------------------------------------------------------------------------------- On 07/20/94 [L-S document 497082, 59 FR 36944, 3360 lines] ENVIRONMENTAL PROTECTION AGENCY 40 CFR Part 80 [AMS-FRL-5007-9] Regulation of Fuels and Fuel Additives: Standards for Reformulated and Conventional Gasoline AGENCY: Environmental Protection Agency. ACTION: Direct final rule. ----------------------------------------------------------------------------- SUMMARY: The Clean Air Act, as amended in 1990 (CAA or the Act), mandated that the Environmental Protection Agency (EPA or the Agency) promulgate regulations requiring that gasoline sold in certain areas be reformulated to reduce vehicle emissions of toxic and ozone-forming compounds and that gasoline sold outside these areas would not be more polluting than it was in 1990. On February 16, 1994, EPA published the final rule establishing performance standards and compliance provisions for conventional and reformulated gasoline (RFG). This direct final rule (DFRM) makes minor corrections, clarifications, and revisions to various provisions in the final reformulated gasoline rule. This action addresses the following issues: Work-In-Progress (WIP) baseline adjustments; JP-4 baseline adjustments; summer/winter season definition for baseline determination; complex model valid range extension for conventional gasoline baselines; valid range limits for aromatics, oxygen, benzene, and RVP; clarifications to the VOC and NOX extrapolations in the complex model; clarifications of seasonal condition inconsistencies; and enforcement corrections/clarifications associated with the reformulated gasoline and anti-dumping regulations, as well as several technical clarifications and typographical corrections. EFFECTIVE DATES: This rule will be effective September 19, 1994 unless notice is received by August 19, 1994 that adverse or critical comments will be submitted or that an opportunity to submit such comments at a public hearing is requested. If such comments or a request for a public hearing are received by the Agency, then EPA will publish a subsequent Federal Register notice withdrawing from this action only those items which are specifically listed in those comments or in the request for a public hearing. See SUPPLEMENTARY INFORMATION for further discussion on submission of public comment. ADDRESSES: Interested parties may submit written comments (in duplicate, if possible) to Public Docket No. A-94-30, at Air Docket Section, U.S. Environmental Protection Agency, Waterside Mall, Room M-1500, 401 M Street SW., Washington, DC 20460. The Agency requests that commenters also send a copy of any comments to David Korotney at the address listed in the FOR FURTHER INFORMATION CONTACT section. Materials relevant to the reformulated gasoline final rule and this direct final rule are contained in Public Dockets A-91-02 and A-92-12, 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: David Korotney, U.S. EPA (RDSD-12), Regulation Development and Support Division, 2565 Plymouth Road, Ann Arbor, MI 48105, Telephone: (313) 668- 4507. To Request Copies of this Action Contact: Delores Frank, U.S. EPA (RDSD-12), Regulation Development and Support Division, 2565 Plymouth Road, Ann Arbor, MI 48105, Telephone: (313) 668-4295. SUPPLEMENTARY INFORMATION: Public Comments For parties that submit adverse or critical comments, notify EPA of intentions to submit adverse comments, or request a public hearing, the Agency requests that commenters identify each of the items at issue by the specific preamble section numbers that discuss those items. For instance, adverse comments on the change to the oxygen valid range limits should include a reference to Section (Item Number) II.A.1 of the preamble. Adverse comments on any of the insubstantial errors in Section I of the preamble should include a reference to the identification code associated with each change in that section. For instance, adverse comments on the paragraph reference change in Sec. 80.41(h)(2)(iii) should include a reference to Item Number I-A. The EPA will withdraw from this direct final action only those specific provision(s) so identified. All other provisions included in today's notice will become effective on September 19, 1994. EPA believes that the use of a direct final rule is appropriate because the changes made are generally minor in nature and all are expected to be non- controversial. The DFRM will allow the Agency to finalize such changes in a timely manner. For instance, many of the changes contained herein clarify issues relevant to the development and auditing of individual baselines which, in general, must be submitted by either June 1, 1994 or September 1, 1994. Likewise, the reformulated gasoline program will commence on December 1, 1994. The clarifications and changes contained herein will promote successful implementation of this program. A copy of this action is available on the OAQPS Technology Transfer Network Bulletin Board System (TTNBBS). The TTNBBS can be accessed with a dial-in phone line and a high-speed modem (PHNZ 919-541-5742). The parity of your modem should be set to none, the data bits to 8, and the stop bits to 1. Either a 1200, 2400, or 9600 baud modem should be used. When first signing on, the user will be required to answer some basic informational questions for registration purposes. After completing the registration process, proceed through the following series of menus: (M) OMS (K) Rulemaking and Reporting (3) Fuels (9) Reformulated gasoline A list of ZIP files will be shown, all of which are related to the reformulated gasoline rulemaking process. Today's action will be in the form of a ZIP file and can be identified by the following title: RFG-DFRM.ZIP. To download this file, type the following instructions and transfer according to the appropriate software on your computer: ownload,

rotocol, xamine, ew, ist, or elp Selection or to exit: D filename.zip You will be given a list of transfer protocols from which you must choose one that matches with the terminal software on your own computer. The software should then be opened and directed to receive the file using the same protocol. Programs and instructions for de-archiving compressed files can be found via ystems Utilities from the top menu, under rchivers/de- archivers. Please note that due to differences between the software used to develop the document and the software into which the document may be downloaded, changes in format, page length, etc. may occur. The remainder of this preamble is organized into the following sections: I. Insubstantial Errors II. Valid Range Limits III. Complex Model IV. Enforcement Corrections and Clarifications V. Summer/Winter Season VI. Baseline Determination Adjustments VII. Public Participation and Effective Date VIII. Statutory Authority IX. Administrative Designation X. Regulatory Flexibility Analysis I. Insubstantial Errors The final rulemaking for the regulation of reformulated and conventional gasoline contained a number of errors, ambiguities, and misrepresentations of Agency intent which are being addressed by this direct final rule. Of these errors, many are minor in both form and effect. The minor errors do not require detailed discussions since they all have at most a negligible effect on compliance with the regulations, and require only simple corrections. Such minor errors took the form of typographical errors, grammatical errors, inadvertent omissions, and inadvertent insertions. The table below lists all the insubstantial errors that are being corrected in this direct final rule. Other errors are more substantial. The more substantial errors and the associated corrections have been discussed on an item-by-item basis in the following sections. Identification code Regulation reference Correction A 80.41(h)(2)(iii) Correct the paragraph reference from 80.101(g) to 80.101(h). B 80.41(j)(2) Correct the paragraph reference from (j)(1)(i) to (j)(1). C 80.41(m)(1) A missing word "of" is inserted into the text. D 80.42(c)(1) In the table, change the valid range limits for "Oxygenate" in volume percent to valid range limits for "Oxygen" in weight percent to match the values already present. E 80.42(a) Correct the definition of exhaust and nonexhaust VOC from nonmethane hydrocarbons to nonmethane, nonethane hydrocarbons. All calculations and equations are correctly based on nonmethane, nonethane hydrocarbons. F 80.42(b)(1)(ii), Add a concluding sentence 80.42(b)(2)(ii), and which clarifies that the use 8042(b)(3)(ii) of methanol and other non- alcohol, non-ether oxygenates in reformulated gasoline is limited to vehicle testing under the Complex Model. G 80.45(b)(3) Correct the inadvertent omission of the first decimal place in the baseline values for exhaust VOC and NOX in Table 3. H 80.45(c)(1)(iv)(A) The footnotes to Table 6 are clarified to indicate that the higher E300 limit can be no higher than 94 percent as described in paragraphs 80.45 (c)(1)(iv)(C)(6) and 80.45(c)(1)(iv)(D)(6). I 80.45(c)(1)(iv)(B) Change the word "and" to "and/or". J 80.45(c)(1)(iv)(C)(11) The paragraph references are corrected from (c)(1)(iv)(C)(8) and (9)" to "(c)(1)(iv)(C)(9) and (10)." K 80.45(c)(1)(iv)(C)(12) Correct the second "E300" to "DE300". L 80.45(c)(1)(iv)(C)(14) The paragraph references are corrected from "(c)(1)(iv)(C)(11) and (12)" to "(c)(1)(iv)(C)(13)." M 80.45(c)(1)(iv)(D)(11) The paragraph references are corrected from "(c)(1)(iv)(D)(8) and (9)" to "(c)(1)(iv)(D)(9) and (10)." N 80.45(c)(1)(iv)(D)(12) Correct "E30' 0" to "E300". O 80.45(c)(1)(iv)(D)(14) The paragraph references are corrected from "(c)(1)(iv)(D)(11) and (12)" to "(c)(1)(iv)(D)(12) and (13)." P 80.45(c)(8)(ii) An extraneous word "for" at the end of the paragraph is removed. Q 80.45(d)(1)(iv)(B) Change the word "and" to "and/or". R 80.45(e)(1)(ii) Correct the toxic emissions baseline values in the equations which were rounded incorrectly to mimic the correct values in Table 5 of 80.45(b)(3): In Phase I, "TOXICS2%=[100%x(TOXICS2- 47.58mg/mi)]/(47.58 mg/mi)" In Phase II, "TOXICS1%=[100%x(TOXICS1- 86.34mg/mi)]/(86.34 mg/mi)" S 80.45(e)(4)(iii) A missing word "an" is inserted into the last sentence. T 80.45(e)(5)(iv) & Add a concluding sentence 80.45(e)(6)(iv) which clarifies that the use of methanol and other non- alcohol, non-ether oxygenates in the Complex Model is limited to augmentation through vehicle testing. U 80.45(e)(9) & 80.45(e)(10) Correct the variable names in the equations to mimic the variable names in the definition list: "HSVOC1" is corrected to "VOCHS1", "DIVOC1" is corrected to "VOCD11", "RLVOC1" is corrected to "VOCRL1", "RFVOC1" is corrected to "VOCRF1", "HSVOC2" is corrected to "VOCHS2", "DIVOC2" is corrected to "VOCDI2", "RLVOC2" is corrected to "VOCRL2", "RFVOC2" is corrected to "VOCRF2". V 80.45(f)(1) The paragraph references are corrected from "(a), (c), and (d)" to "(c), (d), and (e)". W 80.45(f)(1) (i) & (ii) The units for E200 and E300 are corrected from "volume percent" to "percent evaporated". X 80.48(c)(1) Correct the paragraph reference from "(c)(1)(iv) of this section" to "(c)(1)(iv) of this section and 80.49(d)". Y 80.48(c)(1)(v) Revise last sentence to clarify that the model must be re-estimated after dropping the Bi term. Z 80.48(c)(2)(iii) A concluding sentence is added indicating that the centered form of the Complex Model will be made available upon request. AA 80.48(g) Correct "the augmentation petition" to "other augmentation petitions". AB 80.49(a)(5)(i) The "Candidate parameter" entry is deleted from the table. AC 80.49(b)(3)(iii) The paragraph reference is corrected from "(b)(2)(ii) and (b)(3)" to "(b)(2)(ii)". AD 80.59(a) Revise last sentence to clarify that closed-loop systems and adaptive learning components are minimum requirements. AE 80.65(d)(3) A cross-reference to the blendstock accounting requirement in 80.102(e) is corrected. AF 80.66(g)(1) and (g)(2)(ii) Cross-references to 80.45, pertaining to the calculation of per-gallon values for VOC, NOX, and toxics emissions performance reduction, are corrected. AG 80.68(c)(8)(ii)(A) and Cross-references to the (c)(9)(ii)(A) complex model in 80.45 are corrected. AH 80.68(c)(9)(ii)(B) A cross-reference to the annual toxics emissions weighting formula in 80.68(c)(9)(i)(B) is corrected. AI 80.68(c)(10(i) A cross-reference to the NOX emissions reduction percentage in 80.45 is corrected. AJ 80.69(a)(7)(ii) A cross-reference to the applicable correlation ranges in 80.65(e)(2)(i) is corrected. AK 80.69(b)(3) A cross-reference to the oxygen averaging requirements in 80.67(f) is corrected. AL 80.70(j)(11) The spelling of Allegheny County, Pennsylvania, is corrected. AM 80.75(j) A cross-reference to the survey provisions in 80.41(q)(2) is corrected. AN 80.81(h) A cross-reference to the sampling and analysis methodology in 80.46 is corrected. AO 80.90(b)(1) In the equation, correct the variable "BX" to "BZ". AP 80.90(e)(2) The paragraph reference is corrected from "(e)(2)" to "(e)(1)". AQ 80.91(e)(2)(iv) In the equation only, the variable "Njs" is corrected to njs", the variable "ns" is corrected to "Ns", the variable "pj" is corrected to "pjs". AR 80.91(e)(4)(i)(A) Correct the equation to include division by 100 as follows: UV=[AV/(100-OV)]-100 AS 80.91(e)(4)(i)(B) Correct the equation to include division by 100 in two places: UR=[BR- {S(OVixORi)}/100]/[{100- SOVi}/100] AT 80.91(e)(4)(ii)(A) Correct the equation to include division by 100 as follows: AV=UV(100-OV)/100 AU 80.91(e)(4)(ii)(A) Correct the definition of UV from "nonoxygenated parameter value" to "non- oxygenated parameter value". AV 80.91(e)(4)(ii)(B) Correct the equation to include division by 100 as follows: BR={URx[100- S(OVi)]+S(OVixORi)}/100 AW 80.93(a)(3)(ii) Re-word the first sentence to read "Petitions, 'showings' and other associated proof may be