[Federal Register Volume 59, Number 248 (Wednesday, December 28, 1994)]
[Unknown Section]
[Page ]
From the Federal Register Online via the Government Publishing Office [www.gpo.gov]
[FR Doc No: 94-31815]


[Federal Register: December 28, 1994]


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


40 CFR Part 60
40 CFR Part 80

[AMS-FRL-5128-9]


Regulation of Fuels and Fuel Additives: Standards for Deposit 
Control Gasoline Additives

AGENCY: Environmental Protection Agency (EPA).

ACTION: Proposed rule; reopening of comment period.

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

SUMMARY: On December 6, 1993, EPA published a Notice of Proposed 
Rulemaking (NPRM) to require the use of deposit control (detergent) 
additives in all gasoline used in the United States. The NPRM contained 
a proposed interim program which would take effect beginning January 1, 
1995, as well as a proposed performance-based detergent additive 
certification program to be implemented at a later date. The interim 
program was finalized on October 14, 1994. The detergent certification 
program is expected to be finalized by June 30, 1995, and to take 
effect a year later.
    As proposed, the detergent certification program was concerned 
specifically with the control of port fuel injector deposits (PFID) and 
intake valve deposits (IVD). While also acknowledging the potential 
importance of combustion chamber deposits (CCD), EPA did not propose 
any requirements for CCD control because of uncertainty regarding the 
scope of the problem and the current lack of suitable performance test 
procedures and standards. Subsequently, some commenters expressed 
concern that a federal requirement for PFID and IVD control might 
encourage detergent overuse, which could potentially exacerbate CCD 
concerns. These commenters suggested that, as a temporary measure, 
until CCD performance tests become available, EPA should cap detergent 
use by implementing standards to limit the level of unwashed gums in 
additized gasoline. However, other commenters disagreed with this 
suggestion. This document discusses the CCD issue and requests comments 
on alternative approaches for addressing it. Additional comment is also 
sought in regard to selected issues related to the proposed detergent 
certification program and the enforcement provisions of the detergents 
program.

DATES: Written comments on the specific issues discussed in this 
document will be accepted until January 27, 1995. EPA is not soliciting 
new comments on aspects of the original proposed rule that are not 
specifically addressed in this document.

ADDRESSES: Comments on this notice should be submitted in duplicate to: 
EPA Air Docket Section (LE-131); Attention: Public Docket No. A-91-77; 
Room M-1500, 401 M Street S.W., Washington, DC 20460. (Phone 202-260-
7548; FAX 202-260-4000). This docket also contains all other materials 
relevant to this rulemaking. The docket is open for public inspection 
from 8:00 a.m. until 4:00 p.m., except on government holidays. As 
provided in 40 CFR part 2, a reasonable fee may be charged for copying 
docket materials.
    Electronic copies of this and other documents related to this 
rulemaking are available through the Office of Air Quality Planning and 
Standards (OAQPS) Technology Transfer Network Bulletin Board System 
(TTNBBS). Details on how to access this bulletin board are included in 
Section VII of this notice.

FOR FURTHER INFORMATION CONTACT: For general information and 
information related to technical issues contact: Mr. Jeffrey A. Herzog, 
U.S. EPA (RDSD-12), Regulation Development and Support Division, 2565 
Plymouth Road, Ann Arbor, MI 48105; Telephone: (313) 668-4227, Fax: 
(313) 741-7816. For information on enforcement related issues contact: 
Judith Lubow, U.S. EPA, Office of Enforcement and Compliance Assurance, 
Western Field Office, 12345 West Alameda Parkway, Suite 300, Lakewood, 
CO 80228; Telephone: (303) 969-6483, FAX: (303) 969-6490.

SUPPLEMENTAL INFORMATION:

I. Introduction

    Section 211(l) of the Clean Air Act (CAA) specifies that, beginning 
January 1, 1995, all gasoline shall contain additives to prevent the 
accumulation of deposits in engines or fuel supply systems. Such 
deposits can have adverse effects on the emissions of a vehicle as well 
as its fuel economy and driveability. As part of the CAA mandate, EPA 
was tasked with the promulgation of rules establishing specifications 
for such additives.
    In response to these statutory requirements, on December 6, 1993, 
(59 FR 64213) EPA published a Notice of Proposed Rulemaking (NPRM) 
requiring that gasolines contain detergents to control port fuel 
injector deposits (PFID) and intake valve deposits (IVD). The rule was 
proposed to be implemented in two phases. The first phase was an 
interim program focused on registration and record-keeping requirements 
which would apply for the first year. The second phase, to apply 
thereafter, was a certification program requiring detergents to be 
evaluated in a matrix of test fuels using national consensus test 
procedures and prescribed standards for IVD and PFID control.
    The detergents rule is being finalized in two parts. The interim 
program was promulgated October 14, 1994, and appeared in the Federal 
Register on November 1, 1994 (59 FR 54678). Promulgation of the full 
detergent certification program is expected to occur by June 30, 1995. 
This second phase of the rule was delayed for two reasons. First, the 
national consensus procedure EPA proposed for use in testing PFID 
control was not finalized by the American Society for Testing and 
Materials (ASTM) in time to be incorporated into the rule. Since an 
interim program had already been proposed for 1995, EPA saw little 
value in abandoning reliance on the ASTM procedure solely for the sake 
of promulgating the entire rule at once. Second, issues were raised in 
the comments regarding the possible impact of IVD and PFID control 
requirements on combustion chamber deposits (CCD).
    While the NPRM also contained a substantial discussion on CCD, it 
did not propose specific CCD control measures due to a lack of national 
consensus test procedures and performance standards as well as 
uncertainty as to the scope of the CCD problem. Given the need for a 
slight delay in the final rulemaking schedule to allow completion of 
the test procedure development, EPA believed it was important to 
further evaluate the CCD issue prior to finalization of the 
certification program.
    As part of that evaluation effort, this notice seeks public comment 
on concerns raised by auto industry commenters that the gum levels in 
gasoline will increase as a result of the use of IVD and PFID detergent 
additives and that these gums will increase CCD. This could have an 
adverse effect on emissions and vehicle driveability. Additional 
comment is also sought on several points raised in the comments 
regarding the proposed implementation and enforcement schemes under the 
detergent certification program. Each of these areas is discussed 
below.

II. Combustion Chamber Deposit Control

A. Background

    CCDs are composed of fuel and other material which enter the 
combustion chamber, are not fully burned, and are deposited in various 
portions of the engine's combustion chamber rather than leaving the 
chamber with other exhaust products. The formation of CCD is governed 
by the combustion process, combustion temperatures, and the design of 
the combustion chamber itself. Other factors influencing CCD formation 
include fuel and oil composition, engine design/technology and 
operating conditions (e.g., speed, load, coolant temperature), engine 
oil consumption, and fuel/oil additive usage. The technical literature 
indicates that CCD potentially have three effects on engine operations: 
an adverse effect on emission rates, an increase in engine octane 
requirements, and an impact on fuel economy. Additional background 
information on CCD is provided in the NPRM as well as a memorandum to 
the docket entitled, ``Summary Discussion of Combustion Chamber 
Deposits'' (Docket Item IV-B-04).
    Based on the incomplete information available in the literature and 
the lack of available consensus test procedures or standards, EPA did 
not propose a requirement for CCD control related to emissions or 
octane requirement increase (ORI) effects. EPA received comments both 
supporting and opposing this position.
    Vehicle manufacturers expressed concern that the proposed rule, 
which contained IVD/PFID performance standards but no requirements to 
control CCD, may have an adverse impact on vehicle performance. They 
stated that the use of excessive quantities of some types of detergent 
additives, which provide an acceptable level of IVD/PFID control, tend 
to incrementally contribute to CCD. They further held that EPA's rule, 
by requiring a consistent level of IVD/PFID control additives in all 
gasoline, may exacerbate this incremental additive contribution to CCD, 
since there would be at least a slight incentive to over- additize fuel 
to assure compliance. An adverse impact on vehicle NOx and HC emissions 
performance and durability would result.
    Vehicle manufacturers also stated that CCD can cause ORI, and that 
higher engine octane demand might limit future attempts to achieve fuel 
economy improvements through the design of engines with higher 
compression ratios. Concerns regarding the potential impact of CCD on 
vehicle driveability and durability centered around reports that 
detergent additive overuse could result in valve sticking and 
combustion chamber deposit interference (CCDI). These commenters stated 
that CCDI results from the accumulation of CCD on the top of the piston 
and combustion chamber, which causes mechanical interference during 
engine operation.
    Vehicle manufacturers urged EPA to investigate the potential 
adverse impact of some IVD/PFID additives on CCD as soon as possible, 
with the ultimate goal of a standardized CCD control test. As an 
interim measure, until such a performance test is available, several 
commenters urged EPA to, in effect, adopt a cap on additive 
concentration, by setting a maximum unwashed gum level in additized 
gasoline as measured by ASTM test D381. Two vehicle manufacturers 
recommended an unwashed gum standard of 20 mg/100 ml, based on the 
allowable limit for Japanese gasoline. They stated that CCD problems 
are not present in vehicles in Japan and this may be due in large part 
to the unwashed gum standard of 20 mg/100 ml. The American Automobile 
Manufacturers Association (AAMA) recommended an unwashed gum limit of 
70 mg/100 ml, and stated that this value would act as a cap and not 
create a large burden since it exceeds the levels found in most 
gasolines today. These commenters supported their position by stating 
that a relationship exists between the increased use of IVD detergent 
additives, an increase in unwashed gum levels in gasoline, and an 
increased incidence of CCD-related customer complaints.
    One vehicle manufacturer recommended that, as a condition for 
detergent certification, data should be required demonstrating that the 
additive does not increase CCD relative to base gasoline. This 
commenter further stated that it would be appropriate in the interim to 
allow each additive manufacturer to use its own CCD control test 
methods. The commenter stated that, because these test procedures would 
be expected to vary from manufacturer to manufacturer, this approach 
was recommended only as an interim solution until a standardized CCD 
performance test and standard is developed.
    In contrast to the vehicle manufacturers, the petroleum and 
additive industries contended that regulatory control of CCD was not 
appropriate at this time because of a lack of data on the effects of 
CCD and the lack of an adequate performance test and standard. They 
stated that data suggesting a link between additive use and increased 
CCD with associated increased emissions and ORI is too sparse to reach 
sound conclusions. These commenters also stated that engine 
manufacturing tolerances may play a greater role than fuel properties 
in determining the onset and occurrence of CCDI. Commenters from the 
petroleum industry stated that, while negative impacts from 
overtreatment exist, including valve sticking, these occurrences have 
been very rare and usually result from an accidental overtreatment. In 
summary, API stated that there is no immediate, growing CCD-related 
problem in the field, and EPA's planned IVD/PFID control requirements 
are not likely to cause one. They stated that regulatory control of CCD 
should not be attempted until the Coordinating Research Council (CRC) 
has had an opportunity to investigate potential adverse effects of CCD 
and to gain an understanding of all of the factors which can affect 
CCD, including hardware, lubricant, fuel and additive effects.
    Commenters from the petroleum industry all agreed that, even if the 
need to control CCD were demonstrated, setting a cap on the unwashed 
gum level in additized gasoline would not be an appropriate control 
measure. They argued that the data offered to support a correlation 
between unwashed gum levels and CCD formation is unreliable. 
Furthermore, they claimed that the unwashed gum test is highly 
unrepeatable and cannot differentiate different sources of the gum 
content in the gasoline. They also noted that, since many current IVD/
PFID detergents have varying levels of dispersant and carrier oils 
(some having no carrier oil), the unwashed gum level in gasoline can be 
highly misleading as a measure of the amount of additive present.
    Given the disparity in these positions, EPA requested additional 
input from the affected industries. Additional comment was received 
from AAMA, API, Shell Development Company, and the Chemical 
Manufacturer's Association (CMA)\1\ Two technical papers were also 
published in this time frame that presented additional useful data on 
CCD related issues.\2\
---------------------------------------------------------------------------

