Regulation of Fuels and Fuel Additives: Modifications to Standards and Requirements for Reformulated and Conventional Gasoline

SUMMARY: Through the amended Clean Air Act of 1990, Congress mandated

that EPA promulgate regulations requiring that gasoline sold in certain

areas be reformulated to reduce vehicle emissions of toxic and ozone-

forming compounds. The EPA published rules for the certification and

enforcement of reformulated gasoline (RFG) and provisions for non-

reformulated or conventional gasoline on February 16, 1994.

Based on experience gained since the promulgation of these

regulations, EPA is proposing a variety of changes to the regulations

relating to emissions standards, emissions models, compliance related

requirements and enforcement provisions. The proposed changes involve

both the reformulated and conventional gasoline programs. Many of the

changes codify guidance issued by the Agency since the initial adoption

of these gasoline programs. These changes are in the nature of minor

adjustments to the structure of these programs. The emissions benefits

achieved from reformulated gasoline will not be reduced.

DATES: The comment period on this proposed action will close August 11,

1997, unless a hearing is requested, in which case the comment period

will close 30 days after the close of the public hearing. EPA will

conduct a hearing (date and location to be announced) if a request for

such is received by July 18, 1997.

ADDRESSES: Written comments on this proposed action should be addressed

to Public Docket No. A-97-03, Waterside Mall (Room M-1500),

Environmental Protection Agency, Air Docket Section, 401 M Street, SW,

Washington, DC 20460. The Agency requests that commenters also send a

copy of any comments to Marilyn Bennett, U.S. Environmental Protection

Agency, Office of Air and Radiation, at the address listed in the For

Further Information Contact section. Those wishing to notify EPA of

their intent to submit adverse comment or request an opportunity for a

public hearing on this action should contact Marilyn Bennett at (202)

233-9006. Materials relevant to the final rule establishing standards

for reformulated gasoline and anti-dumping standards for conventional

gasoline are contained in Public Dockets--A-92-01 and A-92-12, and are

incorporated by reference.

The preamble, regulatory language and regulatory support document

are also available electronically from the EPA Internet Web site and

via dial-up modem on the Technology Transfer Network (TTN), which is an

electronic bulletin board system (BBS) operated by EPA's Office of Air

Quality Planning and Standards. Both services are free of charge,

except for your existing cost of Internet connectivity or the cost of

the phone call to TTN. Users are able to access and download files on

their first call using a personal computer per the following

information. The official Federal Register version is made available on

the day of publication on the primary Internet sites listed below. The

EPA Office of Mobile Sources also publishes these notices on the

secondary Web site listed below and on the TTN BBS.

Internet (Web)

http://www.epa.gov/docs/fedrgstr/EPA-AIR/

(either select desired date or use Search feature)

http://www.epa.gov/OMSWWW/

(look in What's New or under the specific rulemaking topic)

TTN BBS: 919-541-5742

(1200-14400 bps, no parity, 8 data bits, 1 stop bit)

Voice Helpline: 919-541-5384

Off-line: Mondays from 8 am to 12 Noon ET

A user who has not called TTN previously will first be required to

answer some basic informational questions for registration purposes.

After completing the registration process, proceed through the

following menu choices from the Top Menu to access information on this

rulemaking.

Gateway to TTN Technical Areas (Bulletin Boards)

OMS--Mobile Sources Information

(Alerts display a chronological list of recent documents)

Rulemaking & Reporting

At this point, choose the topic (e.g., Fuels) and subtopic (e.g.,

Reformulated Gasoline) of the rulemaking, and the system will list all

available files in the chosen category in date order with brief

descriptions. To download a file, type the letter ``D'' and hit your

Enter key. Then select a transfer protocol that is supported by the

terminal software on your own computer, and pick the appropriate

command in your own software to receive the file using that same

protocol. After getting the files you want onto your computer, you can

quit the TTN BBS with the oodbye command.

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.

FOR FURTHER INFORMATION CONTACT: Marilyn Bennett, Fuels and Energy

Division, U.S. EPA, 401 M Street, SW (6406J), Washington, DC 20460.

Telephone: (202) 233-9006.

SUPPLEMENTARY INFORMATION:

Regulated Entities

Regulated categories and entities potentially affected by this

action include:

------------------------------------------------------------------------

Category Examples of regulated entities

------------------------------------------------------------------------

Industry.............................. Refiners, importers, and

distributors of motor vehicle

fuel; motor vehicle fuel retail

outlets and wholesale purchaser-

consumer facilities; facilities

that act as independent

laboratories.

------------------------------------------------------------------------

This table is not intended to be exhaustive, but rather provides a

guide for readers regarding entities likely to be regulated by this

action. This table lists the types of entities that EPA is now aware

could be potentially regulated by this action. Other types of entities

not listed in the table could also be regulated. To determine whether

your entity is regulated by this action, you should carefully examine

the applicability criteria of part 80, subparts D, E and F, of title 40

of the Code of Federal Regulations. If you have questions regarding

applicability of this action to a particular entity, consult the person

listed in the preceding FOR FURTHER INFORMATION CONTACT section.

Today's preamble explains the basis for the regulatory changes and

the purpose of the proposed rule. The remainder of this preamble is

organized into the following sections:

I. Corrections of Typographical Errors and Minor Revisions

II. General Fuels Provisions

III. RFG and Anti-dumping Standards/Models

IV. RFG Compliance Requirements

V. Enforcement

VI. Anti-dumping Requirements

VII. Attest Engagements

[[Page 37339]]

VIII. Environmental and Economic Impacts

IX. Public Participation

X. Regulatory Flexibility

XI. Executive Order 12866

XII. Paperwork Reduction Act

XIII. Unfunded Mandates Act

XIV. Statutory Authority

I. Corrections of Typographical Errors and Minor Revisions

Sec. 80.2(w)................ The reference to the cetane index test

method is removed and added as Sec.

80.3(e). As a replacement, a definition

is proposed for ``previously certified

gasoline'' to mean RFG and conventional

gasoline that has been produced or

imported in conformance with applicable

requirements and included in the

refinery, oxygenate blender or importer

compliance calculations.

Sec. 80.2(y)................ The reference to the sulfur content test

method is removed and added as Sec.

80.3(f). This section is revised to

conform to the sulfur test method in

Sec. 80.46(a).

Sec. 80.2(z)................ The reference to the aromatics content

test method is removed and added as Sec.

80.3(g). This section is revised to

limit the test method to use for diesel

fuel only to avoid conflict with the

test method for aromatics content of RFG

in Sec. 80.46(f).

Sec. 80.2(ee)............... Revises the definition of reformulated

gasoline to delete the requirement for a

gasoline marker under Sec. 80.82.

Sec. 80.2(gg)............... Revises definition of gasoline ``batch''

to make this definition apply to

conventional gasoline as well as to RFG.

Sec. 80.41(d)............... Revises chart to replace ``32.6'' for VOC per-gallon minimum

reduction with ``32.6'' and

replace ``-2.5'' with ``-2.5'' for per-gallon minimum NOX

performance reduction (percent).

Sec. 80.45(c)(1)(iv)(B)..... Corrects several small typographical

errors in both the Phase I and Phase II

equations.

Sec. 80.45(c)(1)(iv)(D)(12). Corrects typographical error by changing

``(E300 X 72 percent)'' to ``(E300--72

percent).''

Sec. 80.45 (c)(1)(iv)(D) Corrects typographical error by changing

(13). Phase I coefficients to Phase II

coefficients, i.e. change ``80.32 +

(0.390 X ARO)'' to ``79.75 + (0.385 X

ARO).''

Sec. 80.45(d)(1)(iv)(B)..... Corrects typographical errors to the

equation.

Sec. 80.45(f)(1)(i)......... Corrects the entry for aromatics

``acceptable range'' to read ``0.0--55.0

volume percent.''

Sec. 80.49(a)............... Corrects typographical error. There is a

reference to section 80.43(c), which is

incorrect. The proper reference is to

section 80.49(a)(5)(i).

Sec. 80.49(a)(1)............ Corrects typographical error in formula

at the bottom of the new parameter under

Fuel 2. Changes from ``C+B/2'' to

``(C+B)/2.''

Sec. 80.49(a)(3)............ Corrects typographical error. There is a

reference to Sec. 80.43(c), which is

incorrect. The proper reference is to

Sec. 80.49(a)(5)(i).

Sec. 80.49(b)............... Corrects typographical error. There is a

reference to Sec. 80.43(c), which is

incorrect. The proper reference is to

Sec. 80.49(a)(5)(i).

Sec. 80.50(a)(2)............ Corrects reference to ``extension fuels

per the requirements of Sec. 80.49(a)''

to read ``extension fuels per the

requirements of Sec. 80.49(b).''

Sec. 80.65(e)(2)(ii)(B)..... Revises to apply to importers as well as

refiners.

Sec. 80.65(g)............... Revises to delete heading: ``Marking of

conventional gasoline.''

Sec. 80.68(b)(2)(ii)........ Revises the word ``area'' to read

``area(s)'' to clarify the application

of the equation to a situation in which

more than one area fails a survey or

survey series in a single year.

Sec. 80.69(a)(6)(iv)........ Revises to add reference to Sec.

80.69(e)(2).

Sec. 80.69(e)............... Revises to clarify reference by removing

``who obtains any RBOB in any gasoline

delivery truck'' and adding ``other than

a terminal storage tank blender

specified in Sec. 80.69(c)'.

Sec. 80.69(e)(2)(i)(A)...... Revises to add the word ``to.''

Sec. 80.69(e)(2)(v)......... Corrects reference to Sec.

80.70(b)(2)(i). The correct reference is

to Sec. 80.65(e)(2)(i).

Sec. 80.75(a)............... Revises to require refiners, importers,

and oxygenate blenders to include

notification to EPA of per-gallon versus

average election with the first

quarterly reports submitted each year.

Sec. 80.75(a)(3)............ Revises to add a new Sec. 80.75(a)(3)

which provides a mathematical equation

for converting weight percent oxygen

from an oxygenate to volume percent

oxygenate.

Sec. 80.77(c)............... Revises to add reference to RBOB.

Sec. 80.77(f)............... Revises to add reference to RBOB.

Sec. 80.128(e)(2)........... Revises by changing reference from Sec.

80.69(a)(9) to Sec. 80.69(a)(2).

II. General Fuels Provisions

A. Test Methods (Sec. 80.3; RFG Test Methods Sec. 80.46)

1. Replacement of Lead and Phosphorus Test Methods with Industry

Standard Test Methods (Secs. 80.3 (a) and (b))

40 CFR part 80, appendices A and B, specify the test methods that

are used for determining, respectively, the phosphorus content and the

lead content of gasoline. Today's proposal would remove appendices A

and B and add Secs. 80.3 (a) and (b) which would require the use of

ASTM method D 3231-94 for phosphorus and methods D 3237-90 or D 5059-92

for lead. The phosphorus and lead test methods are used primarily to

determine compliance with the standards under Secs. 80.22 and 80.23,

dealing with the unleaded gasoline program. Also, under

Sec. 80.41(h)(1), RFG may contain no heavy metals. As a result, the

proposed lead test method would be used for determining the presence of

this heavy metal in RFG.

The test methods in appendices A and B of 40 CFR part 80 originally

were adopted from ASTM standard test methods. Over time, however, ASTM

has updated their test methods, while EPA has not. EPA believes the

current ASTM test methods are equivalent to the methods currently in

the regulations, and are more consistent with the test methods

regulated parties normally use for commercial purposes. As a result,

the proposed test methods would be appropriate for determining

compliance with the provisions of 40 CFR part 80.

EPA believes there would be little additional burden on the

regulated industry if the proposed phosphorus and lead test methods

were adopted. Initially, EPA understands that the proposed test methods

are the current industry standard test methods, so most gasoline

testing laboratories already are equipped to conduct the proposed test

[[Page 37340]]

methods. In addition, there is no requirement for regulated parties to

test their gasoline for phosphorus or lead under either the unleaded

gasoline or the RFG regulations, so parties would not be obligated to

use the proposed test methods at all. Rather, the phosphorus and lead

test methods in the regulations are used by EPA to determine if

gasoline meets standards for these metals. EPA or a regulated party

also could use non-regulatory phosphorus or lead test methods. However,

in an enforcement proceeding, the results from non-regulatory test

methods would only constitute evidence of the results that would have

been obtained if the regulatory test method had been conducted on the

gasoline at issue.

2. Reformulated Gasoline Test Methods (Secs. 80.46 (a) Through (g))

In Sec. 80.46, test methods were specified for the measurement of

the regulated properties of reformulated gasoline. Many of the test

methods designated in the original rule were consensus standards,

prepared and maintained by ASTM. Since the original issuance of the

rule, some of these methods have been updated. EPA is now proposing to

replace the current regulatory methods with the updated versions of

these methods for the measurement of sulfur, olefins, and distillation

parameters. In addition, EPA is proposing an alternative test method

(ASTM D 5453-93) for determining the sulfur content in conventional

gasoline until September 1, 1998. This proposed alternative test method

is discussed in Section VI.B.6. The proposed updated methods all are

finally approved ASTM test methods. In addition, ASTM has developed a

method (ASTM D 5599-95) that is the same as the procedure for the

measurement of oxygenates at Secs. 80.46(g) (1) through (8), and EPA

proposes to replace Secs. 80.46(g) (1) through (8) with a reference to

the ASTM method. For the measurement of RVP, EPA proposes to eliminate

the appendix containing EPA Method 3 (appendix E), and designate ASTM D

5191-96 as the required method, with the exception that the correlation

equation as described in EPA Method 3 must be used in place of the

correlation equation described in ASTM D 5191-96. ASTM D 5191-96 is

identical to the RVP test method in appendix E when the correlation

equation from EPA Method 3 is used with the ASTM method. In all cases,

these changes do not amount to a deviation in method, or significant

change in procedure. Most of the ASTM changes revolve around

improvements in quality statements. The inclusion of ASTM D 5599-95 for

oxygenates is the result of ASTM preparing a test method that is

consistent with that previously defined in the Federal Register.

The test method previously designated for benzene, ASTM D 3606, has

been updated since the original publication of the rule. However, the

parameters must be adjusted to allow for the resolution of ethanol and

methanol from the benzene. In addition, the EPA GC/MS method has been

demonstrated through ASTM round-robin testing to be an equivalent

method for the measurement of benzene. Since the use of the EPA GC/MS

method would allow two parameters (benzene and aromatics) to be

performed with a single test, EPA believes the use of the EPA GC/MA

method for the measurement of benzene would result in a reduced burden

to the regulated industry, and, therefore, is proposing to allow its

use as an alternate test procedure for the measurement of benzene in

gasoline.

3. Butane Test Methods (Sec. 80.46(h))

Blendstocks require the same full set of parameter measurements as

reformulated gasoline, since final properties must be extrapolated for

all final blends. When butane designated for blending must be tested,

the designated methods are generally not applicable, since the

properties for butane typically fall outside the scope of the methods.

Therefore, EPA is proposing to designate several test methods

specifically for butane blendstock testing. ASTM D 2163-91 and D 5623-

94 have been identified as suitable methods for the measurement of

light hydrocarbons and sulfur respectively. The Gas Producers

Association (GPA) has developed a method for the measurement of benzene

and aromatics in butane. This method is GPA 2186-95. EPA is not

proposing to designate a method for measuring olefins in butane. No

consensus method currently exists for measuring total olefins in butane

blendstocks. ASTM D 2163-91 will measure the lighter olefins, but not

any heavier ones in the mix. EPA has identified a proprietary method,

known as the Wasson ECE 383-01 method, which measures all of the

olefinic compounds in the blendstock. This method is not a consensus

standard, but is of the type that would be acceptable, due to its

ability to measure total olefins.

4. Volatility Test Methods (Secs. 80.3 (c) and (d))

As discussed above, for the measurement of RVP, EPA proposes to

eliminate the appendix containing EPA Method 3 (appendix E) and

designate ASTM D 5191-96 as the regulatory method, with the exception

that the correlation equation as described in EPA Method 3 must be used

in place of the correlation equation described in ASTM D 5191-96.

