[[pp. 28081-28130]] Acid Rain Program; Continuous Emission Monitoring Rule Revisions

Note: EPA no longer updates this information, but it may be useful as a reference or resource.

[Federal Register: May 21, 1998 (Proposed Rules)]
[Page 28081-28130]
From the Federal Register Online via GPO Access [wais.access.gpo.gov]
[DOCID:fr21my98-43]

[[Continued from page 28080]]

[[Page 28081]]

GCV value of 100,000 Btu per hundred scf, the value of the ``sulfur-to-
heat content'' ratio is 2.0 gr/mmBtu. Therefore, a candidate gaseous
fuel would qualify to use the default SO2 emission rate of
0.0006 lb/mmBtu for part 75 reporting purposes if the 720 hours of
historical data demonstrate that the mean value of the sulfur-to-heat
content ratio is 2.0 gr/mmBtu or less.
To demonstrate that a unit qualifies to use Appendix D when
combusting a gaseous fuel, the designated representative for the
facility would be required to show that the gaseous fuel has a total
sulfur content of 20 grains/100 scf or less. This demonstration would
apply to all gaseous fuels. For gaseous fuels other than pipeline
natural gas, the sulfur content information could come either from
contractual information on the sulfur content based on routine vendor
sampling and analysis or from historic fuel sampling data to show the
gaseous fuel's sulfur content (see Docket A-97-35, Item II-I-9, Policy
Manual, Question 2.15). For gaseous fuels that are produced in batches
or lots with a relatively uniform sulfur content, such as liquefied
petroleum gases, it would be sufficient to provide historical
information on each batch over the past year. This approach was
accepted by the Agency for six units combusting liquefied petroleum gas
(see Docket A-97-35, Items II-C-14 and II-D-22).
In addition to documenting the total sulfur content of the fuel,
the owner or operator would be required to submit certain other fuel-
specific information. As previously noted, for units combusting
pipeline natural gas, a designated representative would be required to
provide contractual information to demonstrate that the natural gas is
supplied under specification and has a hydrogen sulfide content less
than or equal to 0.3 gr/100 scf. And historical data would have to be
provided, as described above, to obtain permission to use the default
SO2 emission rate of 0.0006 lb/mmBtu for a fuel other than
pipeline natural gas. For other gaseous fuels that are not produced in
batches with relatively uniform sulfur content, such as gaseous fuel
generated through an industrial process (e.g., digester gas from a
paper mill), since the sulfur content of the gaseous fuel could be
highly variable, section 2.3.3.4 of today's proposed revisions to
Appendix D would require a minimum of 720 hours of historical data
documenting the sulfur content of the fuel under representative
operating conditions. This information would allow the Agency to
determine how variable the sulfur content is and if the daily sampling
procedure under section 2.3.1 of Appendix D is sufficient to capture
this variability without allowing the underestimation of sulfur
content. If the sulfur variability were too great, continuous sampling
using a gas chromatograph and hourly reporting of sulfur content would
be required under today's proposed rule.
2. Fuel Sampling
(a) Fuel Oil.
Background
Diesel fuel is distillate fuel oil of grades No. 1 or 2. Diesel
fuel is heavily refined and has a much lower sulfur content and greater
consistency than other grades of fuel oil. Section 2.2 of Appendix D to
the May 17, 1995 direct final rule provides three options for sampling
of diesel fuel and two options for sampling of other fuel oils. First,
for all fuel oils, including diesel fuel, daily manual sampling is
allowed. Second, diesel fuel and other fuel oils may also be sampled
continuously using an automated sampler according to ASTM D4177-82
(Reapproved 1990), either using continuous drip sampling or flow
proportional sampling. The samples would then be mixed to form a daily
composite sample. Third, diesel fuel may be sampled ``as-delivered,''
upon receipt of a shipment. These sampling approaches were selected to
ensure that sulfur content values would be as accurate as possible,
would not underestimate SO2 mass emissions, and would
account for any variability in the sulfur content of fuel.
Many utilities have expressed concern about the cost of daily oil
sampling (see Docket A-97-35, Items II-D-18, II-D-20, II-E-13, II-E-
14). Some utilities indicated that for a unit that burns oil every day,
the cost of daily oil sampling is greater than the cost of
SO2 CEMS and flow monitors. Furthermore, industry
representatives provided information indicating that within a given
shipment of fuel oil from a supplier, the variability in sulfur content
is low (see Docket A-97-35, Items II-D-18 and II-D-59). Many companies
already have state or Federal requirements for sampling of fuel from
each truck delivery or in a storage tank on site at the plant whenever
fuel is added to the storage tank (see Docket A-97-35, Item II-D-93).
The storage tank is a tank at a plant that holds oil that is actually
combusted by the unit on that day. In other words, no fuel will be
blended between the time when a fuel lot is transferred to the storage
tank and when the fuel is combusted in the unit. In other cases, such
as EPA's NSPS regulations for industrial boilers under 40 CFR part 60,
subpart Db, companies keep copies of fuel receipts from the supplier to
indicate the sulfur content is below the required sulfur content. Based
upon this information, EPA is proposing to reduce the required sampling
frequency for fuel oil. This would be a significant reduction in burden
and cost of using Appendix D, without causing underestimation of
SO2 emissions.
Discussion of Proposed Changes
Several utilities suggested that the Agency propose to allow
sampling of each delivery of oil (see Docket A-97-35, Items II-D-18,
II-D-20, II-E-13, II-E-22). Under this approach, either a facility or
its supplier would sample each truck or barge containing oil before the
fuel is transferred into a tank at the plant. If a delivery shipped in
a group of trucks were purchased under the same order and were
specified to have the same gross calorific value, density, and sulfur
content, then only one sample would be necessary for the group of
trucks. Samples taken by the supplier would not need to be split and
kept on hand at the site. This approach is currently allowed only for
diesel fuel under section 2.2.1.2 of Appendix D, but would be extended
to apply to all fuel oils under today's proposed rule. This approach
would be particularly useful to a facility that receives large,
infrequent deliveries of fuel or to a facility that already has other
State or Federal regulations requiring sampling of each truck or barge
delivered to the plant.
A similar approach suggested by another industry representative,
allowing facilities to use a sample of oil taken from a tank belonging
to the supplier before the oil is delivered, is also proposed in
today's rulemaking. The supplier could take the sample and the facility
would be able to use that value as long as it keeps records of the fuel
analysis results from the supplier. This approach would be particularly
useful to a facility that receives a delivery of oil from a single
supplier's tank that is shipped in many different trucks. This approach
also would be useful for a small facility that would prefer to rely on
samples taken by the supplier rather than taking its own samples and
paying for their analysis.
Finally, the Agency proposes a third sampling approach, allowing a
facility to sample oil manually from its storage tank at the plant
whenever oil is added to the tank. This approach would yield samples
that are more representative of the oil combusted because it would
include any fuel remaining in the tank as well as all fuel added.
Sampling from the storage tank at the plant would be

[[Page 28082]]

useful to a facility that burns oil infrequently and adds oil to its
storage tank infrequently. It also would be helpful where a facility
already has other State or Federal regulations requiring sampling after
adding fuel to the storage tank.
Both the ``before delivery'' and ``as delivered'' sampling
approaches would require a sample for each ``lot'' of oil;
consequently, a suitable definition of a ``lot'' is needed. For
purposes of determining when an oil sample should be taken for the NSPS
applicable to utility boilers, section 5.2.2.2 of Method 19 in Appendix
A to 40 CFR part 60 relies on a definition of fuel ``lot'' developed by
the American Society for Testing and Materials (ASTM). This definition
states that ``the lot size of a product oil is the weight of product
oil from one pretreatment facility and intended as one shipment (ship
load, barge load, etc.).'' In essence, a lot is a single batch of oil
that has uniform properties and is purchased from a single supplier and
delivered to a buyer. Among those uniform fuel properties are gross
calorific value, density, sulfur content, and viscosity. In today's
rulemaking, EPA proposes to adopt this definition of a lot of oil for
use in the Acid Rain Program.
The Agency also considered whether it is appropriate to keep the
current approach of daily manual oil sampling as an option. Although it
seems unlikely that facilities would choose daily sampling option if
they have the three options of sampling by lot, sampling upon addition
of fuel to a storage tank, or continuous sampling, a utility group has
requested that EPA retain daily manual sampling as an option. The
agency is, therefore, proposing to retain daily manual oil sampling as
an option in Appendix D to allow facilities this additional
flexibility. An industry representative suggested that EPA could define
the oil combusted during a 24-hour period as a lot. For the reasons
discussed below and in the section addressing sulfur content, density,
and gross calorific values used in calculations, EPA is not
incorporating this suggestion in today's proposed rule.
EPA also reconsidered whether it is necessary to require daily
composite samples when samples are taken continuously with an automatic
sampler. In today's proposal, the Agency is proposing that continuous
samples may be composited on a weekly basis rather than daily. The
Agency also considered allowing an even longer compositing period, such
as a month, but is not proposing this option for the reasons discussed
below. A weekly composite sample of oil that is sampled continuously
would be an attractive option for a facility that wants the most
representative and accurate sulfur content data possible. This also
would be a useful option for those few facilities that receive oil via
a pipeline, rather than in discrete lots.
Rationale
Facilities wish to be able to perform less frequent fuel sampling
in order to save money. From the information EPA has examined over the
previous year, the Agency believes that less frequent oil sampling can
be technically justified. Based upon information provided by utilities,
the sulfur content of a lot of oil varies from sample to sample, with a
standard deviation of 0.036 percent S to 0.063 percent S, or 5.62 to
6.85 percent of the average sulfur content for all daily samples
between deliveries (see e.g., Docket A-97-35, Item II-D-18). Density
and gross calorific value of oil in a lot should vary even less than
sulfur content, because sulfur is an impurity in the composition of the
fuel and not an essential physical property of the oil, as is density.
Furthermore, the difference between the sulfur content, density, gross
calorific value, and carbon content of a fuel during the first daily
sample after a new delivery is received and the average sulfur content,
density, gross calorific value, and carbon content for all daily
samples from between two deliveries is extremely small (see Docket A-
97-35, Items II-B-18 and II-D-18 for supporting information).
Therefore, the Agency expects that the variability of fuel
characteristics within a lot is low enough that only a single
representative sample is necessary for the lot. Data have indicated
that there could be a significant difference in sulfur content between
shipments, however (see Docket A-97-35, Items II-B-12, II-B-18 and II-
D-18). The Agency believes that differences between lots, which could
potentially result in the underestimation of SO2 emissions,
can be dealt with by selecting a conservative sulfur content, density,
or gross calorific value that would not be exceeded in any sample,
rather than retaining more frequent sampling requirements. Therefore,
today's proposal incorporates this approach.
Prior to drafting today's proposed rule revisions, EPA requested
comments on removing the option to perform daily manual oil sampling
for Appendix D units. At least one utility group expressed interest in
retaining the option to allow flexibility. The prime benefit to a
facility from continuing to use daily manual sampling would appear to
be that the facility could continue to use the same daily operating
procedures and that reprogramming of a DAHS would not be necessary.
Note that when using the approach of daily manual oil samples, a
facility calculates SO2 mass emissions using the highest
sulfur content in the previous 30 daily oil samples. Therefore, this
approach requires more frequent analysis than either the proposed
weekly composite sample for continuous samples or the proposed sampling
by lot, and provides less accurate and more conservative results. The
Agency believes it would be simpler and less confusing for both the
Agency and for the regulated community to deal with a smaller number of
approaches to sampling and calculating SO2 emissions.
However, the Agency is retaining this option since at least some
affected utilities want the flexibility to continue to use this option.
EPA also considered the suggestion to define a 24-hour period as a
lot in order to allow facilities to continue to perform daily manual
sampling. EPA is not proposing this approach because of the added
complexity, compared to keeping the current language in section 2.2.4
of Appendix D concerning manual daily sampling of oil. If a lot were
defined as an arbitrary 24-hour period, the other requirements in the
current rule (e.g., conservative sulfur, gross calorific value, and
density values used to calculate SO2 mass emission rate and
heat input rate) would need to be retained to ensure that
SO2 emissions were not underestimated. Furthermore, using
the terminology of a ``lot'' for both a delivery and a period of time,
while requiring different treatment of sample data from the two
different types of ``lots,'' could potentially be confusing. It seems
preferable to keep the current language for daily manual samples.
Because the Agency now believes it is appropriate to sample each
fuel lot instead of sampling daily, the Agency reconsidered whether
daily composite samples are necessary when a facility performs
automated continuous sampling. Because continuous samplers take fuel
samples multiple times each hour, they are highly representative of the
oil being burned. Flow proportional samplers take samples automatically
when a certain volume or mass of fuel has passed by, rather than during
a particular time period. Generally, automatic samplers take multiple
samples each hour; however, only one sample per hour is required under
section 2.2.3 of Appendix D of the current rule. Even if the
compositing time period is extended, the composite sample will be
representative of the sulfur content, density, and gross calorific
value of the oil between samples. Therefore, the Agency believes

