[Federal Register Volume 63, Number 98 (Thursday, May 21, 1998)]
[Proposed Rules]
[Pages 28195-28200]
From the Federal Register Online via the Government Publishing Office [www.gpo.gov]
[FR Doc No: 98-11750]


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

ENVIRONMENTAL PROTECTION AGENCY

40 CFR Part 75

[FRL-6007-7]
RIN 2060-AH64


Acid Rain Program: Determinations under EPA Study of Bias Test 
and Relative Accuracy and Availability Analysis

AGENCY: Environmental Protection Agency (EPA).

ACTION: Notice of proposed determinations and proposed rulemakings.

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

SUMMARY: Title IV of the Clean Air Act Amendments of 1990 (the Act) 
authorizes EPA to establish a program to reduce the adverse effects of 
acidic deposition. The Act requires electric utilities affected by the 
Acid Rain Program to install continuous emission monitoring systems 
(CEMS) to measure emissions of sulfur dioxide (SO2), 
nitrogen oxides (NOX), and carbon dioxide (CO2). 
On January 11, 1993, Continuous Emission Monitoring regulations were 
published. They established procedures and requirements for installing, 
certifying, operating, and quality assuring CEMS at Acid Rain affected 
utility units. In response to comments and litigation from 
representatives of the electric utility industry and environmental 
advocacy groups, provisions were incorporated in the CEMS regulations 
requiring EPA to conduct studies, reach determinations, and, if 
necessary, initiate rulemakings on the appropriateness of retaining or 
revising three elements in the CEMS regulations: the bias test, 
relative accuracy test, and the availability trigger conditions of the 
Missing Data Substitution Procedure. This Notice of Proposed Rulemaking 
presents EPA's proposed determinations and consequent proposed rule 
revisions.

DATES: Comments. Comments on the proposed determinations and rule 
revisions must be received on or before July 6, 1998.
    Public Hearing. Anyone requiring a public hearing must contact EPA 
no later than June 1, 1998. If a hearing is held, it will take place 
June 5, 1998, beginning at 10:00 a.m.

ADDRESSES: Comments. All written comment must be identified with the 
appropriate docket number (Docket No. A-97-56) and must be submitted in 
duplicate to EPA Air Docket Section (6102), Waterside Mall, Room M1500, 
1st Floor, 401 M Street, SW, Washington, D.C. 20460.
    Public Hearing. If a public hearing is requested, it will be held 
at the Environmental Protection Agency, 401 M Street, SW, Washington, 
D.C. 20460, in the Education Center Auditorium. Refer to the Acid Rain 
homepage at www.epa.gov/acidrain for more information or to determine 
if a public hearing has been requested and will be held.
    Docket. Docket No. A-97-56, containing supporting information used 
to develop the proposed determinations and rule revisions is available 
for public inspection and copying from 8:00 a.m. to 5:30 p.m., Monday 
through Friday, excluding legal holidays, at EPA's Air Docket Section 
at the above address.

FOR FURTHER INFORMATION CONTACT: Elliot Lieberman at (202) 564 9136, 
Acid Rain Division (6204J), U.S. Environmental Protection Agency, 401 M 
St., S.W., Washington, D.C. 20460; or the Acid Rain Hotline at (202) 
564 9620. Electronic copies of this notice and technical support 
documents can be accessed through the Acid Rain Division website at 
http://www.epa.gov/acidrain.

SUPPLEMENTARY INFORMATION:

I. EPA Studies Under 40 CFR 75.7
    A. Background
    B. Collaborative Field Study
    C. Certification Test Study
    D. Proposed Findings and Conclusions
II. EPA Analyses in Response to 40 CFR 75.8
    A. Background
    B. Relative Accuracy
    C. Availability Trigger Conditions for Missing Data Substitution 
Procedure
III. Proposed Rule Revisions
IV. Administrative Requirements
    A. Executive Order 12866
    B. Unfunded Mandates Act
    C. Paperwork Reduction Act
    D. Regulatory Flexibility