The measurement of alcohols, especially ethanol, for the volatility

rule has been described in detail in appendix F of 40 CFR part 80. In

this appendix, Method 1 describes a water extraction method, and Method

2 details a chromatographic procedure (an older version of ASTM D

4815.) In an effort to harmonize methods, EPA believes it would reduce

the testing burden to allow test methods that are consistent with the

reformulated gasoline rule. As a result, EPA proposes to eliminate

Appendix F and designate ASTM D 5599-95 as the method for the

measurement of alcohols in gasoline for the purpose of complying with

the volatility regulations. Consistent with the reformulated gasoline

rule, the use of ASTM D 4815-94a will be allowed as an alternate as

long as this use is allowed under the reformulated gasoline rule.

5. Diesel Fuel Test Methods (Secs. 80.3 (e), (f), and (g))

When the diesel sulfur rule was originally published by EPA,

several methods were included for the measurement of the regulated

properties. Included in these properties are sulfur concentration,

cetane index, and aromatic content. The current designated test for

sulfur is ASTM D 2622-87, with D 4294-83 being an allowable alternate.

As discussed above, EPA proposes to substitute the current regulatory

test method for sulfur, D 2622-87, with the latest version of this

method, D 2622-94. EPA also proposes to substitute the alternate method

for determining sulfur content in diesel fuel, D 4294-83, with the

latest version, D 4294-90(1995), and substitute the current test method

for cetane index, ASTM D 976-80, with the latest version, D 976-91.

The test for aromatics in diesel had been designated to be ASTM D

1319-88. EPA recognizes that ASTM describes this test as inadequate for

the measurement of the aromatic content in diesel fuel. For some time,

EPA has been performing ASTM D 5186 in parallel with D 1319, and found

D 5186 to be superior in both precision and accuracy. The primary

difficulty in changing from the use of D 1319 to D 5186 to measure

compliance lies in the units reported by the two methods. The

regulation specifies a limit on the aromatic content in volume per-

cent, coincidentally the same units reported

[[Page 37341]]

by D 1319. Unfortunately, D 5186 reports results in mass per-cent. In

order to comply with the regulation, these results must be converted to

volume per-cent. EPA proposes to apply a conversion factor to the

results. The equation to be used for the conversion of mass per-cent

diesel aromatics to volume per-cent diesel aromatics is:

Vol% = (Mass% * 0.916) + 1.33

Where Mass% refers to the output from D 5186-96, the SFC test.

This equation is identical to that used by CARB in their conversion

of mass per-cent results to volume per-cent results for the affirmation

of regulatory limits.

This change should not impose any additional financial burden on

industry, since it is not a required test. The option of measuring

aromatics was originally placed in the rule to allow an alternate to

the requirement that low sulfur fuels meet a minimum requirement of a

40 cetane index. The intent was to regulate aromatic content, and it

was found that some fuels with high napthenic content could actually be

very low in aromatics, yet still not meet the 40 cetane index level.

The option to test for aromatic content would only be exercised if a

fuel fails to meet the required cetane index level, a relatively

infrequent occurrence.

6. Table of Test Methods

The following table sets out the test methods currently required

under the fuels regulations at 40 CFR part 80, and the corresponding

proposed test methods:

----------------------------------------------------------------------------------------------------------------

Parameter Old test New test

----------------------------------------------------------------------------------------------------------------

Reformulated Gasoline:

RVP............................... EPA Method 3.................. ASTM 5191-96, except that equation is as

in Method 3.

Benzene........................... ASTM D3606-92, with exceptions ASTM D3606-96, also with exceptions. In

for Methanol and Ethanol. addition, the use of the EPA GC/MS

Method for the measurement of Benzene

will now be allowed as an alternate.

Distillation...................... ASTM D86-90................... ASTM D86-96.

Aromatics......................... EPA GC/MS Method (80.46)...... EPA GC/MS Method (80.46) (No Change)

alternate is D1319-95a.

Olefins........................... ASTM D1319-93................. ASTM D1319-95a.

Sulfur............................ ASTM D2622-92................. ASTM D2622-94 (ASTM D5453-93 is

alternate for Conventional Gasoline to

9/1/98).

Oxygenates........................ EPA OFID Method (80.46)....... ASTM D5599-95, alternate is D4815*-94a.

Lead Phase Down:

Phosphorus........................ Appendix A.................... ASTM D3231-94.

Lead.............................. Appendix B.................... ASTM D3237-90 (Atomic Absorbance) or

D5059-92 (X-ray).

Volatility:

Alcohol......................... .............................. Consistent with Reformulated Gasoline.

Diesel Sulfur:

Sulfur............................ ASTM D2622-87, or D4294-83.... ASTM D2622-94, or D4294-90 (1995).

Aromatics......................... ASTM D1319-88................. ASTM D5186-96.

Cetane Index...................... ASTM D976-80.................. ASTM D976-91.

Blendstock Tests:

Light Hydrocarbons in Butane...... .............................. ASTM D2163-91.

Sulfur in Butane.................. .............................. ASTM D5623-94.

Benzene and Aromatics in Butane... .............................. GPA 2186-95.

Olefins in Butane................. .............................. Test procedure not specified. Wasson-ECE

383-01 is an example of an acceptable

test procedure.

----------------------------------------------------------------------------------------------------------------

B. Gasoline and Diesel Fuel Sampling Procedures (Proposed

Sec. 80.8)

40 CFR part 80, Appendices D and G, specify sampling procedures for

gasoline and diesel fuel for all motor vehicle fuel programs under 40

CFR part 80, including the programs for unleaded gasoline, gasoline

volatility, diesel sulfur, RFG, and anti-dumping. Today's proposal

would replace the sampling procedures in appendices D and G with the

following ASTM standard practices:

D 4057-95, ``Standard Practice for Manual Sampling of

Petroleum and Petroleum Products;''

D 4177-95, ``Standard Practice for Automatic Sampling of

Petroleum and Petroleum Products;''

D 5842-95, ``Standard Practice for Sampling and Handling

of Fuels for Volatility Measurements;'' and

D 5854-95, ``Standard Practice for Mixing and Handling of

Liquid Samples of Petroleum and Petroleum Products.''

Appendices D and G were adopted from the 1981 version of D 4057.

Over time, however, ASTM has updated D 4057, and these changes are not

reflected in Appendices D and G. For example, appendix D addresses the

collection of samples from a ``tap'' in the shell of a petroleum

storage tank. The current requirement under appendix D, reflective of D

4057-81, requires that taps extend at least three feet into the storage

tank. See, para. 11.3.1.1 of appendix D. However, tap extensions are

necessary only for heavy petroleum products (and not for gasoline and

diesel fuel), and, furthermore, tap extensions are not possible with

floating roof storage tanks that are commonly used today. As a result,

EPA and regulated parties currently agree to waive the tap extension

requirement on a case-by-case basis. Under D 4057-95 sampling tap

extensions are not required for light petroleum products such as

gasoline and diesel fuel, so that if this ASTM procedure were adopted

the tap extension issue would be resolved for all cases.

EPA is proposing to adopt three ASTM methods in addition to D 4057-

95 in order to include procedures that address a broad scope of

sampling situations that are relevant to EPA's motor vehicle fuels

programs. D 4177-95 deals with automatic sampling of petroleum

products, which is relevant under the anti-dumping regulations for

refiners who produce conventional gasoline using an in-line blending

operation where automatic sampling is necessary. Similarly, D 5842-95

deals with sampling and sample handling for volatility measurement,

which is relevant to determining compliance with the volatility

standards in Sec. 80.27 and the RFG standards in Sec. 80.41. Last, D

5854-96 deals with the creation of composite samples, which is relevant

[[Page 37342]]

under the RFG and anti-dumping programs in certain situations involving

imported gasoline where the gasoline from multiple ship compartments is

treated as a single batch.

EPA believes it is appropriate to replace Appendices D and G with

ASTM standard practices. The current ASTM practices reflect up to date

procedures, which if followed would result in improved sample quality

for regulatory purposes. In addition, the adoption of industry standard

procedures would reduce regulatory burden because parties would be able

to follow their customary practices when meeting regulatory

requirements.

III. RFG and Anti-Dumping Standards/Models

A. Standards and Requirements for Compliance (Secs. 80.41 and 80.101)

1. Averaging Per-Gallon Minimum Standards for NOX

(Secs. 80.41 (d) and (f); Sec. 80.68(b)(1)(iv))

Reduction of NOX emissions is a prominent feature of

Phase II of the Reformulated Gasoline Program which goes into effect on

January 1 of 2000 (Phase I provides control at a ``no NOX

increase'' level). The Phase II standard for refiners choosing to

comply on average (requiring a 6.8% reduction from baseline during the

high ozone season) sets the level of NOX emission reduction

required on average by these refiners. Thus, for refiners who choose to

average, the averaging standard effectively controls the overall

environmental benefit contributed to the program by these refiners.

In addition to the average NOX standards, though, there

are also per-gallon minimum reduction standards for refiners that

choose to average (not to be confused with standards for overall

compliance on a per-gallon basis). The averaging minimum standard in

Phase II requires that each gallon (batch) of RFG in the high ozone

season has at least a 3% reduction from the statutory baseline; the

corresponding Phase I standard holds any increase over statutory

baseline for a batch to 2.5%. Less stringent minimum standards apply

outside of the high ozone season in Phase II. The per-gallon minimum

standards are in addition to the year-long average standard of a

refinery's output of a given type of RFG and these minimum standards

set the NOX reduction which must be achieved by each batch

(and therefore each gallon) of RFG.

These NOX per-gallon minimum standards were not put in

place to provide any incremental environmental benefit beyond that

provided by the average standard, but rather to ensure an even

distribution of program benefits from area to area and through time.

This primary reason for the averaging per-gallon minimum standards (for

NOX and other parameters as well) was discussed in the

enforcement section of the preamble to the RFG final rule (Section

VII). An additional but secondary objective of the minimum standard was

to augment the ability of enforcement authorities to detect non-RFG

gasoline being illegally sold in RFG areas. For reasons that will be

discussed more fully below, EPA is proposing to eliminate the per-

gallon minimum standards for NOX and to accomplish the same

objectives that these standards would have accomplished by

substantially expanding the number of area-by-area surveys of RFG

emission performance required to be conducted by refiners choosing to

average. EPA is not proposing any change to the averaging standard for

NOX.

The Problem With the Per-Gallon NOX Minimums

When EPA imposed the per gallon minimum standards, data did not

exist to adequately assess the variability, within refineries' output,

of NOX quality or the factors that affect it across all of

the batches of gasoline produced in a year.

Representatives of the gasoline refining industry (the American

Petroleum Institute (API), the National Petroleum Refiners Association

(NPRA) and representatives of various of their member companies) have

presented data to EPA 1 showing that NOX

performance of actual RFG retail samples varies substantially by octane

grade and from batch to batch 2 within a grade. The

processes involved in gasoline production result in a broad bimodal

3 distribution of NOX quality, with premium

batches showing characteristically lower NOX emissions and

regular batches, with their higher levels of sulfur and olefins,

showing higher NOX emissions. 4 These data on

gasoline produced under the simple model requirements showed a very

substantial proportion of regular grade RFG samples that would have

failed to meet the Phase I minimum reduction standard that applies

beginning in 1998.

---------------------------------------------------------------------------

\1\ Industry representatives met with EPA personnel on January

14, 1997 and presented a graphical analysis which can be found in

the docket for this rulemaking. Docket Number A-97-03, Item Number

II-E-1.

\2\ Since these were retail samples, they could not truly

reflect batch-to-batch variability due to the intermingling of

gasolines from different batches, and even from different sources,

in the distribution system.

\3\ A bimodal distribution here refers to one that has two

distinct frequency peaks or two values around which a large number

of batches will gather.

\4\ Engineering judgment would lead to a conclusion that a broad

distribution of NOX quality differing markedly between

premium and regular gasoline grades would exist in the gasoline

pool. First, NOX quality under EPA's complex model is

primarily a function of sulfur and olefin content in the gasoline.

Thus, differences in either of these properties would result in

differences in NOX quality. Second, in the refinery,

processes which typically contribute large volumes to the regular

gasoline grade are often high in sulfur and olefins, whereas

processes contributing heavily to the premium gasoline pool are

often very low in olefins and sulfur. For example, the fluid

catalytic cracker (FCC) unit in a refinery breaks large molecules

into smaller ones and is the ``workhorse'' of most refineries and

the largest contributor of any refinery unit to the gasoline pool.

The gasoline produced by the FCC unit is highly olefinic, and,

depending upon the crude oil source for the refinery, usually very

high in sulfur. FCC gasoline also possesses octane quality

consistent with regular gasoline. For this reason and since regular

gasoline is typically the highest volume product of U.S. refineries,

most of the product produced by the FCC is used in the production of

regular gasoline. On the other hand, premium gasolines, which differ

from regular grades primarily in the higher octane quality they

possess, contain lower amounts of FCC streams and higher levels of

high-octane aromatic streams produced by catalytic reformers. Such

streams, called reformate, are extremely low in olefins and also

very low in sulfur. Thus, a much lower level of sulfur and olefin

content and therefore, better NOX quality, is found in

the premium pool as compared to the regular pool. (A 1989 study of

blendstocks used to produce U.S. gasoline found FCC blendstocks

possessing an average octane quality of 86.4, an average olefin

content of 29.1 percent, and an average sulfur content of 756 parts

per million (ppm). The same study found that reformate streams,

produced by the reformer, possessed octane quality of 92.6, an

olefin content of less than 1 percent and an average sulfur level of

55 ppm. See ``NPRA Survey of U.S. Gasoline Quality and U.S. Refining

Industry Capacity to Produce Reformulated Gasolines--Part A'',

National Petroleum Refiners Association, 1991 Gasoline Study,

January, 1991, Docket Number A-97-03, Item Number II-B-1.)

---------------------------------------------------------------------------

In order to bring these higher NOX batches of regular

RFG into compliance, the refiners suggested that the industry would

have to incur substantial additional costs in excess of those

calculated in EPA's Regulatory Impact Assessment which EPA relied upon

in adopting the standards for RFG in 1993. That assessment of the costs

of compliance for NOX was based upon the cost of meeting the

average standard, not the per-gallon minimum that applies to refineries

that average, which is the subject of this proposal. They further

argued that in the absence of a substantial enforcement tolerance to

account for the uncertainty of measurement (especially of olefin

levels) in downstream enforcement sampling, the bimodal frequency

distribution would have to be shifted further than would otherwise be

required. While the problem created by the NOX minimum would

already be substantial in Phase I with the 1998 introduction of the

complex model, the

[[Page 37343]]

refiners suggested that Phase II's tighter minimum standard for

NOX in the year 2000 would exacerbate an already very

difficult situation, even given the changes made to refinery processes

in order to be able to comply with the Phase II average standard.

The distribution of retail sample data initially presented by the

refiners in their general meeting with the Agency described the net

result of the product intermingling that occurs in the gasoline

distribution system. By describing all of the nation's gasoline taken

together, these data could suggest the existence of a problem (high

variability with many samples below the minimum reduction standard),

but could not indicate much about how widespread the problem is or show

what types of refineries are likely to be affected. By examining

historical RFG reporting data,5 EPA was able to confirm the

general factual basis of the industry analysis. Specifically, the data

showed a broad distribution of NOX quality with the premium

batches clustered near the high end (high NOX reductions),

while regular batches are more spread out with central tendency nearer

the low end and many batches falling below the Phase I NOX

minimum. Left unanswered by either the industry-supplied information or

EPA's own analysis, was the question of whether refiners could exercise

any control over the variability and shape of the frequency

distribution that was evident in both data sources. In other words, it

was not clear what options were available to refineries to remedy the

problem.