submitted to EPA prior to submittal of the individual baseline (per paragraphs (a)(1) and (a)(2) of this section)". AX 80.93(c)(9) A cross-reference to 80.93 (c)(8) is clarified. AY 80.101(e)(3) & (f)(4) References to California gasoline are clarified. AZ 80.102(b)(1) An extraneous word is deleted. BA 80.102(e)(2)(i) A cross-reference to 80.101(g) is added. BB 80.102(f)(2)(i) A typographical error is corrected. BC 80.125(a) A cross-reference to the reports required by 80.105 is corrected. BD 80.128(e)(2) A cross-reference to the assumptions pertaining to the use of RBOB in 80.69(a)(9) is corrected. BE 80.128(e)(5) A cross-reference to the sampling and testing rates in 80.69(a)(7) is corrected. BF 80.128(g)(3)(iii) A typographical error is corrected. BG 80.129(e) A cross-reference to the sampling and testing rates in 80.69 is corrected. II. Valid Range Limits The valid range limits in both the Simple and Complex Models specify the range for each fuel parameter outside of which the models cannot be used for the evaluation of emission performances. These limits ensure that the models will not be used for extremely high or low fuel parameter values which would compromise the validity of the models. Thus the valid range limits were instituted as a means toward maintaining the accuracy of the compliance calculations, and thus the integrity of the reformulated gasoline and anti- dumping programs. The Agency made every attempt to make the valid range limits as wide as possible to provide flexibility to refiners while maintaining a focus on the need for accurate performance estimates. This was especially true for the conventional gasoline valid range limits, as EPA wanted to avoid, to the extent possible, establishing provisions which would require refiners to reformulate their conventional fuels. To provide additional flexibility, the Agency also allowed the extension of the specified valid range for the Complex Model for conventional gasoline when a refiner's individual 1990 baseline fuel exceeds the valid range in one or more fuel parameters. Since publication of the final rule, the Agency has determined that the flexibility provided to refiners in the valid range limits requires some revision and clarification to avoid unnecessary and costly refinery modifications which have no long term environmental benefit. The changes to the regulations can be separated into two categories: changes to the specified valid range limits, and clarification of the provision for extending the valid range for individual refiner baseline fuels. Both of these topics will be described in detail below. A. Revised Valid Range Limits The valid range limits for the Simple and Complex Models were based on two different sets of data and were developed using different assumptions. The Simple Model valid range limits were determined following the regulatory negotiations held in 1991. The Complex Model valid range limits were based upon an examination of the sufficiency of data in the Complex Model database and the accuracy of extrapolations (See the Regulatory Impact Analysis for the final rule, Section IV.D). Because the valid range limits for the Simple and Complex Models were established independently and through different processes, the valid range limits for the two models were different from one another. Since promulgation of the final rule, the Agency has learned that the specified valid ranges may force refiners to make refinery modifications to comply with the regulations that are unwarranted under the Simple Model, and unnecessary under the Complex Model. Thus EPA is revising the valid range limits for oxygen content, RVP, aromatics content, and benzene content for the Simple Model, and oxygen content for the Complex Model. 1. Change to High End of Oxygen Valid Range The high end of the valid range limit for oxygen in both the Simple and Complex Models was based on the maximum amount of oxygen that an oxygenated fuel was expected to lawfully contain. Of all the oxygenates that will likely be used in the reformulated gasoline program, ethanol has the highest oxygen content at 0.35 grams of oxygen per gram of ethanol. The Agency used this value as a benchmark in determining the high end of the valid range for oxygen, assuming a 10 volume percent ethanol blend. However, since promulgation of the final rule, the Agency has learned that density variations in gasoline blendstocks may result in wide variations in the oxygen content of an oxygenated fuel on a weight percent basis despite the fact that the volume percent remains fixed. For instance, blending 10 volume percent ethanol into a higher density gasoline could produce a blend with an oxygen content as low as 3.4 weight percent, while blending 10 volume percent ethanol into a lower density gasoline could produce a blend with an oxygen content as high as 4.0 weight percent. Since the largest excise tax exemption available to refiners for the use of ethanol in gasoline blends is for oxygenated fuels that contain 10 volume percent, many ethanol blends will contain 10 volume percent ethanol. Thus it is essential that the high end of the valid range for both the Simple and Complex Models be raised to 4.0 weight percent. This change will allow fuels which are already being produced to be evaluated with the Simple and Complex Models. The change will also guarantee that no fuel oxygenated with ethanol will be excluded from the reformulated gasoline program due to an oxygen content that is outside the range of the model, as long as it complies with the volume limits applicable in a waiver issued under section 211(f) of the Clean Air Act. Since fuels with oxygen contents of 4.0 weight percent are already being produced, the change will have no additional impact on vehicle driveability. Also, since the models continue to be accurate between 3.5 and 4.0 weight percent, and the emission standards are not being changed, this change to the high end of the valid range for oxygen content will have no adverse impact on the environment. 2. Change to Low End of RVP Valid Range in the Simple Model The low end of the valid range for RVP in the Simple Model was based on the distribution of data used in the model's development, as well as a consideration of the needs of the reformulated gasoline and anti-dumping programs under the Simple Model. Since VOC emission reductions under the Simple Model are accomplished primarily through limits on maximum RVP levels, the valid range for RVP in the Simple Model only affects toxics compliance calculations. In promulgating the final regulations, the Agency determined that refiners had no incentive to reduce RVP below 6.6 psi for the purposes of complying with the toxics standards since reductions in fuel benzene and aromatics are much more effective in reducing emissions of toxic compounds. In contrast to the Simple Model, the absence of RVP caps and the impact of other fuel parameters on emissions under the Complex Model will likely result in large variations in RVP levels in reformulated gasoline. RVP control will continue to be the primary mechanism through which VOC emissions are reduced because RVP is the most cost-effective fuel parameter to control, and because the RVP effect on VOC in the Complex Model is quite large. As in the Simple Model, the valid range for RVP in the Complex Model was determined from an examination of the distribution of data used in the model's development and the needs of the reformulated gasoline and anti-dumping programs. The low end of the valid range was set at 6.4 psi to accommodate large reductions in RVP while maintaining the accuracy of the Complex Model. Fuel RVPs are expected to reach this low level in Phase II of the reformulated gasoline program, and the potential exists for them to be reached in Phase I as well. California has been developing its own program in which reformulated gasoline must meet more stringent requirements than in the federal program during the years 1996 and 1997 that the Simple Model will be in effect. Given California's more stringent requirements, the potential exists for fuel having an RVP of less than 6.6 psi to be either sold or used as a blendstock in and outside of California. With the low end of the valid range for RVP set at 6.6 psi, these California reformulated gasolines might not be certifiable as federal reformulated gasoline. Also, refiners trying to blend down tanks quickly at terminals in the spring to meet summer volatility requirements may end up with fuels that have RVPs as low as 6.4 psi. Since the low end of the valid range for RVP under the Complex Model is 6.4 psi, the Agency has determined that the low end of the valid range for RVP under the Simple Model should likewise be 6.4 psi. The change from 6.6 psi to 6.4 psi makes the low end of the valid range for RVP consistent throughout Phase I of the federal reformulated gasoline program, and provides an additional element of flexibility for refiners to complement the already established blending and enforcement tolerances. The change should have no effect on the environment, since presumably only cleaner fuels would be allowed certification under the Simple Model. Both the Simple and Complex Models are linear with respect to RVP for all pollutants, and this linear relationship would simply be extended from 6.6 psi down to 6.4 psi. 3. Change to High End of Aromatics Valid Range in the Simple Model It was and is the intention of the Agency to avoid, to the extent possible, establishing regulations which require refiners to reformulate their conventional fuels. The anti-dumping program is designed to ensure that a refiner's or importer's conventional gasoline stays as clean as it was in 1990, and does not require reformulation beyond those levels. Thus the valid range limits for conventional gasoline in both the Simple and Complex Models were designed to be as wide as possible while simultaneously ensuring the accuracy of the models. As an additional level of flexibility, the Agency also allowed for the extension of the valid range for conventional gasoline if a refiner's individual 1990 baseline fuel exceeded the specified valid range limits (see discussion in Section II.B. below). Since promulgation of the final regulations for the anti-dumping program, the Agency has determined that the valid range limits in the Simple Model for aromatics, which are more restrictive than those for conventional gasoline under the Complex Model, could be widened without any detrimental impacts on either the program or on the environment. Without such a change to the regulations, refiners may be forced to make changes to their refineries by 1995 that are not necessary in 1998 when compliance under the Complex Model is mandatory. Thus the Agency is changing the high end of the valid range for aromatics under the Simple Model from 45 volume percent to 55 volume percent. This change will ensure that the Simple Model can be used for as many conventional fuels as possible to show compliance under the anti-dumping program without the need to extend the valid range. The environment will not be adversely impacted since it was the Agency's intent to allow extensions of the valid range when a refiner's 1990 baseline fuel exceeded the specified valid range limits. The change will not affect RFG compliance, since aromatics are controlled by the reformulated gasoline standards for toxics. Driveability will not be affected since fuels with aromatics levels as high as 55 volume percent currently exist in-use. Also, consistency in the high end of the valid range limits for aromatics will be maintained throughout Phase I of the program. Because the relationship between toxic emissions and fuel aromatics levels is linear in the Simple Model, the change will not result in an inaccurate application of the model to higher aromatics levels. 4. Change to Low End of Aromatics Valid Range in the Simple Model The Agency has also determined that the low end of the valid range for aromatics in the Simple Model may not provide refiners with adequate flexibility under the reformulated gasoline program. For the Complex Model, the Agency determined that the relationships between aromatics and emissions could not be trusted at levels below 10 volume percent. However, the Agency determined that a flat-line extrapolation below 10 volume percent, in which no emission benefits or detriments result from lowering aromatics values below 10 volume percent, would provide greater flexibility without compromising the accuracy of the Complex Model equations in this range. Since promulgation of the final rule, the Agency has determined that the Complex Model approach to low values of aromatics can be appropriately applied to the Simple Model as well. Therefore, the Agency is changing the low end of the valid range for aromatics under the Simple Model from 10 volume percent to 0 volume percent, but will not allow any emission benefits in this range. The Agency did not intend to discourage the production of fuels that had very low levels of aromatics, and were, thus, qualitatively considered to be cleaner burning. If the low end of the valid range for aromatics is left at 10 volume percent, fuels with aromatics values of less than 10 volume percent that would otherwise be complying reformulated gasolines under the Simple Model will not be certifiable, despite the fact that those same fuels may be certifiable under the Complex Model. This change is expected to be environmentally benign, as few refiners will have any incentive to reduce aromatics below 10 volume percent for reformulated gasoline under the Simple Model. 