I. EPA Studies Under 40 CFR 75.7

A. Background

    To ensure a consistent level of precision and accuracy in the 
emission measurements obtained across the Acid Rain Program, Part 75 of 
the Acid Rain regulations requires a series of performance tests to be 
conducted on each CEMS both at initial certification and periodically 
thereafter. Among the required performance tests is the relative 
accuracy test audit (RATA) in which a minimum of nine simultaneous 
measurements are taken from a unit's installed CEMS and an EPA approved 
reference method. The paired RATA data are then subjected to two 
statistical tests: The relative accuracy test, which establishes the 
degree of accuracy of the CEMS relative to the reference method; and 
the bias test, which uses a t-statistic to determine if the CEMS 
measurements are consistently lower than the reference method 
measurements. See 40 CFR Part 75, Appendix A and B.
    As stated in the preamble of the January 1993 regulations, EPA 
found that ``both statistical theory and field test results show that 
the bias test is a sound and effective statistical procedure for 
detecting consistent measurement error in the long-term operation of a 
CEMS'' (58 FR 3590, 3627 (1993)). However, at the time of promulgation 
of the Acid Rain regulations, although utilities had extensive 
experience with the relative accuracy test, they had virtually no 
previous experience with the bias test. This unfamiliarity led to 
several concerns with the bias test. Thus, the January 1993 regulations 
committed EPA to conduct field studies to determine ``whether there are 
statistically significant variances'' in the EPA-approved reference 
methods that utilities use to test the performance of the CEMS 
installed under the Acid Rain Program and ``whether the bias test 
should be adjusted to compensate for statistical variances in the 
reference method'' (58 FR 3628).
    In particular, EPA was required to:
    1. Investigate whether there are statistically significant 
variances in the EPA reference methods (Issue #1);
    2. Distinguish between the variability in reference monitor 
readings attributable to measurement error and the variability due to 
the choice of reference monitor among those certified by the Agency 
(Issue #2);
    3. Investigate possible differences in bias test failure rates by 
emission levels (Issue #3); and
    4. Assess whether any adjustments are necessary to properly 
determine measurement bias (Issue #4).
    The regulations called for the completion of a study addressing 
these issues by October 31, 1993. In response, EPA conducted two 
studies. The first was a collaborative field study, involving four 
independent reference method test teams, at Big Rivers Electric 
Corporation's Green Generating Station, Unit 2, in Sebree, Kentucky. 
This location was specifically selected for testing because its 
relatively low range of SO2 emission concentrations (from 56 
ppm to 231 ppm) would allow EPA to examine bias test failure rates at 
SO2 emission levels different from those prevailing in 
previous field studies and consider an industry concern that

[[Page 28196]]

contradictory bias test results were more likely to occur at low, than 
at high, emission concentrations. Field work for this study was 
completed from August 16-31, 1993. Separate data summary (Docket Item, 
A-97-56, II-A-1) and statistical analyses reports summary (Docket Item, 
A-97-56, II-A-2) were completed in March 1994 and September 1996 
respectively.
    The second study involved collection and analysis of bias test 
results from the field tests conducted by affected units under Part 75 
for certification of their CEMS. The certification test data, including 
the bias test, were submitted to EPA from November 1993 to September 
1996. The study results reported here (and contained in Docket Item, A-
97-56, II-A-3) were available in 1997 only after the CEMS at the 
majority of both Phase I and Phase II (lower emitting) units had been 
received and certified by EPA.