[[Page 28083]]

that the compositing period could be extended from a day to as long a
period as a month. However, EPA believes that it is unlikely that any
container for taking samples from an automatic sampler would be large
enough to accommodate all automatic samples taken during a month. In
addition, at least one industry representative suggested that weekly
composite samples were appropriate (see Docket A-97-35, Item II-D-30).
Therefore, in section 2.2.3 of today's proposed rule, EPA would extend
the allowable length of the compositing period for automatic samples to
one week. The Agency believes this will make automatic sampling less
costly, while taking into account the physical limitations of sampling
equipment.
(b) Gaseous Fuels.
Background
Section 2.3 of Appendix D, as revised in the May 17, 1995 direct
final rule, provides only one approach for sampling gaseous fuel: under
section 2.3.1, gaseous fuel sampling must be performed daily.
Relatively few utilities perform daily sampling upon gaseous fuels,
choosing instead to use a default SO2 emission rate for
pipeline natural gas. In part, this is because the vast majority of
gaseous fuel used by power plants is pipeline natural gas. Under
section 2.3.2 of Appendix D, facilities may calculate SO2
mass emissions from pipeline natural gas using a default emission rate
instead of performing fuel sampling. Because of the difficulty and
potential danger of sampling gaseous fuel, gas sampling is generally
conducted by the supplier, rather than by the facility.
Those few utilities combusting gaseous fuels other than pipeline
natural gas have expressed concern about the difficulty and expense of
daily sampling, particularly in comparison to the value of
SO2 allowances for low SO2 emissions from
relatively clean fuel (see, e.g., Docket A-97-35, Items II-E-11, II-E-
20). For gaseous fuels that are delivered in discrete batches or
``lots,'' one would expect the gaseous fuel to behave like an ideal
gas; sulfur should be evenly distributed throughout the batch. On this
principle, the Ohio Environmental Protection Agency allowed a plant to
take propane samples from each discrete delivery, rather than on a
daily basis (see Docket A-97-35, Items II-C-14 and II-D-22).
Discussion of Proposed Changes
Today's proposal incorporates three different sampling approaches
for gaseous fuels: sampling by lot, daily sampling, and continuous
sampling with a gas chromatograph. For gaseous fuel that is delivered
in discrete lots, such as liquefied petroleum gas, the gaseous fuel
could be sampled either daily or for each lot delivered. Any gaseous
fuels other than pipeline natural gas that are not delivered in
discrete lots, such as digester gas or sour natural gas pumped directly
from a field, would, at a minimum, need to be sampled daily. The
samples could be taken either by the supplier or by the facility.
However, if the average sulfur content and sulfur variability of such a
fuel were too high (i.e., mean sulfur content > 7 gr/100 scf and
standard deviation from the mean > 5 gr/100 scf, based on 720 hours of
representative historical data), continuous sampling with a gas
chromatograph and hourly reporting of sulfur content would be required.
Rationale
The approach of sampling upon a lot or discrete delivery of gaseous
fuel is being incorporated into today's proposed rule for the following
reasons. The Agency believes that discrete deliveries are sufficiently
different from pipeline transmission of fuel that a different sampling
approach is appropriate. According to the ideal gas law, all gas within
an enclosed volume is mixed with a consistent composition; therefore, a
single sample should be representative of all gas in the volume.
Although gaseous fuels delivered by lot, such as liquefied petroleum
gas, are higher in sulfur content and have a wider range of sulfur
contents than pipeline natural gas, they still have relatively low
sulfur contents compared to liquid and solid fuels. Thus, less frequent
gas sampling appears appropriate, based on the small difference in the
accuracy of calculated SO2 mass emissions. For this same
reason, the Agency allowed as-delivered sampling for diesel fuel in the
May 17, 1995 direct final rule (see Docket A-94-16, Item II-F-2).
Finally, because of the difficulty of sampling gaseous fuels, EPA
believes that it is less burdensome and less dangerous if gas sampling
is conducted by the gas supplier. It is the Agency's understanding that
the sampling for a gas in a discrete delivery or lot is typically
conducted once for the lot, rather than on a daily basis. Through a
petitioning process, EPA has already allowed one utility to perform
sampling upon a lot or discrete delivery of gaseous fuel (see Docket A-
97-35, Items II-C-14 and II-D-22).
EPA is proposing to require daily or continuous sampling of gaseous
fuels other than pipeline natural gas or the equivalent that are not
shipped in discrete lots, such as sour natural gas pumped directly from
a field, landfill gas, or digester gas. Such gaseous fuels cannot be
guaranteed to be stable in sulfur content. Therefore, proposed section
2.3.3.4 in Appendix D would require a minimum of 720 hours of
representative historical data to characterize the sulfur variability
of such fuels. For the 720 hours of demonstration data, the mean value
and standard deviation of the fuel sulfur content would be calculated.
If the mean value does not exceed 7 gr/100 scf (equivalent to about 10
ppm of SO2 emissions to the atmosphere), daily sampling
would suffice. If the mean value is greater than 7 gr/100 scf, however,
the variability of the sulfur content would be assessed in terms of the
standard deviation. If the standard deviation exceeds 5 gr/100 scf, the
sulfur variability would be considered too high and continuous sampling
of the fuel with a gas chromatograph would be required. If continuous
sampling were required, the owner or operator would have to implement a
quality assurance program for the gas chromatograph. A copy of the QA
plan would be kept on-site, suitable for inspection. For fuel with a
low average sulfur content or a low sulfur variability, daily sampling
would be sufficient. However, for gaseous fuel with a higher sulfur
content, if the sulfur variability were too great, continuous sampling
of the fuel with a gas chromatograph and hourly reporting of sulfur
content would be required.
3. Sulfur, Density and Gross Calorific Value Used in Calculations
(a) Fuel Oil.
Background
The hourly SO2 mass emissions rate due to combustion of
oil is calculated using the mass flow rate of oil combusted and a
sulfur content value from a sample. If a unit's oil flow rate is
measured with a volumetric fuel flowmeter rather than a mass fuel
flowmeter, then it will be necessary to determine the mass flow rate of
oil from the volume of fuel and a density value from an oil sample. The
heat input rate is calculated using the flow rate of oil multiplied by
the gross calorific value (GCV) of a sample.
The sulfur content, density, and GCV used to calculate emissions
and heat input depend upon the oil sampling method used. Some sampling
methods are more accurate than others. For example, for flow
proportional or continuous drip sampling, the actual sulfur content
from a sample is used to calculate SO2 mass emissions.
However,

[[Page 28084]]

when daily manual samples are taken under section 2.2.4 of Appendix D,
a facility must use the highest fuel sulfur content recorded at that
unit from the most recent 30 daily samples, which is not necessarily
the sulfur content of the fuel being burned at any particular time. For
units where diesel fuel is sampled upon delivery, section 2.2.1.2
instructs a facility to calculate SO2 emissions using the
highest sulfur content of any oil supply combusted in the previous 30
days that the unit combusted oil. In daily manual sampling and as-
delivered sampling, conservative sulfur values are used to avoid the
possibility of underestimating SO2 mass emissions due to
variations in sulfur content. Gross calorific values are taken from the
most recent sample, rather than using the highest value in the previous
30 days, because, for natural gas, GCV is more consistent than sulfur
content.
Today's proposed rule includes changes to the sampling frequency
for oil. Therefore, it is also necessary to make corresponding changes
to the sulfur content, density, and GCVs to be used in calculations.
For example, where oil samples would no longer be taken daily, it would
be inappropriate to calculate SO2 mass emissions based upon
a certain number of daily samples. In developing today's proposal, EPA
considered what fuel analysis data values for sulfur content, density,
and GCV would be appropriate and consistent with the approaches for
taking manual samples. The appropriate sulfur content, density, and GCV
values were considered for manual samples taken from a storage tank at
the facility whenever fuel is added to the tank, for samples taken from
each lot before the delivery is transferred from tank trucks or barges,
and for samples taken from the fuel supplier's storage tank.
Discussion of Proposed Changes
EPA has re-evaluated the sulfur content, density, and GCVs to be
used to calculate SO2 mass emissions and heat input based
upon the new oil sampling approaches. For daily manual oil sampling, a
facility would continue to use the highest sulfur content from previous
30 daily samples, and the actual density and GCV. For continuous oil
sampling with an automatic sampler, a facility would continue to use
the actual sulfur content, density, and GCV. For the two new methods of
manual sampling, EPA considered whether conservative or actual values
should be used to calculate emissions and heat input. EPA also
considered whether the same type of calculational value should be used
for sulfur content, density, and GCV. For example, if conservative
sulfur content and density values are used to calculate the
SO2 mass emission rate, should a conservative or an actual
measured GCV be used to calculate the heat input rate?
For manual samples taken from a storage tank at a plant whenever
fuel is added to the tank, EPA considered the following options: (1)
using the highest sulfur content and density from the previous three
samples, and the actual GCV, (2) using the highest sulfur content from
the previous three samples, and the actual density and GCV, (3) using
the actual sulfur content, density, and GCV, (4) using the highest
sulfur content, density, and GCV from the previous calendar year, and
(5) using the maximum sulfur content, density, and GCV allowed by fuel
purchase contract with the fuel supplier. The third, fourth, and fifth
options are incorporated into today's proposal in section 2.2.4.2.
Under this approach, a facility would take a sample from the storage
tank whenever fuel is added to the tank. No blending of fuel would be
allowed from the time the oil is sampled until the fuel is combusted by
the unit. The sample would be analyzed for sulfur content, density, and
GCV. Based on the selected option (3, 4, or 5), the appropriate values
would then be used to calculate the SO2 mass emission rate
and the heat input rate from the date and hour in which the transfer of
oil is complete until the date and hour when oil is again added to the
tank.
EPA considered several different options for the case where a
facility or its supplier would sample each oil delivery (or the
supplier's storage tank) before the fuel is transferred into a tank at
the plant. EPA considered whether or not these values needed to be
conservative and concluded that there was a real possibility of
underestimating SO2 emissions by using the fuel analysis
values from a delivery. The options that EPA considered to avoid the
underestimation were: (1) using the highest sulfur content and density
from all samples taken from oil combusted during the previous 30 days,
and the actual GCV, (2) using the maximum sulfur content, density, and
GCV in the fuel purchase contract specifications, (3) using the highest
sulfur content, density, and GCV from a sample taken in the previous
calendar year, and (4) using the highest sulfur content, density, and
GCV ever recorded for the unit. The second and third options are
incorporated into today's proposed rule in section 2.2.4.3 of Appendix
D.
Under the selected options, a facility or its supplier would need
to sample a delivery of fuel before it is transferred into a storage
tank. The facility would then need to keep records of the fuel
analytical results for three years. The facility would use the
conservative value it selected under option (2) or (3), above, in order
to calculate the SO2 mass emission rate and the heat input
rate. If an as-delivered sample were ever analyzed and found to have a
sulfur content, density, or GCV that exceeded the value being used in
calculations (i.e., the contract specification, or the maximum value
measured in the previous calendar year), then the new sampled value
would be used to calculate the SO2 mass emission rate or the
heat input rate, as follows. For a unit using a default value of the
maximum value measured during the previous calendar year, that new
sample value would become the new default value and would be reported
for the remainder of the current year and the next year, unless
superseded by a higher sampled value. For a unit using a default value
of a contract specification, the new sample value would continue to be
used as the new default value instead of the contract specification
value, unless superseded by a higher sampled value or by a new
contract.
Rationale
EPA considers continuous sampling and the measurement of fuel from
a storage tank at a plant after each addition of fuel to the tank to be
highly accurate methods that will be representative of the fuel
combusted in a unit. However, if samples are taken from the truck or
barge used to ship the fuel, or if samples are taken ``as-delivered,''
the sample values will not necessarily accurately reflect the oil being
combusted by the unit at any particular time (see Docket A-97-35, Item
II-E-22). For example, a storage tank could contain oil with an average
sulfur content of 0.6 percent. Then a new delivery with a sulfur
content of 0.4 percent is received and transferred to the tank. The
``as-delivered'' sample value from the delivery truck would
underestimate the emissions at that time, since the fuel actually
combusted will combine a mixture of the old fuel supply in the storage
tank and the new fuel that is added. Thus, a more conservative sulfur
value should be used to calculate SO2 emissions if samples
are taken from the delivery containers or from a container used by the
oil supplier.
For density and GCV, today's proposal, at the suggestion of some
industry representatives, uses conservative values determined by the
same method for both parameters (see Docket A-97-35, Item II-E-24).
This