---------------------------------------------------------------------------

\5\ Data on the characteristics of gasoline batches as they are

shipped from the refinery are submitted to EPA as part of the

reporting requirements of the RFG regulations. An aggregated

analysis that protects the confidentiality of individual refiners'

data can be found in the docket for this rulemaking. Docket Number

A-97-03, Item Number II-A-5.

---------------------------------------------------------------------------

To provide additional insights, EPA and a refinery expert from the

Department of Energy met separately with individual refiners in order

to look at batch data from single refineries using differing gasoline

production approaches. The refineries represented by the companies EPA

met with comprised a very diverse group. They varied with regard to

size, general level of technology, control over inputs, historical

product slate, and other characteristics. EPA focused the agenda for

these meetings on three basic questions: (1) For each separate

refinery, what is the batch-to-batch distribution of NOX

quality by grade and season, (2) what are the causes of the variability

that is observed in the historical data--which parameters account for

the variability in NOX, and what caused them to vary the way

they did, and (3) how do refinery managers plan to meet the

NOX minimum standards in the absence of a substantial

enforcement tolerance or regulatory relief.

The general picture of the broad bimodal distribution of gasoline

NOX quality by grade that was developed from overall

industry analyses and examination of our own data was generally

confirmed in these more detailed meetings.

As might be expected, individual facilities varied considerably in

the size of the challenge posed by the NOX minimum standards

and they expected to address that problem with varying strategies. The

pattern that emerged from all of these discussions was that refiners

intend to pursue the least capital-intensive solutions wherever

possible, even to the extent of incurring substantial additional

production costs in the short run. Although the strategies articulated

in these meetings 6 did not precisely conform to the pattern

expected by the industry associations (shifting the entire distribution

of NOX quality), they seemed to lead to the same result

economically--excess costs in producing RFG beyond the costs of making

the refinery's average conform to the average standard. Any major

expansion of the RFG program as a result of areas opting into the

program could further increase the costs of meeting the minimum

standard.

---------------------------------------------------------------------------

\6\ Some general examples of the approaches which are likely to

be used to bring sub-minimum batches above the standard include:

Finding another use for the poor NOX quality gasoline or

its components (shifting it to conventional gasoline, if that can be

done without violating anti-dumping standards, or shifting it to

other products) and buying conforming RFG on the spot market to take

its place; reblending the poor NOX quality batches with

clean blendstocks purchased from the outside to make them conform to

the minimum; or simply reducing RFG production.

---------------------------------------------------------------------------

Objectives of the NOX Minimum Standards

The primary purpose of the NOX minimum is to assure an

even temporal and geographic distribution of the program's

environmental benefits. To put this more simply, the minimum is

intended to ensure that no area covered by the RFG program will suffer

from impaired air quality (possibly resulting in an exceedance of the

NAAQS for ozone) as a result of a single refinery's shipping a batch of

high NOX gasoline to an area for which it was a primary

supplier. An additional, though secondary, purpose of the

NOX minimum standards is to provide a tool for detecting the

illegal sale of non-RFG gasoline in areas covered by the program. This

would work by keeping legitimate RFG above the minimum, while illegally

sold non-RFG might fall below the standard.

Avoiding distribution problems. The RFG regulations incorporated

two mechanisms to avoid the unlikely event of an area being

shortchanged on NOX quality due to refinery gate averaging--

the minimum standard and the RFG gasoline quality surveys.7

These surveys were specifically intended to guard against uneven

distribution of benefits. In the event that the surveys find a covered

area to have received less than the intended NOX emission

reduction benefits, the regulations provide for a substantial

tightening of the average standard--an outcome that would be expensive

to the industry and one that it will work hard to avoid. This proposal

includes an increase in the number of surveys to be conducted (an

additional 20 surveys per year) that should improve the surveillance of

gasoline quality on an area-by-area basis.

---------------------------------------------------------------------------

\7\ A program of gasoline quality surveys is required to be

conducted by refiners that wish to comply on average rather than on

a per-gallon basis. The surveys must be done by an independent

contractor in accordance with a statistically sound sampling plan

approved by EPA. The location and timing of surveys is determined by

EPA with minimal advance notice to the industry's contractor. If

survey averages fall short of the criteria set out in the

regulations, the average standards and/or the minimum standards are

made more stringent for subsequent years for all of the refineries

that supplied gasoline to the area(s) where the failure occurred.

---------------------------------------------------------------------------

Detecting non-conforming gasoline. A detailed examination of 1995

and 1996 actual batch-by-batch gasoline quality (NOX

performance) shows that the NOX minimum standard is not a

very useful tool for detecting contamination of RFG by illegally sold

conventional gasoline, since many batches of conventional gasoline,

especially premium grade, are in compliance with the minimum standard.

Minimum standards for other gasoline characteristics (especially oxygen

content and benzene levels) provide far superior capability for

determining if contamination by non-complying gasoline has taken

place.8 The proposed expansion of the survey program would

further enhance these enforcement efforts, since analysis results for

survey samples found to be out of compliance with RFG requirements are

immediately supplied to EPA's enforcement office.

---------------------------------------------------------------------------

\8\ Analysis in support of this conclusion has been placed in

the docket for this rulemaking. Docket Number A-97-03, Item Number

II-A-6.

---------------------------------------------------------------------------

Conclusions and Proposed Regulatory Actions

EPA believes, as a result of the investigations discussed above,

that the averaging minimum standards for NOX

[[Page 37344]]

are likely to be costly to the industry as a whole in both phases of

the program, and will make the 1998 complex model implementation

extremely difficult for a portion of existing refineries. With the

additional costs in question, the overall cost of compliance is likely

to exceed the cost upon which the standards were based (the cost of

meeting the average standard) without providing additional

environmental benefits. By increasing the costs of producing RFG, these

standards may contribute to a higher cost differential between RFG and

conventional gasoline and so pose a significant obstacle to smooth

implementation of Phase II of the program. Since the per-gallon minimum

standards for NOX do not increase the environmental benefit

and their purposes can be as easily served by the RFG surveys, EPA

proposes the elimination of these per-gallon minimum standards.

Since the RFG surveys provide an alternative tool for accomplishing

both of the purposes of the NOX per-gallon minimums, it is

important that the survey program remain adequate to perform these

tasks. The regulations at Sec. 80.68(b)(1) currently prescribe 50

surveys beginning in 1998, with adjustments provided for opt-in of

additional programs and/or potential survey failures. EPA believes that

20 additional surveys would provide significant additional protection

of the NOX quality of gasoline in those RFG covered areas

with limited sources of supply. Accordingly, EPA proposes that the

number of surveys in the initial schedule (Sec. 80.68(b)(1)) for each

year beginning in 1998 be expanded by 20. EPA invites comments on this

proposed change.

2. Clarification That Model Limits Constitute Standards (Proposed

Sec. 80.41(h)(3) and Sec. 80.78(a)(1)(vi); Revised Sec. 80.101(b)(3))

Both the simple and the complex models include restrictions on the

range of parameter values that may be used with these models. See

Secs. 80.42(c) and 80.45(f) for the simple model limits and the complex

model limits, respectively. These parameter range limits are included

because the simple and complex models have not been shown to accurately

predict emissions when parameter values outside the range limits are

used. For this reason, Secs. 80.42(c) and 80.45(f) state that the

models may not be used for fuels with parameter values that are outside

the valid range limits. The complex model specifies different valid

range limits for reformulated versus conventional gasoline. Compare

Sec. 80.45(f)(1)(i) (complex model range limits for reformulated

gasoline) with Sec. 80.45(f)(1)(ii) (complex model range limits for

conventional gasoline).

EPA always has considered the valid range limits to constitute

standards that apply to reformulated and conventional gasoline.

Gasoline subject to simple or complex model standards must be evaluated

for compliance with these standards. Where gasoline has property values

outside the valid range limits, it cannot be evaluated and, therefore,

it is unlawful to produce and sell such gasoline.

Today's proposal would clarify that the valid range limits are

standards, by citing the valid range limits along with the other

standards that apply to reformulated and conventional gasoline. In

addition, EPA is proposing to add a provision to the reformulated

gasoline prohibitions under Sec. 80.78(a) that addresses the valid

range limit standards. This prohibition would clarify that the complex

model valid range limits apply not only to reformulated gasoline when

produced or imported, but throughout the distribution system as well.

The complex model valid range limit standards must be applied

downstream of the refinery or importer because complex model standards

apply throughout the distribution system, i.e., the VOC and

NOX minimum per-gallon emissions performance standards. In

order to evaluate reformulated gasoline for compliance with these

downstream standards, the gasoline must have parameter values that are

within the valid range limits.

EPA is proposing to promulgate the revisions contained in this

rulemaking under the authority of both sections 211 (c) and (k) of the

Act, except for the revisions which would include the valid range

limits as standards under Sec. 80.41 and Sec. 80.101. EPA proposes to

promulgate the revisions concerning the valid range limits under the

authority of section 211(k), but not section 211(c). EPA is proposing

to promulgate the valid range limits as standards solely for the

purpose of ensuring that the models will accurately predict emissions,

and not for the independent purpose of achieving emissions reductions

from the range limits themselves. As a result, EPA believes that it is

not necessary to promulgate the valid range limits as standards under

the authority of section 211(c).

3. Effective Dates for Standard Changes Due to Survey Failures

(Sec. 80.41(p))

Section 80.41(p) states that when a minimum or maximum per-gallon

reformulated gasoline standard is changed to be more stringent as a

result of a survey failure, the effective date for the new standard is

ninety days after EPA announces the new standard. EPA now believes that

additional time is necessary in order to ensure an appropriate

transition to a new standard as a result of the lag time between the

date refiners and importers begin producing gasoline to a new standard,

and the date this gasoline displaces the earlier gasoline through the

distribution system.

For this reason, EPA is proposing a staged introduction to a new

per-gallon standard, that results from a survey failure. The dates the

new standard would be required would be expressed in the number of days

after the date EPA announces the new standard: 60 days for gasoline

produced at a refinery or imported by an importer; 120 days for

facilities downstream of the refinery or importer other than retail

outlets and wholesale purchaser-consumers; and 150 days for retail

outlets and wholesale purchaser-consumers. Under the proposed approach

refiners and importers would have about two months to begin meeting the

new standard, downstream parties such as terminal operators then would

have about two months to transition to the new standard after shipments

of gasoline meeting the new standard begin, and retailers and wholesale

users would have about one month to transition after terminals must

begin shipping gasoline meeting the new standard.

EPA believes the times proposed for these stages are consistent

with current industry practice for transitioning to new standards, such

as the transition to meet the summertime high ozone season standards

each spring. For example, terminals supplying RFG must have gasoline

that meets the VOC-control standard beginning on May 1 each year, and

retailers and wholesale purchaser-consumers in RFG areas must meet the

VOC-control standard beginning about one month later, on June 1.

Refiners must begin producing VOC-controlled RFG early enough

before May 1 that the gasoline distribution system through the terminal

level can transition from non-VOC-controlled gasoline to VOC-controlled

gasoline by May 1. The date when particular refiners begin producing

VOC-controlled RFG each year varies depending on factors such as the

time necessary to transport gasoline from the refinery to the

terminals, and the rate of turnover at the terminal. However, EPA

believes that most long-distance distribution systems are able to

transition within 60 days of the date refiners begin shipping gasoline

meeting the new standard.

[[Page 37345]]

EPA is able to enforce the VOC-control standard at refineries based

on the refiners' batch reports to EPA that identify gasoline batches as

either VOC-controlled or non-VOC-controlled; the VOC-control standards

apply only to batches that are identified as VOC-controlled. However,

there is nothing in the refiners' batch reports to EPA that identifies

the per-gallon minimum and maximum standards to which the gasoline is

subject. As a result, EPA must rely on a date certain on which the new

standard applies at the refinery. Moreover, EPA believes this date must

be sufficiently earlier than the date the new standard applies at the

terminals in order to ensure the availability to terminals of gasoline

meeting the new standard for the terminals' transition. EPA also

believes that 60 days is an appropriate length of time for terminal

transitions, based on experience with VOC-control transitions.

B. Complex Model (Sec. 80.45)

1. Proper E300 Value for the Edge Target Fuel for Use in Complex Model

Extrapolation (Sec. 80.45(c)(1)(iv)(C)(6))

The Complex Model as described in Sec. 80.45 includes provisions

for extrapolations beyond the limits of the data upon which the model

was based. The limits of the data define the ``allowable range'' which

represents the range of fuel parameters within which the Complex Model

equations are directly applicable, and outside of which extrapolation

must be used up to the limits of the model.9 These

extrapolations take the form of intricate equations and a series of

conditions for use of those equations. Among other things, the

conditions associated with extrapolation direct Complex Model users to

determine properties for an ``edge target fuel.'' The edge target fuel

is equivalent in all respects to the target fuel, except that no fuel

parameters are allowed to exceed the limits of the allowable range. In

effect, the edge target fuel represents the point in the multi-

dimensional fuel parameter space where extrapolation begins.

---------------------------------------------------------------------------

\9\ The allowable range of the model is, in fact, a combination

of the limits of the data and additional limitations that may be

imposed by the existence of extreme, or curve turnover points.

---------------------------------------------------------------------------

The Complex Model equation for exhaust volatile organic compounds

(VOC) contained in Sec. 80.45(c)(1) includes a single interactive term.

This term, the product of E300 and aromatics, necessitates that

extrapolations involving E300 include a simultaneous evaluation of the

aromatics level of the target fuel. Thus in paragraph (c)(1)(iv)(C)(6),

Complex Model users are directed to determine whether the mathematical

phrase (80.32 + (-.390xARO)) is greater or less than 94, and to set the

E300 edge target fuel value accordingly. In so doing, users are

determining whether the aromatics-dependent E300 extrema (i.e. curve

turnover) point falls beyond the limits of the available data in the

Complex Model database.

However, the language in paragraph (c)(1)(iv)(C)(6) is misleading.

As currently written, the user is directed to set the E300 value of the

edge target fuel at 94 vol% whenever the value of the phrase (80.32 +

(0.390xARO)) is greater than 94. The Agency's intention, however, was

that this step be taken only if the E300 term is being extrapolated. In

other words, if the target fuel value for E300 falls below the higher

limit for E300 in the allowable range as defined in Table 6,

Sec. 80.45(c)(1)(iv), then E300 is not being extrapolated, and the E300

value of the edge target fuel should be equal to the E300 value of the

target fuel.

To correct this problem, the language in Sec. 80.45(c)(1)(iv)(C)(6)

and its counterpart applicable to Phase II calculations at Sec. 80.45

(c)(1)(iv)(D)(6) would be changed such that Complex Model users will

only set the E300 value of the edge target fuel equal to 94 if the

target fuel value for E300 exceeds the higher limit specified in Table

6, Sec. 80.45(c)(1)(iv).

IV. RFG Compliance Requirements

A. Sampling of Reformulated and Conventional Gasoline (Proposed

Sec. 80.47)

Under Sec. 80.65(e)(1) refiners and importers are required to

collect a representative sample from each RFG batch produced or

imported, and to determine the batch properties based upon analysis of

this sample.10 ``Batch of reformulated gasoline'' is

currently defined in Sec. 80.2(gg) as ``a quantity of reformulated

gasoline which is homogeneous with regard to those properties which are

specified for reformulated gasoline certification.'' Similarly

Sec. 80.101(i)(1)(i)(A) requires refiners and importers of conventional

gasoline to collect a representative sample from each batch produced or

imported, and to determine compliance with the anti-dumping standards

based upon the batch samples.11

---------------------------------------------------------------------------

\10\ Under Sec. 80.69(b) oxygenate blenders who meet the oxygen

standard on average also are required to sample and test each batch

of RFG produced using RBOB, and the discussion in this preamble

section applies to such oxygenate blenders in the same manner as for

refiners of RFG.