B. Collaborative Field Study

    In the collaborative field study at Unit 2 of Big Rivers Electric 
Corporation's Green Generating Station (``Green Unit 2''), four labs 
(i.e., test teams) simultaneously performed Reference Methods 6C (for 
SO2), 7E (for NOX), and 3A (for CO2). 
To test the two general monitoring technologies available for 
performing the reference methods, two of the teams used ``wet-basis'' 
sampling techniques and two used ``dry-basis'' techniques. In the 
``wet-basis'' sampling techniques, a dilution probe is used to extract 
a diluted sample of the effluent from the stack gas. The diluted gas 
sample is then analyzed using an ambient-level analyzer (e.g., pulsed 
fluorescence for SO2, chemiluminescence for NOX, 
and infrared absorption for CO2), which does not require 
removal of moisture from the gas sample. In the ``dry-basis'' sampling 
techniques, a gas sample is extracted from the effluent stream without 
dilution. Moisture is condensed from the gas sample and the resulting 
dry sample is then analyzed using a source-level analyzer (infrared or 
ultraviolet for SO2, chemiluminescence for NOX, 
and infrared for CO2).
    Seventy-two runs of usable data (out of 76 total runs) were 
collected by the four labs. Concurrent measurements were also collected 
from Green Unit 2's SO2, NOX, and CO2 
continuous emissions monitoring systems, previously certified under the 
Acid Rain Program. On 36 of the runs, each lab and the unit's CEMS used 
separate calibration gases as required under 40 CFR Part 75. On the 
other 36 runs, all labs and the plant's CEMS shared common gases when 
calibrating.
    Issues #1 and #2 involve evaluation of the sources of variability 
inherent in EPA's reference methods. In the consideration of these two 
issues only the reference method measurements were analyzed, not the 
unit's CEMS. Issues #3 and #4 involve a comparison of the CEMS and the 
reference method measurements to determine if bias (systematic error) 
is detected in the CEMS measurement. In the consideration of these two 
issues, the unit's CEMS measurements were paired with each of the four 
lab's concurrent reference method measurements. This produced four sets 
of concurrent Relative Accuracy Test Audits (RATA's) which could be 
used in evaluating bias test result consistency across the four labs.
    To address the first two issues concerning the sources and extent 
of variability inherent in the reference methods, the collaborative 
field study employed an experimental design (technically known as a 
``randomized complete block design'') which allowed the quantification 
of the relative variability associated with (i) among-laboratory 
variation, (ii) variation between monitoring technologies (i.e., ``wet-
basis'' or ``dry-basis'' sampling techniques), (iii) the variability 
associated with different calibration gas scenarios (i.e., separate or 
shared calibration gases), and (iv) random error.
    Applying an analysis of variance (ANOVA) statistical procedure to 
the field study data, EPA found that the overall variation in the 
reference methods, considering all the monitoring technologies and 
calibration gas scenarios, was 2.93%RSD (Relative Standard Deviation) 
for SO2, 2.01%RSD for NOX, and 1.59%RSD for 
CO2. Reference method variations below approximately 3%RSD 
are consistent with the findings of an earlier collaborative field 
study, reported in Docket Item, A-97-56, II-A-5, where variations of 
1.4%RSD and 2.9%RSD were found for SO2 and NOX 
respectively. (The variation for CO2 is not available from 
the earlier study since that study did not include CO2 
reference method measurements.) Based on these findings, with respect 
to Issue #1 EPA believes that the statistically significant variances 
in the EPA reference methods are small.
    The analysis in the most recent collaborative study also revealed 
that the range in the Relative Standard Deviation due to the choice of 
reference method monitor (i.e., different analyzers using ``wet-basis'' 
or ``dry-basis'' technology) among allowable reference method 
technologies was very small (below 1%RSD) whether the labs used 
separate or shared calibration gases. Consequently, EPA believes with 
respect to Issue #2 that the variability due to the choice of reference 
method monitor among those available is very small.
    As noted earlier, Issues #3 and #4 require consideration of 
simultaneous measurements by the unit's CEMS along with the four test 
labs. To respond to Issue #3, concerning the consistency of the bias 
test results, the field test data were analyzed to determine how much 
agreement was found among the four labs as to whether the CEM was 
biased or not biased when current provisions of Part 75 are followed. 
In particular, the consistency in bias test results was evaluated by 
counting the number of concurrent RATA's in which agreement among the 
four test teams was 100% (all four labs agree), 75% (three out of four 
labs agree) and 50% (two labs find bias and two find no bias). For each 
pollutant there was never less than 75% agreement among the test teams 
when the reference methods and the installed CEMS were each calibrated 
using independently selected calibration gases, as is required under 40 
CFR Part 75. For NOX and CO2 there was always 
100% agreement. For SO2 there was 100% agreement in bias 
test results in more than 76% of the concurrent RATA's.
    These test results lead EPA to believe that even at a site 
exhibiting low SO2 emission concentrations, there is a high 
degree of consistency in bias test results.