5. Change to High End of Benzene Valid Range in the Simple Model As stated in Section II.A.3, it was and is the intention of the Agency, to the extent possible, not to establish regulations that require the reformulation of conventional gasoline under the anti-dumping program. However, the valid range limits for benzene under the Simple Model may in fact force refiners to reformulate conventional gasoline. The Agency has determined that the valid range limits in the Simple Model for benzene, which are more restrictive than those for conventional gasoline under the Complex Model, can be widened without any detrimental impacts. Without such a change to the regulations, refiners may be forced to make changes to their refineries by 1995 that are not necessary in 1998 when compliance under the Complex Model is mandatory. Thus the Agency is changing the high end of the valid range for benzene under the Simple Model from 2.5 volume percent to 4.9 volume percent. This change ensures that the Simple Model can be used for as many conventional fuels as possible to show compliance under the anti-dumping program without the need to extend the valid range. The change will not affect RFG compliance, since benzene is controlled by the reformulated gasoline standards for fuel benzene content. In like manner to raising the high end of the valid range for aromatics, consistency in the valid range limits will be maintained throughout Phase I of the program. The environment will not be adversely impacted since it was the Agency's intent to allow extensions of the valid range when a refiner's 1990 baseline fuel exceeded the specified valid range limits. As for aromatics, because the relationship between toxic emissions and fuel benzene levels is linear in the Simple Model, the change will not result in an inaccurate application of the model to higher benzene levels. B. Extending the Valid Range for Conventional Gasoline Under the anti-dumping provisions of the final rule, refiners use their individual 1990 baselines to determine compliance with the regulations under both the Simple and Complex Models. Depending on the compliance model being used, the values for particular fuel parameters are restricted by the valid range limits. For instance, if a refiner is using the Simple Model to comply with the anti-dumping regulations, VOC and NOX emissions are regulated through caps on the baseline levels of sulfur, olefins, and T90, while toxic emissions are regulated through an equation giving the benzene fraction of VOC emissions. Since the benzene fraction equation contains only benzene and aromatics as independent variables, the only valid range limits that apply to refiners using the Simple Model to comply with the anti-dumping regulations are those specified in Sec. 80.42(c)(1) for benzene and aromatics content. If, alternatively, a refiner is using the Complex Model to comply with the anti-dumping regulations, NOX and toxics emissions are regulated through the Complex Model. Thus the valid range limits that apply to refiners using the Complex Model to comply with the anti-dumping regulations are those specified in Sec. 80.45(f)(1)(ii) for oxygen content, sulfur content, RVP, E200, E300, aromatics content, olefins content, and benzene content. Section 80.91(f)(2)(ii) allows a refiner to extend the conventional gasoline valid range for the Complex Model if one or more of the fuel parameter values for its individual 1990 baseline fuel falls outside of the valid range specified in Sec. 80.45(f)(1)(ii). However, the regulations did not adequately specify the method, applicability, or limitations of such a valid range extension. As written, the regulations state that, for each baseline fuel parameter value which is outside of the Complex Model conventional gasoline valid range, the Complex Model valid range is extended only for such fuel parameters. The only stated limitation is that such an extension is applicable only to the applicable summer or winter season. The Agency has determined that this portion of the regulations is ambiguous, and requires revision. 1. Applicability of the Provision for Valid Range Extension The Agency has only been made aware of the potential need for extension of the valid range for olefins, aromatics, and benzene. Therefore, a provision for the extension of the valid range has only been made for the high end of the specified valid range for these three fuel parameters. In each case, if the valid range limit is extended, the refiner in question must still be limited by a valid range to eliminate the possibility that the Complex Model will be used for aromatics, olefins, and/or benzene values that are very high, which might compromise the primary objective of the anti-dumping program. As specified in the final regulations, a refiner is allowed to extend the Complex Model valid range for both baseline and compliance emissions calculations, but is not directed as to the specification of any new valid range limits. Such a provision for the extension of the valid range that does not include limitations on fuel parameter values that can be evaluated with the Complex Model would defeat the purposes of specifying a valid range, and was not the Agency's intent. The Agency has determined that the best approach to limiting the extension of the valid range is to allow target fuels to have values at least up to the baseline level. Since the baseline fuel is an "average" fuel of sorts, the Agency has also determined that refiners should be given some flexibility beyond the baseline value. For aromatics, this flexibility will be fixed at a value of 5.0 volume percent. For olefins this flexibility will be fixed at a value of 3.0 volume percent. And for benzene this flexibility will be fixed at a value of 0.5 volume percent. Thus, for example, the extended valid range limit for aromatics would be equal to the individual refiner's baseline fuel value for aromatics, plus 5.0 volume percent. A similar calculation would be necessary for extending the valid range for olefins or benzene. The Agency has no reason to believe that provisions for the extension of the valid range for fuel parameters other than aromatics, olefins, and benzene on either the low or high ends are necessary. For instance, the Complex Model conventional gasoline valid ranges for oxygen, sulfur, aromatics, olefins, and benzene all have 0.0 as their lower limit. Thus no valid range extension would be required on the low end of the valid range for these fuel parameters. Similarly for E300, a fuel can have an E300 value of no higher than 100 percent, which is also the high end of the specified valid range. Other limitations, such as ASTM specifications and the volatility rule, should eliminate the need for valid range extensions in other cases. 2. No-Benefit Limitation of the Provision for Valid Range Extension The Agency continues to believe that the valid range limits specified in Sec. 80.42(c)(1) and Sec. 80.45(f)(1)(ii) identify the fuel parameter values beyond which the compliance models are not considered accurate. Thus the Agency has determined that any extension of the specified valid ranges for conventional gasoline should incorporate flat-line extrapolation. Under flat- line extrapolation, the compliance models provide no emissions benefit or detriment when raising the value of either aromatics or olefins above the values specified in Sec. 80.42(c)(1) and Sec. 80.45(f)(1)(ii). This flat-line extrapolation will apply to both the baseline fuel and any target fuels evaluated with the compliance models under the anti-dumping regulations. 3. Expanding the Applicability of the Valid Range Extension Provision to the Simple Model This direct final rule expands the applicability of the valid range extension provision given in Sec. 80.91(f)(2)(ii) to the Simple Model. However, as noted above, the only fuel parameters having valid range limits under the Simple Model anti-dumping regulations are aromatics and benzene content. The Simple Model valid ranges for both aromatics and benzene are being expanded to be equal to the corresponding ranges for the Complex Model, as described in sections II.A.3-5 of this rule. Thus the new valid range under the Simple Model will be 0-55 volume percent for aromatics and 0-4.9 volume percent for benzene. No valid range extension will be required for either of these fuel parameters on the low end of the valid range. Thus the only relevant valid range extensions that would be necessary for the Simple Model would be for the high end of aromatics and the high end of benzene. III. Complex Model The Complex Model includes separate calculations for exhaust and nonexhaust emissions. The final regulations contained errors in the model descriptions for both exhaust and nonexhaust. The errors in the exhaust portion of the Complex Model were limited to the linear extrapolation methodology, while the errors in the nonexhaust portion arose in the VOC equations themselves. Another error also arose in the calculation procedure for annual average toxics. Each of these errors and the associated corrections will be discussed in detail below. A. Extrapolation The exhaust portion of the Complex Model is a statistically-derived set of equations relating fuel parameters to emissions of VOC, NOX, benzene, 1,3- butadiene, acetaldehyde, formaldehyde, and POM. The Agency determined the conditions under which these exhaust equations must be linearly extrapolated based on the ranges for each fuel parameter within which the equations were considered to be accurate. Linear extrapolation amounts to fixing the slope of the fuel parameter:emission relationship at a constant value. It is used to extend the equations beyond the limits of the data on which they are based, thereby making the reformulated gasoline and anti-dumping programs as flexible as possible. Of the six separate models in the exhaust portion of the Complex Model, all four toxic models are linear, and thus do not require linear extrapolation. Paragraph (c)(1)(iv) of Sec. 80.45 specifies the conditions and limitations of linear extrapolation for VOC, while Sec. 80.45(d)(1)(iv) specifies the conditions and limitations of linear extrapolation for NOX. The details of the linear extrapolation methodology included in these two portions of the final regulations contained a number of errors which require correction. 1. Correct Parenthetical Form of Extrapolation Equations In paragraphs (c)(1)(iv)(B) and (d)(1)(iv)(B) of Sec. 80.45, the linear extrapolation equations contained too many brackets in some regions of the equations, and missing parentheses in other regions. Although these inadvertent omissions and insertions did not change the mathematical nature of the equations, a literal copying of the equations into computer code would result in an error. Thus the Agency has corrected the linear extrapolation equations for both VOC and NOX to contain the correct number of parentheses and brackets in the correct position and order. 2. Correct Missing Sulfur Term in NOX Extrapolation Equation A sulfur term was inadvertently left out of both the Phase I and Phase II NOX extrapolation equations given in paragraph (d)(1)(iv)(B) of Sec. 80.45. This missing sulfur term represents the linear extrapolation of the NOX model with respect to sulfur for high emitters. This correction will have a negligible impact on the emission performance estimates provided by the Complex Model because the NOX equation for high emitters is essentially linear with respect to sulfur./1/ The inclusion of the correct sulfur term in the NOX extrapolation equation will result in the correct application of the edge target fuel to the NOX extrapolation equations. Note /1/ The exhaust portion of the Complex Model includes exponential functions which alters the traditional implications of first- and second- order equations. However, the Complex Model exhaust equations can be referred to and approached as first- and second-order as described in Section IV.D.1 of the Regulatory Impact Analysis for the final rule. 3. Clarify E300 Extrapolation Above 95 Percent In the process of determining the valid range limits for the use of the Complex Model for both reformulated and conventional gasoline, the Agency determined that the emission changes estimated by the exhaust equations were not accurate above an E300 value of 95 percent. However, comments received from the refining industry indicated a need for an E300 valid range that extended up to 100 percent. The Agency concluded that, although the exhaust equation emission change estimates could not be considered accurate above an E300 level of 95 percent, allowing only a zero change in emissions above this E300 level would ensure that refiners could not receive inappropriate benefits for fuels with very high E300 levels. Therefore, the Agency allowed for flat-line extrapolations of all exhaust equations between the E300 values of 95 and 100 percent. However, some portions of the regulations that specified this allowance contained typographical errors which substantially changed the manner in which this flat-line extrapolation for E300 was to be executed. Specifically, paragraphs (c)(1)(iv)(C)(5), (c)(1)(iv)(D)(5), and (d)(1)(iv)(C)(5) of Sec. 80.45 all indicated that the E300 value of the edge target fuel should be held constant at 95 volume percent for target fuels having an E300 value of greater than 95 volume percent. These paragraphs should not have refered to the edge target fuel, but rather to the target fuel for the purposes of determining emissions performances with the Complex Model. These three paragraphs are thus changed accordingly. 