    \1\See the memorandum to the docket entitled ``Summary of 
Additional Comments on Combustion Chamber Deposits'', Docket Item 
IV-E-35. Discussion of additional comments on CCD is also contained 
in the following docket items: IV-E-12, IV-E-27, IV-E-29, IV-E-33, 
IV-E-36, IV-G-37, and IV-E-38.
    \2\``Effects of Gasoline and Gasoline Detergents on Combustion 
Chamber Deposit Formation'', Yasunori Takei et. al., SAE Technical 
Paper Series No. 941893. ``Effects of Fuel and Additives on 
Combustion Chamber Deposits,'' Mitchell Jackson and Sara Pocinki, 
SAE Technical Paper Series No. 941890.
---------------------------------------------------------------------------

B. Considerations for Determining a Course of Action

    EPA is considering the implementation of a CCD control measure at 
this time in view of auto industry concerns that, absent such a 
measure, EPA's planned detergent certification program might contribute 
to CCD-related vehicle performance problems. Any CCD control measure 
which EPA might require must be based on EPA's belief that such a 
measure would prevent a future CCD problem or solve one which now 
exists. In addition, it would have to be practical to implement and be 
founded on a sound understanding of its impacts. If a control program 
is found to be needed, it would preferably be based on a national 
consensus test procedure and performance standard. Because several 
years may be required to develop such a test procedure and standard, 
however, a cap on unwashed gums has been suggested as a surrogate CCD 
control strategy. Thus, to determine whether it is appropriate for EPA 
to implement an unwashed gum standard for additized gasoline as an 
interim measure to control the potential additive contribution to CCD, 
EPA must assess the following questions:
    1. To what extent do some or all IVD/PFID detergents contribute to 
unwashed gum levels in gasoline? Which, if any, classes/types of 
additives are of concern? Under what conditions and how substantial is 
this contribution?
    2. To what extent do gums added to gasoline as part of the 
detergent additive package correlate or contribute to CCD formation?
    3. To what extent is there now a CCD problem associated with the 
voluntary use of IVD/PFID detergents? Will mandatory use of these 
detergents exacerbate the potential concerns?
    4. To what extent do CCD cause vehicle emission and performance 
problems?
    The information found in the literature and presented by the 
industry either in their written comments or in response to EPA follow-
up is helpful in answering these questions. However, more information 
is needed. These points are discussed below, followed by a discussion 
of various alternatives that EPA could adopt regarding CCD control. EPA 
encourages public comment that would help to further the understanding 
of these issues.
1. Additive Contribution to Gasoline Unwashed Gum Levels
    The data summarized in the previous section indicate a trend toward 
increased unwashed gum levels with increased detergent additive use and 
concentration. Public comment is sought on the unwashed gum levels in 
additized gasoline for both the various classes/types of detergent 
additives and individual detergent products. Information is also sought 
on the relative contribution of various detergent additive package 
components to unwashed gums. In addition, EPA is also interested in 
more detail and public comment on claims that some detergents with 
superior CCD control performance cause relatively high unwashed gum 
levels in gasoline.
2. Gum/Additive Package Contribution to CCDs
    The information available seems to support the view that some 
detergent additives contribute to CCD more than others. However, the 
significance of this contribution relative to the potential adverse 
impacts of CCD remains unclear. It is also not completely clear what 
fraction of the additive types currently used contribute to CCD. This 
may be a significant concern if a CCD control measure were considered 
that would preclude the use of a significant fraction of the detergents 
currently being used. EPA understands that PIBA type detergent 
additives that use a mineral-based carrier oil predominate in the 
market. EPA requests comment on several points. First, more data is 
requested on the relationship between gums, the additive packages, and 
CCD. Second, EPA is interested in information on which classes/types of 
additives contribute to this concern more than others and market share 
and cost information on each type. Third, EPA requests comment on 
whether the use of a synthetic-based carrier oil rather than a mineral-
based carrier oil in PIBA detergents would decrease the gum levels and 
improve the CCD impacts of these additives and the economic impacts 
which might result from such a change.
3. Prevalence of CCD-Related Problems Due to Detergent Use and the 
Effect of IVD Control Requirements on CCD
    More information is needed to quantify the scope of the current CCD 
problem as it relates to detergent additives. Additional information is 
needed on the fraction of vehicles with CCD-related performance 
problems as well as estimates of how the problem could grow with 
federal detergent requirements. Some public comment presented the 
viewpoint that the planned IVD/PFID control requirements would not 
result in over-additization because the primary result would be the use 
of deposit control additives by all fuel marketers at levels that are 
currently achieved in the majority of gasoline sold within the U.S. 
However, this comment neglects concerns about possible over-
additization to assure compliance during field enforcement as well as 
for marketing/advertising purposes. EPA encourages additional comment 
on this issue.
4. Potential CCD-Related Vehicle Emissions and Performance Impacts
    A review of the available data on the potential impact of CCD on 
exhaust emissions is suggestive of a potentially significant impact. 
However, in addition to the prevalence issue discussed above, comment 
is requested in a number of areas relative to this point. First, 
comment is requested regarding the role of vehicle technology on the 
sensitivity to CCD. Second, more information is requested regarding a 
potential for threshold effect(s). Is there a level below which no CCD 
effect occurs for emissions, CCDI, ORI, or fuel economy? If thresholds 
exist, do they vary for the different impacts? In what relative order? 
Is there a CCD level above which the level of CCD no longer exacerbates 
the impact? If such thresholds exist, any potential control measure 
would ideally need to result in the reduction of CCD beyond this 
threshold to ensure any meaningful emissions benefit. Since there may 
be a threshold effect for the impact of CCD on fuel economy, the 
relationship of this threshold to the possible threshold for exhaust 
emissions and other impacts would also need to be considered in 
evaluating the overall impact of a CCD control measure. Third, data is 
needed relating these threshold effects to the amount of detergent 
additive used. EPA welcomes the submission of any data which might help 
to clarify these issues.
    The data on CCDI is also not conclusive as to whether this problem 
is related to fuel quality or vehicle manufacturing problems. EPA 
encourages the submission of additional data to further evaluate this 
issue. Of particular value would be data illustrating the extent to 
which out-of-tolerance engine manufacture may play a role.
    The commenters agreed that problems with intake valve sticking 
sometimes result from detergent additive overtreatment. However, 
opinions differed as to the extent of these problems and their 
relevance to the possible necessity of a standard to limit detergent 
additive concentration. The data currently available is limited and EPA 
encourages the submission of additional data to help determine the 
prevalence of this problem. The data on the potential impact on vehicle 
performance associated with oil viscosity increase (OVI) from the use 
of detergent additives is likewise very limited and more data is 
requested in this area.

C. Alternative Actions Which EPA Could Adopt

    Based on the information already available and any additional 
information received in response to this notice, several options will 
be considered. Each of these is discussed below.
1. Option 1: Follow Approach Proposed in NPRM
    The first approach to the CCD issue that EPA is considering is to 
follow the basic strategy outlined in the NPRM. Under this option, EPA 
would defer action on CCD pending the gathering of more data, while 
encouraging responsible actions by industry to develop an industry 
consensus test procedure and standard. The CRC CCD panel may be in the 
best position to lead development of national consensus test procedures 
and standards for CCDs. However, EPA is aware of the disagreement among 
members of the industry regarding the current direction of the CRC CCD 
panel's work. Thus, there is considerable uncertainty as to the 
viability of this option. Members of the CRC's panel from the 
automobile industry have expressed reservations about whether the 
investigative work of the CRC CCD panel will proceed rapidly enough to 
address their concerns regarding the potential impact of CCD on vehicle 
driveability and on in-use emissions. EPA requests comment on whether 
it is appropriate to allow industry to evaluate CCD related issues 
further prior to initiating any potential regulatory action regarding 
CCD control, or if more immediate intervention on the part of the 
Agency is necessary.
2. Option 2: Consideration of a CCD Performance Standard
    As discussed above, the Agency concluded in the proposal that the 
lack of adequate data relating CCD to emissions, ORI, or other engine 
performance indicators, as well as the lack of a suitable test 
procedure, precluded the Agency from proposing a CCD-related detergent 
additive performance standard at that time. However, the Agency stated 
that it may, at a later time, propose test procedures and performance 
standards for the control of CCD (among other deposit-related 
phenomena).
    In light of recent events (e.g., receipt by EPA of written and 
verbal comments concerning issues not considered previously), the 
Agency is again considering the issue of whether or not EPA should 
pursue such a combustion chamber deposit control performance standard 
at this time. A discussion of specific issues related to the 
development of a CCD performance standard is provided in a memorandum 
to the docket entitled, ``Consideration of a Combustion Chamber Deposit 
Control Performance Standard'' (Docket Item IV-B-05).
3. Option 3: Require Data Submission on Unwashed Gums
    As an intermediate measure, EPA could require that data on the 
unwashed gum levels of additized gasoline be submitted as part of 
detergent certification. Under this option, the detergent certifier 
would be required to measure the unwashed gum levels in their 
certification test fuel using the ASTM D-381 test procedures (see next 
section). The only requirement would be that these measurements be 
reported to EPA. Using these data, EPA could assess the difference in 
unwashed gum levels relative to the chemistry of the detergent additive 
package to evaluate whether a correlation exists. This information 
would also be useful in determining what fraction of gasoline is 
treated using additives that cause high unwashed gum levels for the 
purposes of evaluating the potential effect on industry of implementing 
an unwashed gum standard or CCD control program in the future.
    This option would also be beneficial in that it would encourage 
continued industry focus on CCD-related issues and may encourage the 
use of additives with better CCD control performance or less of an 
impact on unwashed gum levels. An unwashed gum reporting requirement 
would have little economic impact on industry, since the testing would 
be conducted concurrent with detergent certification testing, would not 
require a separate submittal to the Agency, and would be of minimal 
cost. EPA requests comments on the potential benefits and usefulness of 
this option in responding to current CCD concerns, and on any 
additional data that EPA should require to be submitted during 
detergent certification to help evaluate whether a CCD control measure 
based on an unwashed gum specification is appropriate.
4. Option 4: Implement an Unwashed Gum Standard
    As a measure to address concerns that detergent over-additization 
would contribute to CCDs, EPA requests comment on whether it would be 
appropriate to implement a limit on the unwashed gum level in additized 
gasoline as part of the requirements for detergent certification. As 
previously discussed, questions remain regarding the need for a full 
regulatory program to control CCD effects, and the potential associated 
impacts on emissions, fuel economy and vehicle driveability. While some 
public comments suggest that limiting the additive contribution to 
unwashed gums would not necessarily always result in a reduction in 
CCD, it is possible that an intermediate measure such as this would 
address the most significant concerns raised by the auto industry 
commenters without levying large direct compliance costs on the fuel 
and detergent additive industries. As perceived by the regulated 
industry the benefits may be very small, but, as discussed below, costs 
would also be very small. On the other hand, this requirement could 
make detergent packages with higher gum levels less attractive and 
could act as a disincentive for research on detergent additive packages 
with high gum levels. If an unwashed gum standard is established, EPA 
asks comment on whether such a standard should only be implemented for 
an interim period, assuming a CCD test procedure/ standard is 
established by regulation.
    a. Unwashed Gum Test Procedure. If an unwashed gum standard is 
adopted, it would likely need to be implemented by requiring testing of 
unwashed gum levels in detergent-additized gasoline using the ASTM D381 
procedure or a derivation of this procedure. Concerns have been raised 
about the repeatability and reproducibility reported for this procedure 
in the ASTM D381 document, and EPA believes it would be useful to 
further explore the reasons for the reproducibility and repeatability 
performance cited in the procedure. In discussions with industry 
experts regarding the ASTM D381 procedure, one member of the panel 
which originally developed the procedure suggested the possibility that 
greater precision might be expected in measuring the unwashed gum 
levels of modern gasoline. The repeatability/reproducibility data in 
the ASTM procedure is based on gasolines of the 1960s and 1970s.\3\ EPA 
requests comment on the extent to which gasoline composition can effect 
the precision that can be achieved in the ASTM test.
---------------------------------------------------------------------------