C. Certification Test Study

    To respond further to Issue #3, EPA analyzed the consistency in 
bias test results across the universe of affected units, by conducting 
a study of the bias test results for all CEMS for which certification 
tests data were submitted under Part 75 between November 1993 and 
September 1996. To see how test results were affected by emission 
levels, the pass/fail rates at different concentrations 
(SO2) and emission rates (NOX) were compared for 
1023 SO2 and 1293 NOX bias tests submitted under 
the Acid Rain Program. This analysis was not performed on 
CO2 monitors, because under Part 75 units are not required 
to perform the bias test on their CO2 monitors.
    Grouping monitors according to the average concentration level (for 
SO2 CEMS) and average emission rate (for NOX 
CEMS), reported by the CEMS during the RATA, the pass/fail rates were 
plotted at regular increasing SO2 emission concentration 
levels and NOX emission rates. The resulting graphs revealed 
that the percentage of passes and fails remained relatively consistent

[[Page 28197]]

across concentration and emission rate categories. For example, for all 
SO2 monitors, 73% (750 out of a total of 1023 monitors) 
passed the bias test. Assigning each tested monitor to one of fourteen 
100 ppm categories, beginning at 0-100 ppm and ending at above 1300 
ppm, showed that the percent of passing monitors in all but three of 
the concentration categories fell between 70 and 90%. The three 
categories whose passing rates were outside this range were 400-500 ppm 
(56% passing), 600-700 ppm (69%), and above 1300 ppm (63%). Thus, there 
was little or no apparent correlation between concentration level and 
bias test failure rates.
    The graphical analysis for SO2 monitors was confirmed by 
calculation of the r-squared value for the data. The r-squared value is 
a measure of the strength of the linear relationship between two data 
sets. R-squared can take on values from zero to one. A high r-squared 
value, i.e., closer to 1 than to 0, would suggest that the bias test 
pass/fail rate is highly correlated with the emission concentration 
level, e.g., that bias test failure is more likely with low emission 
concentration as suggested by utilities. A low r-squared value, i.e., 
closer to 0 than to 1, would suggest the absence of correlation between 
the bias pass/fail rate and the emission concentration level. For the 
plotted SO2 data, the r-squared value was low: 0.0109.
    The same graphical and statistical analysis was performed on the 
certification test data submitted for NOX CEMS. Bias test 
pass/fail rates for 1293 NOX monitors were divided into 
sixteen 0.1 lb/mmBTU categories. Considering all these categories, 67% 
(866 out of a total of 1293 monitors) passed the bias test. A plot of 
the data by emission category showed the bias test passing rate fell 
between 65% and 85% in all but of 3 of the 16 NOX emission 
categories. The three emission rate categories whose passing rates were 
outside this range were not correlated to the measured NOX 
emission rate: 0.1-0.2 lb/mmBTU (47% passing), 0.4-0.5 lb/mmBTU (59%), 
and 1.4-1.5 lb/mmBTU (50%). Again, there was little or no apparent 
correlation between bias test pass/fail rates and emission rate, and 
this was confirmed by the statistical analysis. The r-squared value for 
the NOX data was low: 0.1109.
    Thus, the graphical and statistical analysis performed in the 
certification test study indicates consistent bias test results across 
emission levels.