4. Correct Value of DARO The regulations describing the linear extrapolation methodology for exhaust VOC and NOX contained two other typographical errors that nevertheless were substantial in their effects. The first arose in paragraphs (c)(1)(iv)(C)(9), (c)(1)(iv)(D)(9), and (d)(1)(iv)(C)(9) of Sec. 80.45 in the specification of the value of DARO. The value of DARO should generally be set equal to (DARO-- 18 volume percent) for any target fuel having an aromatics content of less than 18 volume percent. Thus DARO will be negative when the VOC or NOX equation is linearly extrapolated with respect to aromatics at the low end of the valid range. However, for target fuels having an aromatics content of less than 10 volume percent, the VOC extrapolation should be flat-line instead of linear. In this case, DARO should be fixed at a value of -8 volume percent. The value in the regulations was given incorrectly as +8 volume percent. 5. Correct Specification of DE300 The second typographical error that requires explanation and correction involves the specification of DE300 in paragraphs (c)(1)(iv)(C)(13) and (c)(1)(iv)(D)(13). The first sentence in each of these two paragraphs contains two conditions that must be met for DE300 to be set equal to (E300-- 94 percent). The first condition incorrectly states that the E300 level of the target fuel must be less than 94 percent, when in fact the condition should state that the E300 level of the target fuel must be greater than 94 percent. The remainder of both of these paragraphs is correct. 6. Eliminate References to E300 in NOX Extrapolation Finally, the linear extrapolation methodology for NOX in the final rule contained references to the allowable range for E300, despite the fact that the NOx equation is not extrapolated with respect to E300. Thus all references to E300 in paragraphs (d)(1)(iv)(A) and (d)(1)(iv)(B) are removed by this direct final rule. Note that, since all exhaust equations in the Complex Model are flat-line extrapolated for E300 values greater than 95 percent, the flat-line extrapolation for E300 specified in paragraph (d)(1)(iv)(C)(5) is correct (except for the changes described above). B. Nonexhaust Model The equations giving nonexhaust VOC as a function of RVP for the Complex Model were originally proposed in the February 1993 NPRM (58 FR 17175). No changes to those equations were intended for the final rule on reformulated gasoline. However, typographical errors arose in a number of the coefficients in the nonexhaust VOC model when they were entered into the final regulations. These errors would have a small, but not insignificant impact on the VOC emission performances provided by the Complex Model. The errors in the nonexhaust VOC portion of the Complex Model lay in four places. The first was in the sign of the coefficient for RVP in the Phase I running loss equation for Region 1. The second was in the coefficient for RVP in the Phase II hot soak equation for Region 1. The third was in the coefficient for RVP in the Phase II refueling loss equation for Region 1. And the fourth was in the sign of the RVP coefficient for the Phase II running loss equation for Region 2. All the coefficients in the nonexhaust model have been returned to their proper values and signs by this direct final rule. C. Annual Average Calculations Since the averaging standards for toxics under the Complex Model are year- round standards, each refiner who is complying under an averaging scenario must determine the average year-round toxics emissions performance for the fuels that it sold during a given year. Each batch of fuel is uniquely associated with toxic emissions as estimated by the Complex Model. Thus refiners require a method for combining per batch emission performances into a single year-round value. The regulations provided two contradictory methods for combining per batch emission performances into such a single year-round value. Paragraph (g) et al. of Sec. 80.67 directs refiners to weight the emission performances by batch volume and then add them in order to determine a year- round value. In calculating emission performances with the Complex Model, fuels sold in the winter are evaluated with the winter model, while fuels sold in the summer are evaluated with the summer model. Thus this "volume- weighted" approach to determining year-round values correctly leads to an average toxic emissions value for the year. Paragraph (e)(3) of Sec. 80.45 incorrectly directs refiners to weight summer and winter toxic emissions by fixed values to obtain year-round averages. Under this portion of the regulations, summer batches of fuel would be individually weighted by their batch volumes to obtain average summer emission performance estimates. Likewise winter batches would be weighted by their batch volumes to obtain average winter emission performance estimates. However, the fixed weighting of summer and winter emission estimates for the purposes of determining a year-round value would not necessarily mimic a refiner's true ratio of summer to winter fuel. The fixed weightings given in paragraph (e)(3) were used to determine the performance standards, and are not relevant to determining compliance by individual refiners. Therefore, paragraph (e)(3) will be revised to refer to Sec. 80.67(g), and paragraphs (e)(3)(i) and (e)(3)(ii) of Sec. 80.45 will be removed. IV. Enforcement Corrections and Clarifications The following section contains corrections and clarifications to the enforcement portions of the final reformulated gasoline and anti-dumping regulations that were published on February 16, 1994 (59 FR 7716), and to enforcement portions of the preamble of the notice of final rulemaking./2/ The reasons for the changes are listed below. Note /2/ Hereinafter in Part IV of this notice (unless otherwise indicated), references to "final rule" or "final regulations" shall refer to the regulations promulgated in the February 16, 1994, notice of final rulemaking. A. Reformulated Gasoline Regulation (40 CFR Part 80, Subpart D) 1. Measurement of Reformulated Gasoline Fuel Parameters (Sec. 80.46) The table of aromatic compounds in Sec. 80.46(f)(1)(ii)(K) is being revised. The aromatic compounds listed in the final rule are those used by EPA during the development of the analytical method for the rulemaking. In this notice some compounds that no longer are available commercially have been deleted from the list, and several aromatic compounds that are found in commercial gasolines and are available commercially to make reference materials have been added to the list. Only materials of known purity or those specified as 99% pure or greater should be used as calibration standards. The number of calibration levels should be sufficient to bracket the expected concentration of each compound. Two calibration levels were used in the initial evaluation of the test method. In the future, however, EPA probably will use five calibration levels with at least three internal standards used in the standards and samples in order to improve precision. Initially EPA prepared standards by volume. Currently, however, EPA prepares standards, samples, and internal standards by weight. Conversion to volume percent is performed by using the density of the aromatic compound in question. Standards for gasoline that are prepared by weight are considered to be more accurate than standards that are prepared by volume. Neither the use of five calibration levels, nor the preparation of standards by weight, should be viewed as changes to the regulations, but rather as Agency recommendations intended to improve precision. 2. General Requirements for Refiners, Importers and Oxygenate Blenders (Sec. 80.65) a. Designation of Gasoline as Oxygenated Fuels Program Reformulated Gasoline (OPRG) or Non-OPRG. Section 80.65(d)(2)(iii) is revised in order to clarify the categories of gasoline that may be designated as oxygenated fuels program reformulated gasoline, or OPRG. The final rule, at Sec. 80.65(d)(2)(iii), specifies that reformulated gasoline must be designated as OPRG, or not OPRG, to provide a mechanism to ensure reformulated gasoline outside of oxygenated fuels control areas during oxygenated fuels control periods has at least the 2.0 weight percent oxygen content mandated by the Clean Air Act./3/ The final rule requires parties who meet the oxygen standard on average to meet the oxygen standard separately for gasoline not designated as OPRG. If OPRG and non-OPRG gasoline could be averaged together for oxygen purposes, the gasoline in the OPRG areas--where 2.7 weight percent oxygen is required during the oxygenated fuels control period--could be used to offset gasoline with 1.5 weight percent oxygen intended for use in non- OPRG areas. Note /3/ The "oxygenated fuels control areas" are those areas where the use of oxygenated gasoline is required during the winter season pursuant to section 211(m) of the Clean Air Act, 42 U.S.C. 7545(m). Today's revision consists of two parts. First, the regulation is clarified to make specific that the OPRG/non-OPRG designation applies only to reformulated gasoline and not to RBOB. The final rule requires RBOB to be designated as OPRG or non-OPRG, but the OPRG designation for RBOB serves no purpose because RBOB does not become reformulated gasoline until oxygenate has been added. The final rule is unchanged regarding oxygenate blender responsibilities--oxygenate blenders who produce reformulated gasoline by combining RBOB with oxygenate are required to designate the gasoline as OPRG or non-OPRG, and to meet the oxygen standard separately for gasoline not designated as OPRG. The second change regarding OPRG-designated gasoline is the addition of a new provision, at Sec. 80.65(d)(2)(iii)(B), to clarify that reformulated gasoline that contains at least 2.0 weight percent oxygen may be designated as OPRG regardless of whether or not the gasoline is used in an oxygenated fuels program control area during an oxygenated fuels program control period. This change allows terminals that serve both oxygenated fuels areas and non- oxygenated fuels areas to stock a single reformulated gasoline that could be used in both areas, instead of having to stock both OPRG and non-OPRG designated reformulated gasoline. This change also allows all reformulated gasoline that meets the 2.0 weight percent oxygen per-gallon standard to be designated as OPRG, without such gasoline being restricted to use in oxygenated fuel areas during oxygenated fuel control periods. A terminal thus could stock reformulated gasoline that contains 2.7 weight percent oxygen and, therefore, that meets the oxygenated fuels oxygen requirement, and deliver this gasoline into both OPRG and non-OPRG markets. A terminal also could stock reformulated gasoline that contains 2.0 weight percent oxygen for delivery into both OPRG and non-OPRG markets, and splash blend additional oxygenate with those batches of gasoline that are delivered into the oxygenated fuel area during the oxygenated fuel control period./4/ Note /4/ Section 80.78(a)(6) prohibits adding oxygenate to reformulated gasoline, except in the case of reformulated gasoline that is designated as OPRG and is used in an oxygenated fuels program control area during an oxygenated fuels program control period. This change, however, keeps intact the regulatory mechanism for ensuring non-OPRG areas receive reformulated gasoline that contains at least 2.0 weight percent oxygen. Any reformulated gasoline used outside an oxygenated fuels control area during an oxygenated fuels control period that contains less than 2.0 weight percent oxygen (and therefore must be designated as non- OPRG) must be offset with other non-OPRG reformulated gasoline that contains more than 2.0 weight percent oxygen, such that the average oxygen content of the non-OPRG gasoline is greater than or equal to the 2.1 weight percent average standard. b. Designation of Complex Model Gasoline as Meeting NOX Standard on Per- Gallon or Average Basis. Section 80.65(d)(2)(v)(B) is revised in order to clarify that refiners and importers are required to specify whether the NOX standard is being met on a per-gallon basis or on average only for gasoline certified under the complex model. There is no separate NOX standard under the simple model. c. Designation of Reformulated Gasoline Blendstock for Downstream Oxygenate Blending (RBOB). Section 80.65(d)(2)(vi) is being revised in order to clarify that refiners and importers have three options regarding the designation of reformulated gasoline blendstock for downstream oxygenate blending (or RBOB). A refiner or importer has three options for the types of RBOB that may be produced or imported: as suitable for blending with any oxygenate, with ether only, or with a refiner- or importer-specified oxygenate type and amount. The gasoline designation requirements at Sec. 80.65(d)(2)(vi) as promulgated in the final rule did not include the refiner- or importer-specified oxygenate option for RBOB, making this provision inconsistent with the downstream oxygenate provisions at Sec. 80.69(a) that specifies the RBOB options and includes the refiner- or importer-specified option, and the product transfer document requirements for RBOB at Sec. 80.77(i)(2) that includes all three options. In order to correct this inconsistency, today's revision adds the refiner- or importer-specified option to the RBOB designation requirement. d. Assignment of Batch Numbers. Section 80.65(d)(3) requires refiners and importers to assign a unique number to each batch of reformulated gasoline that is produced or imported, and this section includes an example of such a batch number. The numbers contained in the example are being modified to reflect the correct number of digits for the portion of the batch number that states the EPA-assigned facility registration number (five digits) and the sequential batch number (six digits). e. Computer-Controlled In-Line Blending Exemption. Section 80.65(e)(1) requires each refiner or importer to obtain test results for each batch of reformulated gasoline prior to the gasoline leaving the refinery or import facility. Refiners who produce reformulated gasoline using a computer- controlled in-line blending process in which the gasoline is blended directly