, EPA believes it can minimize the environmental detriment that might otherwise occur. EPA requests comment on this proposal. EPA further proposes that augmentations to the model for the effects of a given parameter over a particular range be permitted only once. Whether the emission effects of a parameter are either included in an updated model or not, once the minimum time period for use of a model augmented with the effects of that parameter has expired, the augmentation can neither be used or renewed (even with data from a second identical test program). Further testing would be permitted, however, to provide EPA with the additional data needed to include the effect in a future update to the model. 6. Application of Augmentations The testing process outlined in this section is focused on certifying a specific fuel with a specific concentration of the relevant parameter(s). However, fuel suppliers may wish to produce a range of fuels incorporating parameters for which testing has already been performed without having to repeat the testing process. The Agency recognizes the need to preserve flexibility for fuel suppliers given variations in crude oil feedstocks and the refining process. However, the Agency also recognizes the need to ensure the emission reduction benefits of fuels deemed to be reformulated gasolines are actually achieved in-use. The emission benefits of parameters as determined through testing of particular fuel formulations are difficult to extrapolate to other formulations due to potential interactive and dilution effects. EPA therefore proposes that fuel suppliers be permitted to claim the emission effects of parameters determined through vehicle testing for other fuels subject to the following conditions. First, the concentration of the parameter must not exceed the concentration of the parameter in the candidate fuel for which testing was performed if increasing the concentration of the parameter is beneficial to emissions, or be less than the candidate fuel concentration of the parameter if the opposite is true, since the emission effects of the parameter at such levels would not be known. For example, if testing of an emission-reducing additive at concentrations in excess of 5 percent had never been performed, then that additive would not be permitted at concentrations in excess of 5 percent; further, if a naturally-occurring emission-increasing ingredient had never been tested in reformulated gasoline at concentrations less than 10 percent, then gasolines would not be given credit for any marginal emission benefits of the ingredient at concentrations of less than 10 percent. Second, the parameter may only be introduced into fuels containing parameters for which interactive effects with the parameter in question have been tested as described in section III.B.6. This requirement would help assure that the emissions benefits predicted for a given fuel are actually achieved in-use by preventing fuel suppliers from introducing fuels with unknown and potentially unfavorable interactive effects into the fuel supply. 7. Exclusive Rights, Confidentiality, and Public Comment on Proposed Augmentations The Agency recognizes that the provision of exclusive rights for the use of emission-affecting parameters to fuel suppliers who conduct vehicle testing may encourage more testing than would occur without exclusive rights. However, the Agency also recognizes that provision of exclusive rights may increase the overall cost of the reformulated gasoline program, since cost- saving reformulation methods would not be freely available. The Agency further recognizes that the regulatory burden of administering a system of exclusive rights would be significant and does not believe that the benefits of such a system (in the form of more rapid innovation) would justify its costs (in the form of less-widespread adoption of innovations once discovered and higher administrative costs). Further, there is some question whether EPA would have statutory authority to grant such exclusive rights, and in any case fuel suppliers are able to apply for patents on additives or reformulation process technology independent of any administrative system of exclusive rights for emission effects identified through vehicle testing. Therefore, EPA proposes that any fuel supplier be permitted to utilize any emission effect identified through vehicle testing, subject to the constraints of patent law or other applicable legal restrictions. EPA requests comment on this approach and on whether it might discourage the development of innovative formulations not protected by patents or other applicable legal restrictions. EPA also requests comment on whether the expected benefits of any additional innovations that may be stimulated by the granting of exclusive rights would warrant the regulatory burden and reduced market efficiency associated with a system of exclusive rights. EPA also invites comments on its statutory authority to grant exclusive rights. The Agency also recognizes that, given the costs of vehicle testing and reformulated gasoline production, fuel suppliers may wish to keep vehicle testing results confidential for competitive reasons. However, confidentiality would eliminate the possibility of public comment on proposed augmentations to the model. The Agency anticipates that public comment on proposed updates to the model would be permitted, since model updates would be subject to the rulemaking process. The Agency also proposes that public comment on requests by fuel producers to augment the models through vehicle testing also be permitted. Providing for comment would allow interested parties to review and comment on the testing process employed and to submit supporting or countervailing data. Further, since proposed augmentations to the model would be likely to be considered for inclusion in future updates to the complex model, other fuel suppliers may have a