D. Proposed Findings and Conclusions

    Based on the analyses performed to address Issues #1-3 in the 
collaborative and certification field studies, EPA considered Issue #4, 
concerning the necessity and feasibility of adjustments to the bias 
test. EPA currently believes that the small variability in the 
reference methods (less than 3%RSD across all gas scenarios and monitor 
technologies) indicates that there is very low probability that a 
continuous emission monitoring system will fail the bias test for 
reasons other than the presence of true measurement bias in the CEMS. 
The high level of consistency in bias test results seems to support 
this view.
    Based on these studies, EPA proposes to find that:
    1. The variability attributable to measurement error and to the 
choice of reference monitor technology in the Agency's approved test 
methods for SO2, NOX, and CO2 is low 
(below 3.0% Relative Standard Deviation).
    2. Differences in measurement variability among different allowable 
reference method technologies are small (below 1.0% RSD).
    3. There is a high occurrence of consistency in bias test results.
    4. There is no evidence that bias test failure rates are 
significantly influenced by emission levels.
    Documentation of these proposed findings can be found in four 
docket items: A Collaborative Field Evaluation of EPA Test Methods 6C, 
7E, and 3A (March 1994) (Docket Item, A-97-56, II-A-1) gives a detailed 
description of the collaborative field test activities, site 
characteristics, and equipment employed, presents data obtained in the 
field study, and discusses preliminary findings on the variability of 
the reference methods. A second report, An Operator's Guide to 
Eliminating Bias in CEM System (November 1994) (Docket Item, A-97-56, 
II-A-6) is an independent technical guidance document advising 
environmental technicians on procedures for detecting and correcting 
engineering problems that could produce measurement bias in CEM 
systems. A third report, Statistical Analysis of Reference Method 
Variability and Bias Test Consistency in the Collaborative Field Study 
of EPA Test Methods 6C, 7E, and 3A at Big Rivers Electric Corporation, 
Green Generating Station, Unit 2 (September 1996) ((Docket Item, A-97-
56, II-A-2), focuses on the analysis of the collaborative study field 
data, reports the results of this analysis with respect to the four 
issues that the study was designed to address, and, based on this 
analysis, makes recommendations concerning whether adjustments are 
needed to the bias test. Finally, the graphs and supporting data from 
the certification test study can be found in ``Bias Test Pass/Fail 
Rates at Different SO2 and NOX Emission Levels as 
Reported in Certification Relative Accuracy Test Audits (RATA's) 
submitted through September 1996 under 40 CFR Part 75.'' (December 
1997) (Docket Items, A-97-56, II-A-3 and II-A-4).
    Based on the proposed findings enumerated above, EPA proposes to 
determine that adjustments to the equations in the bias test are 
technically unnecessary to properly determine measurement bias. EPA 
therefore proposes not to initiate a rulemaking to change the bias test 
under Sec. 75.7.

II. EPA Analyses in Response to 40 CFR 75.8

A. Background

    In accordance with a settlement agreement, signed on April 17, 1995 
in Environmental Defense Fund v. Browner, No. 93-1203 and consolidated 
cases (D.C. Cir., 1993), which addressed various CEMS issues, Sec. 75.8 
was adopted as part of the direct final rule, dated May 17, 1995, 
amending the January 11, 1993 rule's CEM provisions. Section 75.8 
required EPA to evaluate the appropriateness of the current relative 
accuracy and availability trigger conditions for missing data 
substitution for SO2, NOX, and CO2 
CEMS and flow monitors. This evaluation was to be based on initial 
certification test data and quarterly report data for the 1993-1996 
period. Using the evaluation, EPA was to determine whether to retain 
the current specifications or propose alternative performance 
specifications. A report evaluating this data was to be prepared by 
July 1, 1997, and EPA is to issue either a notice determining that the 
current rule provisions are appropriate or a notice proposing 
revisions. Any proposal revising the current rule is to be issued by 
October 31, 1997 and finalized by October 31, 1998. The results of 
EPA's evaluations of the current relative accuracy and availability 
trigger conditions are described below.

B. Relative Accuracy

    Relative accuracy is a statistical indicator of how closely the 
measurements by an installed CEM approximate those obtained by a 
concurrently used EPA reference method during a 9-12 run field 
demonstration (known as the relative accuracy test audit (RATA)) that 
must

[[Page 28198]]