    \3\Memorandum to the docket entitled, ``Phone Call with Leo 
Stavinola of Southwest Research Institute (SwRI), Regarding the 
Applicability of the Unwashed Gum Test to Modern Gasoline'', Docket 
Item IV-E-33.
---------------------------------------------------------------------------

    EPA's analysis of 1993-94 unwashed gum data collected under ASTM's 
Interlaboratory Crosscheck Program suggests that the current 
reproducibility of the procedure may be somewhat better than that 
reported in the D381 document.\4\ However, the limited range of 
unwashed gum values represented in the test samples prevents an 
accurate assessment of test variability over the range of unwashed gum 
values likely to be encountered in use.
---------------------------------------------------------------------------

    \4\Report to EPA from George Hoffman, CSC, entitled ``Sample 
Sizes for Unwashed Gums Measurements''.
---------------------------------------------------------------------------

    Not all laboratories using the D381 unwashed gum procedure 
encounter the variability discussed in the ASTM document. EPA contacted 
Southwest Research Institute (SwRI) to further evaluate the suitability 
of the unwashed gum test from the standpoint of repeatability and 
reproducibility.\5\ SwRI stated that the unwashed gum test is used 
routinely by their laboratory to determine whether or not gasoline had 
been detergent additized. In addition, SwRI related that not only was 
this test a useful predictor of additive presence, but it could also 
distinguish additive concentration fairly well. SwRI stated that, in 
their considerable experience in performing the unwashed gums test, 
they have achieved significantly better repeatability than that 
reported in the ASTM procedure. For example, at an unwashed gum level 
of 60 mg/100 ml, they commonly achieve repeatability of results within 
5 mg/100 ml, as compared to the 25 mg/100 ml 
reported in the ASTM procedure. A round robin program conducted by AAMA 
using three fuels and four labs also indicated good reproducibility 
(coefficients of variation of 6 percent, 12 percent, and 2 percent) for 
the three fuels tested. (See memo in public docket at IV-G-39).
---------------------------------------------------------------------------

    \5\Memorandum to the docket entitled ``Phone Call with Karen 
Kohl of Southwest Research Institute (SwRI) Regarding the Unwashed 
Gums Test,'' Docket Item IV-E-32.
---------------------------------------------------------------------------

    SwRI stated that they are aware of several potential problems that 
may partially account for the poor reproducibility reported by ASTM. 
SwRI related that the temperature of the evaporation bath and the flow 
of air to the airjet apparatus must be carefully controlled to limit 
the variability in results. SwRI also related that another potential 
problem which can introduce variability in test results is 
contamination of the oil from the supply pumps. If the repeatability 
reported by SwRI and the reproducibility achieved in the AAMA study 
could also be achieved at other laboratories, it might be possible to 
achieve adequate test precision while minimizing the required number of 
repeat tests.
    EPA requests comment on ways in which the ASTM D381 procedure could 
be modified to limit test variability. In particular, EPA requests 
comment on whether additional specifications are required regarding 
control of the evaporation bath temperature, flow of air to the air jet 
apparatus, and prevention of oil contamination of the fuel sample from 
the supply pumps. Specifically, EPA is considering tighter 
specifications on several elements of the test including the air jet 
apparatus to 600mL/s 30 mL/s, the steam jet apparatus to 
1000 mL/s 50 mL/s, and the evaporation bath well 
temperature to 155 deg.C 2  deg.C. Use of a temperature 
measuring approach other than a thermometer (e.g., thermocouple with 
digital readout) is also being considered. Other similar modifications 
to test specifications will also be considered to the degree that they 
would reduce variability. The submission of test data from the ASTM 
D381 procedure would also be useful in assisting further evaluation 
evaluating test repeatability and reproducibility.
    b. Unwashed Gum Limit. If an unwashed gum standard for additized 
gasoline is adopted, it must be carefully selected based on the 
consideration of several factors. First, the available data suggests 
that the contribution to unwashed gums from sources other than 
detergent additives is generally under 10 mg/100 ml per the ASTM test. 
This consideration is important because the focus of an unwashed gum 
standard must be on the additive contribution to gums. EPA believes 
that it may be reasonable to assume that the base fuel contribution to 
gasoline unwashed gum levels does not exceed 10 mg/100 ml. EPA requests 
data to further evaluate whether this assumption is appropriate.
    Second, the choice of an unwashed gum cap should be based on an 
understanding that such a cap would result in an appropriate level of 
CCD control. This involves consideration of the possible threshold 
effects of CCD-related emissions, fuel economy, and vehicle 
driveability impacts. EPA requests comments on what would be an 
appropriate unwashed gum cap given these considerations.
    There were two separate suggestions in the public comment regarding 
the choice of an appropriate unwashed gum cap. The first suggestion was 
to adopt the unwashed gum cap already in place for Japanese gasoline of 
20 mg/100 ml.\6\ This suggestion was based on the premise that the 20 
mg/100 ml cap has been in force for Japanese gasoline for some time, 
that industry has been able to comply with this cap while meeting other 
deposit control performance needs, and that the presence of such a cap 
might be credited for the absence of CCD-related problems in Japan. EPA 
is concerned about setting a potential standard at this level, because 
it would prevent the use of many current IVD/PFID detergents. Assuming 
absolute precision in unwashed gum measurement, a 20 mg cap would limit 
the additive contribution to unwashed gums to about 10 mg. The 
available data indicates that this is significantly less than the 
contribution to unwashed gums for IVD/PFID detergents currently used in 
the U.S. which typically appears to be between 20 and 60 mg/100 ml and 
can be considerably higher.\7\ EPA's review of AAMA gasoline survey 
data reveals that less than 40 percent of U.S. gasoline samples tested 
had an unwashed gum level below 20 mg/100 ml.\8\ One explanation for 
the apparent large difference in unwashed gum levels in Japanese and 
U.S. gasolines might be that Japanese gasoline tends to require less 
detergent for adequate IVD control. EPA requests comment that would 
help explain the difference between the unwashed gum levels in Japanese 
and U.S. gasoline.
---------------------------------------------------------------------------

    \6\Written comments on the NPRM from Mazda Motor Corporation, 
Docket Item IV-G-17, and from Toyota Motors Inc., Docket Item IV-G-
25.
    \7\Letter to Glenn Passavant, Office of Mobile Sources, from 
James Williams, American Petroleum Institute, Docket Item IV-G-36.
    \8\AAMA unleaded gasoline survey data, 1991-1993, California 
gasoline excluded.
---------------------------------------------------------------------------