\11\ EPA is proposing several changes relative to the sampling

of conventional gasoline that are discussed below in section VI.B.

of this preamble. EPA is also proposing to revise the ``batch''

definition in Sec. 80.2(gg) to apply to conventional gasoline and

not just to RFG. EPA also is proposing to require refiners and

importers of conventional gasoline to separately test each batch,

which would eliminate the current option of testing a number of

batches together using a composite sample.

In addition, EPA is proposing a definition for ``previously

certified gasoline'' to mean RFG and conventional gasoline for which

the refiner, oxygenate blender or importer has met applicable

requirements and standards and that the refiner, oxygenate blender

or importer has included or intends to include in the refinery or

importer compliance calculations.

---------------------------------------------------------------------------

As a result, refiners and importers are required to collect a

representative sample of each gasoline batch. However, EPA has not

previously promulgated requirements for determining when a quantity of

gasoline is homogeneous so that it qualifies as a batch. Today EPA is

proposing such requirements for determining batch homogeneity. In

addition, EPA is proposing procedures whereby an importer of

reformulated or conventional gasoline would be able to treat as a

single batch the gasoline contained in multiple compartments of a ship.

It is important that refiners and importers determine compliance

with the reformulated and conventional gasoline standards using samples

collected from quantities of gasoline that are homogeneous in terms of

the properties relative to these standards. If a quantity of gasoline

is not homogeneous, a sample of that gasoline often will not reflect

the overall average qualities of the gasoline. For example, when a

refiner produces gasoline by combining blendstocks having different

volatilities, unless the tank is thoroughly mixed the gasoline often

will be horizontally stratified, with the higher volatile blendstocks

at the top of the tank and the lower volatile blendstocks at the bottom

of the tank. If a sample is collected of the gasoline at any one spot

in such a stratified tank the sample only will reflect the properties

of the gasoline at that strata. Storage tank sampling techniques such

as ``all level samples'' or ``running samples'' tend to compensate for

stratified product, but these techniques do not assure a truly

representative sample.12

---------------------------------------------------------------------------

\12\ A ``running sample'' of the product contained in a storage

tank is collected by lowering a sample container from the top of the

product to the bottom and then raising the container to the top, at

such a speed that the container is less than full when removed from

the tank. See, 40 CFR part 80, appendix D, para. 11.2.2.2. An ``all

levels sample'' is collected by lowering a stoppered container to

the bottom of the product in a storage tank, removing the stopper

with a cord or chain, and raising the container to the top at such a

speed that the container is less than full when removed from the

tank. See, 40 CFR part 80, Appendix D, para. 11.2.2.1. In theory,

both of these sampling methods obtain product from all strata in the

storage tank somewhat in proportion to the size of the strata.

---------------------------------------------------------------------------

[[Page 37346]]

In the case of RFG, moreover, certain standards must be met on a

per-gallon basis. If any portion of the RFG in a storage tank violates

an applicable per-gallon standard, this gasoline portion is out of

compliance even if the gasoline in the tank would be in compliance if

fully mixed. For example, consider a refinery storage tank containing

RFG designated as simple model, VOC controlled for Region 2. If the

gasoline is stratified by RVP, and the RVP of the upper strata is

greater than the applicable per-gallon maximum standard of 8.3 pounds

per square inch (psi),the gasoline in this upper strata would violate

the applicable per-gallon standard even if the average RVP of the

gasoline in the tank is less than 8.3 psi.13 A single sample

from such a stratified tank may not reflect the violation. Even an

``all levels'' or ``running'' sample of the gasoline in a stratified

storage tank could yield a test result within the standard because to a

certain extent such a sample ``averages'' across the strata, which

would have the effect of masking the violation.

---------------------------------------------------------------------------

\13\ Per-gallon standards must be met by all portions of the

gasoline contained in a storage tank in part because the different

gasoline portions may be distributed without further mixing.

---------------------------------------------------------------------------

As a result, EPA is proposing that refiners and importers would be

required to establish that each quantity of reformulated or

conventional gasoline that will be treated as a batch is homogeneous

before the batch sample is prepared or analyzed.

EPA is proposing two options by which the homogeneity of the

gasoline in a storage tank could be established. Under the first

option, a refiner would collect three separate samples from the storage

tank--upper, middle, and lower spot or tap samples. These samples would

be analyzed for each parameter relevant to applicable standards, and

the gasoline in the storage tank would be considered homogeneous if the

test results agree within the ranges specified in Sec. 80.65(e)(2)(i).

Under the second option for establishing storage tank homogeneity,

the party would demonstrate that it followed tank mixing procedures

that can be shown to result in homogeneity. For example, a refiner

could meet the homogeneity requirement through records that show the

tank mixing procedures used for a batch (tank size and type, volume of

gasoline, the type of tank mixers, the mode of mixer operation if

appropriate, and the duration of mixer operation), together with

historic sampling and testing records demonstrate these procedures

result in complete mixing.

Under this second storage tank option, success of the mixing

procedure must still be confirmed for each batch. However, instead of

requiring analysis for each parameter relevant to applicable standards,

only API gravity analysis of upper, middle, and lower spot or tap

samples would be required. The gasoline would be considered homogeneous

under this option if the demonstrated mixing procedure was performed,

and the API gravity values for the upper, middle, and lower samples do

not differ by more than 0.3 deg. API. Where the configuration of a

storage tank does not permit the collection of upper, middle, and lower

spot or tap samples, the API gravity analysis to confirm the success of

the mixing procedure would be waived.

EPA also is proposing procedures whereby an importer would be able

to demonstrate the gasoline in multiple compartments of a marine vessel

is homogeneous. The importer would collect a ``running'' sample from

each compartment and analyze the samples for each parameter relevant to

applicable standards. The vessel's gasoline would be homogeneous and

could be treated as a batch if the results agree within the ranges

specified in Sec. 80.65(e)(2)(i).

EPA is proposing that for purposes of establishing homogeneity a

party could use test methods other than the methods specified in

Sec. 80.46. The methods in Sec. 80.46 would still be used to establish

the batch properties for ``certifying'' a batch.

EPA also is proposing that in the case of RFG, the gasoline

contained in a storage tank or marine vessel would not be considered

homogeneous if any sample collected to establish homogeneity has a test

result that exceeds an applicable per-gallon standard. Thus, in the

case of standards a refiner or importer is meeting on a per-gallon

basis no test result could violate the per-gallon standard, and in the

case of standards being met on average no test result could violate an

applicable per-gallon minimum or maximum standard.

EPA is proposing additional options by which an importer could

treat the gasoline imported by marine vessel as a single batch without

determining the homogeneity of the gasoline. RFG contained in multiple

compartments of a marine vessel could be certified as a single batch

using a volume weighted composite of samples collected from the

compartments if the entire contents of these compartments is

transferred into a single shore tank. EPA is proposing this option

because it is likely the gasoline from multiple vessel compartments is

completely mixed, i.e., becomes homogeneous, through the process of

being transferred into a shore tank.

Under today's proposal importers also would be allowed to use

composite samples to certify as a single batch the RFG imported by

marine vessel where the gasoline is off-loaded into multiple shore

storage tanks. Under this option, however, the importer would be

required to demonstrate that the RFG off-loaded into each shore tank

separately meets all applicable per-gallon standards, without regard to

any gasoline contained in the storage tank prior off-loading the

imported gasoline (or, ``heel''). Thus, the importer would be required

to sample and test the tank heel prior to off-loading the imported

gasoline and the tank contents after the imported gasoline has been

added, and to mathematically calculate the properties of the imported

gasoline added to the tank.

EPA is proposing that imported conventional gasoline contained in

multiple compartments of a marine vessel could be tested using a volume

weighted composite of samples collected from the compartments with one

limitation. There are no per-gallon standards associated with

conventional gasoline (other than the complex model limit standards, as

is discussed in section III.A.1. of this Preamble), and, as a result,

there are no proposed requirements to separately test vessel

compartment or shore tanks. However, EPA is proposing that each

separate grade of conventional gasoline on a marine vessel (e.g.,

regular, premium) must be treated as a separate batch. EPA believes

that, in general, there is greater variability in the properties of

gasolines of different grades, than of gasolines of the same grade. The

proposed grade limitation on marine vessel compositing for conventional

gasoline would constitute some limit on the range of gasoline

properties that could be included in a single composite sample, which

EPA believes would improve the quality of composite samples. EPA

requests comment on this proposed limitation on the use of composite

samples of imported conventional gasoline.

C. General Requirements (Sec. 80.65)

1. Assignment of Batch Numbers (Sec. 80.65(d)(3))

Section 80.65(d)(3) requires refiners and importers to assign batch

numbers to batches of RFG, RBOB, conventional gasoline, and certain

blendstock that is included in the refiner's compliance

[[Page 37347]]

calculations. The batch numbers are used to identify batches in batch

reports submitted to EPA under Secs. 80.75(a) and 80.105(a).

EPA is proposing to revise Sec. 80.65(d)(3) to require oxygenate

blenders who meet the oxygen standard on average to assign batch

numbers to RFG batches. This would conform Sec. 80.65(d)(3) with the

current reporting requirement at Sec. 80.74(a), that oxygenate blenders

who meet the oxygen standard on average must submit batch reports to

EPA.

2. Clarifications of Requirement to Test RFG and RBOB

(Sec. 80.65(e)(1))

Section 80.65(e)(1) requires refiners and importers to determine

the properties of each batch of RFG that is produced or imported.

Gasoline that complies with the standards in Sec. 80.41 is deemed

certified (Sec. 80.40(a)), hence this process is commonly considered as

``certifying'' each batch. This determination is required for each

parameter relevant to the RFG standards. EPA is proposing two

clarifications of Sec. 80.65(e).

EPA is proposing to add language to Sec. 80.65(e) to clarify that

this section applies to RBOB as well as to RFG, and to add a cross

reference to the requirement in Sec. 80.69(a)(2) that the certified

properties of RBOB are the properties of the RBOB subsequent to

downstream blending with oxygenate, based on test results of a sample

of the RBOB hand blended in the laboratory with the appropriate

oxygenate type and amount. EPA believes the certification of RBOB

already is implicit in Sec. 80.65(e), and that refiners and importers

have been certifying and reporting the properties of RBOB based on the

analysis results of a hand blend, so that the proposed changes would

not change current practices.

EPA also is proposing to clarify that certification testing for RVP

is necessary only for RFG and RBOB that is designated as VOC

controlled, because RVP test results are relevant only to VOC

controlled gasoline. Under the simple model the RVP standard applies

only to VOC controlled product. RVP test results are an input to the

complex model only for VOC controlled gasoline; in the case of non-VOC

controlled gasoline the complex model uses an RVP value of 8.7 psi

regardless of the actual RVP value of the gasoline. This change to

Sec. 80.65(e) also would change the reporting requirement for RVP, to

apply only to VOC controlled RFG and RBOB, because the parameter

reporting requirement in Sec. 80.75(a)(2)(v)(B) cross references the

requirements in Sec. 80.65.

3. Weight Percent Range for Total Oxygen Content (Sec. 80.65(e)(2)(i))

Section 80.65(e)(2)(i) provides a table with ranges for fuel

properties to be used in comparing the refiner's or importer's test

results to the test results obtained from the independent laboratory.

The table at Sec. 80.65(e), however, currently does not include a range

for total oxygen content. The RFG regulations prescribe a standard for

weight percent oxygen, and refiners and importers of RFG are required

to determine and report the total weight percent of oxygen in each

batch of RFG for compliance purposes. It is appropriate, therefore, to

include a range for total oxygen content in the table at Sec. 80.65(e)

for purposes of comparing the refiner's or importer's test results with

the test results obtained from the independent laboratory. A range for

total weight percent oxygen content was unintentionally omitted in the

final rule. As a result, today's rule proposes to add to the table at

Sec. 80.65(e)(2)(i) a 0.10 wt% range for total oxygen content. This

range would be in addition to, and not instead of, the volume ranges

for oxygenates listed in Sec. 80.65(e)(2)(i).

The 0.10 wt % range for total oxygen was derived by multiplying the

values of the oxygenates in the table in Sec. 80.65(e)(2)(i) by the

weight % of the oxygen in the oxygenates and averaging them. EPA

acknowledges that this approach assumes that the density of these

oxygenates is similar to gasoline, but believes that any difference in

density would result in an insignificant increase in the 0.10 wt %

value. EPA continues to believe that this is an appropriate method of

determining an appropriate range for total oxygen content between the

refiner's laboratory and the independent laboratory.

4. Independent Laboratory Requirements (Sec. 80.65(f); Proposed

Secs. 80.72, 80.74(h), and 80.75(n))

Sections 80.65(e) and (f) contain the independent laboratory

requirements for RFG. Under Sec. 80.65(e)(1) each batch of RFG must be

analyzed, either by the refiner or importer, or by an independent lab.

Section 80.65(f) requires each refiner and importer of RFG to designate

an independent lab that must collect a sample from each batch of RFG.

The refiner/importer then has the option of having the independent lab

meet the analysis requirement for all RFG batches (the 100% analysis

option), or of having the independent lab analyze up to 10% of the

samples collected to be identified by EPA (the 10% analysis option).

The 100% analysis option is most often chosen by importers who do not

operate their own company laboratory.

EPA is proposing two categories of changes to the independent

laboratory requirements. The first category of changes would include in

the regulations the guidance EPA previously has issued regarding the

identification of samples for analysis by independent labs, and the

identification of samples the independent lab would send to EPA. The

second category would slightly narrow the criteria by which a

laboratory is considered independent. In addition, EPA is seeking

comment on whether companies that serve as independent laboratories

under the RFG program should be made directly responsible for properly

completing the functions of sample collection, analysis, record keeping

and reporting.

The first category of changes being proposed relate to the

identification of samples to be analyzed under the 10% analysis option,

and the identification of samples to be supplied to EPA under both the

10% and the 100% analysis options.

Sections 80.65(f)(1)(ii) (B) and (C) state that under the 10%

independent analysis option, EPA will identify which samples the

independent lab must analyze. However, the regulations do not specify

the mechanism by which EPA identifies these specific samples. EPA

subsequently provided this sample-identification guidance in

Reformulated Gasoline and Anti-Dumping Questions and Answers (October

3, 1994), titled ``Reformulated gasoline program protocol for use by

independent labs in selecting samples for analysis under the 10%

independent analysis option, and for identifying samples to ship to

EPA.'' This protocol has been in use since it was issued, and EPA has

received no adverse comments from regulated parties regarding this

protocol. Therefore, EPA proposes to incorporate this protocol in the

RFG regulations. See proposed Sec. 80.72.

EPA believes the protocol is an appropriate mechanism for

identifying samples for analysis by independent labs. The protocol

provides an automated system to randomly identify for analysis 10% of

the samples collected by an independent lab in a way that gives

regulated parties no influence over the sample choice.

In addition to identifying the independent laboratory samples to be

analyzed, the proposed protocol also identifies which samples must be

supplied to EPA, including the minimum sample quantity to supply. The

requirement to forward samples to EPA applies to both the 10% and the

[[Page 37348]]

100% analysis options, and, therefore, the proposed sample-shipment

protocol applies to both options. Further, the regulations would

instruct independent labs to send to EPA any sample that, when tested

by the independent lab, is found to violate a per-gallon standard that

applies to the refiner or importer.

The proposal also would specify the quantity of gasoline that

independent labs would be required to supply to EPA. The batch sampling

methodologies of appendix D, in section 12.2, call for sample

containers of one quart as a minimum. Assuming that a single sample is

collected in a one quart container and that the container is filled

only to the minimum 70% level (appendix D requires samples to be 70-85%

full), this would provide a total of approximately 660mL of gasoline.

EPA believes one-half of this quantity, or 330mL, is sufficient for a

laboratory to complete all the testing requirements of the RFG

regulations. Therefore, where an independent lab analyzes an RFG sample

that also must be supplied to EPA, at least half the original sample

volume, or 330mL, would be available for shipment to EPA. Under the

proposed regulation regarding sample quantity, where the independent

lab has not analyzed a sample the lab would be required to supply EPA

with a one quart sample 70-85% full. In the case of a sample that has

been analyzed by the independent laboratory the lab would be required

to supply EPA with a minimum sample volume of 330mL.