[[Page 28085]]

has the advantage of being easy to remember and to program. However, if
greater accuracy is desired, a facility would always have the option of
using actual sulfur content, density, and GCVs if it took samples from
its storage tank after each addition of fuel to the tank, or if it took
continuous, automatic samples.
EPA considered which conservative values would be appropriate for
sulfur, density, and GCV. EPA at first considered using the maximum
value from all oil supplies combusted in the previous 30 days. This is
similar to the current wording of section 2.2.1.2 of Appendix D for
calculation of SO2 emissions from diesel fuel as-delivered
sampling. However, in the process of implementing this provision of
part 75, EPA found this wording was somewhat confusing and issued
policy guidance to clarify section 2.2.1.2 of Appendix D (see Docket A-
97-35, Item II-I-9, Policy Manual, Question 2.9). This policy
essentially directs facilities to keep track of the amount of fuel used
as well as its sulfur content. Because of the more complicated nature
of this accounting, some industry representatives suggested that it
would be simpler to use a conservative default value that would not
require tracking fuel usage (see Docket A-97-35, Item II-E-24). Of the
default values considered, EPA felt that the most appropriate default
values would be the maximum values established by agreement with the
fuel supplier through a contract or the maximum measured value from all
samples in the previous calendar year. Contractual limits should be
higher than or equal to the actual sulfur content, density, or GCV.
Because not all units would necessarily have a fuel contract limiting
oil sulfur content, density, or GCV, EPA is also proposing to provide
the option of using the maximum oil sulfur content, density, or GCV in
the previous calendar year.
The Agency also considered whether the current provisions of 2.2.4
of Appendix D should be retained for calculation of SO2
emissions using the highest sulfur from the previous 30 daily samples
when performing daily manual sampling. As discussed above in Section
III.P.2(a) of this preamble on oil sampling frequency, the Agency is
proposing to retain the option as requested by at least one utility
representative.
(b) Gaseous Fuels.
Background
The vast majority of Acid Rain units which burn gaseous fuels
combust pipeline natural gas. Section 2.3.2 of Appendix D contains a
provision for calculation of SO2 mass emissions from
pipeline natural gas using a default SO2 emission rate in
lb/mmBtu and the heat input rate of pipeline natural gas. However, if a
facility or its supplier is sampling gaseous fuel for sulfur content,
either because it is not pipeline natural gas or because the facility
chooses to use a sampled value, then Appendix D requires the facility
to calculate the SO2 mass emission rate using the sulfur
content of the sample and the volume of gas combusted, and to calculate
the heat input using the GCV of the sample and the volume of gas
combusted (see Equations D-5 and F-20). Because of the nature of
gaseous fuels, they are always measured with a volumetric fuel
flowmeter. The formulas for calculating the SO2 mass
emission rate and the heat input rate use volume directly and do not
require information on gas density. The current provisions of Appendix
D allow a facility to calculate the SO2 mass emission rate
and the heat input rate using the actual value from a daily sample of
gaseous fuel.
When the provisions of section 2.3 of Appendix D were added to part
75 in the May 17, 1995 direct final rule, EPA presumed that virtually
every utility combusting gaseous fuel was combusting pipeline natural
gas. However, the Agency found that utilities were combusting other
types of gaseous fuels. One utility submitted a monitoring plan and a
certification application for fuel flowmeter monitoring systems that
indicated the utility was also using propane liquefied petroleum gas
(LPG) (see Docket A-97-35, Item II-D-6). The utility indicated that it
wished to use the default emission rate factor reserved for pipeline
natural gas in its monitoring plan and later petitioned the Agency
specifically for permission to use the default emission rate factor of
0.0006 lb/mmBtu. In conversations with utility staff, EPA found that
the utility wanted to avoid the expense of additional daily samples and
the trouble of entering daily sulfur values manually into its data
acquisition and handling system (see Docket A-97-35, Items II-E-11, II-
E-20). The Agency eventually approved a revised petition for the
utility that allowed the utility to take propane samples from each
discrete delivery, rather than on a daily basis, where the utility
calculates sulfur dioxide emissions from propane by using the highest
sulfur content recorded during the previous 365 days and reports these
data in its quarterly electronic data report (see Docket A-97-35, Items
II-C-14 and II-D-22).
The Agency found that there were also some utilities burning
gaseous fuels that were by-products of an industrial process (see
Docket A-94-16, Item II-D-71). EPA had concerns that such ``digester
gas'' might have a more variable sulfur content than pipeline natural
gas, since the gaseous fuel would begin with a higher sulfur content
than pipeline natural gas and would not necessarily go through a
process that would reduce and stabilize the sulfur content.
Discussion of Proposed Changes
In today's proposed rule, the provisions for sampling gaseous fuels
are found in section 2.3.1 of Appendix D. For gaseous fuels that are
delivered in discrete lots, a facility would use conservative values
for sulfur content and GCV to calculate the SO2 mass
emission rate and the heat input rate. For the sulfur content value,
the highest sampled sulfur content from the previous calendar year or
the maximum value allowed by contract would be used to calculate the
SO2 mass emission rate. For GCV, the highest of all sampled
values in the previous calendar year or the maximum value allowed by
contract would be used to calculate the heat input rate. If, for any
gas sample, the assumed sulfur content or GCV were exceeded, the
sampled value would become the new assumed value. For units using the
contract value, the sampled value would continue to be used unless a
new (higher) contract specification were put in place or unless an even
higher sampled value is obtained. For units using the maximum value
from the previous year, the sampled value would continue to be used for
the remainder of the current year and for the next calendar year unless
it was superseded by an even higher sampled value.
For any gaseous fuel where daily fuel sampling is required, a
facility would use the highest sulfur in the previous 30 daily samples.
For gaseous fuels other than pipeline natural gas, where daily sampling
of sulfur content is required, the highest GCV from the previous 30
daily samples would be used. For pipeline natural gas, where monthly
sampling of GCV only is required, the actual measured GCV, the highest
of all sampled values in the previous calendar year, or the maximum
value allowed by contract would be used.
For a gaseous fuel that is not produced in batches and that has a
relatively high sulfur content and a high sulfur variability,
continuous sampling with a gas chromatograph would be required. Sulfur
content would be reported as actual measured hourly average values. The
GCV would also be determined on an hourly basis, or,

[[Page 28086]]

alternatively, the highest value in the previous 30 unit operating days
could be reported.
Rationale
For gaseous fuel supplied in discrete deliveries, EPA is proposing
to take the same approach as for fuel oil that is being delivered to a
plant by barge or truck. EPA has already approved this approach with
one utility that combusts liquefied petroleum gas (see Docket A-97-35,
Items II-C-14 and II-D-22). Because a discrete delivery of gaseous fuel
would be maintained in an enclosed chamber with a relatively constant
temperature and pressure, one would expect the gaseous fuel to behave
like an ideal gas. Thus, sulfur and other constituents of the fuel
should be evenly distributed throughout the delivery of fuel. Using
conservative values to calculate the SO2 mass emission rate
and the heat input rate should account for any variability between
deliveries. Furthermore, this reduces the number of changes that would
be made to a data acquisition and handling system to add fuel supply
data.
For gaseous fuel other than pipeline natural gas, where daily fuel
sampling is required, EPA considered leaving unchanged the current
provisions of section 2.3.1 of Appendix D that would allow a utility to
use the actual value from a day's sample to calculate the
SO2 mass emission rate and the heat input rate. However, the
Agency believes that it is appropriate to change the sulfur content
value to be a somewhat conservative historical value. This is because
the Agency has concerns that there may be some gaseous fuels other than
natural gas, such as digester gas, that may have significant
variability in their sulfur content over the course of a day or a
longer period of time. This might result in the underestimation of the
SO2 mass emission rate.
In the case of fuel oil, some industry representatives suggested it
was simplest to determine the appropriate conservative values for
sulfur content, density, and GCV by the same method (see Docket A-97-
35, Item II-E-24). With one exception (for fuels with relatively high
sulfur content and high sulfur variability), today's proposal follows
this suggestion for gaseous fuels. The proposal uses the highest sulfur
content and the highest GCV from the previous 30 daily samples. This is
currently the procedure used to determine the sulfur value used in
calculations from daily manual oil samples. Since this algorithm for
daily manual oil sample calculations is already being used by many
software programmers, it is a good conservative value to use for daily
samples in this case. The Agency notes that currently, the heat input
is calculated using the actual sampled GCV and that this change would
require software reprogramming for units where gaseous fuel is sampled
daily. However, for pipeline natural gas that is sampled monthly for
GCV, facilities could continue to use the actual GCV measured in a
monthly sample. The other two options are more conservative and would
require software changes. The Agency requests comment on the proposal
to use the more conservative GCV value to determine the heat input rate
for gas combustion when gaseous fuel is sampled daily (which differs
from the current procedure in section 2.3.1.3 of Appendix D and section
5.5.2 of Appendix F).
For gaseous fuel that has a relatively high sulfur content and high
sulfur variability, daily sampling is not considered adequate to ensure
that SO2 emissions will not be underestimated. Therefore,
for such fuels, continuous sampling with a gas chromatograph and hourly
reporting of sulfur content would be required. For GCV, which is
expected to be less variable than sulfur content, either the actual
hourly measured value or the highest GCV value obtained in the last 30
unit operating days could be reported.
4. Missing Data Procedures for Sulfur, Density, and Gross Calorific
Value
Background
(a) Fuel Oil. The May 17, 1995 direct final rule included missing
data procedures for missing analytical information on sulfur content,
density, and GCV in section 2.4 of Appendix D. These procedures are
based on a daily sampling frequency. For example, missing sulfur
content, density, or GCV data are to be calculated using the highest
measured sulfur content, oil density, or GCV during the previous thirty
days when the unit burned oil. This was intended to mean that the
substitute data values are to be based on the previous thirty daily oil
samples for which data are available.
In order to ensure that a DAHS is capable of implementing the
missing data procedures required by the rule, Sec. 75.20(c)(7) and
Sec. 75.20(g)(1)(ii) require testing of each DAHS. EPA issued policy
guidance discussing how facilities should report the results of these
tests for units measured with fuel flowmeters. This policy guidance
provided a form checklist that facilities could use to show the results
of their own tests of the missing data substitution procedures (see
Docket A-97-35, Item II-I-9, Policy Manual, Question 15.9). Some
utilities objected to testing the DAHS missing data procedures on the
grounds that they should never miss sample data. In part, this would be
because the facility is required, under section 2.2.5 of Appendix D, to
split its sample and keep a portion. One utility offered to substitute
the maximum potential sulfur content, which would require less
complicated DAHS programming than using the maximum sulfur content of
the previous 30 daily samples.
(b) Gaseous Fuels. Section 2.4.1 of Appendix D, as revised by the
May 17, 1995 direct final rule, provides missing data substitution
procedures for missing sulfur data from daily samples of gaseous fuel.
The DAHS is required to substitute the highest measured sulfur content
recorded during the previous 30 days when the unit combusted gaseous
fuel. As for oil, this was intended to be the highest sulfur value from
the previous 30 daily samples with available sulfur values. Section
2.4.2 of Appendix D requires the substitution of the highest measured
GCV recorded during the previous three months that the unit burned
gaseous fuel when data are missing from a monthly gaseous fuel sample.
As for fuel oil, the missing data procedures for gaseous fuels are
linked to the frequency of fuel sampling.
A utility indicated to EPA that because it receives gas sampling
information from its supplier, it should never have missing data for
GCV. The utility suggested that it should not have to go to the expense
of programming its DAHS for missing data procedures that should never
need to be used. This argument was similar to that used by another
utility when referring to missing data procedures for manual samples of
fuel oil taken upon each delivery.
Discussion of Proposed Changes
EPA proposes to revise the missing data substitution procedures for
both fuel oil and gaseous fuel, in order to simplify them. For any
instance in which the sulfur content, GCV, or density value is missing,
the maximum potential value would be reported until the results of a
subsequent valid sulfur content analysis, GCV determination, or density
measurement are obtained. The proposed appropriate maximum potential
values are specified in the table below. The default values for sulfur
content, GCV, and density of residual oil and diesel fuel were taken
from handbook values (see Docket A-97-35, Item II-A-7). The default
maximum sulfur content values for gaseous fuel are consistent with the
maximum sulfur content allowed under

[[Page 28087]]

the definition of natural gas and the de facto maximum sulfur content
of pipeline natural gas, based on the proposed definition. Thus, any
gas with a sulfur content that did not allow it to qualify as pipeline
natural gas (i.e., greater than 0.30 gr/100 scf) but still allowed it
to be measured following Appendix D procedures (i.e., total sulfur
content not exceeding 20.0 gr/100 scf) would have a default maximum
potential sulfur content of 20.0 gr/100 scf. The default values for GCV
of gaseous fuels were taken from handbook values (see Docket A-97-35,
Item II-I-1). For pipeline natural gas, it is assumed that the gas is
primarily methane (GCV of 1050 Btu/scf) with a small amount of other
hydrocarbons with a higher GCV (see Docket A-97-35, Item II-E-19). For
other gaseous fuels, it is assumed that they are primarily butane (GCV
of 2100 Btu/scf), the hydrocarbon gas with the highest GCV of gases
commercially used for fuel.