to a pipeline, however, will not have the test results for the "batch" prior to the release of at least some, if not all, of the gasoline. To correct the incompatibility between this requirement and the nature of the computer- controlled in-line blending process, the Sec. 80.65(f)(4) exemption from independent sampling and testing for refiners that produce reformulated gasoline using computer-controlled in-line blending equipment is expanded to include an exemption from the Sec. 80.65(e)(1) requirement that refiners have test results for each batch prior to the gasoline leaving the refinery. f. Release of Reformulated Gasoline Certified Under Simple Model as Not VOC-Controlled. Section 80.65(e)(1) is also revised in order to clarify that reformulated gasoline certified under the simple model that is not VOC- controlled may be released from the refinery or import facility after the refiner or importer has test results for oxygen and benzene only. The final rule requires refiners and importers to have RVP test results in hand prior to release of all simple model-certified reformulated gasoline, but the RVP standard applies under the simple model only to VOC-controlled gasoline. g. List of Reformulated Gasoline Properties to be Established by Testing. Section 80.65(e)(2)(i) contains a list of reformulated gasoline properties that must be established by testing. This list in the final rule includes 50% distillation (T-50) and 90% distillation (T-90) (the temperatures in degrees F at which 50% and 90% of a liquid are evaporated). This list is being revised to add terms for E200 and E300 (the percent of a liquid that are evaporated at 200 and 300 degrees F). E200 and E300 are correlated with T-50 and T-90, and may be approximated from T-50 and T-90 measurements using conversion equations. The most accurate way of determining the E200 and E-300 of gasoline, however, is using ASTM-86-90, the distillation test method specified at Sec. 80.46(d). As a result, the list of mandatory testing is being expanded to include the E-200 and E-300 terms. This expansion does not constitute a change from the mandatory testing requirements for reformulated gasoline, because the E200 and E300 terms are established as part of the ASTM-86-90 distillation test that already is required. The correlation ranges for E-200 and E-300, 2.5 vol% and 3.5 vol% respectively, that are included in the Sec. 80.65(e)(2)(i) table are the reproducibility figures for these terms from the ASTM test, and are comparable to the five degrees F correlation range provided for the T-50 and T-90 terms. h. Reconciliation of Test Results. Section 80.65(e)(2)(ii)(A) is being revised to clarify one option for reconciling reformulated gasoline test results from a refiner's or importer's laboratory as compared with test results from an independent laboratory. Under this option, where the refiner's or importer's test result for any parameter is not confirmed, the refiner or importer would use the result for that parameter that is the "worst case" for the refiner or importer ("best case" for the environment). Under this option in the final rule, the smaller of the two results for oxygenates is used for calculating all standards except RVP. This option is being revised to refer to oxygenates as a class without separately naming each oxygenate. This change will keep the option from becoming dated if any new oxygenates are used in reformulated gasoline. The option also is being revised to eliminate the larger oxygenate volume assumption in the case of RVP, because RVP is a parameter that is measured directly. Any oxygenate effect on RVP will be measured in the RVP test and would not be changed by a calculation using the "worst case" oxygenate level. i. Attest Engagement Requirement. Section 80.65(h) is being revised to specify that the attest engagement requirement applies to oxygenate blenders who meet the oxygen standard on average, and not to oxygenate blenders who meet the oxygen standard on a per-gallon basis. This change conforms Sec. 80.65(h) with Sec. 80.69(b)(4) which limits the attest engagement requirement to oxygenate blenders who average. 3. Compliance Surveys (Sec. 80.68) Section 80.68(c)(13) is being revised to clarify when VOC and NOX emissions reduction calculations must be reported to EPA by the surveyor under the gasoline quality survey requirements./5/ VOC emissions reduction calculations must be reported only for surveys during June 1 through September 15 of each year, including simple model surveys where a specified simple model VOC emissions reduction equation is cross-referenced. NOX emission reduction calculations must be reported for all complex model surveys before January 1, 2000, and beginning on January 1, 2000 for surveys outside the period June 1 through September 15. NOX surveys are not required during June 1 through September 15 under Phase II (beginning on January 1, 2000) because the Clean Air Act's restriction on NOX increases is satisfied through the NOX reductions required for VOC-controlled reformulated gasoline under Phase II. The rationale for this approach for NOX surveys is discussed at 59 FR 7773 (February 16, 1994). Note /5/ This notice is also adding the paragraph number for this provision, which was inadvertently omitted in the notice of final rulemaking. 4. Covered Areas (Sec. 80.70) a. Putnam and Orange Counties, New York. In order to correct an oversight in the final rule, Sec. 80.70(d)(3) is being amended to include Putnam and Orange Counties, New York, as part of the New York-Northern New Jersey-Long Island-Connecticut "covered area." These counties are part of the New York City CMSA and are thus appropriately part of the New York City reformulated gasoline covered area. See 57 FR 13444 (April 16, 1992). Putnam and Orange Counties are also included in the New York City CMSA for purposes of the oxygenated fuels program requirements. b. Bullitt and Oldham Counties, Kentucky. Section 80.70(j) is being amended to specify the applicable boundaries for the portions of Bullitt and Oldham Counties, Kentucky, that are nonattainment areas. c. Essex County, New York. The listing of Essex County, New York, is also being amended to include a specific description of the nonattainment area. d. Smyth County, Virginia. Section 80.70(j) is also being amended to delete Smyth County, Virginia, as a covered area for the reformulated gasoline program. This area was expressly excluded from coverage in the Governor's opt-in request and should not have been included in the regulatory list of covered areas. e. Richmond, Virginia. An erroneous reference to the City of Richmond is also being corrected. 5. Reporting Requirements (Sec. 80.75) a. Sulfur, NOX and T90 Averaging Reports Section 80.41(h)(2) of the final rule requires that refiners and importers subject to the simple model meet their 1990 baseline sulfur, olefins and T90 levels on average for the entire year. However, it does not include a requirement to report the average values of these properties to EPA. It was the Agency's intention to have these values reported and so an additional reporting requirement for sulfur, olefins and T90 is being added at Sec. 80.75(b)(2). In addition, the RVP averaging reporting requirements are being renumbered, under Sec. 80.75(b)(1), so that the sulfur, olefin and T90 averaging reporting requirements may be included under Sec. 80.75(b). b. Oxygen Averaging Reports. Section 80.75(f)(2)(ii)(A) is being revised to conform the categories of reformulated gasoline that must be reported separately for oxygen averaging purposes to the categories that are specified in the oxygen averaging section, at Sec. 80.67(h)(1)(v)(A). 6. Registration Requirements (Sec. 80.76) The Agency is making several revisions to the registration requirements in Sec. 80.76(c) of the final rule. EPA is removing the requirement to indicate where off-site records are kept from the refiners', importers' and oxygenate blenders' registration information, and instead requiring that information be submitted for each refinery, oxygenate blending facility, and in the case of importers, operations within each PADD. The registration requirement to indicate what type of gasoline each refinery or oxygenate blending facility will produce (reformulated, RBOB, conventional or blendstocks) has been removed because the Agency believes that it is not necessary for registration purposes. The Agency intended that importers would identify the independent laboratories used to comply with the independent sampling and testing requirements, but there was no specific regulatory text requiring them to do so. This notice requires each importer to provide that information to EPA for its operations in each PADD. 7. Product Transfer Documentation (Sec. 80.77) a. Minimum and/or Maximum Standards. Section 80.77(g)(2), which requires that per-gallon minimum/maximums must be included in product transfer documents, is being revised to clarify the categories of reformulated gasoline for which these values must be specified. Paragraphs (g)(2)(iii), (g)(2)(iv)(A), and (g)(2)(iv)(B) of Sec. 80.77 are being revised to specify that the RVP maximum and the VOC emissions performance minimum must be included only for VOC-controlled reformulated gasoline, because these standards apply only to VOC-controlled gasoline. Paragraphs (g)(2)(iv)(A) and (g)(2)(iv)(B) of Sec. 80.77 also are being revised to specify that the NOX minimum must be included only for reformulated gasoline certified using the complex model. There is no separate NOX standard under the simple model. b. VOC-Controlled Gasoline That Contains Ethanol. Section 80.77(g)(3) is being added to require that product transfer documents must identify any VOC- controlled reformulated gasoline that contains any ethanol as an ethanol- containing reformulated gasoline, so that downstream parties will have sufficient knowledge to avoid violation of the prohibition, at Sec. 80.78(a)(8), against combining VOC-controlled reformulated gasoline produced using ethanol with VOC-controlled reformulated gasoline produced using any other oxygenate during the period January 1 through September 15. c. Complex Model Gasoline Certified Prior to 1998. Section 80.77(h) is being revised to clarify that the product transfer document requirements related to gasoline certified using the complex model before January 1, 1998, apply to reformulated gasoline and RBOB. The final rule makes reference to "gasoline" and RBOB without specifying "reformulated gasoline," which could have caused confusion. 8. Controls and Prohibitions (Section 80.78) Section 80.78(a)(1)(v)(B) and (a)(1)(v)(C), concerning reformulated gasoline prohibited activities, are being revised to clarify that gasoline subject to the per-gallon RVP maximum must have an RVP that is less than or equal to this standard, and that gasoline subject to the VOC and NOX emissions reduction minimum must have emissions reductions that are greater than or equal to these standards. The final rule describes these requirements only in terms of gasoline that is "less than" or "greater than" these standards, while gasoline that equals these standards also is in compliance. 9. Enforcement Exemptions for California Gasoline (Sec. 80.81) a. Definition of California Gasoline. The final rule was intended to extend the California enforcement exemptions to gasoline produced at refineries outside California that produce only California reformulated gasoline and federal conventional gasoline. See 59 FR 7759, col. 3 (February 16, 1994). However, the regulatory definition of "California gasoline" could be read to exclude non-California refineries producing California reformulated gasoline from the enforcement exemptions. Therefore, Sec. 80.81(a)(2)(iii) is clarified to exclude from this definition only gasoline produced by non- California refineries that are also producing reformulated gasoline for sale in covered areas outside California. b. Compliance Demonstration Submittal. Section 80.81(b)(4) requires refiners, importers, and oxygenate blenders to submit the compliance demonstration mandated by Sec. 80.81(b)(3) by May 31, 1996 "along with reports required to be submitted under Sec. 80.75(a)(1)." The quoted language is corrected to make clear that the compliance demonstration should accompany the report for the first quarter of 1996 due to be submitted on that date under Sec. 80.75(a)(1)(i). c. Use of California Sampling and/or Testing Methodologies. Section 80.81(h) allows refiners and importers of California gasoline to use sampling and test methods that are set forth in the California reformulated gasoline regulations, in lieu of those methods prescribed under Sec. 80.46 for the federal reformulated gasoline program. This provision is being revised to clarify that these California-approved sampling and test methods may be used only with California gasoline, and that these methods may not be used to satisfy the sampling and testing requirements for reformulated or conventional gasoline that does not meet the definition of California gasoline. B. Anti-Dumping Regulations (40 CFR Part 80, Subpart E) 1. Standards Applicable to Refiners and Importers (Sec. 80.101) a. Compliance Baseline Formula. Section 80.101(f)(4) currently provides that refiners and importers who use an individual 1990 baseline, and who increase their gasoline production volume above a certain level, must calculate a compliance baseline for each averaging period. EPA has combined the separate formulas for "Veq" and "CBi" in Sec. 80.101(f)(4) into a single, simpler formula. In addition, EPA has specified that the denominator of the compliance baseline equation includes the volume of California gasoline. The final rule excluded the volume of California gasoline produced in 1995 and thereafter from the compliance baseline equation, but this exclusion