    AAMA suggested a significantly higher unwashed gum cap than that 
observed in Japan. AAMA contended that their suggested cap of 70 mg/100 
ml took into account the base gasoline contribution to gums and the 
variability in the ASTM test procedure, and would allow the use of 
detergents capable of IVD, PFID and CCD control. They further stated 
that compliance with this cap could be achieved using current additive 
technology that results in unwashed gum levels of 35 to 40 mg/100 ml. 
AAMA's suggestion is based on the premise that an appropriate level of 
control of the detergent additive contribution to CCD can be achieved 
by preventing the use of the highest concentrations of detergent 
additives and unnecessary over-additization. EPA's review of AAMA 
gasoline survey data reveals that more than 90 percent of U.S. gasoline 
samples analyzed had unwashed gum levels below 70 mg/100 ml. A review 
of the same data also showed that approximately 35 percent of U.S. 
gasoline had unwashed gum levels in excess of 40 mg/100 ml. Thus, 
depending on the precision of the unwashed gum test, a 70 mg/100 ml 
unwashed gum cap could result in disqualification for use of detergent 
packages/concentrations used in approximately 10 to 35 percent of 
gasoline in the U.S.
    EPA requests comment on the suitability of an unwashed gum cap of 
70 mg/100 ml. In particular, EPA requests comment on the potential 
benefit of implementing such a standard, and the ability of industry to 
meet IVD/PFID deposit control requirements with additives that comply 
with a 70 mg/100 ml unwashed gum standard.
    Given concerns expressed by some commenters about variability, EPA 
proposes that certification or data submission requirements be based on 
the average of five unwashed gum tests (ASTM D381). With regard to an 
unwashed gum standard, EPA asks for comment on the minimum number of 
unwashed gums tests which should be specified to demonstrate compliance 
and, as an option to industry, the maximum number which could be 
allowed.
    c. Incorporation of Unwashed Gum Standard Into the Detergent 
Certification Program. If EPA adopts an unwashed gum standard or 
requires submission of unwashed gum levels, the primary compliance 
focus would be placed on unwashed gum testing of the properly additized 
test fuel(s) used during detergent certification. One option would be 
to require the measurement of the unwashed gum level of a test fuel 
blended with the minimum treatment rate of the detergent package as 
determined by certification testing for IVD/PFID control. To limit 
variability in the base fuel contribution to unwashed gums, the test 
fuel would be required to comply with other specifications regarding 
gasoline composition (i.e., aromatic, olefin, T-90 and oxygenate 
content) prescribed for detergent certification testing. For the 
detergent additive to be certified, the unwashed gum level in this test 
(or the average for a series of tests) would need to be below the 
unwashed gum standard.
    One of the major issues raised by the auto industry commenters was 
concern that over-additization could lead to increased gum levels and 
CCD-related problems. While there may be no economic incentive to over-
additize, it is likely to occur. It may occur accidentally, some fuel 
manufacturers may choose to slightly over-additize to ensure compliance 
with the volume accounting reconciliation requirements of EPA's 
enforcement program, or it may occur for other reasons such as 
advertising. One way to accommodate these possibilities and yet also 
address the concerns of auto makers is to set the concentration of the 
additive in the test fuel during testing for unwashed gums at a level 
higher than the certified treat rate. For example, requiring 
certification to the unwashed gum standard (or testing of unwashed gums 
under Option 3) at a level of 110 percent of the certified treat rate 
would allow the assessment of the impact of a slight over-additization 
without creating an unreasonable obstacle to compliance. This would 
also allow for some variation in the unwashed gum level of the 
unadditized gasoline. Yet another approach with a similar effect would 
be to reduce the proposed standard by 10 percent. That is, instead of 
20 mg/100 ml or 70 mg/100 ml, the standard would be set at 18 mg/100 ml 
or 63 mg/100 ml with testing conducted at the certified treat rate.
    The detergent registration and certification programs require that 
each gasoline contain detergent in an amount at least equivalent to the 
minimum certified treat rate. No maximum limit is anticipated. While 
testing at 110 percent of the minimum certified treat rate would 
address slight over-additization when the minimum treat rate is the 
actual target concentration, it will not be representative for fuels 
additized at even higher rates. EPA, therefore, asks comment on an 
option which would require those fuel manufacturers who add detergent 
at greater than 110 percent of the minimum treat rate to pass the 
unwashed gum standard using fuel treated at the rate actually being 
used.
    As an ancillary enforcement tool, EPA might also require the 
submission of test data on the unwashed gum levels of in-use fuels from 
detergent additive blenders. In-use testing of unwashed gum levels 
might also be conducted by EPA. EPA does not anticipate that such in-
use testing would be conducted routinely, but rather that it would be 
used in cases where there was a suspicion of violation, or would be 
conducted randomly to encourage compliance. Penalties for noncompliance 
with an unwashed gum standard would be similar to those for 
noncompliance with minimum concentration requirements. EPA requests 
comments on the potential certification and enforcement requirements 
described above as well as input on other options.
    d. Costs and Benefits. The potential economic impacts of 
implementing an unwashed gum standard or data submission requirement 
vary with the provisions of the potential program, but in general would 
be similar. First, EPA needs to further evaluate the number of unwashed 
gum tests that would be required for each detergent certification. If a 
single unwashed gum test was performed during certification testing, 
the additional cost for each certification would be approximately $100 
for a total cost to industry of approximately $40,000 for the 
approximately 400 detergents certified in the first year of the 
program. (This assumes that these tests are not already conducted for 
other reasons. If multiple tests are used to reduce variability 
concerns raised in association with the ASTM D381 test, the total cost 
would be somewhat higher. If, for example, five tests were conducted as 
proposed above, the total cost for all detergents would be $200,000 in 
the first year. Recurring annual costs would vary between $6,000 and 
$30,000 assuming 15 percent new certifications per year.
    While it may be difficult to quantify the emission benefits, it is 
possible to estimate whether such a program could be cost effective. 
One approach would be to calculate the amount of emissions that would 
need to be prevented by an unwashed gum standard for the control 
measure to be cost-effective. If the costs were low, the amount of 
emissions that would need to be prevented could be relatively small and 
yet the program could still be quite cost effective. Using the first 
year cost of $200,000 discussed above, only 200 tons of HC/NOx emission 
reductions would be required to achieve a nominal cost effectiveness of 
$1,000 per ton. Given the large number of miles travelled per year by 
gasoline-fueled vehicles, only a tiny fraction of the fleet would have 
to get a tiny benefit for this level of benefits to be achieved. For 
example, less than a hundredth of one percent of vehicles would have to 
achieve a 0.01 g/mi benefit in order to achieve this benefit. This 
level of benefits and more appears quite conceivable. The potential 
negative impact of fuel economy that may accompany a reduction in CCDs 
complicates consideration of the potential economic impact of the 
program, as do potential adverse impacts on companies manufacturing 
detergent additives with high unwashed gum levels. EPA requests comment 
on the approach to determine the cost- effectiveness of CCD control 
outlined above and on a method by which any potential negative impact 
on fuel economy might be accounted for. EPA also asks comment on 
whether it is possible to reduce emissions or prevent emission problems 
without adversely affecting fuel economy. It should be noted that 
section 211(l) of the CAA does not require EPA to justify deposit 
control measures on a cost-effectiveness basis. However, EPA will give 
the costs and benefits of any potential CCD control measure due 
consideration.
5. Other Options
    a. Use of a Standard Based on Thermogravimetric Analysis of 
Detergent Additives as an Interim CCD Control Measure. EPA requests 
comment on whether a standard based on thermogravimetric analysis (TGA) 
of detergent additives could be used as an interim measure to limit the 
additive contribution to CCD. Such a standard would be based on the 
premise that, to limit the additive contribution to CCD, a detergent 
additive must decompose at temperatures encountered in the combustion 
chamber. Any additive which does not decompose and burn in the 
combustion chamber may provide material for the formation of CCD. 
Thermogravimetric analysis is one method of determining the thermal 
characteristics of materials, and might be useful in determining what 
portion of detergent additive packages remain unburned and may 
contribute to the formation of CCD.
    One potential TGA-based standard would require that, as part of 
detergent certification testing, a TGA analysis would be performed on 
the detergent additive package. For a detergent to be certified, the 
residue of the detergent package left at the specified temperature 
would be required to not exceed a specified fraction by weight of the 
original sample. Specifying a testing temperature is a critical element 
in a potential TGA-based CCD control standard. Recent research 
indicates that the surface temperature in the combustion chamber may be 
the most important physical parameter in the formation of CCD.\9\ This 
research indicates that, with high surface temperatures, CCD will reach 
stabilized levels earlier and the total amount of deposits will be 
lower. It further indicates that the critical surface temperature 
beyond which no additional fuel-derived CCD form is 310  deg.C, and the 
critical surface temperature beyond which no oil-derived deposits form 
is approximately 60  deg.C higher. This research may be useful in 
helping to determine the proper test temperature specification for a 
TGA-based CCD control standard. EPA requests comment on an appropriate 
temperature specification.
---------------------------------------------------------------------------

    \9\``A Physical Mechanism for Deposit Formation in a Combustion 
Chamber'', Society of Automotive Engineers (SAE) Technical Paper 
Series No. 941892.
---------------------------------------------------------------------------

    Comment is also requested on the allowable fraction of the additive 
package which could remain as residue after the completion of the TGA 
analysis. EPA believes that, for a TGA-based standard to provide a 
meaningful measure of CCD control, the specification on the allowable 
amount of residue would need to be set in the range of 2 percent of the 
original sample. EPA requests comment on whether, alternately, it would 
be more appropriate to place an absolute specification on the allowed 
weight of detergent residue for a given quantity of additized gasoline, 
and what an appropriate specification would be. Comment is also 
requested on what procedural guidelines EPA should implement for the 
TGA test.
    The detergent additive package on which a TGA analysis was 
performed would be required to contain the maximum concentration of the 
detergent-active components covered under a registration. Use of these 
components at higher concentrations to fulfill gasoline detergency 
requirements would be prohibited. To enforce such a requirement, EPA 
could require that samples of detergent additive be tested by the 
additive manufacturer or submitted by the additive manufacturer for 
analysis by EPA. A violation would occur if this sample failed to 
satisfy the specified TGA-based limit. Alternately, a violation would 
occur if, through fourier transform infrared spectroscopy (FTIR) 
analysis, a sample was found to contain higher concentrations of 
detergent-active components than those specified in the registration.
    b. Use of Existing CCD Control Performance Data. One commenter 
suggested that, as an interim measure, EPA require vehicle/engine CCD 
control test data to be submitted by each detergent certifier without 
detailing any requirements regarding the acceptability of the data. EPA 
believes that such a requirement, while potentially encouraging 
responsible consideration of CCD control by industry, would not provide 
a meaningful measure of an additive's CCD control performance. EPA 
believes that CCD control performance testing is too ill defined to 
allow EPA to implement guidelines regarding its acceptability. The 
Agency encourages public comment on this and other options that EPA 
might adopt to control the potential adverse impacts of detergent 
additive overuse.

D. Relationship to California Detergent Additive Program

    Under EPA's proposed detergent certification requirements, a 
detergent certified for use under the state of California's detergent 
program (Title 13, section 2257 of the California Code of Regulations) 
could be used to satisfy federal detergency requirements for gasoline 
sold in California. The detergent additive certification requirements 
currently implemented by the state of California include IVD/PFID 
control requirements but do not include any unwashed gums or other CCD 
control provisions. Therefore, if EPA were to implement an unwashed gum 
standard or reporting requirement, a detergent certified by the 
California Air Resources Board (CARB) would need to satisfy these 
requirements in addition to complying with CARB's detergent 
certification program in order to be eligible for use in complying with 
federal detergency requirements in the state of California.