The proposed regulations state that samples supplied to EPA should

be sent to an address to be specified by EPA. This address would be the

following: United States Environmental Protection Agency, National

Vehicle and Fuel Emissions Laboratory (NVFEL), Fuels and Chemical

Analysis Branch 2565 Plymouth Road, Ann Arbor, MI 48105, (313) 668-

4200.

EPA is not proposing to include this address in the regulations in

order to facilitate an address change if it becomes necessary. If there

is an address change for samples shipped to EPA, regulated parties

would be notified through individual letters, a Federal Register

notice, or some other appropriate means.

The second change being proposed would revise one criteria used to

determine if a laboratory is ``independent.'' Section

80.65(f)(2)(iii)(A), and proposed Sec. 80.72(b)(2)(I), specify that in

order to be considered independent a laboratory may ``not be operated

by any refiner or importer * * *.'' EPA now believes this independence

requirement is too stringent, and should apply only in the case of

refiners and importers of RFG.

Laboratories used to satisfy the independent sampling and testing

requirements are required to be independent in order to increase the

credibility of the laboratories' test results, as discussed at 59 FR

7765 (February 16, 1994). The independent sampling and testing

requirement applies only to refiners and importers of RFG, however, and

as a result EPA believes refiners and importers who operate a

commercial laboratory, but who produce or import no RFG, should be

allowed to serve as independent laboratories under the RFG program. EPA

is proposing that this definition of ``independence'' would not apply

if any RFG is produced or imported within a common corporate structure.

Thus, if a parent corporation has a subsidiary corporation that is a

refiner or importer of RFG, no other subsidiary of that parent

corporation could be considered independent.

Finally, EPA is seeking comment on whether companies that serve as

independent laboratories should be made regulated parties under the RFG

program.

Section 80.65(f)(3) describes the sample collection and reporting

procedures, and requires that each refiner or importer shall ``cause

its designated independent laboratory'' to carry out these procedures.

Under these procedures the independent lab collects a representative

sample from the RFG batch, determines the batch volume and other

information about the batch, reports test results to EPA, and supplies

samples to EPA upon request. A refiner or importer whose independent

lab fails to properly carry out these procedures would have failed to

meet the independent lab requirements, which would constitute a

violation of the RFG requirements by the refiner or importer.

EPA requests comments on whether the regulations should be revised

to provide that a laboratory that undertakes to act as an independent

lab under the RFG program becomes responsible to properly carry out the

independent lab requirements, in order to allow better monitoring and

enforcement of the independent lab requirements. For example, currently

there is no requirement for independent labs to retain records, which

creates potential difficulties when EPA attempts to audit and inspect

independent labs.

Under this approach, where an independent lab failed to properly

carry out an independent lab procedure, the independent lab would be

liable for a violation of the RFG regulations. In addition, the refiner

or importer for whom the lab is performing the independent lab function

would have failed to meet the independent lab requirement which would

constitute a violation of the RFG regulations. Under this approach, the

independent lab would also be required to retain records and submit

reports to EPA.

The authority to regulate laboratories that serve as an independent

labs under the RFG program is based on Clean Air Act sections 114(a),

208(a), 211(c), and 211(k). Analysis of RFG by independent laboratories

is critical to enforcement of the RFG standards, for reasons that are

discussed at 59 FR 7765 (February 16, 1994). In order for independent

laboratory sampling and testing to serve a useful purpose, however, the

independent lab must properly perform the procedures. EPA believes

independent labs would be more likely to take the steps necessary to

ensure the required procedures are properly performed if there were

regulatory consequences that applied directly to the independent

laboratory, and not just indirectly through sanctions against the

refiner or importer.

The current regulations state that a lab that is debarred,

suspended, or proposed for debarment pursuant to the Governmentwide

Debarment and Suspension regulations cannot serve as an independent lab

under the RFG program. An independent lab that fails to properly carry

out the required procedures could be the subject of a suspension or

debarment action by EPA. EPA requests comment on whether the suspension

or debarment sanction is adequate to ensure that independent labs

properly perform required procedures, in the absence of regulatory

liability.

In addition, EPA requests comment on whether regulations should be

proposed that would require labs to be accredited in order to carry out

the RFG independent lab requirements. EPA has not previously proposed a

lab accreditation requirement because of the likelihood that refiners

and importers would use only labs the refiners and importers are

convinced are fully capable of properly performing the independent lab

requirements. However, EPA has received comments that an accreditation

requirement could result in greater certainty that labs have the

equipment, training, and internal procedures necessary to properly

carry out the independent lab requirements, that could assist refiners

and importers in selecting independent labs.

Therefore, EPA requests comments on whether lab accreditation would

be appropriate for the RFG program;

[[Page 37349]]

whether accreditation should be performed by EPA or by an independent

body, and which independent body or bodies should be considered; the

accreditation criteria that would be appropriate; the estimated costs

of an accreditation program; and any other considerations EPA should

include as part of a lab accreditation proposal.

5. Compliance Audits (Sec. 80.65(h) and Sec. 80.105(c))

EPA proposes to modify Secs. 80.65(h) and 80.105(c) to make clear

that the attest requirement applies separately to each refinery

operated by a refiner, or the gasoline imported by an importer. The

amended rules clarify EPA's intent that refiners and importers of RFG,

RBOB, and conventional gasoline, and oxygenate blenders who blend RBOB

and meet the oxygen standard on average, must perform a separate attest

engagement for each facility at which such gasoline or product is

produced. In the process of issuing the Final Rule, EPA considered and

rejected the suggestion that parties be able to aggregate multiple

facilities within one attest engagement. Such an aggregation would

adversely skew the effect of the random sampling protocol described in

Sec. 80.127 by increasing the population of batches subject to random

sampling, and by potentially spreading the samples drawn over several

facilities. The effect, therefore, would be to produce less than the

95% confidence level for each facility that the attest engagement is

designed to accomplish.

6. Calculations Involving Previously Certified Gasoline (Sec. 80.65(i);

Sec. 80.78(a); Sec. 80.101(e))

Under Secs. 80.65(i) and 80.101(e)(1) refiners are required to

exclude from a refinery's compliance calculations gasoline that was not

produced at that refinery and gasoline that was produced at that

refinery but was included as part of another batch, sometimes called

``previously certified gasoline,'' or ``PCG.'' These requirements are

included in order to prevent double counting of PCG. Section

80.101(g)(3) provides the procedure by which refiners are required to

calculate the properties of blendstock that are combined with PCG to

produce conventional gasoline. However, the procedure in

Sec. 80.101(g)(3) is appropriate only for the simple model anti-dumping

standards, and there is no procedure specified for excluding PCG from

RFG compliance calculations. As a result, EPA is proposing procedures

for excluding PCG from the complex model compliance calculations for

both RFG and conventional gasoline. In addition, the procedures EPA is

proposing would allow refiners to use conventional gasoline to produce

RFG or RBOB, and to reclassify RFG with regard to VOC control and OPRG.

The procedures EPA is proposing would require refiners to determine

the volume and properties of each batch of PCG used in the refinery

operation, along with the designations of the gasoline: RFG, RBOB or

conventional gasoline; and for RFG, the designations relative to VOC

control and OPRG. The volume and properties of each PCG batch would be

reported to EPA as a negative batch using the same designations as when

received by the refiner. The PCG then would be used by the refiner as

another blendstock in the refinery operation, and any gasoline produced

using the PCG would be sampled and tested and included in compliance

calculations without regard to the PCG content. Gasoline produced using

the PCG could have the same designations as the original PCG batch, or

different designations. Thus, the proposed procedures would allow a

refiner to reclassify conventional gasoline as RFG, or to reclassify

RFG with regard to VOC control and OPRG.

Under the current regulations refiners are prohibited from

reclassifying gasoline in certain ways. For example, Sec. 80.78(a)(10)

prohibits any person from reclassifying conventional gasoline as RFG.

However, EPA understands that prohibitions against reclassifying

gasoline, such as Sec. 80.78(a)(1), constrain the operational

flexibility for regulated parties, and that such prohibitions should be

imposed only where necessary. EPA believes the PCG proposal allows

greater flexibility without compromising the environmental goals or

effective enforcement of the RFG program, and the PCG proposal is

appropriate for this reason.

In the case of standards that are met on average a refiner who uses

PCG would meet each average standard based upon the net average

properties of gasoline in the relevant averaging pool, 14

consisting of the positive volume and properties of all gasoline

produced in that averaging pool and the negative volume and properties

of all PCG in that averaging pool. In addition, each averaging pool

would be required to have a net ``positive'' gasoline volume--each

averaging pool's volume of gasoline produced would have to be greater

than the volume of PCG.

---------------------------------------------------------------------------

\14\ Compliance with each average standard is based on the

average property or emissions performance of the subset of the

gasoline produced at a refinery that is relevant to that standard,

sometimes called an ``averaging pool.'' For example, the averaging

pool for anti-dumping standards is all conventional gasoline

produced during an averaging period. In addition, certain RFG

standards must be separately met by more than one averaging pool.

For example, under Sec. 80.67(g) the RFG NOx emissions performance

standard must be met by the averaging pool of all RFG and RBOB that

is VOC controlled, and separately by the averaging pool of all RFG

and RBOB that is not VOC controlled.

---------------------------------------------------------------------------

Consider, for example, Refiner A who has elected to meet the VOC

emissions performance standard on average at Refinery X. In this

example a batch of PCG, designated as RFG, VOC controlled for Region 1,

is used to produce RFG at Refinery X. This PCG would be included as a

negative batch in Refinery X's VOC emissions performance compliance

calculations for the ``VOC controlled for Region 1'' averaging pool,

regardless of whether the PCG was used to produce RFG with this or with

another designation. 15 Refiner A nevertheless would be

required to meet the VOC standard for the ``VOC controlled for Region

1'' averaging pool, and the net volume of gasoline in this averaging

pool would have to be greater than zero.

---------------------------------------------------------------------------

\15\ EPA is proposing that a ``negative'' batch would be

included in the ``Actual Total'' calculation in Sec. 80.67(g)(1)(ii)

by multiplying the ``Vi'' term (the batch volume) times

minus 1.

---------------------------------------------------------------------------

In a case where a refiner has elected to meet a parameter or

emissions performance standard on a per-gallon basis, and a batch of

RFG or RBOB is produced using previously certified RFG, the value of

the per-gallon standard the refiner would be required to meet for this

batch would be the more stringent of: (1) The per-gallon standard that

applies to the refinery under Sec. 80.41; or (2) the value for that

parameter or emissions performance for the previously certified RFG

used to produce the batch. If previously certified conventional

gasoline is used, however, use of this PCG would not affect the per-

gallon RFG standards.

Consider again the example of Refiner A, and in this example

Refiner A has elected to meet the benzene standard on a per-gallon

basis at Refinery X. Under Sec. 80.41(c), and in the absence of

applicable survey ratchets, the benzene per-gallon standard is 1.00

volume percent (vol%). Also, in this example the batch of previously

certified RFG has a benzene content of 0.85 vol%. In consequence, any

RFG produced at Refinery X using any amount of this PCG would be

subject to a benzene per gallon standard of the more stringent 0.85

vol%.

Any previously certified conventional gasoline used to produce RFG

or conventional gasoline would be included in the compliance

calculations

[[Page 37350]]

for the gasoline produced. In addition, the previously certified

conventional gasoline would be included, as a negative batch, in the

refinery's anti-dumping compliance calculations.

Finally, any previously certified RFG or conventional gasoline

would be included as a negative volume for purposes of calculating a

refinery's compliance baseline under Sec. 80.101(f).

The proposed approach is summarized in the following table.

----------------------------------------------------------------------------------------------------------------

Gasoline produced standards

Previously certified gasoline (PCG) Gasoline produced type -------------------------------------------------

type Per-gallon Average

----------------------------------------------------------------------------------------------------------------

RFG or RBOB.......................... RFG.................... More stringent of: Include PCG in

Sec. 80.41 compliance

per gallon standards; calculations as

or negative batch.

PCG All RFG pool

properties volumes for average

standards must be

positive.

Conventional Gasoline (CG)........... RFG or RBOB............ Sec. 80.42 per gallon Include PCG in

standards. CG compliance

calculations as

negative batch.

CG pool volume

must be positive.

CG \1\............................... CG..................... None................... (Same as above).

----------------------------------------------------------------------------------------------------------------

\1\ Includes RFG used to produce CG, because previously certified RFG may be ``downgraded'' to previously

certified CG.

EPA believes the approach proposed for addressing PCG is

appropriate because it would provide regulated parties with

significantly additional flexibility, with no apparent risk of adverse

environmental consequences. The additional flexibility would result

from the ability for regulated parties to more easily use previously

certified gasoline in refinery operations.

At the time the RFG regulations were promulgated EPA was concerned

that the overall quality of the various gasoline pools could be

degraded if refiners were able to reclassify conventional gasoline into

RFG, or to reclassify certain categories of RFG into other categories.

For example, if a refiner could reclassify conventional into RFG, it

would be possible for a refiner to produce very ``clean'' conventional

gasoline and include this gasoline in its anti-dumping compliance

calculations, and then reclassify this same gasoline into RFG with very

little or no additional blending. This would enable the refiner to meet

the anti-dumping standards using gasoline that, in reality, will be

used as RFG. One effect of this type of ``gaming'' would be to degrade

the quality of the conventional gasoline pool, with consequent adverse

environmental effects.

As a result of these concerns, EPA included provisions in

Sec. 80.78 that prohibit parties from combining certain categories of

gasoline. For example, Sec. 80.78(a)(10) prohibits parties from

combining RFG with conventional gasoline to produce RFG, in part in

order to address the ``gaming'' concern described above.

However, the proposed PCG accounting procedure would allow refiners

to reclassify conventional gasoline into RFG in a manner that avoids

the potential for adverse environmental effects from ``gaming.'' This

is true because reclassifications using PCG may occur only at a

refinery, and the PCG must be included, as a negative batch, in the

refinery's compliance calculations for the gasoline pool that

corresponds to the PCG's designations when first produced. Consider

again the example of ``gaming'' involving very ``clean'' conventional

gasoline, described above. Under the PCG proposal any of the very

``clean'' conventional gasoline used as PCG would have to be included

in the refinery's anti-dumping compliance calculations as a negative

batch, this pool would have to meet the anti-dumping standards, and the

pool volume would have to be positive. This would require the refiner,

in effect, to produce other conventional gasoline that is equal in

quantity and quality to very ``clean'' conventional gasoline used as

PCG, that would offset the loss of this gasoline to the conventional

gasoline pool. Thus, under the proposal there would be no net change in

the quality of the conventional gasoline pool.

This same logic would allow refiners to reclassify RFG with regard

to VOC control and OPRG.

In the case of RFG standards that are met on a per-gallon basis, a

different approach would be used to ensure no degradation in the

quality of the overall RFG pool as a result of the PCG proposal, since

averaging calculations are performed only where standards are met on

average. The approach proposed, as discussed above, would prohibit the

receiving refiner from degrading the quality of any previously

certified RFG batch with regard to any standard the receiving refiner

meets on a per-gallon basis, by setting the per-gallon standard at the

parameter value of the PCG if it is more stringent than the normal per-

gallon standard.

As a result, EPA is proposing to specifically allow refiners to

change the classifications of RFG and conventional gasoline under the

PCG procedures. In addition, EPA is proposing to revise the

prohibitions in Sec. 80.78 to reflect the PCG proposal. In proposed

revisions to Secs. 80.78(a) (5) and (7) parties would be allowed to

combine RFG or RBOB with blendstock under the terms of the PCG

proposal.