Maximum Potential Default Values for Sulfur Content, Density, and GCV Data
----------------------------------------------------------------------------------------------------------------
Parameter Fuel Maximum potential default value
----------------------------------------------------------------------------------------------------------------
Sulfur content.......................... residual oil............... 3.5 percent by weight.
diesel fuel................ 1.0 percent by weight.
pipeline natural gas....... 0.30 gr/100 scf.
gaseous fuels with sulfur 20.0 gr/100 scf.
content greater than
pipeline natural gas.
GCV/heat content........................ residual oil............... 19,500 Btu/lb.
diesel fuel................ 20,000 Btu/lb.
pipeline natural gas....... 1100 Btu/scf.
gaseous fuels with sulfur 2100 Btu/scf.
content greater than
pipeline natural gas.
Oil Density............................. residual oil............... 8.5 lb/gal,
diesel fuel................ 7.4 lb/gal.
----------------------------------------------------------------------------------------------------------------

Rationale
(a) Fuel Oil. It seems possible that a facility might occasionally
miss a sample taken with an automatic sampler, and thus, would have
missing data. Therefore, today's proposal includes a provision for
substitution of missing sulfur content, density, and GCV data from
continuous, automatic sampling.
Based upon comments from some utilities, it seems relatively
unlikely that both a facility and its supplier would miss performing a
sample during a delivery. Both a facility and its fuel supplier will
want to verify that the fuel delivered is actually supplying the heat
content that it is supposed to, either under a contract or a fuel
specification; thus, both a facility and its fuel supplier will have an
incentive to ensure sampling takes place for a delivery. Furthermore,
if samples taken by a facility are split, then there should generally
be the ability to provide analytical data for that fuel, even if test
results were somehow lost. Because the event of missing fuel samples is
unlikely for as-delivered samples, EPA believes that it would be
appropriate to establish a simple, conservative value that could easily
be substituted in a data acquisition and handling system. This would be
easier to program than using historical values that require tracking
fuel usage over an extended period of time.
EPA is specifically proposing the most conservative (maximum
potential) values for missing data purposes. This would ensure that
substituted missing data values would be less advantageous to a
facility than taking samples and using sulfur content, density, and GCV
data from samples. In addition, several utilities suggested to EPA that
this was a reasonable approach (see Docket A-97-35, Item II-E-24).
(b) Gaseous Fuels. As mentioned previously, gas sampling is
generally performed by fuel suppliers because of the difficulty and
potential danger of opening up a pressurized pipe containing a highly
flammable gas. It seems extremely unlikely that a fuel supplier would
not have information available on the sulfur content or GCV of gaseous
fuel, since industrial customers will purchase fuel or agree to a
contract based upon these characteristics. The exception to this might
be gaseous fuel manufactured through an industrial process that is not
produced specifically for sale as a fuel, such as digester gas. In
today's proposed rule, EPA is using the same reasoning as above for
missing manual fuel oil sample data and is using the same basic
substitution approach for missing sulfur content and GCV data for
gaseous fuel.
EPA considered keeping the existing missing data substitution
procedures from sections 2.4.1 and 2.4.2 of Appendix D for missing data
from gaseous fuel. This would have the advantage of requiring no
reprogramming of software for facilities already following the existing
procedures. EPA also considered using the maximum sulfur content or GCV
from the previous calendar year, the same procedure proposed in today's
rule for calculation of SO2 mass emission rate or heat
input, for discrete deliveries of gas or for manual samples of oil
taken from a delivery truck or barge. However, using the proposed
maximum value would require little reprogramming and would greatly
simplify the missing data procedures. In policy guidance, the Agency
has indicated it would accept a simplified DAHS for units using the
procedures of Appendices D and E. In particular, these policies endorse
manual entry of fuel analytical data, simplified missing data
procedures for fuel flowmeters, and a DAHS that uses commercial
spreadsheet software instead of a specialized custom software for
purposes of part 75 (see Docket A-97-35, Item II-I-9, Policy Manual,
Questions 14.72 and 14.73). In keeping with the policy of allowing
Appendices D and E units to use commercial spreadsheet software, EPA
has proposed what it believes to be the simplest possible missing data
substitution procedure for missing sulfur content and GCV data. In
addition, using the proposed maximum potential sulfur content or GCV
would ensure that substituted missing data values are more conservative
than the values normally used to calculate the SO2 mass
emission rate and the heat input rate.

[[Page 28088]]

5. Installation of Fuel Flowmeters for Recirculation
Background
The current provisions of section 2.1.1 of Appendix D require the
use of an additional ``return'' fuel flowmeter when some fuel is
recirculated, i.e., initially sent toward a unit and then diverted away
from the unit without being burned. This additional fuel flowmeter is
required, regardless of the amount of fuel being diverted.
At least one utility has requested to use only the fuel flowmeter
measuring fuel leaving the oil tank without a second fuel flowmeter to
measure any fuel diverted away by the recirculation fuel line. The
utility argued that using a single fuel flowmeter would result only in
the overestimation of SO2 emissions, since the utility would
measure a larger amount of fuel usage. This would allow the facility to
avoid the expense of installation, certification, and quality assurance
testing on a fuel flowmeter on the recirculation fuel line. Since the
proportion of fuel being recirculated was minimal, the utility was
willing to use a more conservative SO2 emissions calculation
in exchange for devoting fewer resources for the testing and
maintenance of the recirculation line fuel flowmeter.
Discussion of Proposed Changes
In today's proposal, EPA proposes to allow facilities to use only a
fuel flowmeter on the main fuel line from the oil tank if the amount of
oil recirculated is demonstrated to be less than 5.0 percent of total
fuel usage for each hour during the year.
Rationale
EPA believes that it is reasonable not to require installation,
certification and quality assurance of secondary fuel flowmeters in
cases where the amount of fuel to be combusted is a small proportion of
the total fuel used, and where knowing the exact volume of the
recirculated fuel makes little difference in the calculation of
emissions and heat input. EPA has allowed one utility to use an
estimate of the maximum oil usage at start-up, rather than requiring
the utility to install a return line oil flowmeter to measure the
startup fuel flow rate.
At first, EPA considered making the installation of a fuel
flowmeter on a recirculation fuel line optional. Presumably, if the
cost in lost SO2 allowances were greater than the cost of
installing and maintaining a fuel flowmeter, then a facility would
choose to use a fuel flowmeter on the recirculation fuel line. However,
many fuel flowmeters used under Appendix D for determining the
SO2 mass emission rate and the heat input rate are also used
to estimate the NO_X emission rate in lb/mmBtu under Appendix
E to part 75. The Appendix E procedures estimate hourly NO_X
emission rates using a correlation between measured NO_X
emission rates and heat input rates. The correlation is established
during a testing period. Therefore, subsequent to the test period, if
the hourly heat input values should become less accurate, it could
result in the estimated NO_X emission rates becoming less
accurate. Such loss in accuracy could occur if the heat input rates
during the initial testing period were based upon subtraction of
measured volumes or masses of recirculated fuel from the total fuel
flow rates, and then the facility later began estimating, rather than
measuring, the recirculated fuel volumes or masses. The potential
inaccuracy would increase if the proportion of recirculated oil to the
total flow rate of oil varies over time. The NO_X emission
rate can sometimes increase with increases in the heat input rate and
can sometimes decrease with increases in the heat input rate, depending
on the particular type of boiler; in addition, when certain types of
control equipment are installed, the NO_X emission rate may
not have any relationship with the heat input. Thus, an overestimation
of the heat input rate would sometimes result in the overestimation and
sometimes result in the underestimation of the NO_X emission
rate under Appendix E. For these reasons, EPA believes that there needs
to be some limits on the cases where a facility can choose not to use a
return fuel flowmeter.
In today's proposed rule, EPA is proposing that a facility may
choose to use only a fuel flowmeter on the main fuel line from the oil
tank and not install a return meter in those cases where the previously
measured proportion of oil from the recirculation line is less than or
equal to 5.0 percent of the unit's total oil usage during each hour of
the year. EPA believes that an error of 5.0 percent in the heat input
rate should be small enough that it will not significantly affect
accounting for the NO_X emission rate under Appendix E. An
analysis of emissions data from a gas-fired Appendix E unit with a
higher than average NO_X emission rate for gas (0.157 lb/
mmBtu) showed that a 5.0 percent increase in heat input would change
the quarterly average NO_X emission rate by only 3.17 percent
(0.152 vs. 0.157 lb/mmBtu) (see Docket A-97-35, Item II-B-19). At the
same time, EPA believes that an average proportion of 5.0 percent of
total fuel usage should provide relief for the most extreme situations
where it might cost more to perform quality assurance testing on a
return fuel flowmeter than the value of the allowances saved by
monitoring with the return flowmeter.
The Agency also considered whether it would be more appropriate to
determine the proportion of recirculated fuel on an hourly average
basis or on an annual average basis to determine if the returned fuel
was less than 5.0 percent of total fuel usage. The Agency concluded
that the proportion of fuel could be determined only if a return fuel
flowmeter were already installed on the recirculation fuel line. Thus,
there would appear to be little advantage to basing the proportion of
fuel on an annual basis. Hourly average fuel flow rate would also be
more directly related to the heat input rate used to calculate hourly
NO_X emission rate under Appendix E. EPA notes this is not
fully consistent with the objective of revising this provision, i.e.,
to exempt facilities from installation and operation of additional fuel
flowmeters. Therefore, the Agency believes it is better to base the
reduced fuel flow rate monitoring requirement either on actual
historical fuel flowmeter data or on some other method, as yet unknown,
that would yield a reasonable estimate of the average proportion of
fuel recirculated to the total amount of fuel used. At this time, the
Agency is unaware of what other methods could provide a reasonable
estimate of the average proportion of fuel recirculated to the total
amount of fuel used, either on an hourly or an annual basis.
Accordingly, the Agency would allow facilities to suggest methods
through the petitioning process of Sec. 75.66.
6. Fuel Flowmeter Testing
(a) Fuel Flowmeter Accuracy Tests.
Background
Sections 2.1.5 and 2.1.6 of Appendix D, as revised by the May 17,
1995 direct final rule, refer to calibration and recalibration of fuel
flowmeters. Section 2.1.5.2 gives procedures for a test of the
flowmeter accuracy by comparing a candidate flowmeter against another
flowmeter that has already been calibrated according to specified
procedures. If a flowmeter does not meet the specified accuracy, then
it would need to be recalibrated by adjusting it, then retested to
ensure it is reading accurately.
Some utilities have found confusing the terminology of
``calibration'' for a test that compares measurements from two
different flowmeters. Generally, the