III. Other Issues for Additional Comment

A. Formulation of Certification Test Fuels

    The discussion in this notice regarding the formulation of 
certification test fuels pertains specifically to measures which EPA 
might take to account for yet-to-be-identified nonoxygenate fuel 
severity factors. Also discussed is the potential difference in the 
deposit forming tendency of ethanol produced by different processes, 
and measures that EPA might take to account for this variability. Other 
issues related to test fuel definition will be addressed in the 
detergent certification final rule.
1. Accounting for Unidentified Nonoxygenate Fuel Severity Factors
    a. Approach Proposed in the NPRM. Under the proposed requirements, 
certification testing for IVD/PFID control would normally be conducted 
using a matrix of four test fuels, although testing in as few as two 
fuels would be allowed under certain circumstances. Test fuels would be 
required to meet or exceed minimum specifications regarding the levels 
of the following four nonoxygenate fuel parameters which EPA believes 
affect the deposit forming tendency of gasoline (referred to as fuel 
severity factors): olefins, sulfur, aromatics, and T-90 distillation 
point. To account for the deposit forming tendency of oxygenates, one 
fuel would be required to contain 10 percent ethanol, and another would 
be required to contain 15 percent MTBE. The NPRM included provisions 
allowing other fuel parameters to be used to define certification test 
fuels if their effect on IVD/PFID fuel severity could be demonstrated.
    In the NPRM, EPA discussed concerns that specifications on these 
fuel severity factors may not completely define a gasoline's deposit 
forming severity. If this were the case, detergent certifiers might be 
able to locate or specially blend certification test fuels that meet 
the proposed compositional requirements but are still inappropriately 
mild in their deposit forming tendency. To help account for unknown 
factors in gasoline composition that may affect fuel severity, EPA 
proposed that gasoline samples for certification testing must be drawn 
from normal production gasoline stock (finished commercial gasoline) 
taken from normally operating refinery and/or terminal facilities. In 
addition, the multiple test fuels were to be drawn from separate 
production/distribution facilities. This requirement would increase the 
certainty that unknown severity factors will be represented by ensuring 
that various refinery stocks are tested. It would also tend to limit 
the opportunity to select test fuels from refineries that, for 
unidentified reasons, tend to produce gasoline with a relatively low 
deposit forming tendency.
    b. Public Comment on the Proposal. Comments submitted by API and 
CMA stated that requiring commercial fuels for testing would be 
extremely burdensome and would have impractical consequences with 
regard to the logistics and mechanics of obtaining non-additized fuel. 
Amoco, API, CMA, Nalco, Texaco, and Ashland stated that EPA should 
allow the use of refinery blend stocks that meet ASTM D4814 
specifications to formulate certification fuels. It was also suggested 
that the finished certification test fuels should meet ASTM D4814 
specifications. One commenter stated that, to help alleviate concerns 
regarding fuel blending, the identity of each blending component should 
be documented regarding its source and properties.
    AAMA stated that there might be characteristics that effect a 
gasoline's deposit forming tendency which are not represented by 
consideration of the five parameters proposed to define certification 
test fuels. AAMA acknowledged that there are no data available to 
determine which additional parameters would be appropriate to include 
in the test fuel specifications. However, AAMA stated that 
certification test fuels could be made more representative by requiring 
that the major refinery streams be represented in the blending stocks 
used to formulate these fuels. Such a practice would tend to help 
ensure that yet-to-be-identified fuel severity factors are represented 
in the certification test fuels. AAMA stated that requiring more than 
one certification test fuel would allow for the inclusion of more 
refinery streams in the formulation of certification test fuels, 
thereby enhancing the representativeness of the testing.
    Comments from API and CMA requested that EPA allow certification 
testing on a single fuel, formulated from refinery blend stocks, that 
contains each nonoxygenate fuel severity factor at its highest level in 
the proposed test fuel matrix, plus 10 percent ethanol. These 
commenters stated that the increase in gasoline deposit forming 
tendency resulting from the addition of 10 percent ethanol is more than 
enough to address concerns about the potential deposit forming tendency 
of MTBE. Hence, testing on a second MTBE-containing fuel should not be 
required. On the other hand, commenters from the automobile industry 
stated that testing on both ethanol- and MTBE-containing fuels should 
still be required to account for differences in their tendency to form 
deposits.
    c. Alternative Approaches that EPA Might Adopt. EPA is considering 
several approaches, in addition to those proposed in the NPRM, to help 
ensure the representativeness of the certification test fuels. Under 
the first option, specifications on fuel parameters in addition to the 
five that were proposed to be used to define certification test fuels 
(olefins, sulfur, aromatics, T-90 distillation point, and oxygenate 
content) would be added to better define the test fuels deposit forming 
tendency. The NPRM discussed other fuel parameters that might have an 
impact on a gasoline's tendency to form deposits. Several of these are 
refinements on the nonoxygenate fuel parameters already considered.
    The limited data available indicates that diolefins may affect a 
gasoline's tendency to form deposits more significantly than do mono-
olefinic species. Certain species of sulfur, such as ditertiary butyl 
disulfide, may also have more of an impact on a gasoline's deposit 
forming severity than do other species of sulfur. In addition, limited 
data suggests that heavier, polycyclic aromatic species may have a 
greater impact on fuel severity than do other aromatic species. These 
additional fuel parameters might be used in place of, or in addition 
to, the less specific fuel parameters proposed in the NPRM (total 
olefins, total sulfur, and total aromatics). Data were also discussed 
in the NPRM suggesting that fuel nitrogen content might be useful in 
helping to define the deposit forming tendency of certification test 
fuels.
    Nevertheless, the data available appears to be insufficient to 
positively identify these fuel parameters as additional severity 
factors. Also, since levels of such fuel parameters as diolefins, 
ditertiary butyl disulfide, polycyclic aromatics, and fuel nitrogen are 
not commonly measured or reported, it would be difficult to determine 
what levels of these additional fuel parameters would be appropriate to 
require in certification test fuels. As noted earlier, the proposed 
test fuel specifications on the levels of sulfur, olefins, aromatics, 
and T90 were based on national gasoline survey data. EPA requests 
comments on what additional fuel parameters could be used to help 
better define the severity of certification test fuels, and on the 
levels of such fuel parameters that would be appropriate to require in 
the test fuels to ensure adequate deposit control.
    Given that it may not be feasible to use specifications on 
additional fuel parameters to better define the severity certification 
test fuels, EPA is considering several other alternatives to improve 
their representativeness. One such approach would be to specify the 
refinery blendstocks which must be used in the formulation of 
certification test fuels. Due to their potential impact on the deposit 
forming tendency of finished gasoline, EPA believes that specifications 
on the minimum proportions of the following gasoline blendstocks to be 
used in the formulation of certification test fuels may be useful in 
helping to better define test fuel severity: full range fluid 
catalytic-cracked (FCC) naphtha, full range reformate, full range 
straight run naphtha, and light coker naphtha.
    The relative proportions of these blendstocks required to be used 
in the formulation of certification test fuels could be based on the 
fraction at which they are used in blending finished gasoline in the 
U.S. In keeping with the proposed certification test fuel 
specifications, which require that each of the four nonoxygenate fuel 
severity factors must be represented at levels greater than their 
respective national average levels, it may be appropriate to require 
that the subject blendstocks be represented in the certification test 
fuels at proportions that exceed their average representation in 
finished U.S. gasoline. Based on this approach, appropriate 
specifications on the minimum percentage of the subject refinery 
blendstocks which must be used in formulating the certification test 
fuels are likely to fall within the following ranges: 25 to 40 percent 
full range FCC naphtha, 25 to 40 percent full range reformate, 1 to 3 
percent full range straight run naphtha, and 1 to 3 percent light coker 
naphtha.\10\ These specifications would account for approximately 50 to 
85 percent of the gasoline blendstocks used in formulating 
certification test fuels. The remaining blending components used in 
formulating the test fuels could be drawn from any type of refinery 
blendstock.
---------------------------------------------------------------------------

    \10\For a review of the quantities of manufactured gasoline 
blendstocks blended into U.S. gasoline during the period of April 1 
through September 30, 1989, see the National Petroleum Refiners 
Association Survey of U.S. Refining Industry Capacity to Produce 
Reformulated Gasolines, part A, January 1991.
---------------------------------------------------------------------------

    EPA requests comment on the usefulness of this approach in 
accounting for yet-to-be-identified fuel severity factors and in 
preventing the use of certification test fuels with inappropriately 
mild deposit forming severity. EPA specifically requests comment on 
which refinery blend stocks should be included and on the proportions 
at which they should be represented in the test fuels. EPA also 
requests comment on the number of test fuels which would be needed to 
adequately represent all the relevant refinery blend stocks given the 
need to meet EPA's proposed fuel compositional specifications 
(regarding minimum required levels of olefins, sulfur, aromatics, T-90, 
and oxygenates). EPA believes that a single test fuel may not be 
adequate for this purpose. In addition, the Agency requests comment on 
the extent to which previous commenters concerns regarding potential 
logistical problems associated with EPA's proposal to require testing 
of finished commercial fuels also apply to the approach of formulating 
test fuels from specified proportions of particular refinery blend 
stocks.
    Another approach EPA is considering would involve raising the 
levels of the four nonoxygenate fuel parameters required in the 
certification test fuels. This approach may be useful in helping to 
limit the variability in fuel severity for a given fuel specification, 
because the extent of this variability may tend to decrease as levels 
of these fuel parameters increase. EPA is considering requiring that 
the 65th percentile levels of these four nonoxygenate fuel parameters 
must be present in the certification test fuels after the addition of 
oxygenates, rather than before their addition as was proposed. This 
approach is being considered due to public comment on the NPRM which 
indicated that allowing the levels of these fuel parameters to be 
diluted by the addition of oxygenate may inappropriately reduce test 
fuel severity. If EPA were to take such an approach, the resulting 
increase in fuel severity may tend to reduce concerns over the 
variability in fuel severity due to the effect of yet-to-be identified 
fuel parameters. EPA requests comment on the extent to which 
variability in test fuel severity would continue to be a concern if 
such an approach were adopted.
    EPA also requests comment on whether it would be appropriate to 
increase the required levels of the nonoxygenate fuel parameters still 
further to limit the variability. To this end, EPA is considering 
requiring that each of the nonoxygenate fuel parameters must be 
represented in the certification test fuels at their respective 70th 
percentile levels after the addition of oxygenate. Such comment should 
also address the potential that such an upward adjustment in test fuel 
severity may cause higher concentrations of detergent additive to be 
used, with an associated heightening of concerns regarding the additive 
contribution to the formation of CCD.
    Another approach EPA is considering would require that the deposit 
forming tendency of test fuels be demonstrated through testing of the 
non-additized fuels prior to their use for certification testing 
purposes. Under this approach, each test fuel would be evaluated for 
its tendency to form intake valve deposits using the BMW 318i test 
procedure (as adopted by EPA for certification testing). The cost of 
implementing such a requirement would naturally increase with the 
number of test fuels required, and its adoption would be most 
economical if only a single test fuel formulation were required. For a 
fuel to be acceptable for use in certification testing, a specified 
deposit weight would need to be accumulated during pre-certification 
testing of the fuel in an unadditized state. EPA is considering several 
alternative test fuel evaluation criteria. Under the first alternative, 
at least 150-250 mg of intake valve deposits (on average) would be 
required to be accumulated over the course of 10,000 test miles for a 
test fuel to be acceptable. Under the second alternative, at least 80-
120 mg of IVD would be required to be accumulated over the course of 
5,000 test miles. Specific comment is also sought on point values 
within the two mg ranges presented above.
    EPA requests comment on whether this approach is a necessary and 
useful measure to ensure the representativeness of certification test 
fuels. Comment is requested on whether it would be useful to require 
that fuels meet compositional specifications if EPA were to require 
that they be tested to demonstrate adequate deposit forming severity 
prior to use for certification purposes. Specific comment is requested 
on the level of IVD that should be required to be generated for a test 
fuel to be acceptable as having an adequately severe deposit forming 
tendency. EPA also requests comment on whether it is necessary to 
require demonstration of a test fuel's tendency to form fuel injector 
deposits (PFID) and the appropriate test/standard which might be used 
to accomplish this purpose. One option under consideration is use of 
the Chrysler 2.2L vehicle in the ASTM test with the fuel required to 
generate a flow restriction in the range of 10-20 percent. Specific 
comment is requested on this option including a point in the range and 
also on whether a bench rig test may be sufficient to screen a test 
fuel for its tendency to form PFID. Comment is also requested on other 
measures which may be implemented to ensure the representativeness of 
certification test fuels.
2. Accounting for Potential Variability in the Deposit Forming Severity 
of Ethanol
    EPA proposed that the ethanol used in formulating certification 
test fuels must be of fuel-grade quality. Specially processed ethanol 
would not be acceptable for use in formulating certification test 
fuels. This requirement was proposed to account for the deposit forming 
tendency of other substances commonly found in fuel-grade ethanol. EPA 
now believes that depending on the processing method used to produce 
fuel-grade ethanol, various levels and types of impurities may be 
present. These varying levels of these impurities, such as fatty acids, 
might potentially effect the deposit forming severity of fuel-grade 
ethanol to different degrees. If this were the case, certification 
testing could be conducted using fuels blended with fuel-grade ethanol 
that has an inappropriately low tendency to increase gasoline forming 
tendency.
    EPA requests comment on the extent to which this is a concern and 
on potential approaches that EPA might take to ensure that the ethanol 
used during certification testing is representative of in-use ethanol 
blending stock in its tendency to from deposits. One approach would be 
to require that ethanol used for certification testing purposes must 
contain a minimum level of impurities. Specifications might also be 
necessary on the type of impurities which must be present. The maximum 
allowed levels of impurities specified in commercial standards for 
fuel-grade ethanol might be useful in determining the levels of 
impurities that must be present in ethanol used for certification 
testing. Potable and chemical grade ethanol as well as fuel grade 
ethanol which receives additional processing not common to all fuel 
grade ethanol could not be used in testing. EPA also requests comment 
on the extent to which the concerns discussed above may also apply to 
MTBE used in certification testing. EPA also asks for comment on 
whether the restrictions discussed above for ethanol should also be 
applied to MTBE.