significant interest in evaluating the impact of the proposed augmentation on their fuels and, in some cases, may undertake additional testing to confirm or disprove the proposed emission effect. The Act provides the Agency with 180 days to act on requests for fuel certification, which the Agency interprets to include verification of vehicle test results once a petition to augment the model is complete. EPA believes that this time is sufficient to permit public comment on vehicle test results. The Agency recognizes that provision for public comment implies that vehicle testing results could not be treated as confidential business information; however, EPA believes the potential gains in the quantity and quality of data used to determine augmentations are significant, and outweigh the potential benefits from additional testing that might be encouraged by treating the information as confidential. EPA requests comment on the proposals outlined above regarding non-exclusivity of rights to use emission effects established through vehicle testing and the opportunity for public comment. B. General Vehicle Test Program Requirements 1. Seasonal Variation in Testing Requirements In order to be certified as reformulated, a gasoline must meet VOC emission requirements in the high ozone season; separate toxic emission requirements in summer and winter or on an averaged year-round basis; and NOx emission requirements and the oxygen, benzene, and heavy metal content requirements year-round (see section III of the NPRM.) As discussed in Section II of this notice, the Agency does not have sufficient data to model winter emissions. While differences between the effects of fuel parameters under summer and winter conditions beyond those discussed in Section II may exist, the Agency does not have any evidence to date to suggest that they are significant. Therefore, EPA will apply the model developed for summer emissions to winter fuels as well for purposes of determining their VOC emissions. The Agency is concerned that allowing winter testing for some fuel parameters while modeling the effects of other parameters based on summer emission data creates the possibility of "gaming" the testing process. Fuel suppliers could use the summer model to determine the effects of parameters that would behave unfavorably under winter conditions and use winter testing to determine the effects of parameters that would behave favorably under winter conditions. This possibility may result in fuels being certified for winter use (through a combination of winter testing and summer modeling) that result in smaller emission reductions in-use than are intended by the Act or than would occur by using the summer model. Therefore, EPA proposes at this time that all testing be performed under summer ambient conditions. The Agency requests comment on this proposal, on whether winter testing should be permitted, and on the potential for gaming if winter testing were permitted. 2. Pollutants To Be Measured When testing to augment the simple model (i.e., fuels containing oxygenates at levels beyond those covered by the model), EPA proposes that only the exhaust emissions of carbon monoxide (CO), carbon dioxide (CO2), hydrocarbons, and nitrogen oxides (NOx) be reported. While only the NOx measurement would be used to determine whether the oxygenate at the levels in question increases NOx emissions, the Agency believes the reporting of the other emission measurements would be necessary for test validation purposes and would add little, if any, cost to the test program. To the extent testing is performed to augment the complex model, EPA proposes that it be performed to determine the emission effects on all the pollutants covered by the reformulated gasoline certification requirements, including toxics. (As discussed above, carbon monoxide and carbon dioxide emissions should be measured to permit validation of test results.) Failure to have such a requirement could allow fuel producers to "game" the certification requirements by permitting them to utilize the modeling option for one pollutant when it would be advantageous and the test results for another pollutant when it would be advantageous. Certified reformulated gasolines may then not meet all of the applicable emission reduction requirements in-use. For example, the model augmented by test results may indicate that a fuel meets the VOC requirement but fails the toxics requirement, while the model alone may indicate that the fuel meets the toxics requirement. Allowing the petitioner to claim the toxics emission effects predicted by the model while claiming VOC benefits determined through testing would ignore fuel effects on toxics that may not be addressed by the model. Testing costs could be significantly reduced if only VOC and NOx emissions were measured by testing, and toxics emissions were allowed to be modeled. However, since the testing option could only be used when the candidate fuel's parameters fall outside of the range of the model, EPA believes that seldom will adequate information be available to allow toxics emissions to be adequately modeled if adequate information was not available to do so for VOC and NOx. If a fuel parameter is expected to affect VOC or NOx and is not covered by the model, toxics emissions may very well be affected and should be measured. However, the Agency proposes that automatic testing for dilution and interactive effects be limited to NOx and VOC emissions. As discussed in section II, toxics emissions are largely (1) due to specific precursors contained in the fuel and (2) otherwise proportional to VOC emissions. Therefore, EPA expects that any dilution or interactive effects for toxics emissions should result from such effects on VOC emissions. However, EPA reserves the right to require that toxics be measured during such testing when evidence exists that adverse dilution and interactive effects may exist for toxics