[[Page 28089]]

term ``calibration'' is used to refer to adjustments made to a
flowmeter to ensure it is reading accurately. However, the type of test
described in section 2.1.5.2 is more like a relative accuracy test
audit than a calibration, in that it checks the flowmeter accuracy by
comparing the fuel flowmeter readings against readings from an outside
standard.
Discussion of Proposed Changes
To alleviate the confusion surrounding flowmeter testing, today's
proposal introduces the term ``flowmeter accuracy test.'' This
terminology is used in sections 2.1.5 and 2.1.6 of Appendix D.
Rationale
EPA believes that the term ``flowmeter accuracy test'' more clearly
reflects the nature of the test that is performed. Introducing this new
term also will clarify that the word ``calibration'' refers to
flowmeter adjustments, rather than to a comparative test between a
candidate flowmeter and a reference meter.
(b) Methods for Fuel Flowmeter Accuracy Testing.
Background
Section 2.1.5.1 of Appendix D, as revised by the May 17, 1995
direct final rule, includes a list of standards and procedures that may
be used to determine if a flowmeter is sufficiently accurate for use
under the Acid Rain Program. However, because of the large number of
different brands and kinds of fuel flowmeters, there are also many
manufacturers' procedures that are not explicitly permitted under part
75. Consequently, many Acid Rain certification applications for units
with fuel flowmeters have contained petitions under Secs. 75.23 and
75.66 for approval of other fuel flowmeter testing procedures. Among
those methods was AGA Report No. 7 for turbine flowmeters. This method
was incorporated by reference into part 75 in the November 20, 1996
final rule. In addition, another standard method that EPA approved
through petitions is American Petroleum Institute (API) Section 2,
``Conventional Pipe Provers,'' from Chapter 4 of the Manual of
Petroleum Measurement Standards, October 1988 edition (see reproduction
of this document in Docket A-97-35, Item II-D-10 (Attachment B)).
In the process of implementing part 75, many utilities have
commented on the problems of testing and calibrating fuel flowmeters.
Unlike CEMS or stack flow monitors, it is not always possible to
perform an accuracy test with the fuel flowmeter remaining in the pipe
where it is installed. Utilities have stated that certain fuel
flowmeters are extremely difficult to remove, send out for testing,
recalibrate, and then reinstall (see Docket A-97-35, Item II-E-22). In
addition, removing a fuel flowmeter from in-line may require stopping
flow of the fuel and possibly shutting down the unit, with negative
economic consequences (see Docket A-97-35, Item II-E-8). In addition,
if a facility needs to operate a unit while the flowmeter is being
tested at a laboratory, then no flow data will be available for the
fuel measured by the flowmeter unless the facility has a backup fuel
flowmeter. Utilities have petitioned for alternative quality assurance
procedures for fuel flowmeters in order to avoid the inconvenience and
expense of removing the fuel flowmeter and testing it (see Docket A-97-
35, Item II-D-9). Because of this, the Agency has been evaluating
various ways of testing a fuel flowmeter in-line (that is, still
installed in the pipe in its regular position).
Some utilities have suggested that an alternative way to check fuel
flowmeter accuracy would be to compare over time the ratio of the fuel
flowrate to unit output (``load''), measured either in electrical
generation in MWe or in steam flow in 1000 lb/hr (see Docket A-97-35,
Item II-E-21). A fuel flow-to-load comparison could be used to
determine if fuel flowmeter readings are still similar to the readings
obtained the last time the fuel flowmeter was tested against an outside
method. A significant change in the amount of fuel used at a load level
would call into question the validity of fuel flow readings from a
flowmeter. A fuel flow-to-load comparison could provide this check
without removal of the fuel flowmeter from its installed location,
which would be of considerable benefit to facilities.
Discussion of Proposed Changes
EPA is proposing to incorporate by reference the standard: American
Petroleum Institute (API) Section 2, ``Conventional Pipe Provers,''
from Chapter 4 of the Manual of Petroleum Measurement Standards. The
Agency also specifically requests comment on any other voluntary
consensus standards from standard setting organizations, such as API,
AGA, ASME, or ISO, that would be appropriate for incorporation by
reference into part 75. Any suggested methods should also be submitted
to the Agency as part of the comments to assist in the Agency's
evaluation.
Section 2.1.7 of Appendix D to today's proposed rule includes
provisions for an optional, supplemental quality assurance test for
fuel flowmeters using a ratio of the fuel flow rate and the unit load.
The fuel flow rate-to-load ratio comparison test would provide an
additional way to meet the requirement to periodically test fuel
flowmeter accuracy. This test would serve as a supplement to more
rigorous fuel flowmeter tests. These more rigorous tests include the
standards incorporated by reference under section 2.1.5.1 of Appendix D
that require the fuel flowmeter to be taken out of line and shipped to
a laboratory, and the ``master meter'' comparison procedures under
section 2.1.5.2 of Appendix D. For orifice-, nozzle-, and venturi-type
flowmeters, the more rigorous tests would include an inspection of the
primary element and an accuracy test on the transmitters or
transducers. If a facility performed and passed regular quarterly fuel
flow-to-load ratio testing, then it would need to perform the more
rigorous checks on monitor performance only once every 20 calendar
quarters (five years).
The fuel flow-to-load ratio test would require a facility to
establish a baseline period from a period of time when the fuel
flowmeter is known to be operating properly. After establishing this
baseline of accurate fuel flow data (or heat input rate data), a
facility would calculate the fuel flow-to-load ratio (or ``gross heat
rate'' (GHR)) during the baseline period. In each ``flowmeter operating
quarter'' that the fuel flowmeter operates after the baseline period is
completed, the facility would calculate the fuel flow-to-load ratio (or
GHR) for each hour the fuel flowmeter is used to report data. The
facility would compare the hourly fuel flow-to-load ratio (or GHR) to
the fuel flow-to-load ratio (or GHR) during the baseline period in
order to calculate the absolute value of the percentage difference for
each hour. Next, the facility would calculate the average percentage
difference for the quarter. If the percentage difference exceeded the
specified limits for the test, the fuel flowmeter would fail the test.
The key elements of the fuel flow rate-to-load evaluation are discussed
in the following paragraphs.
(1) Use of Gross Heat Rate-to-Load Ratio. Today's proposed rule
would allow a facility the option of calculating either the ratio of
the fuel flow rate to the gross generation in MWe or the steam flow
rate in thousands of pounds of steam per hour (``fuel flow-to-load
ratio'') or the ratio of the heat input rate to the gross generation in
MWe or the steam flow rate in thousands of pounds of steam per hour
(``GHR''). In order to allow a meaningful comparison, a facility would
use one of these two ratios consistently, both in calculating

[[Page 28090]]

an initial baseline ratio and in calculating hourly ratios during a
particular quarter. Equations D-1c and D-1e describe the calculation of
the fuel flow-to-load ratio for the baseline period and for hourly
values during a calendar quarter, respectively. For the GHR, the
respective equations are Equations D-1d and D-1f. These equations are
found in proposed sections 2.1.7.1 and 2.1.7.2 of Appendix D.
(2) Baseline Period for Fuel Flow-to-Load Ratio. The provisions for
calculating the baseline fuel flow-to-load ratio or gross heat rate are
found in section 2.1.7.1 of today's proposed rule. EPA is proposing
that the owner or operator of a facility would establish a baseline of
fuel flow rate (or heat input rate) data following a flowmeter accuracy
test under either section 2.1.5.1 or 2.1.5.2 of Appendix D, or
following both a transmitter or transducer accuracy test under section
2.1.6.1 of Appendix D and an inspection of a primary element for an
orifice-, nozzle-, or venturi-type fuel flowmeter under section
2.1.6.6. Throughout section 2.1.7 of today's proposed rule, these are
referred to as ``the most recent quality assurance procedure(s).'' The
baseline period of fuel flow rate (or heat input rate) data for a fuel
flowmeter to be tested under section 2.1.7 would use the first 168
hours of quality assured data measured by that flowmeter following the
most recent quality assurance procedure(s) for which: (1) only the fuel
measured by that fuel flowmeter is combusted (i.e., no co-firing of
fuels occurs); (2) the load is relatively stable and not ``ramping''
rapidly up or down; and (3) the load is sufficiently above the minimum
safe, stable operating load (unless low-load operation is normal for
the unit).
Today's proposal includes a limit to the length of time over which
the baseline period could extend. The baseline period of 168 hours
could not extend for longer than the end of the second calendar quarter
following the calendar quarter in which the most recent quality
assurance procedure(s) was performed. For orifice-, nozzle-, and
venturi-type fuel flowmeters, two quality assurance procedures would be
required: both a transmitter or transducer accuracy test under section
2.1.6.1 of Appendix D and an inspection of a primary element, such as
an orifice plate. For practical purposes, this means that the
transmitter or transducer accuracy test and the primary element
inspection would have to be completed either in the same calendar
quarter or in consecutive calendar quarters. If there were not 168
hours of quality-assured fuel flowmeter data from hours when a single
fuel is combusted, then the fuel flowmeter would not be allowed to be
tested using the fuel flow-to-load ratio as a supplement to other
quality assurance tests.
The 168 hours of quality-assured fuel flowmeter data next would be
averaged and divided by the average load, in megawatts or 1000 lb
steam/hr, during the same 168 hours to determine the baseline fuel
flow-to-load ratio (see Equation D-1c). Alternatively, the facility
could instead calculate the gross heat rate by averaging hourly heat
input rate during the 168 hours of the baseline period and by dividing
the average heat input rate by the average load during the same 168
hours (see Equation D-1d).
In cases where the fuel flowmeter is located on a common pipe
header, one fuel flow rate measurement could be associated with the
load from several units that receive fuel from the common pipe header.
In order to analyze the fuel flow-to-load ratio for a flowmeter on a
common pipe header, the load from all units receiving fuel from the
common pipe header would have to be combined for each hour, averaged
over the baseline period of 168 hours, and compared to the average fuel
flow rate during the baseline period. If a single unit receives fuel
from multiple pipes, each pipe with its own fuel flowmeter, then the
flow rates from all fuel flowmeters would have to be added together to
obtain the average fuel flowrate for the unit to be divided by the unit
load.
(3) Data Preparation and Analysis. In each flowmeter operating
quarter following the final quarter of the baseline period, all hourly
fuel flowmeter data would be compared to the load. A flowmeter
operating quarter would be a calendar quarter in which the unit
combusts the fuel measured by the fuel flowmeter for at least 168
hours. For each hour in which the fuel is combusted, the owner or
operator would calculate the fuel flow-to-load ratio (or GHR) (see
Equation D-1e for the fuel flow-to-load ratio and Equation D-1f for the
GHR). Hourly fuel flow rates on common pipe headers would be compared
to the sum of the loads from all units receiving fuel from the common
pipe header. For units with multiple pipes and multiple fuel
flowmeters, the total hourly fuel flow rate for the fuel would be
compared to the unit load.
Next, the facility would compare the hourly fuel flow-to-load
ratios (or GHRs) to the baseline fuel flow-to-load ratio (or GHR). The
absolute value of the percentage difference would be calculated for
each hour using Equation D-1g. Then the facility would calculate the
average value of the percentage difference for the quarter, using each
hourly percentage difference in Equation D-1h.
The quarterly average of the hourly percentage difference values
next would be compared to the limitation. For either the fuel flow-to-
load ratio or the GHR, Ef, the quarterly average of the hourly
percentage difference values would need to be no greater than 10.0
percent, unless the average of the hourly loads used for the analysis
was 50 MWe (or 500 klb/hr of steam), in which
case the limit on Ef would be 15.0 percent. If a fuel flowmeter were to
fail to meet this limit when using all data in the flowmeter operating
quarter, then the facility would have the option of excluding certain
hours. Otherwise, a failure to meet the 10.0 percent (or 15.0 percent,
if applicable) limit would be considered a failure of the fuel flow-to-
load ratio test.
(4) Optional Data Exclusions. As mentioned above, if a fuel
flowmeter's data would not meet the 10.0 percent (or 15.0 percent, if
applicable) limit on the quarterly average of the percentage difference
values, then a facility could opt to exclude certain hours of
unrepresentative fuel flow rate (or heat input rate) data and then
reanalyze the smaller set of data. The types of data that EPA proposes
as non-representative would be the same as the hours excluded during
the baseline period, including: (1) hours when the unit combusts
multiple fuels measured by multiple fuel flowmeters, such as co-firing
of gas and residual oil or co-firing of residual oil and diesel fuel;
(2) hours when the unit load is rapidly rising or falling, sometimes
referred to as ``ramping,'' to such a degree that the load in a given
hour differs by more than 15.0 percent from the load
during either the previous hour or the hour afterwards; or (3) hours in
which the unit load is in the lower 10.0 percent of the unit's
operating range, unless operation at those low levels is considered
normal for the unit. The facility would proceed to analyze the
remaining quarterly fuel flow rate or heat input rate values, provided
that there are at least 168 hours remaining for the quarter after
excluding non-representative hours. If less than 168 representative
hours remained after excluding the allowable hours, then a flow-to-load
or GHR test would not be required for that flowmeter for that flowmeter
operating quarter. If the fuel flowmeter data still failed to meet the
10.0 percent (or 15.0 percent, if applicable) limit on the quarterly
average of the percentage difference values after excluding the
allowable