rendered this equation invalid for refiners of California gasoline because the numerator of this equation includes gasoline produced for the California market in 1990. In order to constitute a valid comparison of the volume of gasoline produced in 1990 versus the volume produced in 1995 and thereafter, both the numerator and the denominator of the compliance baseline equation must contain the volume of gasoline produced for the California market. This approach for including California gasoline in the compliance baseline equation is equivalent to the requirement that reformulated gasoline produced for use in covered areas outside the State of California must be included in the denominator of the compliance baseline equation. The definitions of the factors "DBi," "CBi," and "Bi," are clarified to bring them in conformance with the terms used in the complex model calculations under Sec. 80.101(g). b. Compliance Calculations. Section 80.101(g), entitled "Compliance Calculations," is restructured to reverse the order of the simple model calculation formula currently in Sec. 80.101(g)(1)(i) and the formula for determining the average value for the parameter being evaluated currently in Sec. 80.101(g)(1)(ii). This organizational change is necessary because in order to perform the simple model calculation for exhaust benzene emissions, the refiner or importer must first determine the average value for certain parameters. In addition, and for purposes of clarity, the heading "Simple Model Calculations" is inserted as a new paragraph (i) under Sec. 80.101(g)(1), and the heading "Complex Model Calculations" is inserted as a new paragraph (ii) under Sec. 80.101(g)(1). These changes restructure the compliance calculations in a logical sequence that will make this section clearer for compliance purposes. The formula definition of "SGi" in Sec. 80.101(g)(1)(i)(A) is amended to clarify that the specific gravity term only applies to calculations involving sulfur. The definitions of the factors "BZ" and "AR" under Sec. 80.101(g)(1)(i)(B), pertaining to compliance calculations for exhaust benzene emissions under the simple model, are clarified as to how these are calculated by expanding them to read "calculated per paragraph (g)(1)(i)(A) of this section." Section 80.101(g)(1)(iii), (g)(1)(iv), and (g)(1)(v), pertaining to complex model calculations, are consolidated and simplified as a new Sec. 80.101(g)(1)(ii). c. Sampling and Testing. Section 80.101(i)(1), which concerns requirements for sampling and testing of conventional gasoline and other products to which the compliance standards apply, is being revised to delete the requirement that such sampling and testing be conducted prior to the gasoline or product leaving the refinery. This change is necessary because this requirement interferes with the ability of refiners and importers to do composite sampling and in-line blending. 2. Controls Applicable to Blendstocks (Sec. 80.102) Since gasoline produced for and marketed in California is subject to that State's stringent reformulated gasoline standards, it is not necessary to include such gasoline, or applicable blendstocks used in the production of such gasoline, in EPA's blendstock tracking requirements beginning in 1995. Thus, the definitions of the volume of gasoline, "Vg" and "Vg,i," in the averaging period blendstock-to-gasoline ratio in Sec. 80.102(d)(1)(i) and the running cumulative compliance period blendstock-to-gasoline ratio in Sec. 80.102(d)(2)(i), respectively, are amended by adding an explicit exclusion for California gasoline. Similarly, an additional subparagraph (v) has been added under Sec. 80.102(d)(3) to exclude applicable blendstocks used to produce California gasoline in the blendstock ratio calculations. These corrections are necessary to make clear that the volumes of California gasoline, and applicable blendstocks used to produce California gasoline are not part of the ratio calculations.6 Note 6 The baseline blendstock-to-gasoline ratio does include both gasoline and blendstock produced for the California market. This baseline ratio nevertheless is a valid basis for comparison with the compliance period blendstock-to-gasoline ratios (that exclude California gasoline and blendstock) because the baseline ratio represents a refiner's or importer's actual 1990 gasoline and blendstock volumes. 3. Record Keeping Requirements (Sec. 80.104) Under Sec. 80.104(a)(2)(ix), refiners and importers are required to retain documents to demonstrate that blendstocks were transferred for other than gasoline blending purposes as a basis for excluding such blendstocks from tracking. There are various other bases specified in Sec. 80.102(d)(3) for excluding blendstocks from tracking (e.g., exported, transferred as a feedstock) for which document support is not required in the final rule. EPA believes that the document support requirement should apply to all excluded blendstocks. Therefore, Sec. 80.104(a)(2)(ix) is expanded to require the retention of documents which demonstrate any of the specified bases for the exclusion of blendstocks from blendstock tracking. 4. Reporting Requirements (Sec. 80.105) Section 80.105(a)(2) in the final rule requires refiners and importers to report the overall volume of applicable blendstock produced or imported and transferred to others. This provision is being clarified to require separate reporting for those applicable blendstocks that are, and those that are not, excluded under Sec. 80.102(d)(3). C. Preamble The preamble of the February 16, 1994, notice of final rulemaking contains two errors in the sections discussing reformulated gasoline and anti-dumping enforcement provisions. The following corrections should be noted for these sections of the preamble: ** On page 7759, in the second column, in the 12th and 13th lines of the second full paragraph, "(March 1, 1996, through February 29, 1996)" should read "(March 1, 1996, through February 28, 1998)." ** On page 7800, in the second column, in the 4th and 5th lines of the carryover paragraph, "section 211(k)(c) of the Act" should read "section 211(k) and (c) of the Act." V. Summer/Winter Season Section 80.91 of the Reformulated Gasoline Final Rule (59 FR 7716) defines summer and winter data and sampling requirements as follows: paragraph (d)(1)(i)(A) states that "Data shall have been obtained for at least three months of the refiner's or importer's production of summer gasoline and at least three months of its production of winter gasoline." The regulation goes on to define a summer month as "any month during which the refiner produced any gasoline which met the federal summer gasoline volatility requirements. Winter shall be any month which could not be considered a summer month." Several comments received by EPA since the rule was published indicate that this present definition severely restricts or eliminates the winter period in some cases, and inappropriately allocates winter data to the summer calculation. Further, the current rule inadvertently precludes the use of actual data in some calculations, even when such per batch actual data is available. Considering that the goal of the baseline is to most accurately reflect actual 1990 gasoline composition, the rule will be modified to more correctly allocate parameter data. Provision will be made for the use of actual RVP data to define summer and winter gasoline. When such Method 1, per batch actual data is not available, summer and winter months will be redefined to better approximate the seasonal gasoline fuel parameter and emission values. This modification to the reformulated gasoline regulation will satisfy several comments received since publication in the Federal Register. EPA will redefine summer and winter months, for use when Method 1 actual batch data is not available, as follows. A summer month will be redefined as any month during which more than 50 percent (by volume) of the gasoline produced by a refiner met the federal summer gasoline volatility requirements. Winter shall remain defined as any month which could not be considered a summer month. This will correct situations in which small quantities of summer volatility gasoline are produced early in the year. Originally, data from a month in which even small quantities of summer volatility fuel was produced was considered a summer month. With this modification, such months in which the majority of fuel was winter volatility would be correctly allocated as a winter data month. Further, for any month for which both winter and summer gasoline were produced, if actual RVP (Method 1) data are available, that data will be accurately divided between the summer and winter calculations, as appropriate. If such per batch data is not available, all data for that month would be considered either summer or winter, based on the production volumes. When actual per batch data is used, for minimum data requirement purposes, a month is considered equivalent to 4 weeks of seasonal data. Therefore, 12 weeks of data sampling on summer volatility fuel satisfies the minimum three months of data collection required. If a refiner, such as the California refiners, still cannot provide three months of winter data, they may petition for less than minimum data, under the provisions outlined in Sec. 80.91(d)(1)(C). VI. Baseline Determination Adjustments A. Work-In-Progress (WIP) Adjustment The final reformulated gasoline rule provided criteria for allowing work- in-progress baseline adjustments (section 80.91(e)(5)). Work-in-progress (WIP) refers to one or more major capital changes or commitment(s) made by a refiner prior to or in 1990. A WIP adjustment allows a refiner to modify its baseline volume and fuel parameter values (which affect emissions) to account for the WIP. In order to obtain a WIP modification, a refiner must petition EPA and EPA must approve the petition. As indicated in the preamble to the final rule, EPA believes that the criteria for a WIP adjustment should be fairly stringent, as the adjustment was intended only for those for whom a significant investment had already been made in order to comply with another government mandate. Additionally, a broad program of adjustments could indicate that EPA exceeded its equitable discretion under Alabama Power, as discussed in the Regulatory Impact Analysis (RIA) accompanying the final rule. In the final rule, EPA required that a refiner meet each of three criteria in order to qualify for a WIP baseline adjustment. A fourth criterion was also required to minimize environmental harm due to WIP adjustments, via simple model parameter caps and emissions caps for both the simple and complex models. Since promulgation of the final rule, EPA has re-evaluated this fourth criterion of the WIP provisions. EPA intended to allow WIP adjustments to relieve severe hardship where the adjustment did not allow emissions to increase significantly relative to the Clean Air Act baseline. The methods EPA chose to constrain WIP adjustments had inconsistent impacts depending on simple model or complex model use. The modification proposed in this Direct Final Rulemaking would ensure that the fuel parameter constraints on WIP adjustments under the simple model would be more consistent with the emissions performance constraints under the complex model. Specifically, WIP adjustments will be permitted to exceed the simple model parameter caps, but only to the extent the baseline still complies with the complex model emission caps. Without this change, a WIP adjustment would be more constrained in 1995 than would be the case in 1998, possibly requiring a refiner to make processing changes in 1995 that would not be necessary in 1998. While the emissions and parameter caps were set to minimize environmental harm due to the WIP (realizing that a WIP adjustment will actually increase baseline emissions relative to 1990) EPA believes this modification will increase compliance flexibility while maintaining the environmental goals of the program. With regard to the effect of WIP adjustments on reformulated gasoline compliance, the simple model caps stated in the regulations apply to reformulated gasoline as well as to conventional gasoline. As stated in the final reformulated gasoline rule, when the simple model is used for compliance, the WIP-adjusted annual average baseline values for sulfur, olefins and T90 are the actual WIP-adjusted values of those parameters, provided they do not exceed the unadjusted baseline values or the simple model parameter caps given in section 80.91(e)(viii)(B). However, baseline parameters may now exceed these caps if the WIP adjusted baseline does not result in exhaust emissions of VOC, toxics, and NOX which exceed the emission levels specified in Sec. 80.91(e)(5)(vii)(B), namely 105% of the annual average statutory value. Based on questions received since promulgation of the final rule, two changes in the language of the Work-In-Progress (WIP) provisions are made to section 80.91(e)(5) to further clarify certain aspects of the WIP adjustment not explicitly addressed in the final rule. The regulatory language dealing with the emissions and parameter caps is unclear as to whether the caps apply to the actual values or to the change in emissions or parameter values. In addition, there appeared to be some confusion over what was meant by "adjusted" baseline. Paragraphs 80.91(e)(5)(vi) and 80.91(e)(5)(vii) have been modified to clarify agency intent. B. JP-4 Adjustment In the final rule for reformulated gasoline, EPA allowed adjustments for specific extenuating circumstances. Baseline fuel parameters, volumes and emissions of a refinery can be adjusted due to the occurrence of specific extraordinary or extenuating circumstances which caused its 1990 gasoline production to be different than it would have been had the circumstance