B. ASTM IVD and PFID Control Test Procedures

    In the NPRM, EPA proposed test procedures to evaluate IVD and PFID 
control that were largely based on draft procedures under evaluation by 
the American Society for Testing and Materials (ASTM) at the time of 
the proposal. EPA also proposed that the ASTM test procedures might be 
adopted in the detergent certification final rule if they were 
finalized by ASTM in time, and there were no changes that would require 
further public notice and comment.\11\ These test procedures have 
recently been finalized by ASTM and EPA anticipates their adoption 
under the detergent certification final rule.\12\ EPA requests comment 
this approach. EPA proposed an alternate IVD control standard at a 
5,000 mile test length, in addition to the traditional 10,000 mile 
standard. Comment is requested on what addition to the ASTM IVD control 
procedure would be necessary to allow the use of a 5,000 mile test 
length as well as the 10,000 mile test length specified by ASTM.
---------------------------------------------------------------------------

    \11\Drafts of the ASTM IVD and PFID test procedures were placed 
in the public docket for review.
    \12\The ASTM procedures are designated as follows: ASTM test 
method D 5500, ``Standard Test Method for Evaluation of Unleaded 
Automotive Spark-Ignition Engine Fuel for Intake Valve Deposit 
Formation'', and ASTM test method D 5598, ``Standard Test Method for 
Evaluating Unleaded Automotive Spark-Ignition Engine Fuel for 
Electronic Port Fuel Injector Fouling''.
---------------------------------------------------------------------------

C. Applicability of Gasoline Detergency Requirements

    The gasoline detergency requirements implemented by the interim 
detergent registration rule apply to all gasoline, leaded and unleaded, 
highway and off-road, including both reformulated and conventional 
gasolines, oxygenated gasoline, and the gasoline component of M85 and 
E85, as well as marine fuel and gasoline for military purposes. (M85 is 
a mixture of 85% methanol and 15% gasoline. E85 is a mixture of 85% 
ethanol and 15% gasoline.) Gasoline service accumulation fuel is also 
required to comply with gasoline detergency requirements, as is the 
gasoline component of alcohol blend service accumulation fuel. As noted 
in the detergent registration final rule, the following types of 
gasoline are exempted from compliance: aviation fuel, racing fuel, 
emissions certification fuel, and gasoline used for research and 
developmental purposes. Gasoline detergency requirements apply to all 
gasoline other than those types for which an exemption was specifically 
prescribed.
    As an example, although not specifically mentioned in the 
regulation, since it is to be used in highway motor vehicles factory 
fill fuel must comply with gasoline detergency requirements.\13\ 
Factory fill fuels must comply with these requirements despite the 
exemption from general fuel registration requirements noted for these 
fuels in 40 CFR 79.4(a)(3).
---------------------------------------------------------------------------

    \13\Factory fill fuels must comply with these requirements 
despite the exemption from general fuel registration requirements 
noted for these fuels in 40 CFR 79.4(a)(3).
---------------------------------------------------------------------------

    An automobile manufacturer recently suggested to EPA that certain 
detergent additives used in the gasoline portion of E85 may contribute 
to the formation of deposits in vehicle fuel pumps in E85 vehicles.\14\ 
EPA requests comment on the potential incompatibility of certain 
detergent additives for use in the gasoline portion of E85, and on the 
regulatory measures which EPA could take to address such a problem if 
it exists. Comment is also requested on the extent to which such 
concerns might also apply to M85 fuel.
---------------------------------------------------------------------------

    \14\Memorandum entitled: ``Phone Call with Gerald Barnes of 
General Motors Regarding the Potential Incompatibility of Certain 
Detergent Additives for use in the Gasoline Portion of E85 fuel.'', 
Docket Item IV-E-34
---------------------------------------------------------------------------

D. Aftermarket Detergent Additives

    The proposed detergent certification program did not cover the use 
of aftermarket detergent additives.\15\ EPA requests comment on whether 
any requirements regarding the deposit control efficiency of 
aftermarket additives should be implemented, and the authority under 
which EPA could act to implement such requirements. Specifically, EPA 
requests comment on what requirements might be necessary on aftermarket 
additives to limit their potential contribution to the formation of 
CCD. This is of particular concern since EPA believes that the 
manufacturers of such aftermarket additives advise their use at very 
high concentrations to remedy PFID/IVD related problems. As discussed 
earlier, potential interim CCD control options involve limiting the 
concentrations used of detergents which have a tendency to form CCD.
---------------------------------------------------------------------------

    \15\Aftermarket detergent additives are marketed directly to 
gasoline consumers for addition directly into the vehicle fuel tank.
---------------------------------------------------------------------------

E. Detergent Additives Certified for Use in California Phase-2 
Reformulated Gasoline

    The interim detergent registration rule specified that data used to 
support the certification of detergent additives under California's 
detergent additive program, specifically for use in California Phase-2 
Reformulated Gasoline, will be acceptable for use in demonstrating the 
performance of detergents used to comply with Federal detergent 
gasoline requirements only for gasoline sold within the state of 
California. During a workshop on the interim requirements for deposit 
control additives, held by the American Petroleum Institute (API) on 
October 26, 1994, commenters from the audience suggested that EPA 
should allow California Phase-2 reformulated gasoline that is additized 
with detergents certified under California's detergent additive program 
to be sold outside of California. EPA agrees that this approach is in 
keeping with the goals of this program. The addition of such a 
provision would also allow improved flexibility in the gasoline 
distribution system.
    Therefore, EPA proposes to amend the provisions of the interim 
detergent rule to provide that gasoline additized within the state of 
California in accordance with the requirements of the California Air 
Resource Board's (CARB's) detergent rule may also be sold outside of 
the state of California. Specifically, EPA proposes to amend section 
80.141(e)(1) of the regulatory text to state that, under the interim 
program, CARB detergent certification data specific to California 
phase-2 reformulated gasoline will be accepted as adequate support of 
detergent effectiveness for detergent gasoline that is blended within 
the state of California in accordance with CARB's detergent program.

F. Allowed Variation in the Concentration of Detergent-Active 
Components Within a Single Registration for a Detergent Additive 
Package

    During the workshop on the interim requirements for deposit control 
additives held by API, commenters from the audience requested 
clarification of the allowed variation in the concentration of 
detergent-active components in a detergent additive package under a 
single additive registration. They stated that it is not possible to 
comply with the requirements of section 80.141(c)(2), which state that 
no variation is allowed in the concentration of any of the detergent-
active components under a single variation, given variability in 
manufacturing process. They further stated that upward variation in the 
concentration of these components should be allowed, since such 
variation would not compromise the additive's deposit control 
efficiency. Indeed, as suggested by these commenters, EPA intended the 
language of section 80.141(c)(2) to prevent downward variation in the 
concentrations of the detergent-active components from the minimum 
concentrations reported in the subject additive registration. It was 
not intended to prevent an upward variation in the concentration of 
these components. Therefore, to clarify this provision, EPA proposes to 
amend section 80.141(c)(2) to state that upward variation in the 
concentration of detergent active components would be permitted, 
provided that such variation is specified in the registration, and that 
such variability does not change the minimum recommended concentration 
reported to be necessary to control deposits.