[[Page 28091]]

hours, the flowmeter would fail the fuel flow-to-load ratio test.
(5) Consequences of Failing Fuel Flow-to-Load Ratio or GHR Tests.
There would be two primary consequences of failing a fuel flow-to-load
ratio or a GHR test. First, the data from the fuel flowmeter would no
longer be considered quality-assured. Thus, the facility would need to
invalidate data from the fuel flowmeter following the test. Proposed
section 2.1.7.4 of Appendix D specifies that the missing data
procedures of section 2.4.2 of Appendix D would be used to substitute
for the invalid data (unless a different fuel flowmeter is available
that has been tested for accuracy and has been demonstrated to meet the
accuracy specification), beginning with the first hour the fuel
measured by the fuel flowmeter is used during the quarter following the
flowmeter operating quarter in which the meter fails the fuel flow-to-
load ratio test. Second, in order to establish that the fuel flowmeter
is again operating properly and providing quality-assured data, the
facility would perform a fuel flowmeter accuracy test according to
sections 2.1.5.1 or 2.1.5.2 of Appendix D or, for orifice-, nozzle-,
and venturi-type flowmeters, a transmitter or transducer accuracy test
according to section 2.1.6.1 of Appendix D. In addition to the
transmitter or transducer test, orifice-, nozzle-, and venturi-type
fuel flowmeters would need to be further tested following a failed
flow-to-load or GHR test in order to ensure that the problem causing
the failure of the fuel flow-to-load ratio was a problem with the
transmitters or transducers.
Once the orifice-, nozzle-, or venturi-type flowmeter has been
recalibrated and passes a transmitter or transducer accuracy test
according to section 2.1.6.1 of Appendix D, the facility would perform
a shortened version of the fuel flow-to-load ratio test. The shortened
version of the test would use six to twelve hours of data following the
passed transmitter or transducer accuracy test. If the fuel flowmeter
passed the abbreviated fuel flow-to-load ratio test, then its data
would be considered valid, beginning with the time and date of the
passed transmitter or transducer accuracy test. However, if the fuel
flowmeter were to fail the abbreviated fuel flow-to-load ratio test,
then it would be necessary for the facility to inspect the primary
element for corrosion or damage. Furthermore, data would be considered
invalid until the orifice-, nozzle-, or venturi-type fuel flowmeter
passes an inspection of the primary element. Although data from the
flowmeter would be considered quality-assured after successful
completion of all required accuracy testing, visual inspections and
diagnostic tests, the baseline would have to be re-established no later
than the end of the second flowmeter operating quarter following the
quarter in which the quality assurance tests are completed.
Rationale:
EPA is proposing to incorporate by reference the standard: American
Petroleum Institute (API) Section 2, ``Conventional Pipe Provers,''
from Chapter 4 of the Manual of Petroleum Measurement Standards,
October 1988 edition. The Agency has already approved this method of
fuel flowmeter testing in response to a petition (see Docket A-97-35,
Item II-C-6). This is also a standard agreed to by API that is
traceable to NIST standards. The Agency has a general policy of
approving standards from technically knowledgeable groups such as the
Organization for International Standards (ISO), the American Society
for Testing and Materials (ASTM), the American Society of Mechanical
Engineers (ASME), the American Gas Association (AGA), the Gas
Processors Association (GPA), and API. EPA would also be willing to
incorporate additional standards by reference if commenters supply a
copy for consideration.
The Agency recognizes that it is difficult and sometimes costly to
take a fuel flowmeter out from its installation location to be tested
(see Docket A-97-35, Item II-E-22). Today's proposed rule would provide
the flexibility of an additional approach for testing fuel flowmeters
where they are installed. Today's proposal for a fuel flow rate-to-load
comparison test would allow facilities to assure the quality of their
fuel flow rate data without taking a fuel flowmeter out of line.
Several industry representatives suggested that a fuel flow rate-to-
load comparison was a useful approach to quality assuring data (see
Docket A-97-35, Items II-E-22, II-E-23). Some industry representatives
felt that a fuel flow rate-to-load ratio was straightforward and even
more representative than a stack flow rate-to-load ratio (see Docket A-
97-35, Item II-E-23).
In general, utilities have indicated that the idea of a fuel flow-
to-load ratio is an appropriate quality assurance test for fuel
flowmeters (see Docket A-97-35, Items II-D-30, II-D-41, II-E-33). Use
of the fuel flow-to-load ratio was first suggested to the Agency as an
alternative to annual orifice inspections (see Docket A-97-35, Item II-
E-22). One utility mentioned that the fuel flow-to-load ratio test
would be most useful if it allowed them to stretch the time between
transmitter or transducer accuracy tests on orifice-, nozzle-, and
venturi-type fuel flowmeters, as well as primary element inspections
and fuel flowmeter accuracy tests performed in-line against a ``master
meter'' or performed in a laboratory (see Docket A-97-35, Item II-D-
49).
Utilities have also indicated that they would prefer the provisions
of the fuel flow-to-load ratio test to be as similar as possible to the
stack flow-to-load ratio test in today's proposed rule (see Docket A-
97-35, Item II-E-33). This would be easier for facilities to comply
with because they would need to learn fewer new procedures, they could
use the same equations and algorithms in computer software or hand
calculations, and they could report information in a similar format. To
the extent possible, the Agency has incorporated this suggestion in
today's proposed rule. However, because monitoring with fuel flowmeters
is not identical to monitoring with stack volumetric flow monitors,
there are some differences in the procedures and in the data to be
recorded and reported.
Today's proposed rule would allow the quarterly fuel flow-to-load
ratio test as an optional supplement to flowmeter accuracy tests under
section 2.1.5.1 or 2.1.5.2 of Appendix D, transmitter or transducer
accuracy tests under section 2.1.6.1 of Appendix D for orifice-,
nozzle-, and venturi-type fuel flowmeters, and visual inspections of
the primary element required under section 2.1.6.6 of Appendix D for
orifice-, nozzle-and venturi-type fuel flowmeters. These more rigorous
fuel flowmeter quality assurance procedures would still be required at
least once every 20 calendar quarters (five years), even if the
procedures of section 2.1.7 of Appendix D were followed. The Agency has
proposed a quarterly fuel flow-to-load ratio test for several reasons:
(1) this is consistent with the provisions of the proposed volumetric
stack flow-to-load ratio test in today's proposed rule; (2) the test
involves examining data more closely when preparing quarterly reports;
and (3) a quarterly test allows facilities to find problems in fuel
flowmeter data before an entire year has passed. The Agency also
considered requiring the fuel flow-to-load ratio to be used more
frequently than quarterly, perhaps daily; however, this would require
facilities to spend far more time and effort in evaluating data at
different times during the quarter than they may do currently,
particularly for small, infrequently operated units. In addition, many
utilities claim that fuel

[[Page 28092]]

flowmeters tend to be stable, and therefore little change would be
expected over short time periods such as a day (see Docket A-97-35,
Item II-E-33).
EPA is proposing that the optional fuel flow-to-load ratio test
could serve as a supplement to other quality assurance procedures for
fuel flowmeters for up to 20 calendar quarters (five years). EPA is
proposing a time period of 20 calendar quarters for the following
reasons. First, it is similar to the current provision in section
2.1.5.2 of Appendix D, which allows a reference fuel flowmeter to be
accuracy tested as seldom as once in five calendar years if comparison
with an in-line ``master'' flowmeter shows less than a 1.0 percent
difference in their flow rates. Second, a five-year test cycle offers
certain administrative advantages. For instance, fuel flowmeters used
to provide heat input data for the heat input-versus-load correlation
of Appendix E could be accuracy-tested before each Appendix E test
(i.e., once every five years). In addition, a five-year period would
ensure that fuel flowmeters are tested by the time the unit's operating
permit is renewed. The 20 calendar quarter (five-year) period is
consistent with the provisions for reduced three-level flow RATAs for
stack flow monitors. The 20 calendar quarter (five-year) period between
tests is also consistent with the proposed time between quality
assurance tests for fuel flowmeters that are used very infrequently.
Repeating the periodic quality assurance procedures for fuel flowmeters
at least every five years would catch slow, long-term changes in heat
rates mentioned by a facility and would allow a facility to update its
baseline data periodically (see Docket A-97-35, Item II-D-49). Finally,
allowing the option of a 20 calendar quarter (five-year) period between
more rigorous quality assurance procedures would be safer and less
costly than annual testing, while, in coordination with quarterly fuel
flow-to-load ratio testing, still providing assurance of the quality of
the data.
(1) Use of Gross Heat Rate or Flow-to-Load Ratio. Today's proposed
rule would allow a facility the option of calculating either the ratio
of the fuel flow rate to the gross generation in MWe or the steam flow
rate in thousands of pounds of steam per hour (``fuel flow-to-load
ratio'') or the ratio of the heat input rate to the gross generation in
MWe or the steam flow rate in thousands of pounds of steam per hour
(``gross heat rate'' or ``GHR''). One utility suggested that, because
the load is created based upon a number of factors in addition to the
fuel flow rate, such as the gas heat rate (i.e., gross calorific
value), a ratio of the heat input to the unit load would be a better
test than the ratio of the fuel flow rate to the unit load (see Docket
A-97-35, Item II-D-50). In addition, some utilities pointed out that
the Agency allows facilities to use either a stack flow-to-load ratio
or a heat input-to-load ratio (gross heat rate) as a diagnostic test on
stack volumetric flow monitors, through Policy Manual Question 13.15
(see Docket A-97-35, Item II-I-9). The Agency agrees that the heat
input-to-load ratio (GHR) is also a technically appropriate check on
the performance of fuel flowmeters. Therefore, today's proposal
includes options for both the fuel flow-to-load ratio and the GHR.
(2) Baseline Period for Fuel Flow-to-Load Ratio or GHR. When using
this type of comparison test, it is important to establish a baseline
of reliable data to which hourly data can later be compared. For the
stack volumetric flow-to-load ratio, the baseline of reliable data
consists of data from the reference method for flow, Method 2 of
Appendix A to 40 CFR part 60. However, there is no universally
applicable test for flowmeters that is performed in-line with a
reference method while the unit is operating, parallel to the flow
RATA. EPA asked several utilities what could be a source of baseline
data to which the fuel flowmeter could later be compared. One utility
suggested using fuel flowmeter readings during a time when the unit is
operating at a steady load, such as when the unit undergoes Appendix E
testing for a NO_X-versus-heat input correlation or when a
NO_X CEMS undergoes a normal level RATA (see Docket A-97-35,
Item II-D-41). A second utility recommended that the baseline be
established just after performing a transmitter calibration, i.e.,
after performing a quality assurance test on the fuel flowmeter (see
Docket A-97-35, Item II-D-49). The Agency believes that using fuel
flowmeter data taken immediately following a flowmeter quality
assurance test would be most likely to be accurate and representative
of proper operation of the fuel flowmeter. Flowmeter quality assurance
tests might include any of the methods incorporated by reference in
section 2.1.5.1 of Appendix D; meter testing against a certifiable
``master'' meter under section 2.1.5.2 of Appendix D; or transmitter or
transducer accuracy testing under section 2.1.6.1 of Appendix D, and
inspection of a primary element for an orifice-, nozzle-, or venturi-
type fuel flowmeter under section 2.1.6.6 of Appendix D. This approach
is proposed in today's rule.
The utilities supporting the idea of using fuel flowmeter data
taken immediately after a flowmeter quality assurance test have
suggested that it would be important to have a fairly large number of
hours in the baseline, on the order of 100 or more, to ensure that the
baseline period is representative of typical operation (see Docket A-
97-35, Item II-E-33). In today's rule, EPA is proposing to use the
first 168 hours of quality assured data measured by that flowmeter for
which: (1) only the fuel measured by that fuel flowmeter is combusted;
(2) the unit load is not significantly ``ramping'' up or down; and (3)
the unit load is safely above the minimum safe, stable load. The Agency
believes that a baseline period containing 168 hours of data is
sufficiently long to be representative of different unit operating
conditions that may occur later. This specific time period is
consistent with the minimum number of hours that a unit combusts a fuel
before the quarter counts toward the deadline for the next quality
assurance test, and with the minimum number of hours that a unit
combusts a fuel before a quarter needs to be evaluated using the fuel
flow-to-load ratio. Certain hours would be excluded from the baseline
(i.e., periods of co-firing, unstable, or low load), because the fuel
flow-to-load ratio or GHR would tend to be less reliable during those
periods.
Today's proposal would also limit the baseline period so that it
may extend no more than two quarters beyond the quarter in which the
flowmeter passes its accuracy tests. The Agency has concerns that if
the baseline data were to extend longer than this, the performance of
the fuel flowmeter might degrade. In order for the baseline data to
reflect fuel flow rate data that are most likely to be accurate, the
Agency is proposing that the fuel flow rate or heat input rate data
used in the baseline period must either be obtained in the calendar
quarter in which the quality assurance procedure is performed, or
within two calendar quarters after the QA test. The Agency considered
limiting the time period to the same calendar quarter as the quality
assurance procedure or to one flowmeter operating quarter beyond the QA
test. However, because a quality assurance procedure may be conducted
at any time during a quarter, it could be difficult for a facility to
collect 168 hours of fuel flowmeter data after a quality assurance
procedure in the same calendar quarter or even (for infrequently
operated units that ramp