Under the proposed PCG procedures it would be important that any

gasoline claimed as PCG actually is used in a refinery's operation--

otherwise, these procedures could cause a degradation in gasoline

quality. For example, consider a refinery that received a batch of

relatively ``dirty'' conventional gasoline. If this gasoline is

classified as PCG, is used in the production and compliance

calculations of conventional gasoline, and is added to the anti-dumping

compliance calculations as a negative batch, there would be no net

effect of the ``dirty'' PCG on the refinery's overall anti-dumping

compliance calculations. If, however, the refiner never used the PCG as

a component for gasoline production, yet included the ``dirty'' PCG as

a negative batch in compliance calculations, the refinery's

conventional gasoline pool would appear ``cleaner'' than it was in

reality.

As a result, EPA is proposing record keeping and attest

requirements that would apply in the case of any refiner who uses the

PCG option, that would include records demonstrating the storage and

movement of the PCG from the time it is received at the refinery until

it is used in the production of gasoline. The proposed attest

procedures would require the auditor to verify that PCG was used to

produce gasoline at the refinery, and that the PCG batch report to EPA

is consistent with the refiner's sampling and testing of the PCG, and

the PCG product transfer documents, when received at the refinery.

[[Page 37351]]

7. Requirements for Imported Gasoline (Sec. 80.65(j))

Section 80.65(j) ``Requirements for imported gasoline,'' is

proposed as an addition to the general requirements of Sec. 80.65(e) to

qualify import certifications. This is in response to importer and

independent laboratory questions regarding certification of import

cargoes. The Agency has received questions regarding where and when

imported gasoline must be certified, and how to treat gasoline destined

for multiple ports. The Agency has issued policy guidance in response

to these questions in Reformulated Gasoline and Anti-Dumping Questions

and Answers. Today's regulatory revision is somewhat more restrictive

than the Reformulated Gasoline and Anti-Dumping Questions and Answers

policy guidance, in that batch certification would have to comply with

the U.S. Customs Service requirements for imported gasoline. The

original intent of the RFG regulation was to follow the normal import

industry practices as regulated by the U.S. Customs Service. Some

allowances were provided in the Reformulated Gasoline and Anti-Dumping

Questions and Answers guidance that may not conform with the U.S.

Customs Service regulations and today's proposal reverses any changes

that may have occurred.

The first requirement proposed in Sec. 80.65(j) is that batch

certification sampling be conducted at the time and place permitted

under U.S. Customs Service regulations, 19 CFR 151.42, and as specified

in the new Sec. 80.47 Sampling of reformulated and conventional

gasoline, which is discussed above. Section 80.47 provides specific

sampling procedures for reformulated and conventional gasoline, and

refers to Sec. 80.8 Sampling Methods for the general sampling

procedures that apply.

This requirement reflects the majority of guidance provided in

Reformulated Gasoline and Anti-Dumping Questions and Answers. For

instance, the guidance provides that when an import vessel off-loads

its cargo at more than one U.S. Customs Service port, then it must

certify the cargoes off-loaded in the separate ports as different

batches. The reason for this is that there is no mechanism for EPA to

enforce or even to find out about possible additions to a certified

batch when a vessel leaves the port where it was sampled. Today's

proposal also requires separate batch certifications for separate entry

ports as governed by the U.S. Customs Service regulations. However, in

Reformulated Gasoline and Anti-Dumping Questions and Answers, an

exception to this guidance is provided for multiple ports within a

given harbor area, such as the New York City harbor area, wherein a

single batch may be off-loaded at multiple Customs ports within the

harbor. Today's proposal will not include this exception because it

does not conform with U.S. Customs Service regulations. EPA relies on

U.S. Customs Service records for enforcement of the EPA fuels

regulations. By following the Customs Service regulations EPA maximizes

the usefulness of this enforcement tool. It also minimizes regulatory

confusion by conforming the EPA requirements with an existing

regulatory requirement of the U.S. Customs Service.

U.S. Customs Service regulations for imported petroleum products

allow for sampling once an import vessel is docked and ready to off-

load its cargo, although under 19 CFR 151.42, Controls on unlading and

gauging, each port director independently establishes the methods of

control. As such, the protocols for sampling an import vessel could

vary from port to port and could also depend on the type of import

vessel (for instance, ship, barge, rail car). EPA requests comments on

the requirement to follow the U.S. Customs Service procedures during

batch certification. EPA will retract any conflicting guidance that

remains in Reformulated Gasoline and Anti-Dumping Questions and Answers

after final revisions to this regulation are promulgated.

The second and final requirement of proposed Sec. 80.65(j) is that

batch size could be no larger than a ``line item,'' or a single item of

merchandise, of an entry summary under U.S. Customs Service

requirements specified at 19 CFR part 141, subparts D, E, and F, and

part 142, subparts A and B. These subparts of the Customs Service

regulations specify the documentation required for import cargoes. This

documentation must differentiate merchandise by listing or invoicing

items subject to different duty rates (19 CFR 141.61(e)), and it must

list or invoice items of varying commercial value separately (19 CFR

141.86). Therefore, it is EPA's understanding that the Customs Service

regulations require quantities of gasoline imported on a single vessel

to be distinguished on the basis of their differences in commercial

values or potential for differences in commercial value. For instance,

different grades (segregated in different tanks) would be entered as

separate line items. Also, gasoline from different sources but of the

same grade, would normally be entered as separate line items due to

their potential for the separate sources not meeting the agreed upon

commercial specifications. Limiting batch size to U.S. Customs Service

entry ``items'' serves two functions: (1) It adjusts the EPA

requirements to fit better with the existing regulatory standards of

the U.S. Customs Service, and (2) it puts a limit on the variations of

RFG property values within a batch (that could lead to inaccurate

sample representation as discussed above in the preamble to Sec. 80.47,

regarding homogeneity determination).

D. Compliance on Average (Sec. 80.67)

1. Transfer of Oxygen and Benzene Credits (Sec. 80.67(h)(1)(iv))

Section 80.67(h)(1)(iv) permits the transfer of credits directly

from the refiner, importer, or blender who generates them to the

refiner, importer, or blender who uses the credits for compliance

purposes. EPA has received several inquiries with regard to whether

transfers within the same company are included in the language of this

section. It is the Agency's intention that the refiner, importer, or

blender may properly transfer legitimate credits within the company or

outside of the company. As a result, EPA is proposing to modify

Sec. 80.67(h)(1)(iv) to clarify that credit transfers may be either

intercompany or intracompany.

E. Compliance Survey Requirements (Sec. 80.68)

1. Method of Computation for Averages in Survey Series (Secs. 80.68

(c)(9)(I)(B) and (ii)(B), (c)(10), (c)(11), (c)(12) and (c)(13))

The RFG surveys were designed to deter and detect situations where

the flexibility afforded refiners through averaging gasoline

characteristics at the refinery gate (as opposed to averaging each

refinery's contribution to the gasoline in a particular covered area)

results in a covered area obtaining gasoline that on average differs in

relevant qualities from the average gasoline quality that would occur

if averaging was required separately for each covered area. The surveys

are conducted by an industry association according to a statistical

sampling plan approved by EPA and involve sampling gasoline from retail

outlets. If the gasoline in an area fails to meet standards set forth

in the regulations for a particular parameter, the standards for that

parameter are made more stringent and the number of surveys that must

be conducted in the following year is increased.

[[Page 37352]]

Some of the gasoline characteristics evaluated by the survey are

chiefly of interest because of their role in causing or contributing to

ambient ozone levels. Surveys for these parameters (e.g., VOC surveys)

are passed or failed based upon the average of results from a week-long

survey. Other parameters (like benzene and toxics) are of concern

because of their cumulative effects over a longer period of time.

Surveys for these latter characteristics are passed or failed based

upon the average of a year-long series of one-week surveys. This

discussion is primarily concerned with how the average of such a series

of one-week surveys should be computed.

Under the current regulations, determining the average for each

survey series 16 involves computation of a simple average

17 of parameter values from each gasoline sample across all

of the samples gathered during the year (without any consideration of

which week-long survey the sample was a part). If all of the individual

week-long surveys had equal sample sizes, this approach to computation

would yield as good a representation of the fuel supply as the timing

and distribution of the week-long surveys throughout the year

permitted.18 Practical considerations involved in the design

and conduct of an efficient overall survey operation, though, dictate

some substantial variations in sample size among the week-long surveys.

One such effect, and probably the most important one, stems from the

fact that high-ozone season surveys for ozone precursors must yield a

confidence interval on the mean small enough to meet the precision

requirements of the regulations (Sec. 80.68(c)(13)(iii)) for each

individual survey. Since practical considerations dictate that surveys

for the various parameters be conducted concurrently (i.e., each

gasoline sample is analyzed for all parameters covered by the survey

program), this situation results in larger-than-necessary sample sizes

in the summer for non-ozone precursor parameters. Outside the summer

ozone season there is no need to maintain precision standards for each

individual survey, but only for the annual series of such surveys. In

the interest of efficiency, the survey manager may be expected to cut

back on sample sizes during these times at the beginning and end of the

calendar year. As a result, the simple average substantially

overrepresents summertime gasoline.

---------------------------------------------------------------------------

\16\ Sec. 80.68(c)(9)(i)(B) for toxics; (10)(ii) for

NOX; (11) for benzene; and (12) for oxygen.

\17\ In the case of toxics, the computation introduces weights

for the season (high-ozone season or outside of high-ozone season)

since the statistical model used to compute the emissions is

different in the two seasons. The weights substantially correct the

overemphasis on summer that affects other non-ozone-related

parameters, as discussed in the remainder of the text.

\18\ While the design for each of the individual week-long

surveys is probabilistic, a variety of considerations prevent EPA

from distributing the surveys in a perfectly random manner with

respect to time. The overall sampling approach for survey series

thus departs, to some extent, from a purely probabilistic design.

---------------------------------------------------------------------------

An additional reason for altering the prescribed approach to

computing averages of series has to do with the weights attached to

each sample to handle either lack of pre-survey information about an

individual retail outlet's throughput or the situation where an outlet

with unusually high throughput is located in a covered area with

relatively few outlets and is consequently selected into the sample

with certainty. For both situations the sample is not self-weighting

and weights must be computed to properly represent the outlet's

gallonage in the sample. The current approach, the simple average,

requires that such weights be computed two different ways, once for the

outlet's inclusion in the week-long survey for ozone-related parameters

and then again for the annual average computation for non-ozone-related

parameters. The latter set of weights cannot be computed until the

year's data collection is complete, leaving some uncertainty up to the

end of the year as to the status of survey results in areas where

throughput data are not available for most outlets. This particular

problem is a characteristic of the sample design approach currently

being used by the industry survey organization, but that approach or

some variant of it is likely to be used in any thorough attempt to meet

the survey requirements in the regulations.

Both the distortion and the difficulty in computing weights, as

discussed above, can be eliminated by changing the method by which the

average of each survey series is computed for a given parameter in a

given RFG covered area. Instead of averaging all of the measurements on

individual gasoline samples in the survey series, we are proposing the

following: (1) That the measurements for each week-long survey in an

area be averaged, regardless of the sample size, to create a set of

means of week-long surveys, and then (2) that all of the resulting

individual survey averages for the area be averaged, themselves, across

all of the surveys in the series. This approach removes a significant

source of distortion, simplifies calculations, and improves the

representativeness of the number that we use to make the important

decision on whether the gasoline in an area has passed or failed a

survey series.

2. Clarification of Applicability of Survey Precision Requirements

(Sec. 80.68(c)(13)(iii))

The intent of the survey precision requirements set forth in the

regulations (Sec. 80.68(c)(13)(iii)) was to ensure that errors (in

either direction) in survey or survey series pass/fail determinations

would be unlikely. Without these requirements survey managers would be

able to trade off risk of inappropriate survey failure against survey

costs, and the environment would not be protected against the increased

risk of errors in the other direction resulting from insufficient

sampling.

Thus the precision requirements should apply to the body of data

that serves as the basis of each pass/fail determination. As currently

written, the regulations attach the precision requirements exclusively

to individual surveys without making it clear that for certain survey

parameters (for example, oxygen under the simple model) the pass/fail

determination is made against a year-long series of surveys rather than

against a single survey. The regulations would therefore be altered to

attach the precision requirements to the appropriate body of data for

each determination--to the individual survey where the parameters being

evaluated are ozone-related and to the survey series for other

parameters.

F. Downstream Oxygen Blending (Sec. 80.69)

1. Refiner ``Hand-Blending'' of RBOB (Secs. 80.69 (a)(2), (a)(8) and

(a)(9))

Under Sec. 80.65(c)(1) refiners and importers are required to meet

all RFG standards for RBOB, except for the oxygen standard. Under

Sec. 80.65(c)(2) the oxygen standard for RBOB is met by the oxygenate

blender. Section 80.69(a)(2) requires refiners and importers to

determine the non-oxygen properties of RBOB by blending the appropriate

type and amount of oxygenate with a sample of the RBOB (sometimes

called a ``hand blend''), and testing the properties of the resulting

RFG. Under Sec. 80.69(a) an RBOB refiner or importer is allowed to hand

blend the amount of oxygenate actually used by the oxygenate blender

only if, inter alia, a quality assurance program is carried out over

the oxygenate blending operation. In the absence of such a quality

assurance program, under Sec. 80.69(a)(8) specified types and amounts

of oxygenate must be hand blended.

EPA is proposing to revise Sec. 80.69(a)(2) to provide additional

guidance regarding the type and amount of oxygenate that must be hand

blended, and to move the hand blending

[[Page 37353]]

instructions from Sec. 80.69(a)(8) to Sec. 80.69(a)(2) in order to

improve the organization of this section. The additional guidance would

apply in the case of ``refiner specified'' RBOB (i.e., neither ``any

oxygenate'' nor ``ether only'') for which the refiner or importer has

specified options for more than one oxygenate type, or for a range of

oxygenate volumes. EPA is proposing that the hand blend for such RFG

must be formulated with the most conservative options. For example,

where an RBOB specification allowed ethanol and other oxygenates, the

hand blend would have to be formulated using ethanol, because ethanol,

as compared with other oxygenates at the same weight percent oxygen,

generally results in RFG with worse emissions performance.

Section 80.69(a)(9) specifies that where RBOB is designated as

``refiner specified'' but the quality assurance program is not

completed, the hand blend must be formulated with 4.0 vol% ethanol. EPA

is proposing to merge this paragraph with Sec. 80.69(a)(2).

2. Deletion of Secs. 80.69 (a)(4) and (a)(10)

Section 80.69(a)(4) requires refiners of RBOB to determine

properties of the RBOB, which would allow downstream parties to

determine if any contamination had occurred and thereby ensure that the

RFG produced using the RBOB would meet applicable standards. Section

80.69(a)(4) was included in the final reformulated gasoline regulation

to facilitate quality assurance programs by downstream parties who

handle RBOB, particularly where RBOB from a specific refinery travels

as a segregated product. However, EPA believes that, in practice, most

RBOB is being transported in a fungible manner. As a result, there is

little value to Sec. 80.69(a)(4) and EPA is proposing to delete this

requirement. EPA believes that downstream parties may conduct fully

adequate quality assurance programs over RBOB by hand blending the

oxygenate type and amount specified for the RBOB and testing the hand

blended sample to determine compliance with applicable standards.

Section 80.69(a)(10) requires refiners and importers of RBOB to

include in the RBOB blending specifications a range of oxygenate types

and amounts for all RBOB. This requirement was included in the RFG rule

because at the time the RFG regulations were promulgated it was not

clear the types of RBOB regulated parties would choose to produce. As a

result, the regulations were structured to accommodate a wide variety

of RBOB types. In practice, however, refiners and importers of RBOB

have chosen to produce only a limited slate of RBOB types, and mainly

only two types: generic ``ether-only'' RBOB, and RBOB with blending

instructions that are specific to the refiner and that is shipped in a

segregated manner. As a result, EPA now believes that Sec. 80.69(a)(10)

creates a burden on refiners and importers of RBOB, yet provides little

or no benefit to oxygenate blenders or to the environment. Accordingly,

EPA is proposing to eliminate this requirement.