and not VOC and NOx emissions. Furthermore, as discussed more fully in section III.D, the Agency proposes that in most cases duplicate testing not be required for the measurement of toxics emissions. This would also reduce the testing costs associated with evaluating the toxics emission impacts of the fuel parameter in question. To better optimize the test program for the particular fuel parameter being evaluated, the Administrator may approve a request to waive certain of the pollutant measurement requirements proposed in this section. Any such waiver would have to be obtained in advance. A request for such a waiver should include an adequate justification for the requested change, including the rationale for the request and supporting data and information. Such a request must justify the reason that measurement of certain pollutants is clearly not necessary, and identify those pollutants for which additional testing may be warranted. An example might be a petition that reducing the concentration of a certain high molecular weight paraffin decreased VOC emissions even though the overall concentration of similar paraffins remained the same. In this case the petitioner may be able to justify a reduced need for toxics measurement, since the concentration of one high molecular weight paraffin relative to that of another would not be expected to impact toxics concentrations in the exhaust. However, given the uncertainty associated with such a fuel change significantly affecting VOC emissions, a greater amount of testing may be justified for VOC emissions to enable a greater degree of statistical confidence in the test results. As a result, the fuel supplier may be able to justify to EPA that a greater amount of testing for VOC emissions and a lesser amount of testing for toxics emissions may be warranted. 3. Types of Emissions to be Monitored Under this rulemaking, when testing oxygenates to augment the simple model, the only pollutant of interest is NOx. EPA therefore proposes that such testing involve testing for exhaust emissions only, since NOx is present only in exhaust emissions. However, when testing to augment the complex model, NOx, VOC, and toxics emissions are all relevant to determining the parameter's emission effects; the latter two pollutants occur in both exhaust and nonexhaust emissions. Fuel parameters that affect nonexhaust emissions are likely to have an exhaust emission effect as well, while the opposite is not necessarily true. As a result, combining testing for some emission types with modeling for other emission types would reduce the cost of vehicle testing while not compromising the integrity of the testing process, while combining testing for some pollutants or seasons with modeling for other pollutants or seasons might compromise the integrity of the testing process. EPA therefore proposes that the testing option be coordinated with the modeling option such that a fuel producer could (1) test for all emission types (exhaust, evaporative, running losses, and refueling); (2) test for exhaust, evaporative, and running loss emissions and model refueling emissions, or (3) test for exhaust emissions only and model evaporative, running loss, and refueling emissions. For example, the producer would likely choose to test for all emission types if the parameter in question were expected to favorably affect all emission types. However, if the parameter in question were expected to favorably affect exhaust and running emissions but not to affect refueling emissions, the producer would likely choose to model refueling emissions while testing for the other emission types. If the parameter in question were expected to affect exhaust emissions only, the fuel producer would likely choose to test for exhaust emissions while modeling evaporative, running loss, and refueling emissions in order to reduce the cost of the test program. If the fuel supplier wishes to model nonexhaust emissions for a fuel or fuels undergoing exhaust emission testing, the fuel supplier would have to demonstrate that the candidate fuel's nonexhaust emissions can be determined accurately by the complex model. Limitations on the applicability of the complex model will be included in the complex model rulemaking. If the fuel supplier cannot demonstrate compliance with these limitations for the fuel or fuels in question, then nonexhaust emission testing would have to be conducted. By allowing nonexhaust emissions to be modeled under appropriate circumstances even though exhaust emissions are determined through testing, EPA believes that the candidate fuel's emissions would be more accurately determined, and testing resources could be focused on those emission effects which the model predicts with the least degree of certainty (i.e., exhaust emissions). The model will be based on emission testing results from a large number of vehicles, resulting in greater accuracy from using the model to predict nonexhaust emissions than from a vehicle testing program if the fuel can be modeled accurately. Additionally, by freeing resources for testing, the results from testing could then be used to improve the models over the long run. 4. Test fuels To isolate the effects of compositional changes on emissions, EPA proposes that a candidate-baseline fuel be defined and produced for each candidate fuel. The candidate-baseline fuels would help ensure that emission effects identified through vehicle test programs reflect the emission effects of the parameter in question rather than the normal testing variability associated with the emission effects of other parameters. The candidate-baseline fuels also would more closely reflect the properties of the fuels found in-use and would more closely reflect the properties of the fuels for which the parameter effects would be claimed. The Clean Air Act baseline fuel would not satisfy these requirements; therefore, EPA proposes that the candidate- baseline