be performed periodically for each CEMS under Part 75. Relative 
accuracy is expressed as a percent deviation of the CEMS results from 
the reference method results. The lower the relative accuracy value for 
a CEMS, the closer its measurements are to the reference method. Under 
40 CFR Part 75, Appendix A, Sec. 3, and Appendix B, Sec. 2.3.1, all 
SO2, NOX, and CO2 CEMS are required to 
have in a RATA a relative accuracy of 10%. Those that have a superior 
relative accuracy of 7.5% or less have one year to undergo their next 
RATA. Those that have a relative accuracy equal to or less than the 
required 10% but greater than 7.5% must undergo their next RATA within 
six months. The tighter specification of 7.5% is referred to as the 
``reduced frequency standard,'' while the 10% specification is known as 
the ``normal frequency standard.'' For flow monitors the normal 
frequency standard is 15%, while the reduced frequency standard is 10%. 
On January 1, 2000 the normal and reduced frequency standards for flow 
monitors will be lowered to correspond to the standards for the 
pollutant CEMS, i.e., 10% and 7.5% respectively.
    The evaluation of initial certification test data submitted for 
1993-1996 showed that the average relative accuracy was 3.42% for the 
965 SO2 CEMS installed under the Acid Rain Program, 3.62% 
for 1272 NOX CEMS, 3.28% for 1097 CO2 CEMS., and 
6.88% for 1070 flow monitors. This means that for all pollutants and 
flow, the average relative accuracy was below the reduced frequency 
standard. Furthermore, 91.3% of all SO2 CEMS, 94.1% of all 
NOx CEMS, 96.3% of all CO2 CEMS, and 91.9% of all flow 
monitors met their respective reduced frequency standard. See Docket 
Item, A-97-56, II-A-7 for a complete analysis of the certification test 
relative accuracy results.
    A similar evaluation was performed on the relative accuracy test 
results reported in quarterly reports for the 1994-1996 period. This 
analysis showed that the average relative accuracy over the three years 
of data was 3.49% on 2802 SO2 RATAs, 3.67% on 3935 
NOX RATAs, 3.06% on 2736 CO2 RATAs, and 5.78% on 
3019 flow RATAs. Like the certification test results, the data in the 
quarterly reports indicate that for each type of monitor, the average 
relative accuracy was below the reduced frequency standard. In 
addition, on 96.2% of the SO2 RATAs, 96.0% of the 
NOX RATAs, 97.9% of the CO2 RATAs, and 93.5% of 
flow RATAs, the monitors met their respective reduced frequency 
standard. A complete analysis of the quarterly report relative accuracy 
test results can be found in Docket Item, A-97-56, II-A-8.
    The relative accuracy test results obtained by these installed CEMS 
imply that no appreciable improvement in achieved relative accuracies 
could be expected unless the relative accuracy standard were brought 
down to or below these currently achieved average relative accuracies. 
However, studies cited above (Docket Item, A-97-56, II-A-2 and II-A-5) 
of the variability of the reference methods for SO2, 
NOX, and CO2 suggest that such reduced relative 
accuracy standards might be beyond the technological limits of current 
monitoring technology since they approach the variability inherent in 
the reference methods themselves. Thus, tightening the relative 
accuracy standards further for these CEMS is unlikely to produce a 
corresponding improvement in the achievable relative accuracy.
    Moreover, the existing regulations already provide that the normal 
and reduced frequency relative accuracy standards for flow monitors 
will be tightened to the same levels as for the other CEMS beginning in 
the year 2000. In light of the already low average relative accuracy 
(reflecting high monitor accuracy) for flow monitors, there is little 
or no basis at this time for concluding that any further tightening 
would be appropriate. In addition, EPA believes that the results of the 
tightening in 2000 should be evaluated before any further tightening is 
contemplated.
    Therefore, based on the evaluation required under Sec. 75.8, the 
Agency proposes to conclude that the current performance specifications 
for relative accuracy are appropriate at this time.