not occurred. However, the Agency's objective is not to establish a broad adjustment program. Allowable circumstances include unforeseen, unplanned downtime of at least 30 days of one or more gasoline blendstock producing units due to equipment failure or natural cause beyond the control of the refiner, or for non-annual maintenance (turnaround) downtime which occurred in 1990. These types of adjustments reflect instances where the 1990 baseline truly deviated from the otherwise expected baseline (historic and future), had the incident not occurred. EPA also expects that allowed adjustments will have minimal environmental impact while relieving a large regulatory burden. In keeping with that policy objective, EPA promulgated provisions to permit baseline adjustments for certain refiners which produced JP-4 jet fuel in 1990, upon petition and approval. As discussed in the RIA for the reformulated gasoline final rule, EPA believes that it has authority to allow such adjustments due to the discretion afforded EPA by Congress. Additionally, Alabama Power v. Costle /7/ gives EPA "case-by-case discretion" to grant variances or even dispensation from a rule where imposition of the requirement would result in minimal environmental benefit but would extremely burden a regulated party. Today's action changes two portions of the provisions for JP-4 adjustment: the multiple refinery requirement and the JP- 4 to gasoline production ratio. In the final reformulated gasoline rule, JP-4 baseline adjustments are generally limited to single-refinery refiners because such refiners have no way to aggregate baselines /8/ so as to reduce the combined burden of a JP-4 phaseout and the anti-dumping requirements on their operations. In some cases, if no relief were granted in this area, the viability of a refinery could be at stake. Note /7/ Alabama Power Company v. Costle, 636 F.2d 323.357 (D.C. Cir 1979). Note /8/ A refiner with more than one refinery may determine an aggregate baseline, i.e., a conventional gasoline compliance baseline, which consists of the volume-weighted emissions or fuel parameters, as applicable, of two or more refineries. 1. Multiple-Refinery Requirement The final reformulated gasoline rule also promulgated baseline adjustment provisions for multi-refinery refiners where each refinery produced JP-4 in 1990. This adjustment provision assumes that multi-refinery refiners are predominantly in the business of fuel production and thus possess the means to offset the refinery's JP-4 volume and associated fuel parameter increases with fuels volumes at other locations. The adjustment also assumes that refiners with multiple-refineries have process units offering various processing options which support an average (or typical) fuel production operation. Today's action modifies the JP-4 baseline adjustment multiple refinery requirement. Every refinery of a multiple-refinery is no longer required to have produced JP-4. Such multi-refinery refiners are allowed to average their 1990 JP-4 production to 1990 gasoline production ratio across all of their refineries. However, all refineries of a multi-refiner must still meet the other two criteria specified for the JP-4 baseline adjustment in the reformulated gasoline final regulations. Specifically, only refiners that will not produce reformulated gasoline and that meet the 1990 JP-4 to gasoline production ratio are allowed to make a JP-4 baseline adjustment. The Agency is amending the requirement stipulating that each refinery of multiple-refineries produced JP-4 in 1990 because, essentially, the same environmental impact and economic hardship is expected regardless of whether a single refinery or all refineries of a multiple-refinery produced JP-4. 2. JP-4 Baseline Adjustment Ratio The final reformulated gasoline rule promulgated baseline adjustment provisions which stipulate that 1990 JP-4 production must have constituted a specified portion of a refiner's 1990 fuel production in order for a significant enough burden to exist to justify permiting baseline adjustments. Additionally, a baseline is neither unrepresentative of 1990, nor incalculable, because of post-1990 changes in crude availability, fuel specifications, fuel markets, etc. EPA is permitting baseline adjustments for certain refiners which produced JP-4 jet fuel in 1990 because, as discussed in the Regulatory Impact Analysis for the reformulated gasoline final rule, EPA believes that it has a limited authority to allow such adjustments in certain extreme cases. The final reformulated gasoline regulations require that the ratio of the refinery's 1990 JP-4 production to its 1990 gasoline production equal or exceed 0.5. Upon further evaluation of the baseline adjustment provisions, the Agency has concluded that the JP-4 to gasoline production ratio, as promulgated in the RFG final rule (0.5), is unnecessarily restrictive. Based on responses from affected refiners, under the December 1993 criteria only two refiners in the United States would likely qualify for the adjustment. In other words, very few refiners under contract to produce JP-4 will have the relief intended by the provision. Today's action alters the refiner's 1990 JP-4 production to 1990 gasoline production ratio from 0.5 to 0.2. EPA believes this revised ratio indicates that a significant amount of the refinery feedstock used for JP-4 production would have to be converted in order to produce gasoline. Altering the JP-4 to gasoline production ratio will allow additional small refiners to qualify for the adjustment and not be forced to operate from a drastically less competitive position or be driven out of business. Based on feedback EPA has received, changing the ratio from 0.5 to 0.2 raises the number of affected refiners from 2 to 6. If large refineries had such a ratio also, the regulatory burden would be just as great. Also, it would be more difficult to argue de minimis environmental impact for large refiners. In reality none do, such that the ratio is only an issue for small refiners. Several different ratio options were suggested by commenters during the reformulated gasoline rulemaking as to what minimum portion of a refinery's 1990 production JP-4 should have constituted for the circumstance to be extenuating, as follows: JP-4 production to total refinery production, 20%; JP-4 production to gasoline production, 5%; JP-4 production to gasoline production, 75%; and, JP-4 production to gasoline plus JP-4 production, 10%. EPA's 0.2 JP-4 to gasoline ratio is in line with some of the commenters suggestions. At less than a 1990 JP-4 to gasoline production ratio of 0.2, EPA believes the impact on benzene and aromatics may make it more costly for refiners to comply with the regulations, though it is unlikely that such refiners will be forced out of business or experience some similar extreme burden. For example, it is expected that no economic hardship will occur at a JP-4 to gasoline ratio of 0.1. Thus, the Agency discretion recognized in Alabama /9/ to grant variances or even dispensation from a rule where imposition of the requirement would result in minimal environmental benefit but would extremely burden a regulated party, would not apply. Note /9/ Alabama Power Company vs. Costle, 636 F.2d 323.357 (D.C. Cir 1979). While the adjusted emission baselines of those approved for both amended JP-4 adjustments are likely to be higher than their actual 1990 baselines (primarily due to increased benzene and aromatics) EPA expects minimal negative environmental affects. First, the number of refineries meeting the criteria is still expected to be quite small. Second, the total production of all such refineries is also small. Thus, not very much additional gasoline will be affected by any baseline adjustments for JP-4 than if the criteria were less stringent or adjustments were not allowed at all. The modification of the multi-refinery requirement and the reduction of the ratio requirement to 0.2 both provide necessary flexibility to refiners and allow additional refiners (that are simultaneously burdened by the JP-4 phaseout and the anti- dumping provisions) regulatory relief. As stated in the Regulatory Impact Analysis (RIA) for the reformulated gasoline final rule, JP-4 baseline adjustments will be allowed only for those refiners which will not produce reformulated gasoline, which is the most critical factor in assessing environmental impact. While the anti-dumping requirements, in general, apply to all conventional gasoline whether or not reformulated gasoline is also produced, in these specific cases no dumping will occur due to reformulated gasoline production. The intent of Congress with regard to the anti-dumping program will be met while not unduly burdening those that meet the specified criteria. Since both the unadjusted and adjusted baselines must be determined, if a refiner granted such an adjustment subsequently produces reformulated gasoline, its conventional gasoline compliance would immediately be subject to its original unadjusted baseline. VII. Public Participation and Effective Date The Agency is publishing this action as a direct final rule because it views the changes contained within as non-controversial and anticipates no adverse or critical comments. This action will be effective September 19, 1994 unless the Agency receives notice by August 19, 1994 that adverse or critical comments will be submitted, or that a party requests the opportunity to submit such oral comments pursuant to section 307(d)(5) of the Clean Air Act, as amended. If such notice is received regarding a change to a particular regulatory provision, the provision in question will be withdrawn before the effective date by publishing a subsequent Federal Register notice withdrawing the direct final rule for the identified provision. Parties that submit adverse or critical comments, notify EPA of intentions to submit such comments, or request a public hearing within the allotted time period should identify the specific provision(s) at issue by specifying the preamble section numbers that discuss the provision(s). For instance, comments on the change to the oxygen valid range limits should include a reference to Section (Item Number) II.A.1 of the preamble. Comments on any of the insubstantial errors in Section I of the preamble should include a reference to the identification code associated with each change in that section. For instance, adverse comments on the paragraph reference change in Sec. 80.41(h)(2)(iii) should include a reference to Item Number I-A. The EPA will withdraw from final action only those specific provision(s) identified by the commenters or persons who notify EPA of their intent to comment or who request an opportunity to submit oral comments. All provisions in today's action that are not commented upon or for which EPA does not receive notice as described above will become effective September 19, 1994. VIII. Statutory Authority The statutory authority for the actions finalized today is granted to EPA by Sections 114, 211 (c) and (k) and 301 of the Clean Air Act, as amended; 42 U.S.C. 7414, 7545 (c) and (k), and 7601. IX. Administrative Designation Pursuant to Executive Order 12866, (58 FR 51735 (October 4, 1993)) the Agency must determine whether the regulatory action is "significant" and therefore subject to OMB review and the requirements of the Executive Order. The Order defines "significant regulatory action" as one that is likely to result in a rule that may: (1) Have an annual effect on the economy of $100 million or more or adversely affect in a material way the economy, a sector of the economy, productivity, competition, jobs, the environment, public health or safety, or State, local, or tribal governments or communities; (2) Create a serious inconsistency or otherwise interfere with an action taken or planned by another agency; (3) Materially alter the budgetary impact of entitlements, grants, user fees, or loan programs or the rights and obligations of recipients thereof; or (4) Raise novel legal or policy issues arising out of legal mandates, the President's priorities, or the principles set forth in the Executive Order. Pursuant to the terms of Executive Order 12866, it has been determined that this direct rule is not a "significant regulatory action". X. Regulatory Flexibility Analysis The Regulatory Flexibility Act (RFA) of 1980 requires federal agencies to examine the effects of the renewable oxygenate regulation and to identify significant adverse impacts of federal regulations on a substantial number of small entities. Because the RFA does not provide concrete definitions of "small entity," "significant impact," or "substantial number," EPA has established guidelines setting the standards to be used in evaluating impacts on small businesses./10/ For purposes of the renewable oxygenate requirement for reformulated gasoline, a small entity is any business which is independently owned and operated and not dominant in its field as defined by SBA regulations under section 3 of the Small Business Act. Note /10/ U.S. Environmental Protection Agency, Memorandum to Assistant Administrators, "Compliance with the Regulatory Flexibility Act," EPA Office of Policy, Planning, and Evaluation, 1984. In addition, U.S. Environmental Protection Agency, Memorandum to Assistant Administrators, "Agency's Revised Guidelines for Implementing the Regulatory Flexibility Act," Office of Policy, Planning, and Evaluation, 1992. The Agency believes that the interpretations, clarifications, and corrections published in today's final action are unlikely to have a significant economic impact on a substantial number of small entities. In fact, the revisions contained herein are designed to promote successful implementation of the reformulated gasoline program for all regulated parties. List of Subjects in 40 CFR Part 80 Environmental protection, Air pollution control, Fuel additives, Gasoline, and Motor vehicle pollution. Dated: June 27, 1994. Carol M. Browner, Administrator. 