G. Enforcement Issues

    EPA has several enforcement issues for which comments are requested 
prior to the issuance of the certification final rule. EPA believes 
these issues are integral to successful enforcement of the detergent 
program, and will be addressing these issues in the certification final 
rule.
1. Meters
    a. Required Use of Meters on Automated Additization Systems.
    In the NPRM, the proposed mass balance (now volumetric additive 
reconciliation, or ``VAR'') formulas provided for three different 
additization systems: hand blending, meters on every injector, and a 
metering system that did not have meters on every injector. EPA 
requested comments on whether the certification rule should require, 
for the purpose of measurement accuracy, that automated additization 
equipment be equipped with meters on every injector.
    EPA received no comments specifically addressing this issue. 
However, API commented that the final automated VAR formulas should be 
flexible enough to permit the use of presently existing automated 
additization equipment, which includes meters on every injector, 
metered systems without meters on every injector, and automated systems 
without any meters that measure detergent use through tank inventory 
gauging.
    The VAR formula in the interim program final rule does not require 
automated detergent blenders to have metered measurement systems, in 
deference to lead time concerns. However, as was proposed in the NPRM, 
EPA does intend to require some form of metered measurements for 
automated detergent blenders in the certification program final rule. 
EPA is still interested in receiving comments about the value of 
requiring systems with meters on every injector, as opposed to 
permitting the use of metered systems that measure the flow from many 
injectors on one meter.
    A metered measurement system will be required in the certification 
program final rule because meters are unquestionably a more accurate 
measurement system than tank inventory gauging. Meters measure the 
actual flow of product going through them, while inventory gauges 
merely determine drop in tank volume. As API commented in its analysis 
of automated additization system errors, inventory measurement systems 
are subject to error both in measuring inventory as well as in 
measuring additions to inventory. API's own member survey, submitted to 
EPA as comment, suggested that metered systems would need a smaller 
accuracy tolerance range than inventory systems, i.e., an 8 to 10 
percent tolerance based on metered system errors, as opposed to the 
greater 10 to 15 percent tolerance range based on inventory system 
errors. (Parenthetically, an enforcement tolerance was not provided in 
the interim program final rule and is not anticipated in the 
certification program final rule. This issue was addressed at length in 
the preamble to the interim program final rule.)
    The final detergent certification program rule, expected to be 
effective in mid-1996, will not have the lead time issues associated 
with implementation of the interim program. EPA therefore reconfirms 
its proposal that, under the final certification program, all automated 
detergent blenders using the automated formula be required to have 
metered measurements. Furthermore, to assure accuracy of the VAR 
measurements, EPA proposes that all metered systems must include meters 
on every injector. This more stringent proposal recognizes that not all 
metered systems can accurately establish whether gasoline is being 
appropriately additized. If the system merely has a master meter which 
measures, in the aggregate, the flow of detergent from the tank into a 
multitude of injectors, it is not possible to determine whether any 
particular injector is operating properly and dispensing the proper 
amount of detergent. To ensure that each injector is dispensing the 
appropriate detergent concentration, a meter would have to be installed 
on each injector.
    Obviously, EPA wants to fulfill its statutory mandate of ensuring 
the proper additization of gasoline to prevent deposits, and believes 
it is necessary to require all automated systems to be equipped with 
meters on every injector. However, the Agency is also concerned about 
the costs of such a requirement, and is therefore very interested in 
receiving comments about the number of automated blenders that would 
need to upgrade under such a requirement, the cost that would accrue to 
such blenders, as well as the aggregate cost. The Agency would then be 
interested in comparing these costs with the value received in 
additional additization accuracy expected to result from this 
requirement.
    b. Precision of Metered Measurements. The NPRM proposed that VAR 
detergent measurements for automated blenders be determined and 
recorded to one tenth of a gallon. API commented that detergent volume 
measurement should only be required to be recorded at the gallon level, 
since scaling back the determination to precision of a gallon would be 
necessary to accommodate all present additization systems. In deference 
to the concern that some parties would be unable to upgrade to systems 
with precision to at least one tenth of a gallon in time for the 
January 1, l995 implementation date of the interim program rule, the 
interim program requires VAR detergent measurements to be recorded only 
to the nearest gallon.
    EPA is now proposing that, under the certification program final 
rule, the VAR detergent measurements for automated blenders must be 
recorded to a precision of at least one tenth of a gallon. This 
provision is consistent with the proposal that all automated blenders 
using the automated VAR formula must be equipped with metered detergent 
measurement equipment. It is EPA's understanding that metered equipment 
is easily able to measure to the one tenth of a gallon precision 
standard, and it should not be difficult for regulated parties to meet 
this higher level of precision once the provisions of the certification 
program go into effect. Comments are requested about the measurement 
capabilities of metered measurement equipment and about the 
reasonableness of requiring recording accuracy to one tenth of a 
gallon.
2. The Use of a Per-Gallon Minimum Detergent Requirement within VAR 
Compliance Periods for Automated Detergent Blenders
    In the NPRM, EPA proposed that all detergent blenders conduct 
regular product reconciliations to determine the accuracy of their 
additizations. Hand blenders were proposed to perform this 
reconciliation on a per-batch basis, whereas automated blenders, not 
having the uniform capability to easily determine per-batch usage, were 
proposed to perform the reconciliations on a weekly basis.\16\ The 
detergent reconciliation proposed for automated blenders was, in 
effect, an averaging procedure, permitting automated blenders to 
calculate the accuracy of their additization throughout the compliance 
period on an averaged basis.
---------------------------------------------------------------------------

    \16\In fact, pursuant to comments, the interim program requires 
only a monthly reconciliation period for automated blenders.
---------------------------------------------------------------------------

    To limit the ability to average detergent usage, EPA also proposed 
in the NPRM that any intentional alteration of the detergent 
concentration within the compliance period would terminate the period 
and necessitate the start of a new period. Under this proposal, 
intentional compensation for under-additization discovered in the 
period would not be permitted. However, in its comments, API expressed 
concern that, without a significant enforcement tolerance to determine 
compliance with the VAR standard over the entire period, automated 
detergent blenders in some circumstances would be required to make 
concentration adjustments in order to stay in compliance. As explained 
in detail in the interim program final rule, EPA did not believe it was 
appropriate to create a VAR standard enforcement tolerance, but did 
decide to permit limited intentional compensation within the compliance 
period for discovered under-additizations. The permissible adjustment 
was limited to 10 percent above the concentration initially used in the 
period. The purpose of this limitation was to prevent blenders from 
attempting to compensate for substantial under-additizations (and thus 
avoid liability for such under-additizations) by means of intentionally 
over-additizing, to a significant degree, for the rest of the period.
    However, the Agency still has concerns that permitting averaging 
over the automated blender VAR compliance period will tolerate 
unacceptable amounts of per-gallon under-additization of gasoline. 
Depending on the amount of additized gasoline measured in the 
compliance period, it is clear that even the 10 percent adjustment 
limitation could permit some quantity of gasoline being non-additized 
or significantly under-additized. Even if serious under-additization 
occurred early in the period, the averaged compliance standard might 
still be attained by the blender through the permissible upward 
adjustment of the detergent concentration rate for the rest of the 
period.
    Naturally, EPA would like to ensure that all gasoline is additized 
at an effective detergent concentration rate, as established during 
certification testing. The VAR compliance procedures included in the 
interim program go a considerable way toward accomplishing this goal, 
but only on an average for the compliance period. Individual loads of 
gasoline may still be severely out of specification. Consequently, for 
the final certification program rule, EPA is proposing that, within the 
monthly compliance period for automated blenders, each load of the 
product must be additized at a rate that is at least 90 percent of the 
certified detergent treat rate. This 90 percent figure would allow for 
some lapse in equipment efficiency while providing assurance that each 
load in the averaging period will approximate the certification 
standard. The reasonableness of the proposed 90 percent minimum 
requirement is supported by on an API member survey which was submitted 
as a comment to EPA. This survey indicated that many blenders (73 per 
cent of proprietary systems, and 37 percent of non-proprietary systems) 
can apparently attain a monthly blending accuracy to within 10 per cent 
of the goal.
    The minimum per-gallon requirement within the averaged VAR 
compliance standard would not only provide greater assurance of across-
the-period accurate additization, but would also provide EPA with a 
useful enforcement tool. With this per-gallon minimum, EPA would be 
able to sample and test additized gasoline for detergent program 
compliance at the facilities of any regulated party. Under the interim 
program, on the contrary, the testing of additized product may reveal 
complete non-additization without necessarily establishing a violation, 
since there is no minimum per-gallon requirement under the interim 
period's VAR procedures. Although extensive sampling and testing of 
additized product is not presently contemplated, the ability to conduct 
such testing on a case-specific basis and to establish violations based 
on such testing would be very useful.
    EPA is requesting comments on the proposed per-gallon minimum 
requirement within the VAR compliance procedures. The Agency is 
particularly interested in learning whether the 90 percent minimum 
requirement is considered attainable, what problems would be associated 
with this minimum requirement, what records would be useful in 
monitoring compliance with the requirement, and what the costs would be 
in attaining compliance with a minimum standard. Comment is also 
requested on an alternate per-gallon minimum requirement, with 
rationale to support its adoption instead of a 90 percent minimum 
requirement. If commenters believe the per-gallon minimum concept is 
not appropriate, EPA requests suggestions for alternative means to 
assure greater additization accuracy throughout the VAR compliance 
period. One such alternative option being contemplated by the Agency is 
the previously proposed weekly compliance period, which would obviously 
permit less extensive averaging than the longer, monthly period.
3. Presumptive Liability for VAR Violations
    Under the interim program final rule, only detergent blenders will 
be held liable for VAR violations. However, EPA stated in the preamble 
to interim program rule that this issue may be revisited in the 
detergent certification final rule. Although some commenters asserted 
that detergent blenders should be the only parties liable for VAR 
violations, EPA believes that parties other than, or in addition to 
detergent blenders, could cause VAR violations. Therefore, limiting 
liability for VAR violations to detergent blenders may allow culpable 
parties to evade liability for VAR violations that they have caused.
    For example, a party may cause VAR violations by providing 
inaccurate or incomplete blending instructions to detergent blenders, 
or by conspiring with detergent blenders to sell under-additized 
gasoline that violates VAR requirements. The latter scenario could come 
about if a retailer attempted to save money by intentionally purchasing 
cheaper, unadditized or under-additized gasoline from a terminal which 
was intentionally violating VAR requirements to produce such gasoline. 
In this scenario, the collusion would occur because the retailer would 
benefit from the lower cost of its purchased gasoline, and the terminal 
would benefit from the sale of the misadditized product to the 
retailer.
    Further, the newly proposed per-gallon minimum requirement within 
the VAR procedures would make it possible for EPA to effectively 
discover VAR per-gallon minimum violations downstream, at retail 
outlets or at other regulated parties receiving product from the 
detergent blenders. When such violations are found, EPA needs the 
ability to hold these parties liable for such VAR violations. As shown 
in the preceding paragraph, it clearly could be in the financial 
interest of these parties to cause VAR violations by colluding with 
detergent blenders to sell VAR violating, less expensive gasoline.
    EPA believes that parties who cause VAR violations should not 
escape liability for those violations. Therefore, EPA is considering as 
an option under the detergent certification final rule, that all 
parties in the misadditized product's distribution system, including 
detergent blenders but excepting upstream carriers, will be 
presumptively liable for such VAR violations. As is typically the case, 
carriers upstream to where a violation is found will be liable for such 
violations if EPA can establish that they caused the violation.
    The rationale for imposing a presumptive liability scheme for VAR 
violations would be the same as that which supports presumptive 
liability for other violations in the detergent program, and for 
violations of other EPA fuels programs such as reformulated gasoline, 
gasoline volatility, and lead contamination. Typically, many parties 
handle and control gasoline, detergent and detergent-additized post-
refinery component, which are, often times, fungible products. Under 
these circumstances, it will be difficult for EPA to determine who, in 
addition to the detergent blender, might have caused the VAR 
violations. When multiple parties potentially may have caused a 
violation, the Agency needs a presumptive liability scheme to 
effectively enforce its regulatory program.
    All presumptively liable parties would have the right to establish 
an affirmative defence to such liability. EPA expects that the vast 
majority of VAR violations will be caused by detergent blenders, and in 
these cases, the blenders will be the only parties liable.
    EPA requests comments on the issue of presumptive liability for VAR 
violations, including comments regarding EPA's legal authority to 
impose such liability. Parties disagreeing with EPA's proposed approach 
are urged to submit alternative options to address the problem of 
multiple party causation of VAR violations. One alternative approach 
that EPA is considering, and is requesting comment on, is the 
imposition of an affirmative duty (with attendant penalties) on 
upstream parties transferring title to detergent or custody to those 
who will physically do the blending, to transfer accurate written 
blending instructions for use of that detergent. In addition, a new 
liability section would be created which would hold any regulated party 
liable for a VAR violation if EPA could establish that the party caused 
the violation.
4. Right of Entry to Inspect the Premises of Regulated Parties in the 
Detergent Distribution System
    Current regulations permit EPA to enter the premises of gasoline 
refiners, retailers, wholesale purchaser-consumers, distributors, and 
importers, and to make inspections, take samples, and conduct tests at 
such facilities to determine compliance with EPA requirements. See 40 
C.F.R. 80.4. EPA is today proposing to amend this provision to include 
manufacturers, distributors, and carriers of detergent additives.
    To investigate possible violations of detergent specifications 
promulgated under Sections 211(l) and 211(c), and to ensure compliance 
with those specifications, it is necessary for EPA to have the ability 
to collect samples of detergent additives at the facilities of the 
detergent manufacturer, distributor, and carrier. Section 211(l) 
directs EPA to establish, by regulation, specifications for detergent 
additives that must be used to prevent the accumulation of engine or 
fuel system deposits. EPA must be able to sample the detergent additive 
before it is blended into gasoline to ensure that manufacturers produce 
detergents that comply with EPA regulations, and to ensure that 
distributors and carriers of detergents do not take actions that alter 
detergents in such a way that the requirements are not met. Further, 
this approach is consistent with the currently existing provision in 
Part 80 under which EPA may enter the facilities of gasoline refiners 
and other parties, and make inspections, take samples, and conduct 
tests at such facilities to determine compliance with the requirements 
of that part.
    Therefore, under EPA's broad authority under Section 301(a) to 
promulgate such regulations as are necessary to carry out the Agency's 
functions under the Act, and EPA's authority under Sections 114, 
211(c), and 211(l), EPA is proposing to amend the currently existing 
right of entry provision in 40 C.F.R. 80.4 to include detergent 
manufacturers, distributors, and carriers. Also, see Dow Chemical Co. 
v. United States, 476 U.S. 227 (1986), recognizing EPA's broad 
authority to use reasonable means to carry out its investigatory 
function under the Act, including but not limited to Section 114. EPA 
invites comment on this proposed amendment.
5. The Use of Multiple Detergent Concentrations by Automated Detergent 
Blenders
    Under the proposed regulatory approach published in the NPRM, 
automated detergent blenders were not permitted to include several 
detergent concentrations in one VAR compliance record. Any change in 
concentration under the NPRM would have necessitated the start of a new 
VAR period. The purpose of this limitation was to prevent the attempt 
by blenders to compensate for significant under-additization within a 
compliance period by changing the rate of additization in the rest of 
the period.
    The Agency received a comment protesting this prohibition against 
the use of several detergent set rates in one VAR record. The commenter 
argued that this restriction would penalize blenders with sophisticated 
equipment that could automatically change set rates depending on the 
product being additized. According to this commenter, such a blender 
would be required to create a new VAR record every time his equipment 
automatically switched concentration rates to accommodate different 
grades of product being additized.
    EPA does not intend to discourage the use of sophisticated 
additization equipment by detergent blenders. Therefore, under the 
interim program EPA did not prohibit the use of several set rates on a 
single VAR record. However, blenders are required to meet the following 
conditions under the interim program to ensure additization accuracy: 
no concentration set rate can be set below the detergent's lowest 
additive concentration (LAC); each initial set rate used in the period 
must be recorded on the VAR record, along with the product to be 
additized with each rate; no adjustments to the initially set rates can 
be made in the period above 10 percent of the initial rate; the blender 
must maintain records of all adjustments to the set rates.
    These requirements protect additization accuracy, while at the same 
time, provide industry with the flexibility to use sophisticated 
additization equipment capable of automatically varying set rates. The 
Agency continues to believe that these VAR requirements are useful. 
However, further restrictions are necessary to ensure additization 
accuracy under the detergent certification final rule.
    Under the interim program, automated blenders can measure on a 
single VAR record product being additized by the same detergent under 
different set rates, as long as they meet the conditions stated above. 
Since VAR compliance under the interim program is an averaged 
compliance, there is no way to determine under this procedure if the 
product being additized at the lower rate is actually meeting its LAC 
requirements, because the product being additized at the higher rates 
would conceal any failure of the lower product to meet the averaged 
standard. This procedure of multi-rate averaging is thus not conducive 
to assuring the additization of all product.
    Consequently, EPA is proposing that, under the detergent 
certification final rule, automated blenders may continue to blend 
their detergent at multiple rates in one VAR period, but product being 
blended at each set rate must have its own VAR record. This would 
ensure that product additized at each rate meets the LAC standard. 
Further, the Agency is also proposing that a blender choosing to use 
different set rates within a single VAR period must have equipment that 
can accurately measure detergent use at each set rate, (e.g. meters 
which can switch to measuring different rates, meters on individual 
injectors measuring different rates, or other effective systems). This 
latter requirement will ensure that the accuracy of each set rate's 
additization can be effectively measured, not merely estimated or 
assumed.
    Comments are requested about these proposed changes to automated 
blender VAR requirements. Specifically requested is information about 
the cost associated with implementation of these changes, such as the 
number of blenders they would impact, the number of blenders that would 
be forced to upgrade their equipment to accommodate these changes, and 
the cost of such upgrades. If commenters are opposed to the 
implementation of these changes, they are urged to submit alternative 
plans to address the problem of ensuring proper additization of product 
additized at lower concentrations when multiple concentrations are 
being measured.
6. Regulation of Imported Additized Gasoline
    Under the NPRM, importers of gasoline are regulated parties, 
subject to the requirements of the detergent rule. They are thus 
subject to product transfer document requirements and the prohibitions 
against the transfer and sale of nonconforming product. They are 
permitted to import either additized or nonadditized product. If they 
additize the gasoline they import, they are considered detergent 
blenders, and are subject to VAR requirements for the gasoline. If they 
import additized gasoline but are not detergent blenders, they are not 
required to create and maintain VAR records establishing VAR 
compliance. However, if the product they import is sampled and shown to 
be under- additized, the importer would be presumptively liable for the 
violation. As a practical matter, it would be extremely difficult for 
the importer to effectively meet its affirmative defence requirements, 
unless it would provide EPA with the VAR records documenting proper 
additization of the product.
    The Agency believes the control over imported gasoline needs to be 
strengthened in regard to gasoline which is additized prior to 
importation. As was extensively discussed in the NPRM, enforcement of 
the detergent regulation will primarily be based on record review, 
since sampling and testing of additized gasoline for compliance will be 
difficult. Agency review of VAR records to determine accurate 
additization is thus critical to the success of the program. It is just 
as important for EPA to be able to review VAR records for imported 
gasoline which is additized, as it is to review those records for 
domestic product, given the sampling and testing limitation. Under the 
interim program, however, Agency review of VAR records for imported 
additized gasoline will be limited to those situations in which 
violations have already been discovered through testing or otherwise. 
This means that, in regard to imported gasoline, the Agency is deprived 
of its prime measures to determine detergent program violations.
    To correct this deficiency, EPA is proposing that the detergent 
certification final rule will amend the definition of detergent blender 
to include importers of additized gasoline within the definition. If 
importers choose to import additized product, they must be held 
accountable for its additization accuracy as detergent blenders. If 
not, those selling domestically additized gasoline will be at a 
disadvantage to importers, since VAR records establishing violations 
will not be generally available for EPA review. Including importers of 
additized product within the definition of detergent blender will make 
the importers responsible for the creation and maintenance of the VAR 
records verifying additization accuracy, as well as with compliance 
with the other VAR requirements that apply to domestically additized 
product. EPA requests comments about the proposed amendment of the 
definition of detergent blender.