C. Availability Trigger Conditions for Missing Data Substitution 
Procedure

    In 40 CFR 75.30-75.38 (Subpart D) a missing data procedure is 
prescribed for calculating emissions when valid data are not being 
supplied by a unit's continuous emissions monitoring system. The 
missing data procedure is a multi-tiered computational routine for 
deriving a substitution value from values previously recorded, or the 
highest potential values, by the monitor. The procedure is based on the 
premise that the lower the annual monitor availability and/or the 
longer the gap in recorded data, the more conservative the value to be 
substituted.
    In concert, two trigger conditions determine the conservativeness 
of the substituted value. The first trigger condition is annualized 
monitor availability, i.e., the percentage of the immediately preceding 
8760 unit operating hours in which valid, quality assured data was 
obtained. The second trigger condition is the length of the current 
period during which valid data are not being produced. Current 
availability trigger conditions include three tiers: (1) less than 90% 
availability, (2) equal to or greater than 90% but less than 95% 
availability, and (3) 95% or greater availability.
    To determine if retaining the current availability trigger 
conditions is appropriate, the Agency analyzed the annual percent 
monitor availability (PMA) as reported in the 1994-1996 quarterly 
emission reports. The PMA indicates the proportion of the operating 
hours in each year that the monitor was providing valid, quality 
assured measurements. High PMAs would indicate that current trigger 
conditions are providing a sufficient incentive for keeping monitors 
operating properly.
    The evaluation of the quarterly report data for 1994-1996 showed 
that the average PMA for SO2 CEMS was 94.7% in 1994, 96.7% 
in 1995, and 97.2% in 1996. For the same three year period it was 
91.8%, 94.1%, and 95.8% for NOx CEMS, and 95.0%, 96.3%, and 97.0% for 
flow monitors. As a rule, separate percent monitor availabilities for 
the CO2 CEMS are not routinely reported, since 
CO2 CEMS usually serve as diluent components in 
NOX systems. However, the average PMA for CO2 CEMS in a 
given year must be at least as good as the corresponding average of the 
reported NOX PMAs. Not only are the average PMAs above the 
95% availability trigger level, but they have also consistently 
increased in each successive year of the Acid Rain Program. To 
appreciably improve monitor availabilities would require increasing the 
third tier availability trigger up to or above the high average 
availabilities currently being achieved. EPA believes that such an 
increase in the required availabilities would be close to or beyond the 
limits of what is reasonable to expect from current CEMS technology 
when properly operated under the conditions prevailing in utility 
stacks. A complete summary of the PMA's submitted in the 1994-1996 
quarterly reports can be found in Docket Item, A-97-56, II-A-9.
    Moreover, any tightening of the availability trigger conditions 
would require reprogramming of most affected units' data acquisition 
and handling systems, which automatically calculate and record the 
appropriate substitution values for periods when valid CEMS data are 
not available. Given the current high levels of monitor availability, 
there is little or no basis for finding that adjusting the trigger 
conditions would improve availability sufficiently to

[[Page 28199]]

justify the reprogramming costs that such a change would impose.
    Therefore, based on the evaluation required under Sec. 75.8, the 
Agency proposes to determine that retaining the current performance 
specifications for availability trigger conditions is appropriate at 
this time.

III. Proposed Rule Revisions

    Having completed the studies and evaluations required in 40 CFR 
75.7 and 75.8 and in light of EPA's determinations proposed above for 
retaining current rule provisions for the bias test, relative accuracy, 
and availability trigger conditions, EPA proposes revising Part 75 to 
delete Secs. 75.7 and 75.8.

IV. Administrative Requirements

A. Executive Order 12866

    Under Executive Order 12866, 58 FR 51735 (1993), the Administrator 
must determine whether the regulatory action is ``significant'' and 
therefore subject to Office of Management and Budget (OMB) review and 
the requirements of the Executive Order. The Order defines 
``significant regulatory action'' as one that is likely to result in a 
rule that may:
    (1) Have an annual effect on the economy of $100 million or more or 
adversely affect in a material way the economy, a sector of the 
economy, productivity, competition, jobs, the environment, public 
health or safety, or State, local, or tribal governments or 
communities;
    (2) Create a serious inconsistency or otherwise interfere with an 
action taken or planned by another agency;
    (3) Materially alter the budgetary impact of entitlements, grants, 
user fees, or loan programs or the rights and obligations of recipients 
thereof; or
    (4) Raise novel legal or policy issues arising out of legal 
mandates, the President's priorities, or the principles set forth in 
the Executive Order.
    Pursuant to the terms of Executive Order 12866, it has been 
determined that this proposed rule is a ``significant regulatory 
action'' because the rule seems to raise novel legal or policy issues. 
As such, this action was submitted to OMB for review. Any written 
comments from OMB to EPA, any written EPA response to those comments, 
and any changes made in response to OMB suggestions or recommendations 
are included in the docket. The docket is available for public 
inspection at the EPA's Air Docket Section, which is listed in the 
ADDRESSES section of this preamble.