3. Refiner Evaluation of RFG Produced by Oxygenate Blender

(Sec. 80.69(a)(7))

In the case of a refiner of RBOB conducting oversight over the RFG

produced at a downstream oxygen blending facility, the refiner of the

RBOB is required to calculate the non-oxygen parameter values for the

RFG produced using the RBOB. To do so, the refiner may use either the

oxygen blending assumptions under Sec. 80.69(a)(2) or the actual oxygen

blending levels if the refiner meets the contractual and quality

assurance testing requirements specified in Sec. 80.69(a)(5) through

(7).

The quality assurance provisions of Sec. 80.69(a)(7) require the

refiner to use sampling and testing to ensure that the RFG produced by

the downstream oxygen blender meets ``applicable standards.'' The

applicable standards are not further specified in that paragraph.

EPA is proposing to amend Sec. 80.69(a)(7) to require the refiner

to evaluate the RFG produced by an oxygenate blender for the oxygenate

type and oxygen amount, but not for other RFG standards. The principal

purpose of the Sec. 80.69(a)(7) oversight program is to ensure that the

oxygenate blender uses the proper type and amount of oxygen, to support

the refiner's RBOB compliance calculations. Other sections of the

regulations address quality assurance sampling and testing for all

standards that apply at all downstream locations, including at

oxygenate blending facilities. See, for example, Sec. 80.79(c), which

requires quality assurance sampling and testing as an affirmative

defense for violations of downstream standards. As a result, EPA

believes it is most appropriate to require sampling and testing only

for oxygenate type and amount under Sec. 80.69(a)(7).

4. Oxygenate Blending Instructions (Sec. 80.69(b)(1))

Under Sec. 80.69(b)(1) oxygenate blenders are required to blend

with RBOB only the type and amount of oxygenate that is specified for

the RBOB. EPA is proposing to amend this section to provide additional

guidance to oxygenate blenders regarding this blending. In addition,

EPA is proposing regulations that would specify the allowed quantity of

de minimis, extraneous, oxygenates that may be present in an oxygenate

blending operation.

EPA is proposing oxygenate blending requirements under

Sec. 80.69(b)(1) that are in accord with the RBOB blending

instructions. In addition, EPA is proposing language that would clarify

that the minimum oxygenate volume that could be used is the minimum

volume specification for the RBOB, and that the oxygenate blender is

free to add additional oxygenate up to the maximum oxygen standard

under Sec. 80.41(g).

EPA understands that when RBOB is blended with oxygenate to produce

RFG at oxygenate blending facilities, the RFG may contain de minimis

amounts of oxygenate other than the principal oxygenate that is

blended. These oxygenates may result, for example, when RBOB is shipped

via pipeline adjacent to RFG (that necessarily would contain

oxygenate), and these products are imperfectly segregated. Also, when

an oxygenate is produced it is normal that de minimis amounts of other

oxygenates also are produced and that remain present in the principal

oxygenate.

EPA believes that de minimis quantities of oxygenate that are in

addition to the principal oxygenate used to produce RFG do not degrade

the quality of RFG beyond a trivial amount. As a result, EPA is

proposing regulations that would specifically allow de minimis amounts

of incidental oxygenate, and that would specify the oxygenate amounts

that would be considered de minimis. However, EPA is also proposing

that these incidental oxygenates could not have been intentionally

blended, because the purpose of this proposed provision is to address

inadvertent oxygenate anomalies and not to provide additional oxygenate

blending options.

5. Every-Batch Sampling and Testing Requirement for Splash Blenders

(Proposed Sec. 80.69(b)(5))

Under Sec. 80.69(b)(4), an oxygenate blender who meets the oxygen

standard on average is required to sample and test each batch of RFG

produced to determine the batch's oxygen content, and assign a number

to the batch for reporting purposes. This every-batch sampling and

testing requirement was intended to be applied regardless of

[[Page 37354]]

whether the oxygenate blending is carried out in a large terminal tank

or through blending in trucks (sometimes called splash blending).

Every-batch sampling and testing is required in order to give the

oxygenate blender the best information with which to calculate the

average oxygen content of RFG produced. EPA believes that oxygenate

blenders, like other parties who produce RFG, should use adequate

procedures to determine, with great certainty, the oxygen content of

RBOB produced. This is particularly true of parties who meet the oxygen

standard (or other standards) on average, because, in part, any errors

in calculating average oxygen content could result in the transfer to

other parties of invalid oxygen credits. Every-batch sampling and

testing provides this certainty.

However, EPA believes that this every-batch sampling requirement

adds significant difficulties in the case of oxygenate splash blenders.

As a result, EPA is proposing to add Sec. 80.69(b)(5) which would allow

oxygenate splash blenders to meet the oxygen standard on average

without conducting every-batch sampling and testing under certain

conditions. These conditions, which are described below, would require

the oxygenate blender to use procedures that give certainty about

oxygen use, and that, taken together, EPA believes are as effective as

every-batch sampling.

a. Computer-controlled oxygenate blending required. Under the

proposal, the oxygenate blending would have to be carried out using

computer-controlled in-line or sequential blending that operates in

such a manner that the volumes of oxygenate and RBOB are automatically

dispensed when a particular grade of gasoline is selected for loading

into a truck, and where no operator instructions are required regarding

the oxygenate-RBOB proportions when an individual truck is loaded.

Thus, this alternative averaging approach would not be available where

the oxygenate and RBOB are separately metered into a truck, regardless

of whether the separate metering occurs at the same terminal or at

different terminals.

b. Oxygenate blender must operate blending equipment. The oxygenate

blender would be required to be the party who operates the computer-

controlled in-line or sequential blending equipment. Thus, this

alternative averaging approach would not be available to a party who

receives delivery of splash blended RFG into trucks at a terminal if

the terminal is not operated by that party, regardless of whether the

receiving party is a registered oxygenate blender.

c. Compliance calculations. The oxygenate blender would be required

to base its compliance calculations on the volumes and properties of

RBOB and oxygenate used during a period no longer than one calendar

month. In calculating the oxygen content of the RFG produced, the

oxygenate blender would be required to use either assumptions regarding

the specific gravities of the oxygenate and RBOB blended, or the

oxygenate blender would be required to measure the measured specific

gravities of all oxygenate and RBOB blended in the blending operation.

Similarly, with regard to the denaturant content of the ethanol (if

used), an oxygenate blender would be required to use a denaturant

content of 5 vol% and to support this value with documents from the

ethanol supplier and a quality assurance program, or the oxygenate

blender would be required to determine the denaturant content of all

ethanol used through sampling and testing.

d. Quality assurance sampling and testing. An oxygenate blender who

meets the oxygen standard on average using these procedures would be

required to conduct a program of quality assurance sampling and testing

of the RFG produced, using the procedures and at the frequencies

specified under Sec. 80.69(e)(2).

e. Attest procedures (Secs. 80.129 and 80.134). Under Sec. 80.65(h)

any oxygenate blender who meets the oxygen standard on average is

required to commission an annual attest engagement, to be conducted

under the terms of subpart F. EPA is proposing to add attest procedures

that would apply in the case of an oxygenate splash blender who meets

the oxygen standard on average under the proposed procedures. In

addition, EPA is proposing record keeping requirements that would apply

to such an oxygenate blender. The records which would be kept are those

EPA believes are necessary to an EPA auditor, or an independent

auditor, to ensure the proposed procedures were properly completed.

G. References to Renewable Oxygenate Requirements (Sec. 80.83)

On August 2, 1994, EPA promulgated regulations that would have

required the use of ``renewable'' oxygenates to meet a portion of the

oxygenate standard for RFG. See, 59 FR 39290 (August 2, 1994). However,

implementation of the renewable oxygenate requirements was stayed

effective September 13, 1994, as a result of a legal challenge filed in

the United States Court of Appeals for the DC Circuit. See, 59 FR 60715

(November 28, 1994). The Court of Appeals ultimately held that the

renewable oxygenate requirements for RFG are invalid, as they are not

authorized under sections 211 (c) or (k) of the Clean Air Act. American

Petroleum Institute v. EPA, 52 F.3rd 1113 (D.C. Cir. 1995).

This proposal would remove the regulatory language covered by that

decision.

The proposed changes relating to renewable oxygenates are shown in

the following table.

Sec. 80.2(ss)............... Current paragraph is deleted because it

applies only to renewable oxygenate

requirements. A new paragraph (ss) is

proposed which would define ``tank

truck.''

Sec. 80.2(tt)............... Paragraph is deleted because it applies

only to renewable oxygenate

requirements.

Sec. 80.2(uu)............... Paragraph is deleted because it applies

only to renewable oxygenate

requirements.

Sec. 80.65(d)(2)(vi) (C) Paragraphs are deleted because they apply

through (E). only to renewable oxygenate

requirements.

Sec. 80.83.................. Current section is deleted because it

applies only to renewable oxygenate

requirements. A new section 80.83 is

proposed which would provide procedures

for handling gasoline treated as

blendstock.

Sec. 80.128(e)(2)........... Paragraph is revised to delete language

that applies only to renewable oxygenate

requirements.

Sec. 80.128(e)(6)........... Paragraph is deleted because it applies

only to renewable oxygenate

requirements.

Sec. 80.129(a).............. Paragraph is revised to delete language

that applies only to renewable oxygenate

requirements.

Sec. 80.129(d)(3)(iii)...... Paragraph is deleted because it applies

only to renewable oxygenate

requirements.

[[Page 37355]]

In certain cases, the deleted text is replaced by regulatory

language discussed elsewhere in this proposal.

H. Covered Areas (Sec. 80.70)

Under Clean Air Act Sec. 211(k)(10)(D), any ozone nonattainment

area that is reclassified as Severe becomes an RFG covered area. This

inclusion in the RFG program occurs one year following the date of

reclassification.

Effective June 1, 1995, the Sacramento, California, ozone

nonattainment area was reclassified from Serious to Severe (60 FR 20237

(April 25, 1995)). Sacramento, therefore, became a covered area as of

June 1, 1996. Today's proposal would update the list of RFG covered

areas in Sec. 80.70 to include Sacramento.

I. Record Keeping Requirements (Sec. 80.74)

1. Clarification of test results record keeping (Secs. 80.74(a) and

80.104(a))

Sections 80.74(a)(2)(iii) and 80.104(a)(2)(i) require regulated

parties to keep the results of tests conducted of reformulated and

conventional gasoline. Parties have asked EPA to clarify this

requirement, and in particular have asked whether these regulations

require parties to keep copies of all documents that contain test

results.

In order to clarify these requirements, EPA is proposing changes to

Secs. 80.74(a)(2)(iii) and 80.104(a)(2)(i), that would specify that

parties are required to keep the original result for each test

performed. Thus, for example, where a test is performed using a testing

apparatus that automatically generates a printed document containing

the test result, this printed document must be kept. Where a test is

performed using an apparatus that does not generate a print out EPA is

proposing that the party would be required to keep the first recorded

test result, such as the chemist's laboratory log book.

In addition, EPA is proposing that parties would be required to

keep any other record that contains a test result that is not identical

to the original result. A non-identical test result could occur where a

party determines that an original test result is in error because of

laboratory error, for example. In such a case, the party would be

required to keep both the original test result and the corrected test

result. This proposed requirement would allow EPA, during the course of

an audit or inspection, to review changes that are made to test

results, to determine if the changes are appropriate.

2. Records To Be Kept by Refiners and Importers (Proposed

Sec. 80.74(b)(7))

EPA is proposing to add Sec. 80.74(b)(7) which would require

retention of records that reflect the physical movement of gasoline

treated as blendstock (GTAB) from the point of importation to the point

of blending to produce reformulated gasoline. (See Preamble Section

V.C. concerning the proposed requirements for GTAB.)

3. Records To Be Kept by Independent Laboratories (Proposed

Sec. 80.74(h))

EPA is proposing to add Sec. 80.74(h) which would require

laboratories serving as independent laboratories under proposed

Sec. 80.72 to retain records as required under Secs. 80.74(a)(2) and

80.72(c)(1).

J. Product Transfer Documentation (Secs. 80.77 and 80.106)

Product transfer documentation (PTD) requirements at Secs. 80.77

and 80.106 are intended to insure that on each occasion that any person

transfers custody or title of any RFG, RBOB or conventional gasoline,

other than when gasoline is sold or dispensed for use in motor vehicles

at a retail outlet or wholesale-purchaser-consumer facility, the

transferor produce, and provide to the transferee, documents that

contain certain information. This information would enable the

transferee to know enough about the gasoline being received to meet the

requirements of the RFG program. In addition, the PTD documents, which

parties are required to keep under Secs. 80.74(a)(1) and

80.104(a)(2)(vi) and (vii), help EPA identify the source of any

gasoline found to violate applicable standards.

EPA today is proposing to amend Secs. 80.77 and 80.106 to clarify

the following PTD requirements.

1. Introductory Text of Secs. 80.77 and 80.106

Section 80.77 requires a transferor to provide PTDs to the

transferee on each occasion involving a transfer of custody or title of

RFG or RBOB. Section 80.77 does not distinguish between transfers of

custody and transfers of title concerning the timing necessary for

transfer of PTD information. EPA, however, believes the two situations

may differ in this regard. In the case of transfers of custody, the PTD

information should be transferred before, during, or immediately

following the actual transfer because the transferee will have custody

of the gasoline in question and must know how to handle it. However,

since transfers of title do not always involve the physical handling of

the gasoline, EPA believes a transferee should have the option to rely

on the custody transferee to properly handle the gasoline (e.g., where

the custody transferee is a common carrier pipeline.) Therefore, in the

case of title transfers, EPA believes there is little need for the

required PTD information to be transferred at the time of the transfer

of the product. Accordingly, EPA is proposing, in the case of title

transfers, to allow up to thirty days in which to transfer the required

information. EPA believes this timing would allow parties to transfer

the required information using documents that are transferred as a part

of normal business dealings, and as a result would ease the burden of

meeting the PTD requirements.

The introductory text of Sec. 80.77 excludes from the PTD

requirements gasoline sold or dispensed for use in motor vehicles at a

retail outlet or wholesale purchaser-consumer facility. Section 80.106

does not contain this exclusion, which EPA believes was an inadvertent

omission when the final rule was promulgated. Accordingly, EPA is

proposing to revise Sec. 80.106 to conform to Sec. 80.77 in this

regard. EPA is also proposing to modify the introductory text of

Sec. 80.77 to clarify that this exclusion applies to gasoline sold or

dispensed at a retail outlet or wholesale purchaser-consumer facility

for use by any ultimate consumer, and not only for use in motor

vehicles.

In addition, EPA now believes that the PTD information is of little

value when conventional gasoline is delivered to a retailer or

wholesale purchaser-consumer in a conventional gasoline area.

Accordingly, EPA is proposing to exclude from the PTD requirements

transfers of conventional gasoline to retailers and wholesale-purchaser

consumers in conventional gasoline areas. Note, however, that the PTD

requirements of Sec. 80.106 would continue to apply for all other

transfers of conventional gasoline. Note also that the PTD requirements

of Sec. 80.77 for RFG and RBOB would continue to apply to all transfers

of RFG and RBOB (other than when the gasoline is sold or dispensed by a

retail outlet or wholesale purchaser-consumer facility for use by

ultimate consumers), including transfers in which RFG is delivered to a

retail outlet or wholesale purchaser-consumer.

2. Identification of the Gasoline (Sec. 80.77(f) and Sec. 80.77(g)(3)).

EPA is proposing to amend Sec. 80.77(f) to delete reference to

conventional gasoline, since the requirements of Sec. 80.77 do not

apply to conventional gasoline. EPA is proposing to amend

[[Page 37356]]

Sec. 80.77(g)(3) to delete reference to RBOB. This section requires

parties to identify whether the product contains ethanol, and RBOB, by

definition, does not contain oxygenate. In addition, EPA is proposing

to add references to RBOB to Secs. 80.77 (c) and (f) to specify that

these PTD requirements apply to RBOB as well as to RFG and conventional

gasoline.