IV. Public Participation

    EPA has encouraged full participation of the regulated industry and 
other interested parties in the development of the rule to implement 
the statutory requirements and continues to do so. A public workshop 
was held on February 13, 1992 to initiate open discussion of the 
relevant issues and EPA met with numerous industry representatives 
separately to obtain their input.
    The Notice of Proposed Rulemaking (NPRM) was published on December 
6, 1993 (58 FR 64213) and a public hearing was held in Ann Arbor, 
Michigan on January 11, 1994. Comments on the NPRM were accepted until 
March 11, 1994. EPA received 31 written comments on the NPRM before the 
close of the initial comment period and has received additional 
comments since that time particularly with respect to the issues 
discussed in this notice. EPA encourages additional comment on the 
issues discussed in this notice through the close of the current 
comment period. (For the date on which the comment period closes see 
the Dates section in this notice.) EPA does not intend to hold an 
additional public hearing to discuss the issues raised in this notice.

V. Electronic Copies of Rulemaking Documents

    Electronic copies of this notice, and other documents associated 
with the detergent certification rule are available on the Office of 
Air Quality Planning and Standards (OAQPS) Technology Transfer Network 
Bulletin Board System (TTNBBS). Instructions for accessing TTNBBS and 
downloading the relevant files are described below.
    TTNBBS can be accessed using a dial-in telephone line (919- 541-
5742) and a 1200, 2400, or 9600 bps modem (equipment up to 14.4 Kbps 
can be accommodated). The parity of the modem should be set to N or 
none, the data bits to 8, and the stop bits to 1. When first signing on 
to the bulletin board, the user will be required to answer some basic 
informational questions to register into the system. After registering, 
proceed through the following options from a series of menus:

(T) GATEWAY TO TTN TECHNICAL AREAS (Bulletin Board)
(M) OMS
(K) Rulemaking and Reporting
(3) Fuels
(4) Detergent Additives

    At this point, the system will list all available files in the 
chosen category in chronological order with brief descriptions. The 
following five ``.ZIP'' files are currently available:

``DCA__PRE.ZIP'' (Preamble from the Notice of Proposed Rulemaking)
``DCA__IFP.ZIP'' (Preamble to the final rule on the Interim 
Requirements for Deposit Control Additives)
``DCA__IFR.ZIP'' (Regulatory text for the final rule on the Interim 
Requirements for Deposit Control Additives)
``DCA__RIA.ZIP'' (Regulatory Impact Analysis)
``DCA__RCN.ZIP'' (Notice to Reopen the Comment Period)

    File information can be obtained from the ``READ.ME'' file. Choose 
from the following options when prompted:

ownload, 

rotocol, xamine, ew, ist, elp or to exit. To download a file, e.g., filename.ZIP, the user needs to choose a file transfer protocol appropriate for the user's computer from the options listed on the terminal. The user's computer is then ready to receive the file by invoking the user's resident file transfer software. Programs and instructions for de-archiving compressed files can be found under ystems Utilities from the top menu, under rchivers/de-archivers. Please note that due to differences between the software used to develop the document and the software into which the document may be downloaded, changes in format, page length, etc. may occur. TTNBBS is available 24 hours a day, 7 days a week except Monday morning from 8-12 EST, when the system is down for maintenance and backup. For help in accessing the system, call the systems operator at 919-541-5384 in Research Triangle Park, North Carolina, during normal business hours EST. V. Statutory Authority The statutory authority for the regulation of gasoline detergent additives is granted to EPA by sections 211 (c) and (l) of the Clean Air Act, as amended, 42 U.S.C. 7414, 7545 (c) and (k), and 7601. List of Subjects in 40 CFR Part 80 Environmental protection. Fuel additives, Gasoline detergent additives, Gasoline motor vehicle pollution, Penalties, Reporting and recordkeeping requirements. Dated: December 19, 1994. Richard D. Wilson, Acting Assistant Administrator for Air and Radiation. [FR Doc. 94-31815 Filed 12-27-94; 8:45 am] BILLING CODE 6560-50-P