fuel for augmentation of the simple model contain 25 volume percent aromatics, 1 volume percent benzene, and no oxygenates; have an RVP of 8.1 psi, and have Clean Air Act section 211(k)(10)(B)(i) baseline gasoline levels of all other parameters, including the parameter in question. EPA further proposes that the candidate-baseline fuel for augmentation of the complex model contain 25 volume percent aromatics, 1 volume percent benzene, and 2.0 weight percent oxygen in the form of MTBE (2.0 percent oxygen in the form of the parameter being tested if it is an oxygenate other than MTBE); have an RVP of 8.1 psi, and have Clean Air Act section 211(k)(10)(B)(i) baseline gasoline levels of all other parameters, including the parameter in question. If the parameter is not specified for CAA baseline gasoline, EPA proposes that the level of the parameter in the candidate-baseline fuel be comparable to the level found in gasoline representative of in-use reformulated gasolines; EPA further proposes that petitioners be required to obtain approval for the candidate-baseline level of this parameter from the Agency prior to beginning their vehicle test programs. Such approval would depend in part on the use of an appropriate basis for determining "representative" gasoline. EPA further proposes that the candidate fuel be compositionally identical to the candidate-baseline fuel except for the level of the parameter in question and, to the extent necessary to compensate for changes in the level of the parameter in question, the level of paraffins. The level of the parameter in question would be zero for parameters neither defined in CAA baseline fuel, nor present in representative in-use reformulated gasolines. If the parameter is defined in CAA baseline fuel, then it would have to be present in the candidate-baseline fuel at CAA baseline fuel levels. If the parameter is not defined for CAA baseline fuel but is found in representative in-use reformulated gasolines, then it would have to be present in the candidate-baseline fuel at the levels found in such representative gasolines. EPA further proposes that petitioners be permitted to request the Administrator to establish alternative levels for the parameter in question in the candidate-baseline fuel as part of their initial petition in order to expedite the determination of the candidate-baseline fuel properties. EPA proposes that for all candidate-baseline fuels, paraffin content be altered to balance changes in the levels of other fuel constituents to best isolate the effects of the fuel parameter being varied in concentration. Paraffin content is proposed to balance other fuel composition changes since paraffin effects on emissions are thought to be more neutral than the effects of other, more complex major constituents of CAA baseline gasoline (such as olefins and aromatics) due to their straight chain molecular form. EPA requests comment on the proposed definition of the candidate-baseline fuel and on the use of paraffin levels to balance changes in other fuel components. In determining the composition of candidate-baseline fuel, two other issues also would have to be addressed. First, non-compositional properties of the candidate and candidate-baseline fuels, such as RVP and T90, may differ as a natural result of compositional differences between the two fuels. EPA proposes that the complex model be used to compensate for such differences when evaluating vehicle testing results. Second, variations due to blending may cause properties not included in the complex model to vary between the candidate and candidate-baseline fuels, and such properties may have significant emission effects not predicted by the model. Hence EPA proposes that the properties of the candidate-baseline fuel be required to be the same as those of the candidate fuel within the tolerances defined in Table III-1. Failure to meet this requirement would reduce the certainty that emission effects found in vehicle testing are due to the parameter in question and not due to emission effects of parameters included in the complex model that differ from the effects predicted by the model. However, if a petitioner could show that it is not feasible to meet all such tolerances for the petitioner's candidate-baseline fuel due either to (1) naturally-resulting changes in fuel parameters arising from changes in the parameter(s) in question or (2) blending technology limitations, EPA would consider waiving the relevant tolerances. However, the request must come prior to the start of the test program. The Agency further proposes to use the complex model (including prior test results used to augment the model where appropriate) to adjust for differences between the candidate and candidate-baseline fuels. Due to the difficulty in accurately measuring the initial boiling point (IBP) and the fact that its value tends to be controlled by the RVP and the 10% distillation point, EPA proposes that no limitations be placed on IBP blending tolerances for testing purposes. EPA further proposes that a minimum octane requirement of 87 (measured by the (R+M)/2 method) be met for all fuels used in vehicle testing. Table III-1.--Fuel Parameter Blending Tolerances for Candidate-Baseline Fuel Blending, Relative to the Candidate Fuel Tolerance Parameter /1/ Sulfur, ppm +/-25 Benzene, vol percent +/-0.3 RVP, psi +/-0.3 10%, deg.F +/-5 50%, deg.F +/-5 90%, deg.F +/-5 End Point, deg.F +/-20 Oxygenates, vol percent +/-1.5 Aromatics, vol percent +/-2.7 Olefins, vol percent +/-2.5 Saturates, vol percent +/-2.0 /1/ Letter to Paul Machiele, EPA, from Robert H. Pahl, Manager, Fuels and Lubricants, Phillips Petroleum Co., May 13, 1991. Blending tolerances for detergent additives have not been defined since the measurement methods for such additives have not yet been determined. EPA requests comment on the appropriateness of including such tolerances for detergent additives and the appropriate tol