B. Unfunded Mandates Act

    Title II of the Unfunded Mandates Reform Act of 1995 (UMRA), Pub. 
L. 104-4, establishes requirements for federal agencies to assess the 
effects of their regulatory actions on State, local, and tribal 
governments and the private sector. Under section 202 of the UMRA, EPA 
generally must prepare a written statement, including a cost-benefit 
analysis, before promulgating a proposed or final rule that includes a 
federal mandate that may result in expenditure by State, local, and 
tribal governments, in aggregate, or by the private sector, of $100 
million or more in any one year. Section 205 generally requires that, 
before promulgating a rule for which a written statement must be 
prepared, EPA identify and consider a reasonable number of regulatory 
alternatives and adopt the least costly, most cost-effective, or least 
burdensome alternative that achieves the objectives of the rule. The 
provisions of section 205 do not apply when they are inconsistent with 
applicable law. Moreover, section 205 allows EPA to adopt an 
alternative other than the least costly, most cost-effective, or least 
burdensome alternative if the Administrator explains why that 
alternative was not adopted. Finally, section 203 requires that, before 
establishing any regulatory requirements that may significantly or 
uniquely affect small governments, EPA must have developed a small 
government agency plan. The plan must provide for notifying any 
potentially affected small governments, enabling officials of affected 
small governments to have meaningful and timely input in the 
development of EPA regulatory proposals with significant federal 
intergovernmental mandates, and informing, educating, and advising 
small governments on compliance with the regulatory requirements.
    Because this proposed rule is estimated to result in the 
expenditure by State, local, and tribal governments or the private 
sector of less than $100 million in any one year, the Agency has not 
prepared a budgetary impact statement or specifically addressed the 
selection of the least costly, most cost-effective, or least burdensome 
alternative. Because small governments will not be significantly or 
uniquely affected by this rule, the Agency is not required to develop a 
plan with regard to small governments.
    As discussed above, the proposed rule would eliminate two sections 
requiring studies and evaluations by EPA of certain existing regulatory 
provisions and would not include any other changes to the existing 
regulations. The proposed rule therefore would not change in any way 
the expenditure by State, local, and tribal governments or the private 
sector, or the effect on small governments, resulting from the existing 
regulations.

C. Paperwork Reduction Act

    This action proposing revisions to the continuous emission 
monitoring regulations would not impose any new information collection 
burden. OMB has previously approved the information collection 
requirements contained in the continuous emission monitoring 
regulations, 40 CFR part 75, under the provisions of the Paperwork 
Reduction Act, 44 U.S.C. 3501, et seq. Note, however, that the Agency 
is proposing other revisions to the continuous emission monitoring 
regulations in a separate action in today's Federal Register and that 
those revisions would result in a change to the current information 
collection burden.
    Burden means the total time, effort, or financial resources 
expended by persons to generate, maintain, retain, or disclose or 
provide information to or for a Federal agency. This includes the time 
needed to review instructions; develop, acquire, install, and utilize 
technology and systems for the purposes of collecting, validating, and 
verifying information, processing and maintaining information, and 
disclosing and providing information; adjust the existing ways to 
comply with any previously applicable instructions and requirements; 
train personnel to be able to respond to a collection of information; 
search data sources; complete and review the collection of information; 
and transmit or otherwise disclose the information.

D. Regulatory Flexibility

    The Regulatory Flexibility Act, 5 U.S.C. 601, et seq., generally 
requires federal agencies to conduct a regulatory flexibility analysis 
of any rule subject to notice and comment rulemaking requirements 
unless the agency certifies that the proposed rule will not have a 
significant economic impact on a substantial number of small entities. 
Small entities include small businesses, small not-for-profit 
enterprises, and small governmental jurisdictions.
    As discussed above, the proposed rule would eliminate two sections 
requiring studies and evaluations by EPA and would not include any 
other changes to the existing regulations. The proposed rule therefore 
does not change in any way the potential impacts on small entities 
resulting from the existing regulations. Therefore, I hereby certify

[[Page 28200]]

that this action will not have a significant economic impact on a 
substantial number of small entities.

List of Subjects in 40 CFR Part 75

    Environmental protection, Air pollution control, Carbon dioxide, 
Continuous emissions monitors, Electric utilities, Nitrogen oxides, 
Reporting and recordkeeping requirements, Sulfur dioxide.

    Dated: April 27, 1998.
Carol M. Browner,
Administrator.

    For the reasons set out in the preamble, part 75 of title 40, 
chapter 1 of the Code of Federal Regulations is proposed to be amended 
as follows:

PART 75--[AMENDED]

    1. The authority citation for part 75 continues to read as follows:

    Authority: 42 U.S.C. 7601 and 7651, et seq.

    2. Section 75.7 is removed and reserved.
    3. Section 75.8 is removed and reserved.

[FR Doc. 98-11750 Filed 5-20-98; 8:45 am]
BILLING CODE 6560-50-P