3. Elimination of PTD Requirement for Inclusion of Registration Numbers

(Sec. 80.77(j) and Sec. 80.106(a)(1)(vi))

Sections 80.77(j) and 80.106(a)(1)(vi) require, in the case of

transferors and transferees who are refiners, importers or oxygenate

blenders, that the EPA assigned registration number of those persons be

included on the PTDs. EPA received comments that this requirement is

overly burdensome in certain circumstances, particularly downstream of

the refiner/importer/oxygenate blender where such information may not

be readily available. Based on experience with the program, EPA

believes that this requirement has only limited value as a means of

identifying and tracking the gasoline, and that EPA will be able to

adequately enforce the regulations without inclusion of the assigned

registration number on the transfer documents. As a result, EPA is

proposing to eliminate the requirements to include registration numbers

in PTDs.

4. Use of Product Codes (proposed Sec. 80.77(j))

The petroleum industry historically has used product codes to

identify product type in business transactions involving the transfer

of title or custody of petroleum products. For example, all pipelines

that transport refined petroleum products use codes to identify the

various types of petroleum products that are transported. These product

codes are well-known to persons who operate a pipeline, or who supply

products to or receive products from a pipeline. These pipeline codes

are used as a shorthand for the myriad petroleum products moving

through the distribution system, and make product identification

easier.19 In addition, product codes are used to identify

petroleum products in many of the documents used to memorialize

transfers of title and custody in normal business dealings, in part

because the codes occupy less space on the documents than the full

product names would require.

---------------------------------------------------------------------------

\19\ For example, Colonial Pipeline product code A1 means:

gasoline; RFG; VOC-controlled for Region 1; non-OPRG; simple model;

87 octane; benzene maximum of 1.18 vol%; oxygen minimum of 1.5 wt%

and maximum of 2.7 wt%; RVP maximum of 7.4 psi; and no heavy metals.

---------------------------------------------------------------------------

EPA is proposing to add Sec. 80.77(j) to allow the use of product

codes for certain information required on PTDs to accommodate this

practice, but under conditions that would ensure that the codes would

satisfy the goals of the PTD requirement. In particular, EPA is

proposing that product codes could be used to satisfy PTD requirements

related to identifying the product type (i.e., RFG, RBOB or

conventional gasoline); for RFG and RBOB, the designations and minimum

and maximum standards; and for RBOB, the oxygenate blending

specifications. Product codes, used to meet these PTD requirements

would have to fully reflect the PTD requirements. Thus, a product code

that referred to ``VOC controlled RFG,'' without more, would not meet

the requirement in Sec. 80.77(g)(1)(i) to separately identify RFG that

is VOC controlled for Region 1 and Region 2. Similarly, where product

codes are used to identify minimum and maximum standards, as required

in Sec. 80.77(g)(2), the product codes would have to reflect the actual

numerical value for the minimum and maximum standards.

In addition, EPA is proposing that the codes would have to be

standardized throughout the distribution system in which they are used,

and that transferees would have to be given the information necessary

to know the meaning of the product codes.

EPA is not proposing that product codes could be used to satisfy

PTD requirements unrelated to product types. It is EPA's understanding

that product codes used in normal business practice are limited to

product types. In addition, EPA believes that other PTD information,

such as the name and address of the transferor and transferee, volume

of product, and date of transfer, is included in full text in documents

historically used to memorialize transfers of petroleum products.

In addition, EPA is not proposing that product codes could be used

for transfers of gasoline to truck carriers, retail outlets, or

wholesale purchaser-consumer facilities. EPA believes that these types

of regulated parties may not be sufficiently familiar with product

codes to know their full meaning. This belief is based, in part, on

EPA's experience in enforcing compliance with the RFG requirements by

truck carriers, retailers and wholesale users. EPA has found that in

most cases where codes were used to supply required PTD information to

these parties, the parties did not know the meaning of the product

codes even where the gasoline supplier had previously provided the

information necessary to interpret the product codes.

V. Enforcement

A. Prohibitions (Sec. 80.78)

1. Clarification of Prohibitions (Sec. 80.78(a) (1) through (4))

Sections 80.78(a) (1) and (2) prohibit activities that could result

in the use of non-RFG in RFG covered areas. Specifically, these

sections prohibit the manufacture and marketing of gasoline represented

to be RFG unless the gasoline meets the requirements for federally

certified RFG, and prohibit the distribution and sale of non-RFG for

use by ultimate consumers in RFG covered areas. EPA believes, however,

that the current text of Sec. 80.78(a) should be made clearer with

regard to the scope of these prohibitions. As a result, EPA is

proposing to revise the introductory text of Sec. 80.78(a)(1) and

Sec. 80.78(a)(2), to clarify these prohibitions. In addition, EPA is

proposing to delete Sec. 80.78(a)(3), since this section refers to a

conventional gasoline marker and the regulations currently do not

require a marker for a conventional gasoline. EPA is also proposing to

revise Sec. 80.78(a)(4) for purposes of consistency with the revised

text of Secs. 80.78(a) (1) and (2).

2. Addition of ``Causation'' of Prohibited Activities

(Sec. 80.78(a)(10))

Section 80.78(a) prohibits certain conduct on the part of parties

who are engaged in gasoline industry activities such as gasoline

manufacturing and selling, distributing, dispensing, supplying,

storing, or transporting. Under this subsection, however, parties

currently are liable for ``causing'' prohibited conduct only in the

case of gasoline that is transported in violation of the regulations.

EPA now believes there are other situations where a party may, in

fact, cause another to commit a prohibited act, and in those cases, the

causing party also should be liable for the violation. For example, a

distributor who delivers to a retail outlet reformulated gasoline that

fails to meet one or more standards would have caused the retailer to

sell and offer for sale prohibited gasoline.

As a result, EPA is proposing that parties would be liable not only

for committing prohibited actions, but also for causing another party

to commit a prohibited act.

3. Transition from Simple Model to Complex Model in 1998

Under Sec. 80.41(i), refiners and importers of both reformulated

and conventional gasoline have the option of using either the simple

model or the

[[Page 37357]]

early complex model prior to January 1, 1998. Particularly in the case

of producers of reformulated gasoline, EPA believes that most parties

will elect the simple model standards. Beginning on January 1, 1998,

however, refiners and importers must meet the complex model standards

for all reformulated and conventional gasoline produced or imported. As

a result, in January 1998, it will be necessary for parties to

transition from the simple model to the complex model, yet the current

regulations do not specify how regulated parties should accomplish this

transition. Therefore, EPA now is proposing the manner in which this

transition would occur.

Under the proposal, any gasoline produced or imported during

calendar year 1997, through December 31, 1997, would be subject to the

simple or early complex model standards in the same manner as during

calendar years 1995 and 1996. Thus, any simple or early complex model

standards that are met on an annual average basis for 1997 would be met

for all gasoline produced during calendar year 1997.

Any gasoline produced or imported beginning on January 1, 1998,

would be subject to the complex model standards. Thus, conventional

gasoline produced during calendar 1998 would be subject to the annual

average anti-dumping complex model standards specified in section

101(b)(3), and reformulated gasoline produced during calendar 1998

would be subject to the Phase I complex model standards specified in

Secs. 80.41 (c) and (d).

However, beginning on January 1, 1998, the gasoline located in the

distribution system would be a mixture of gasoline produced to meet the

simple model standards and gasoline produced to meet the complex model

standards. In the case of reformulated gasoline, such a mixture may not

meet certain standards that apply at downstream locations or that are

evaluated under the gasoline quality surveys, i.e., the toxics and

NOX emissions performance standards.20 As a

result, EPA is proposing that gasoline quality surveys conducted during

the period January 1, 1998, through March 31, 1998, will not include

evaluation for toxics or NOX emissions performance. During

this period, however, EPA would continue to enforce the complex model

standards for oxygen and benzene content that apply at downstream

locations, and gasoline quality surveys conducted during this period

would include evaluations for oxygen and benzene content. Beginning on

May 1, 1998, all applicable complex model standards would be enforced

at all locations, and gasoline quality surveys would evaluate with all

complex model standards.

---------------------------------------------------------------------------

\20\ There is no simple model NOX standard, so that a

mixture of simple model and complex model gasoline could fail to

meet the complex model NOX standard. Similarly, a mixture

of simple and complex model gasoline could not be evaluated for

compliance with either the simple model or the complex model toxics

emissions performance standards.

The standards for oxygen and benzene content are the same under

the simple and complex models, so that a mixture of simple and

complex model reformulated gasoline could be evaluated for

compliance with these standards. The standards for VOC and

NOX emissions performance are not evaluated for

downstream compliance until the beginning of the high ozone season

on May 1 each year, and as a result should not be affected by the

transition from the simple to the complex model in early 1998.

---------------------------------------------------------------------------

EPA believes that the three month period, January through March

1998, would be sufficient time for parties to transition the gasoline

at all locations in the distribution system from gasoline produced to

meet simple model standards to gasoline produced to meet complex model

standards. This transition period is similar to the time necessary to

transition to the VOC-control standards each Spring; i.e., terminals

are able to complete their transition to the new standard about 60 days

after refiners begin producing gasoline to the new standard, and retail

outlets complete their transition during the next 30 days.

4. Amount of Oxygenate Permitted to be Added to RBOB (Sec. 80.78(a)(7))

Section 80.78(a)(7) requires that RBOB may be blended only with

oxygenate of the type and amount, or within the range of amounts,

specified by the refiner or importer at the time the RBOB was produced

or imported. Today's proposal revises Sec. 80.78(1)(7) to clarify that

parties may add oxygenate amounts in excess of the minimum required by

the refiner or importer up to the amount allowed by the oxygen maximum

standard under Sec. 80.41(g).21

---------------------------------------------------------------------------

\21\ The amount of oxygen added also may not exceed the maximum

amount allowed under section 211(f) of the Clean Air Act. The

maximum amount allowed under section 211(f) is the amount that is

substantially similar to gasoline used in the motor vehicle

certification process, or allowed under a waiver granted under

section 211(f)(4). In 1991, EPA issued an interpretative rule

increasing the maximum amount of oxygen that EPA believes is allowed

under the substantially similar criteria of section 211(f) from 2.0

to 2.7 wt% oxygen. See 56 FR 5352 (February 11, 1991). Ethanol is

allowed in amounts up to 10% volume pursuant to a waiver granted

under section 211(f)(4). See 44 FR 20777 (April 6, 1979).

---------------------------------------------------------------------------

5. Categories of Gasoline Use within Covered Areas that are Exempt from

RFG Requirements (proposed Sec. 80.78(a)(11))

Section 211(k)(5) of the Clean Air Act describes the scope of the

requirement to use RFG in the reformulated gasoline covered areas:

(5) Prohibition.--Effective beginning January 1, 1995, each of

the following shall be a violation of this section:

(A) The sale or dispensing by any person of conventional

gasoline to ultimate consumers in any covered area.

* * * * *

This statutory prohibition on the sale or dispensing of conventional

gasoline in RFG covered areas is not restricted to gasoline used to

fuel motor vehicles, but rather applies to all gasoline sold or

dispensed within an RFG covered area to any consumer, regardless of the

use. The prohibition, therefore, would include gasoline sold or

dispensed for uses such as in motor vehicles, boats, construction

equipment, recreational vehicles, and lawn and garden equipment.

EPA is proposing to exempt parties from this prohibition in the

following limited situations: gasoline used for research, development

and testing purposes; aviation gasoline sold or dispensed for use in

aircraft, including gasoline that has properties identical to motor

vehicle gasoline that is sold or dispensed solely for use in aircraft;

and gasoline sold or dispensed for use in racing vehicles.

EPA recognizes that there may be facilities located within an RFG

covered area that conduct beneficial research, development, and testing

programs which require the use of conventional gasoline. Today's

proposed rule, therefore, contains provisions for obtaining an

exemption from the prohibitions at Sec. 80.78(a)(1) for persons

distributing, transporting, storing, selling or dispensing conventional

gasoline used for research, development, and testing purposes within

RFG covered areas.

To be exempted from the prohibitions at Sec. 80.78(a)(1) for

research, development or testing under today's proposed rule, the

gasoline: would have to be properly identified in product transfer

documents as conventional gasoline to be used only for research,

development, or testing (as applicable); could not be sold to or from

retail gasoline outlets; could not be sold to or from wholesale

purchaser-consumer facilities unless the wholesale purchaser-consumer

is associated with the research, development, or testing; and would

have to be covered by an

[[Page 37358]]

annual research notification to EPA that includes information that

describes the purpose and scope of the program. EPA believes that these

are the least onerous requirements for industry which also will ensure

that non-RFG gasoline is used only for a legitimate research,

development, and testing purpose. Parties should be aware, however,

that the exemption proposed in today's rule would not exempt gasoline

used for research, development, and testing from complying with any

federal conventional gasoline requirements.

Under today's proposal, any person distributing, transporting,

storing, selling or dispensing aviation and racing gasoline would be

required to clearly identify the gasoline as not reformulated to be

exempted from the prohibitions at Sec. 80.78(a)(1). If any of the

restricted gasoline were used in a manner inconsistent with the

restriction, a violation of the prohibited activity would have

occurred, and any person selling, dispensing, or using the gasoline

would be liable for the violation.

EPA is proposing that the racing vehicle exemption would apply only

in the case of vehicles that are used exclusively as racing vehicles in

races that are sanctioned by generally recognized race sanctioning

bodies. In addition, the exception would apply only in the case of

vehicles that do not meet the definition of ``motor vehicle'' under

Clean Air Act section 216(2) and section 85.1703 22 and that

are not registered or licensed for use on or operated on public roads

or highways. Examples of generally recognized race sanctioning bodies

include the National Association for Stock Car Auto Racing, the Sports

Car Club of America, the National Hot Rod Association, the American

Motorcyclist Association, and the American Power Boat Association. The

racing vehicle exemption applies to use of racing vehicles during

practice and qualifying for, and competition in sanctioned races, and

applies to motorcycles and boats used exclusively in sanctioned races.

---------------------------------------------------------------------------

\22\ Under Sec. 85.1703 a vehicle is a ``motor vehicle'' if it

is self propelled and capable of transporting a person or materials,

unless the vehicle meets one or more of the following criteria: (1)

A maximum speed of not more than 25 miles per hour; (2) the absence

of features customary for street use, such as a reverse gear, a

differential, and required safety features; or (3) the presence of

features that render the vehicle highly unsuitable for street use,

such as tracks.

---------------------------------------------------------------------------

The rationale for the proposed exemption for aviation gasoline used

to fuel aircraft is based on safety considerations. Aviation gasoline

must satisfy performance criteria that are relevant to the safe

operation of aircraft, and this safety consideration outweighs the very

limited potential for adverse environmental effects from conventional

gasoline used in this manner. In addition, aircraft emissions normally

would not be confined to the covered area where the aircraft is fueled,

and could occur in significant part outside any RFG covered area. The

rationale for the proposed exemption for racing gasoline is based on

the special performance requirements for true race vehicles and the

limited volumes of gasoline involved. The environmental impact from

these exemptions is trivial or minimal, and the burden from refusing

these exemptions is potentially significant. EPA believes the

exemptions are warranted under these limited circumstances. See Alabama

Power Company v. Costle, 636 F.2d 323, 357 (D.C. Cir.1979).

Nevertheless, EPA requests comments on whether the racing vehicle

exemption would cause increased air pollution in RFG covered areas that

is not trivial, and if so, whether such an environmental effect would

make the racing vehicle exemption inappropriate.

6. Changing Service of Gasoline Storage Tanks (Sec. 80.78(a) (12) and

(13))

Section 80.78(a) requires the segregation of several categories of

gasoline. These categories include the following:

Reformulated gasoline may not be mixed with conventional gasoline and

sold as reformulated gasoline.

Reformulated gasoline blendstock for oxygenate blending (RBOB) may not

be mixed with reformulated gasoline or conventional gasoline, and

RBOB's that have different oxygen requirement

This text is long and has been trimmed here. Open the source document for the complete record.