[[Page 28093]]

up and down often) in the next calendar quarter.
For orifice-, nozzle-, and venturi-type fuel flowmeters, two
quality assurance procedures would be required prior to collecting the
baseline data: (1) a transmitter or transducer accuracy test, and (2)
an inspection of a primary element. The Agency considered whether these
two quality assurance procedures should be separated and whether the
baseline period could simply be based upon a time period after the most
recent quality assurance procedure. The Agency believes that the
baseline period data would be more reliable if they were taken shortly
after completing both quality assurance procedures for orifice-,
nozzle-, and venturi-type fuel flowmeters. Using the same time period
for both tests simplifies administration of the fuel flow-to-load ratio
test. EPA also notes that a unit does not need to be operating in order
to perform the tests; thus, it should not be burdensome for a facility
to plan to coordinate the two quality assurance procedures.
(3) Data Preparation and Analysis. The proposed procedures for data
preparation and analysis for the fuel flow-to-load ratio are similar to
those for the volumetric stack flow-to-load ratio. Equations of the
same form as those for the stack volumetric flow-to-load ratio are used
to calculate the hourly fuel flow-to-load ratio, the hourly absolute
value of the percentage difference between the baseline fuel flow-to-
load ratio and the hourly fuel flow-to-load ratio, and the quarterly
average percentage difference. Common pipe headers would be treated in
the same way as common stacks. If there were multiple units associated
with a single fuel flowmeter or flow monitor, the total load from all
units would be summed before the flow rate data are divided by the load
data to calculate the flow-to-load ratio. Fuel flowmeters on multiple
pipes would be treated in the same way as multiple stacks associated
with a single unit. If there are multiple fuel flowmeters or flow
monitors associated with a single unit, the flow rates from all fuel
flowmeters for the same fuel or all flow monitors would be added
together before the flow rate data are divided by the load data to
calculate the flow-to-load ratio.
Certain aspects of the volumetric stack flow-to-load ratio test are
not the same for the fuel flow-to-load ratio test. For example, the
volumetric stack flow-to-load ratio test requires the facility to
screen out those hours when the unit operates further than 10.0 percent
away from the average load during the most recent normal-load flow
RATA. As was discussed previously, there is no equivalent of an in-line
flow RATA for fuel flowmeters. EPA does not believe that there is a
need to screen out hours for the fuel flow-to-load test when the unit
operates at a load somewhat less than or greater than normal. Some
facilities have indicated that the fuel flow-to-load ratio or GHR based
on fuel flow readings is less variable over different loads than the
volumetric stack flow-to-load ratio (see Docket A-97-35, Items II-E-33
and II-D-98). However, preliminary evidence has also indicated that the
fuel flow-to-load ratio or GHR can be significantly different at very
low operating loads than at other load levels (see Docket A-97-35, Item
II-A-5). For this reason, EPA is proposing to allow hours in which the
unit load is within the lower 10.0 percent of the range of operation to
be excluded from both the baseline data and the quarterly flow-to-load
or GHR analysis, unless such low loads are considered normal for the
unit.
Another feature of the volumetric stack flow-to-load ratio test
that differs from the fuel flow-to-load ratio test is the treatment of
bias-adjusted data. Fuel flow rate data are never adjusted for bias.
There is no bias test for fuel flowmeters. Bias-adjustment of data is
an issue for the volumetric stack flow-to-load ratio test because bias-
adjusted data has already been adjusted to make it more consistent with
the value of the reference method data. Thus, bias-adjusted volumetric
stack flow data must meet a stricter quarterly average percentage
difference of 10.0 percent from the reference flow-to-load ratio,
whereas the allowable difference is 15.0 percent when unadjusted
volumetric stack flow data are used. (See discussion of stack flow-to-
load test in Section III.M. of this preamble.) EPA notes that since the
same fuel flow meter is used to produce both the baseline data and the
quarterly data, the fuel flow-to-load ratio is more closely analogous
to the use of bias-adjusted volumetric flow data. Therefore, the limit
on the quarterly average percentage difference from baseline for fuel
flow rate data should be at least as stringent as that for bias-
adjusted volumetric flow data (10.0 percent). Information provided by
facilities on the gross heat rate derived from fuel flow rate data have
shown less variability than the corresponding stack heat rate (see
Docket A-97-35, Item II-D-98). Based upon this information, EPA is
proposing a limit of 10.0 percent on Ef, the quarterly
average percentage difference from the baseline for the quarterly flow
rate-to-load or GHR evaluation. EPA considered whether it would be
appropriate to set a different limit for smaller units, as was done for
the stack flow-to-load test. Analysis of some preliminary fuel flow-to-
load data has shown that for lower loads (e.g., < 50 MWe), the flow-to-
load ratio is quite sensitive to small changes in load (see Docket A-
97-35, Item II-A-5). This indicates that it would be appropriate to set
a higher limit for smaller units. Therefore, today's rule proposes a
limit of 15.0 percent on the value of Ef when the quarterly
average load used for the data analysis is 50 megawatts or less (or
500 klb steam per hour). The Agency solicits comment on the
15.0 percent limit for loads less than or equal to 50 megawatts.
(4) Optional Data Exclusions. As for volumetric stack flow
monitors, if a fuel flowmeter's data would not meet the limit on the
percentage deviation from the baseline, then a facility could opt to
exclude certain hours of unrepresentative fuel flow rate (or heat input
rate) data and then reanalyze the smaller set of data. The hours of
data that EPA proposes to view as non-representative for fuel
flowmeters are: (1) hours when the unit combusts multiple fuels; (2)
hours when the unit load in a given hour would differ by more than
15.0 percent from the load during either the previous hour
or the subsequent hour; or (3) hours when the load is very close to the
minimum safe, stable load (unless operation in that range is normal).
The baseline period for fuel flowmeters and the data used for the
quarterly flow-to-load or GHR analyses would include only those hours
when a single fuel is combusted--the fuel measured by the fuel
flowmeter. If the quarterly fuel flow rate data included hours when
multiple fuels are co-fired, the fuel flow-to-load ratio or GHR for the
fuel flowmeter being tested would be biased low. This could result in a
failure of the flow-to-load test or GHR evaluation. Today's proposed
rule would also allow a facility to exclude from the baseline data and
the quarterly analyses those hours that are not representative because
the unit's load is changing rapidly. Specifically, hours could be
excluded when the unit load in a given hour would differ by more than
15.0 percent from the load during either the previous hour
or the hour afterwards. There will be a lag in the time between when
electricity is generated and registered as load and the time that the
fuel flowmeter measures the fuel that is combusted to generate the
load. Therefore, during an hour when the load changes rapidly, the fuel
flow rate will not necessarily be changing by the same amount or in the

[[Page 28094]]

same direction. At least one utility has suggested that the Agency
consider such an exclusion for the proposed fuel flow-to-load ratio
test (see Docket A-97-35, Item II-D-41).
In general, the fuel flow is directly proportional to load, with a
linear graphical relationship. However, this is not always the case at
extremely low loads (see Docket A-97-35, Items II-E-33, II-D-98).
Therefore, today's proposed rule would allow certain low-load hours to
be excluded from the flow-to-load baseline and quarterly data analyses.
Specifically, loads in the lower 10.0 percent of the ``range of
operation'' of the unit, (as that term is defined in proposed section
6.5.2.1 of Appendix A in today's proposal) could be excluded, unless
such loads are considered normal for the unit.
Today's proposed rule, in section 2.1.7 of Appendix D, would also
exempt a fuel flowmeter from the fuel flow-to-load ratio test in a
quarter when a more rigorous quality assurance test is performed. This
is unlike the volumetric stack flow-to-load ratio, which is required
each QA operating quarter, including quarters when the flow monitor is
tested with a RATA (provided, of course, that sufficient data for the
analysis are obtained after the RATA).
(5) Consequences of Failing the Fuel Flow-to-Load Ratio Test. The
consequences of failing the fuel flow-to-load ratio test would be
similar to the consequences of failing quality assurance tests in
general for fuel flowmeters. Data from the fuel flowmeter would no
longer be considered quality assured. Because the fuel flow-to-load
ratio test is only performed at the end of a quarter, the facility
would invalidate data from the fuel flowmeter beginning with the first
hour in the quarter after the quarter in which the meter fails the fuel
flow-to-load ratio test. In order to establish that the fuel flowmeter
is operating properly and providing quality assured data again, the
facility would perform a flowmeter accuracy test or (for orifice-,
nozzle-, and venturi-type flowmeters) a transmitter or transducer
accuracy test. The Agency believes it is appropriate to perform an
accuracy test if the fuel flow-to-load ratio test is failed, because in
such cases the facility has had the benefit of postponing the accuracy
test based upon the assumption that the fuel flowmeter has continued to
measure accurately and consistently with its operation during the
baseline period.
Note that for orifice-, nozzle-, and venturi-type fuel flowmeters,
a transmitter/transducer test alone would not suffice to demonstrate
that the flowmeter is back in control. The owner or operator would
still need to ensure that the cause of the failed fuel flow-to-load
ratio test was a problem with the transmitters or transducers rather
than a problem with the primary element. Sudden changes in flowmeter
performance are likely to be caused by a problem with transmitters (see
Docket A-97-35, Item II-D-33). However, it cannot be assumed that the
transmitters are solely responsible for degradation in monitor
performance. In order to verify that the primary element is not
contributing additional error to the fuel flow measurements because of
corrosion, a facility would conduct an abbreviated (6 to 12 hour)
version of the fuel flow-to-load ratio test, similar to the diagnostic
test for volumetric stack flow monitors in Policy Manual Question 13.15
(see Docket A-97-35, Item II-I-9). The Agency believes that this
abbreviated fuel flow-to-load ratio test would provide additional
assurance that the fuel flowmeter is indeed operating properly. In
addition, it would be more timely than waiting for another calendar
quarter to pass to repeat the fuel flow-to-load ratio. The abbreviated
test would also be less burdensome than removing the primary element
from the fuel pipe. EPA believes the abbreviated fuel flow-to-load
ratio test strikes a reasonable balance by providing some additional
quality assurance in a timely manner. If the orifice-, nozzle-, or
venturi-type fuel flowmeter failed the abbreviated fuel flow-to-load
ratio test, then it would appear that the primary element may also have
a problem. Therefore, upon failure of an abbreviated fuel flow-to-load
ratio test, the facility would be required to inspect the primary
element and to repair or replace it, as necessary.
The rules for data validation upon failure of the fuel flow-to-load
ratio are not parallel with the procedures for data validation
following failure of the volumetric stack flow-to-load ratio test in
that there is no conditional validation of data. A number of utilities
have emphasized that they wish to spend less time and effort preparing
and evaluating quarterly reports for units using Appendix D, which are
generally smaller and less frequently operated than coal-fired units or
oil-fired units that choose to use CEMS (see Docket A-97-35, Item II-E-
33). The concept of conditional data validation for fuel flowmeters is
not consistent with this objective, because it would introduce
additional complexity into the process, would require significantly
more time and resources to quality-assure the data, and might require
additional DAHS programming. Therefore, the Agency is not proposing the
use of conditional data validation for fuel flowmeters.
(c) Fuel Flowmeter Quality Assurance Testing Frequency
Background
Section 2.1.6.1 of Appendix D, as revised by the May 17, 1995
direct final rule, requires regular quality assurance
``recalibrations'' (accuracy tests) of fuel flowmeters at least
annually (once every four calendar quarters). For fuel flowmeters that
were not used on a regular basis, such as fuel flowmeters used to
measure the usage of emergency fuel or backup fuel, or flowmeters
installed on peaking units, owners or operators are allowed to do
flowmeter accuracy tests once every four quarters when the unit
actually combusts the fuel measured by the flowmeter, rather than once
every four calendar quarters. Flowmeters can be retested either by
using one of the methods incorporated by reference in section 2.1.5.1
of Appendix D to part 75 or by an in-line comparison of the fuel
flowmeter against a ``master'' fuel flowmeter using the procedure in
section 2.1.5.2 of Appendix D.
Some utilities have expressed concern about the annual fuel
flowmeter testing requirement (see Docket A-97-35, Items II-D-20, II-E-
13, II-E-14). In many cases, it is neither practical nor cost-effective
to modify the fuel pipes (e.g., to install a parallel length of pipe)
to allow installation of a master fuel flowmeter for comparison
testing. Thus, most utilities must remove a fuel flowmeter from the
pipe and return it to a laboratory or to the manufacturer to be
retested. In some cases, especially for oil flowmeters, this can be
difficult.
Some utilities have raised the issue of whether there should be a
minimum time period that a fuel flowmeter is used before a quality
assurance test is required. For instance, a utility might test its
unit's burners once each quarter for a few hours to ensure that the
unit can be operated when needed and may not operate for the rest of
the quarter. Under the current rule, the fuel flowmeter would have to
be quality assurance tested after four such operating quarters, even
though the flowmeter was only used for a few hours in those calendar
quarters.
Discussion of Proposed Changes
Today's proposed rule includes a provision that only those calendar
quarters in which the fuel measured by the fuel flowmeter is combusted
for at least 168 hours would count toward determining the next quality
assurance test deadline. The 168-hour time period