40 CFR part 80 is amended as follows: PART 80--REGULATION OF FUELS AND FUEL ADDITIVES 1. The authority citation for part 80 continues to read as follows: Authority: Sections 114, 211 and 301(a) of the Clean Air Act as amended (42 U.S.C. 7414, 7545 and 7601(a)). 2. In Sec. 80.41, paragraphs (h)(2)(iii), (j)(2), and the introductory text to paragraph (m)(1) are revised to read as follows: Sec. 80.41 Standards and requirements for compliance. * * * * * (h) * * * (2) * * * (iii) In the case of a refiner that operates more than one refinery, the standards specified under this paragraph (h)(2) shall be met using the refinery grouping selected by the refiner under Sec. 80.101(h). * * * * * (j) * * * (2) The aromatics value which, together with the values for benzene, RVP, and oxygen determined under paragraph (j)(1) of this section, meets the Simple Model toxics requirement specified in paragraphs (a) or (b) of this section, as applicable; * * * * * (m) * * * (1) On each occasion that a covered area fails a NOX emissions reduction survey conducted pursuant to Sec. 80.68, except in the case of Phase II complex model NOX standards for VOC-controlled gasoline, the NOX emissions reduction requirements for that covered area beginning in the year following the failure shall be adjusted to be more stringent as follows: * * * * * 3. Section 80.42 is amended by revising definitions "EXHVOCS1" through "REFVOCS2" in paragraph (a) introductory text; by adding a concluding sentence to paragraphs (b)(1)(ii), (b)(2)(ii), and (b)(3)(ii), and adding paragraph (b)(4); and by revising the table in paragraph (c)(1) to read as follows: Sec. 80.42 Simple emissions model. (a) * * * EXHVOCS1=Exhaust nonmethane, nonethane VOC emissions from the fuel in question, in grams per mile, for VOC control region 1 during the summer period. EXHVOCS2=Exhaust nonmethane, nonethane VOC emissions from the fuel in question, in grams per mile, for VOC control region 2 during the summer period. EXHVOCW=Exhaust nonmethane, nonethane VOC emissions from the fuel in question, in grams per mile, during the winter period. EVPVOCS1=Evaporative nonmethane, nonethane VOC emissions from the fuel in question, in grams per mile, for VOC control region 1 during the summer period. EVPVOCS2=Evaporative nonmethane, nonethane VOC emissions from the fuel in question, in grams per mile, for VOC control region 2 during the summer period. RLVOCS1=Running loss nonmethane, nonethane VOC emissions from the fuel in question, in grams per mile, for VOC control region 1 during the summer period. RLVOCS2=Running loss nonmethane, nonethane VOC emissions from the fuel in question, in grams per mile, for VOC control region 2 during the summer period. REFVOCS1=Refueling nonmethane, nonethane VOC emissions from the fuel in question, in grams per mile, for VOC control region 1 during the summer period. REFVOCS2=Refueling nonmethane, nonethane VOC emissions from the fuel in question, in grams per mile, for VOC control region 2 during the summer period. * * * * * (b) * * * (1) * * * (ii) * * * Oxygen in the form of methanol or non-alcohol, non-ether oxygenates shall not be evaluated with the Simple Model, but instead must be evaluated through vehicle testing under the Complex Model per Sec. 80.48. * * * * * (2) * * * (ii) * * * Oxygen in the form of methanol or non-alcohol, non-ether oxygenates shall not be evaluated with the Simple Model, but instead must be evaluated through vehicle testing under the Complex Model per Sec. 80.48. * * * * * (3) * * * (ii) * * * Oxygen in the form of methanol or non-alcohol, non-ether oxygenates shall not be evaluated with the Simple Model, but instead must be evaluated through vehicle testing under the Complex Model per Sec. 80.48. (4) If the fuel aromatics content of the fuel in question is less than 10 volume percent, then an FAROM value of 10 volume percent shall be used when evaluating the toxics emissions equations given in paragraphs (b)(1), (b)(2), and (b)(3) of this section. (c) * * * (1) * * * Fuel parameter Range Benzene content 0-4.9 vol %. RVP 6.4-9.0 psi. Oxygen content 0-4.0 wt %. Aromatics content 0-55 vol %. * * * * * 4. Section 80.45 is amended by: a. revising Table 3 in paragraph (b)(3); b. revising Table 6 in (c)(1)(iv)(A); c. revising the first sentence and the equations for Phase I and II in paragraph (c)(1)(iv)(B), and revising (c)(1)(iv)(C)(5); d. revising the second sentence in paragraph (c)(1)(iv)(C)(9); e. revising paragraphs (c)(1)(iv)(C)(11) and (12); f. revising the first sentence in paragraph (c)(1)(iv)(C)(13) and revising paragraph (c)(1)(iv)(C)(14); g. revising paragraph (c)(1)(iv)(D)(5), revising the second sentence in paragraph (c)(1)(iv)(D)(9), and revising paragraphs (c)(1)(iv)(D)(11) and (12); h. revising the first sentence in paragraph (c)(1)(iv)(D)(13), and revising paragraph (c)(1)(iv)(D)(14); i. revising the equation for VOCRL1 in paragraph (c)(3)(i); j. revising the equations for VOCHS1 and VOCRF1 in paragraph (c)(3)(ii); k. revising the equation for VOCRL2 in paragraph (c)(4)(ii), and revising paragraph (c)(8)(ii); l. revising paragraph (d)(1)(iv)(A) and Table 7, revising the first sentence and the equations for Phase I and II in paragraph (d)(1)(iv)(B), revising paragraph (d)(1)(iv)(C)(5), and revising the second sentence in paragraph (d)(1)(iv)(C)(9); m. revising the Phase I equation for "Toxics 2%" and the Phase II equation for "Toxics 1%" in paragraph (e)(1)(ii); n. revising paragraph (e)(3) introductory text and removing paragraphs (e)(3)(i) and (e)(3)(ii); o. revising the last sentence in paragraph (e)(4)(iii); p. adding a concluding sentence to paragraphs (e)(5)(iv) and (e)(6)(iv); q. revising equations "HSBZ1" through "RFBZ1" in paragraph (e)(9) and equations "HSBZ2" through "RFBZ2" in paragraph (e)(10); and r. revising paragraph (f)(1). The additions and revisions are set out to read as follows: Sec. 80.45 Complex emissions model. * * * * * (b) * * * (3) * * * Table 3.--Baseline Exhaust Emissions Phase I Phase II Summer Winter Summer Winter Exhaust (mg/ (mg/ (mg/ (mg/ pollutant mile) mile) mile) mile) VOC 446.0 660.0 907.0 1341.0 NOx 660.0 750.0 1340.0 1540.0 Benzene 26.10 37.57 53.54 77.62 Acetaldehyde 2.19 3.57 4.44 7.25 Formaldehyde 4.85 7.73 9.70 15.34 1,3-Butadiene 4.31 7.27 9.38 15.84 POM 1.50 2.21 3.04 4.50 * * * * * (c) * * * (1) * * * (iv) * * * (A) * * * Table 6.--Allowable Ranges of E200, E300, and ARO for the Exhaust VOC Equations in Paragraphs (c)(1)(i) and (ii) of This Section Phase I Phase II Fuel Lower Higher Lower parameter limit limit limit Higher limit E200 33.00 65.83 33.00 65.52 E300 72.00 Variable/1/ 72.00 Variable /2/ ARO 18.00 46.00 18.00 46.00 /1/ Higher E300 limit=lower of 94.0 or 80.32+[0.390x(ARO)]. /2/ Higher E300 limit=lower of 94.0 or 79.75+[0.385x(ARO)]. (B) For fuels with E200, E300, and/or ARO levels outside the ranges defined in Table 6, YVOC(t) shall be defined as: For Phase I: YVOC(t)=100%x0.52x[exp(v1(et))/exp(v1(b))-1] +100%x0.48x[exp(v2(et))/exp(v2(b))-1] +{100%-0.52x[exp(v1(et))/exp(v1(b))] x{[(0.0002144xE200et)- 0.014470]xDE200} +{[(0.0008174xE300et)-0.068624 -(0.000348xAROet)]xDE300} +{[(-0.000348xE300et)+ 0.0323712]xDARO}]} +{100%x0.48x[exp(v2(et))/exp(v2(b))] x[{[(0.000212xE200et)- 0.01350]xDE200} +{[(0.000816xE300et)-0.06233 -(0.00029xAROet)]xDE300} +{[(-0.00029xE300et)+ 0.028204]xDARO}]} For Phase II: Y VOC(t)=100%x0.444x[exp(v1(et))/exp(v1(b))-1] +100%x0.556x[exp(v2(et))/exp(v2(b))-1] +{100%x0.444x[exp(v1(et))/exp(v1(b))] x[{[(0.0002144xE200et)- 0.014470]xDE200} +{[(0.0008174xE300et)-0.068624 -(0.000348xAROet)]xDE300} +{[(-0.000348xE300et)+ 0.0323712]xDARO}]} +{100%x0.556x[exp(v2(et))/exp(v2(b))] x[{[(0.000212xE200et)- 0.01350]]xDE200} +{[(0.000816xE300et)-0.06233 -(0.00029xAROet)]xDE300} +{[(-0.00029xE300et)+ 0.028204]xDARO}]} * * * * * (C) * * * (5) If the E300 level of the target fuel is greater than 95 volume percent, then the E300 value of the target fuel shall be set equal to 95 volume percent for the purposes of calculating VOC emissions with the Phase I equation given in paragraph (c)(1)(iv)(B) of this section. * * * * * (9) * * * If the aromatics level of the target fuel is less than 10 volume percent, then DARO shall be set equal to -8 volume percent. * * * * * (11) If neither of the conditions established in paragraphs (c)(1)(iv)(C)(9) and (10) of this section are met, then DARO shall be set equal to zero. (12) If the E300 level of the target fuel is less than 72 percent, then DE300 shall be set equal to (E300-72 percent). (13) If the E300 level of the target fuel is greater than 94 volume percent and [80.32+(0.390xARO)] also is greater than 94, then DE300 shall be set equal to (E300-94 volume percent).* * * (14) If neither of the conditions established in paragraphs (c)(1)(iv)(C)(12) and (13) of this section are met, then DE300 shall be set equal to zero. (D) * * * (5) If the E300 level of the target fuel is greater than 95 volume percent, then the E300 value of the target fuel shall be set equal to 95 volume percent for the purposes of calculating VOC emissions with the Phase II equation given in paragraph (c)(1)(iv)(B) of this section. * * * * * (9) * * * If the aromatics level of the target fuel is less than 10 volume percent, then DARO shall be set equal to -8 volume percent. * * * * * (11) If neither of the conditions established in paragraphs (c)(1)(iv)(D)(9) and (10) of this section are met, then DARO shall be set equal to zero. (12) If the E300 level of the target fuel is less than 72 percent, then DE300 shall be set equal to (E300)x72 percent). (13) If the E300 level of the target fuel is greater than 94 volume percent and [80.32+(0.390xARO)] also is greater than 94, then DE300 shall be set equal to (E300 -94 volume percent). * * * (14) If neither of the conditions established in paragraphs (c)(1)(iv)(D)(12) and (13) of this section are met, then DE300 shall be set equal to zero. * * * * * (3) * * * (i) * * * VOCRL1=[0.00279 x (RVP2)]+[0.1096 x RVP] -0.7340 * * * * * (ii) * * * VOCHS1=[0.006654 x (RVP2)] -[0.08094 x RVP]+0.2846 * * * * * VOCRF1=[0.004767 x RVP]+0.011859 (4) * * * (ii) * * * VOCRL2=[0.016169 x (RVP2)] -[0.17206 x RVP]+0.56724 * * * * * (8) * * * (ii) The total winter VOC emissions performance of the target fuel in percentage terms from baseline levels shall be given by the following equation during Phase II: VOCW%=[100% x (VOC -1.341 g/mi)] / (1.341 g/mi) (d) * * * (1) * * * (iv) * * * (A) The equations in paragraphs (d)(1)(i) and (ii) of this section shall be used within the allowable range of SUL, OLE, and ARO for the appropriate Phase, as defined in the following Table 7: Table 7.--Allowable Ranges of SUL, OLE, and ARO for the NOX Equations in Paragraphs/(d)(1)(i) and (ii) of This Section Phase I Phase II Fuel Low High Low High parameter end end end end SUL 10.0 450.0 10.0 450.0 OLE 3.77 19.0 3.77 19.0 ARO 18.0 36.2 18.0 36.8 (B) For fuels with SUL, OLE, and/or ARO levels outside the ranges defined in Table 7 of paragraph (d)(2)(iv)(A) of this section, Ynox(t) shall be defined as: For Phase I: YNOX (t)=100% x 0.82 x [exp (n1(et))/exp (n1(b)) -1] +100% x 0.18 x [exp(n2(et))/exp(n2(b)) -1] +{100% x 0.82 x [exp(n1(et))/exp(n1(b))] x [{[(0.00000133 x SULet)+0.000692] x DSUL} +{[(-0.000238 x AROet)+0.0083632] x DARO} +{[(0.000733 x OLEet) -0.002774] x DOLE}]} +{100% x 0.18 x [exp(n2{et))/exp(n2(b))] x [{0.000252 x DSUL} + +{[(-0.0001599 x AROet)+0.007097] x DARO} +{[(0.000732 x OLEet) -0.00276] x DOLE}]} For Phase II: Ynox(t)=100% x 0.738 x [exp(n1(et))/exp(n1(b)) -1] +100% x 0.262 x [exp(n2(et))/exp(n2(b)) -1] +{100% x 0.738 x [exp(n1(et))/exp(n1(b))] x [{[(-0.00000133 x SULet)+0.000692] x DSUL} +{[(-0.000238 x AROet)+0.0083632] x DARO} +{[(0.000733 x OLEet) -0.002774] x DOLE}]} +{100% x 0.262 x [exp(n2(et))/exp(n2(b))] x [{0.000252 x DSUL}+ +{[(-0.0001599 x AROet)+0.007097] x DARO} +{[(0.000732 x OLEet) -0.00276] x DOLE}]} * * * * * (C) * * * (5) If the E300 level of the target fuel is greater than 95 volume percent, then the E300 value of the target fuel shall be set equal to 95 volume percent for the purposes of calculating NOX emissions with the equations given in paragraph (d)(1)(iv)(B) of this section. * * * * * (9) * * * If the aromatics level of the target fuel is less than 10 volume percent, then DARO shall be set equal to -8 volume percent. * * * * * (e) * * * (1) * * * (ii) * * * TOXICS2% = [100% x (TOXICS2 - 47.58 mg/mi)] / (47.58 mg/mi) * * * * * TOXICS1% = [100% x (TOXICS1 - 86.34 mg/mi)] / (86.34 mg/mi) * * * * * (3) The year-round toxics performance in VOC Control Regions 1 and 2 shall be derived from volume-weighted performances of individual batches of fuel as described in Sec. 80.67(g). (4) * * * (iii) * * * If the E300 value of the target fuel is greater than 95 volume percent, then an E300 value of 95 volume percent shall be used when evaluating the equations in paragraphs (e)(4)(i) and (ii) of this section. (5) * * * (iv) * * * Oxygen in the form of methanol or non-alcohol, non-ether oxygenates shall not be evaluated with the Complex Model, but instead must be evaluated through vehicle testing per Sec. 80.48. (6) * * * (iv) * * * Oxygen in the form of methanol or non-alcohol, non-ether oxygenates shall not be evaluated with the Complex Model, but instead must be evaluated through vehicle testing per Sec. 80.48. * * * * * (9) * * * HSBZ1 = 10 x BEN x VOCHS1 x [(-0.0342 x MTB) + (-0.080274 x RVP) + 1.4448] DIBZ1 = 10 x BEN x VOCD11 x [(-0.0290 x MTB) + (-0.080274 x RVP) + 1.3758] RLBZ1 = 10 x BEN x VOCRL1 x [(-0.0342 x MTB) + (-0.080274 x RVP) + 1.4448] RFBZ1 = 10 x BEN x VOCRF1 x [(-0.0296 x MTB) + (-0.081507 x RVP) + 1.3972 * * * * * (10) * * * HSBZ2 = 10 x BEN x VOCHS2 x [(-0