[[Page 28095]]

is roughly equivalent to one week of operation while combusting the
fuel measured by a particular fuel flowmeter. A calendar quarter in
which the fuel measured by a fuel flowmeter is combusted for 168 hours
or more would be called a ``flowmeter operating quarter.'' For example,
if a unit combusted oil for 200 hours in the first calendar quarter of
the year, 10 hours in the second calendar quarter, 250 hours in the
third calendar quarter, and 100 hours in the fourth calendar quarter,
then only the first and third calendar quarters would be considered
flowmeter operating quarters for the oil flowmeter. Only the first and
third calendar quarters would count toward determining the deadline for
the next required oil flowmeter accuracy test.
In today's proposed rule, each fuel flowmeter would need to be
accuracy tested at least once every four flowmeter operating quarters.
However, the deadline for testing infrequently-used meters could not be
extended indefinitely. No more than 20 calendar quarters (five years)
would be allowed to elapse between successive flowmeter accuracy tests,
regardless of the number of ``flowmeter operating quarters'' that have
elapsed since the last test. The interval between successive quality
assurance tests could also be extended for up to 20 calendar quarters
if the quarterly fuel flow rate-to-load procedures in proposed section
2.1.7 of Appendix D were implemented.
Rationale
In evaluating the frequency of fuel flowmeter accuracy testing, EPA
considered simply extending the less strict requirement for fuel
flowmeter quality assurance testing for peaking units, backup fuel, and
emergency fuel to apply to all units and all fuel flowmeters. Thus,
quality assurance testing would be required once every four quarters in
which the unit combusted the fuel measured by the flowmeter.
One industry representative recommended that the Agency require
fuel flowmeter calibrations once every four unit operating quarters,
where a unit operates at least 168 hours in the quarter (see Docket A-
97-35, Item II-E-13). This approach would treat all fuel flowmeters the
same, whether they were used for primary, emergency, or backup fuel.
Another utility suggested that the Agency consider creating some
sort of diagnostic test comparing the flow rate of the fuel flowmeter
to the unit load (generation) to determine whether the fuel flowmeter
readings are degrading over time, rather than specifying a particular
frequency for accuracy testing (see Docket A-97-35, Item II-E-22).
Although this suggestion was originally referring to problems with
corrosion of an orifice plate, such a test could also be used for other
types of fuel flowmeters as a check on the quality of fuel flowmeter
data.
The Agency also considered extending the typical time between
accuracy tests to the equivalent of two years. This time was suggested
by a member of the AGA subcommittee responsible for the drafting of AGA
Report No. 7 for turbine-type flowmeters (see Docket A-97-35, Item II-
E-17). The Agency also considered extending the typical time between
accuracy testing to 12 calendar quarters--the equivalent of three
years. Three years is the period of time that records must be retained
in a file at the source under Sec. 75.54 (or proposed Sec. 75.57).
The Agency also considered allowing fuel flowmeters to continue for
up to five calendar years between accuracy tests. This is similar to
the current provision in section 2.1.5.2 of Appendix D, which allows a
reference fuel flowmeter to be accuracy tested as seldom as once in
five calendar years, if the in-line comparison with a master fuel
flowmeter shows a 1.0 percent or less difference in their flow rates. A
five-year test cycle offers certain administrative advantages. For
instance, fuel flowmeters used to provide heat input data for the heat
input-versus-load correlation of Appendix E could be accuracy-tested
before each Appendix E test (i.e., once every five years). In addition,
the five calendar-year period would ensure that fuel flowmeters are
tested by the time the unit's operating permit is renewed. Facilities
might find this time cycle easier to determine than a time period based
upon a number of calendar quarters. However, test data would need to be
retained for five years, rather than for three years, the recordkeeping
period for most records under part 75. However, the Agency is not
proposing this option because five years is far too long a period of
time to allow a unit to continue with no checks at all upon the quality
of its data. Such an approach would allow the use of data from a fuel
flowmeter that potentially had been reading inaccurately for the
previous five years.
Another option that EPA evaluated was to establish different fuel
flowmeter quality-assurance testing frequencies depending on the fuel
measured by the fuel flowmeter. Under this approach, oil flowmeters
would need to be tested every four calendar quarters in which oil was
combusted. Gas flowmeters would only need to be tested once every five
years. The two fuels would be treated differently because units emit
less NO_X and far less SO2 when combusting gas
than when combusting oil. In addition, gaseous fuels, particularly
pipeline natural gas, should be less corrosive; therefore, a gas
flowmeter should be less likely to degrade than an oil flowmeter.
EPA believes that today's proposed approach to reducing the fuel
flowmeter quality assurance testing frequency takes into account many
of the concerns raised by utilities. All unit types and fuel types
would have the same frequency of testing. This would avoid confusion
that could follow from an approach that set different requirements for
fuels or units that are used less frequently. A group of utilities had
indicated that they prefer a more consistent approach (see Docket A-97-
35, Item II-E-13). Under today's proposal, infrequently-used fuel
flowmeters (e.g., meters for backup fuel or emergency fuel) would only
need to be calibrated once every five years. When a facility renews its
operating permit, the permitting agency could verify that all fuel
flowmeters have been tested at least once in the previous five years.
The minimum period of 168 hours of fuel flowmeter usage which
defines a ``flowmeter operating quarter'' is consistent with the
definition of a ``QA operating quarter'' in Appendix B in today's
proposed rule for the quality assurance of CEMS. The Agency believes
that using a consistent minimum number of hours in a calendar quarter
for both CEMS and fuel flowmeters will make implementation easier for
facilities and air regulatory agencies. In addition, 168 hours should
be a sufficiently long period of time to ensure that short-term usage
of backup fuel or emergency fuel or short-term tests of a unit do not
trigger unnecessary quality assurance testing.
Today's proposed rule would also provide more flexibility in the
methods that could be used for fuel flowmeter quality assurance
testing. As discussed above in Section III.P.2 of this preamble, a new
testing procedure has been proposed that would allow a facility to test
flow rate-to-load ratio of the fuel flowmeter while leaving it
installed. Thus, the Agency believes that the overall burden of fuel
flowmeter testing has been significantly reduced. In addition to the
reduced frequency of testing discussed above, the Agency believes the
less burdensome testing procedures should address concerns of the
regulated community.
The Agency requests comment on whether facilities would prefer to
base

[[Page 28096]]

the frequency of fuel flowmeter quality assurance testing on the type
of fuel used or the amount of time the fuel flowmeter is used. Under
the first approach, gas flowmeters would receive greater regulatory
relief. Under the second approach, which is being proposed in today's
rule, infrequently-used flowmeters (typically oil flowmeters) would
receive greater regulatory relief.
(d) Orifice, Nozzle, and Venturi Visual Inspections
Background
Section 2.1.6 of Appendix D, as revised in the May 17, 1995 direct
final rule, created special provisions for the ongoing quality
assurance testing of orifice fuel flowmeters. Orifice-,
nozzle-, and venturi-type fuel flowmeters are designed and installed
within a set of physical specifications, such as the orifice diameter
(see Docket A-97-35, Item II-D-13). Maintaining these physical
specifications determines the flowmeter's ability to read accurately.
Thus, it is not necessary to take an orifice-, nozzle-, or venturi-type
flowmeter out of line and send it to a laboratory to determine its
accuracy.
After installation of an orifice-, nozzle-, or venturi-type
flowmeter is complete, the two major factors that contribute to error
in flow readings are: drift in the transmitters (or transducers) which
determines the total pressure, differential pressure and temperature,
and corrosion of the primary element (e.g., the orifice plate) itself.
Quality assurance testing of the transmitters is discussed in the next
section of the preamble. In order to identify cases where error might
result from corrosion of the orifice plate, the May 17, 1995 direct
final rule added a requirement for an annual visual inspection of the
orifice plate. If an orifice plate fails the inspection, then the
facility must perform a test on the transmitters during the next
calendar quarter. A procedure for visual inspections is given in
Appendix B of part 2 of American Gas Association (AGA) Report No. 3,
which is one of the accepted standards for installation and use of
orifice flowmeters.
Some facilities have expressed concern with the frequency of visual
inspections (see Docket A-97-35, Items II-D-20, II-E-13, II-E-14). This
process must be done either with a tool, such as a boroscope, or else
the primary element must be removed from the pipe and lifted out to be
inspected. In the case of large, heavy orifices, it is necessary to use
a crane to remove the orifice. Fuel must not be flowing through the
pipe while the orifice plate is being removed (see Docket A-97-35, Item
II-E-8).
The current provisions of Appendix D to part 75 do not explicitly
state the consequences of failing a quality assurance test. Section
2.1.5.1 of Appendix D states that if a fuel flowmeter exceeds the
flowmeter accuracy of 2.0 percent of the upper range
value, then the flowmeter may not be used under part 75. Section
2.1.5.2 states that if a fuel flowmeter's accuracy exceeds
2.0 percent of the upper range value, then the flowmeter must be
recalibrated to meet that accuracy, or it must be replaced with another
flowmeter that meets the specification. Neither section explicitly
states the impact upon the validity of data if a test is failed.
However, if fuel flowmeter systems are to be treated parallel with
continuous emission monitoring systems under Sec. 75.21(e)(2), the
consequences of failing a quality assurance test for a fuel flowmeter
or an inspection of the primary element should result in the monitor
being considered out-of-control and the data being considered invalid.
In section 2.1.6.1 of Appendix D, the specific consequence of
failing a visual inspection of the primary element is that the
transmitters must be tested in the following calendar quarter, rather
than waiting until the regular annual calibration is required. However,
no mention is made of any mandatory corrective action(s) to eliminate
the corrosion problem.
Discussion of Proposed Changes
Section 2.1.6.6 of Appendix D in today's rulemaking proposes to
require visual inspections of primary elements (i.e., orifice, nozzle
or venturi) at the frequency recommended by the manufacturer or once
every three years, whichever is more frequent. The Agency solicits
comment on the proposed frequency of visual inspections.
The proposed rule would also explicitly require repair or
replacement of the primary element and invalidation of data when a
visual inspection is failed. Once the primary element is replaced or
repaired, the new or repaired primary element would have to demonstrate
that it meets the overall flow rate accuracy of 2.0
percent of the upper range value. This could be demonstrated by showing
that the new or repaired primary element meets the design and
installation requirements of AGA Report No. 3 or ASME MFC-3M, the same
methods required for initial certification. Alternatively, the flow
rate accuracy could be demonstrated by testing the fuel flowmeter
against a reference fuel flowmeter using the provisions of section
2.1.5.2 of Appendix D. Finally, whenever a primary element is repaired,
the fuel flowmeter transmitters would also have to be tested before the
fuel flowmeter is used to provide quality assured data.
Rationale
During the process of reviewing certification applications for
units using orifice flowmeters, the Agency learned of one plant where
orifice corrosion was a serious problem. This utility had an orifice
flowmeter which had been installed in the 1960's. This utility did not
have documentation of the standard used to install the orifice as a
demonstration of the meter's accuracy. In order to qualify for
certification, the utility inspected the orifice. The utility personnel
discovered that the orifice had been completely eaten away and was
incapable of reading the flow rate (see Docket A-97-35, Item II-E-22).
The utility replaced the orifice before it was able to have its fuel
flowmeter certified. In addition, it was required to invalidate the
flow rate data from the orifice meter and substitute for the missing
data. Based upon this experience, the Agency believes that corrosion of
an orifice can be a problem, and that in severe cases of corrosion,
replacement of the orifice is necessary.
Despite this, many utilities have expressed concern over the
difficulty of removing an orifice from place for visual inspection (see
Docket A-97-35, Items II-D-20, II-E-13, II-E-14), because removal
requires halting the flow of gas through the pipeline in order to
remove the orifice, which can be expensive (see Docket A-97-35, Item
II-E-8).
Utilities have provided the Agency with several suggestions for
reducing the frequency of primary element inspections. One industry
group recommended that the Agency reduce the inspection frequency to
once every five years, to be coordinated with renewal of the plant's
operating permit under title V of the Act (see Docket A-97-35, Items
II-D-20, II-E-13, and II-E-14). One utility representative mentioned
that most orifice manufacturers recommend an inspection once every
three years; thus, he recommended that the Agency require visual
inspections the earlier of once every three years or the time period
specified by the manufacturer (see Docket A-97-35, Item II-D-41).
Another utility suggested that the Agency consider creating some sort
of diagnostic test comparing the flow rate of the fuel flowmeter to
unit load (generation) to determine whether the fuel flowmeter readings
are degrading

[[Page 28097]]

over time, rather than specifying a particular time period (see Docket
A-97-35, Item II-E-22).
EPA agrees that it would be helpful to facilities to reduce the
frequency of visual inspections from their current annual frequency.
Having considered all of the options suggested by the utilities, the
Agency is proposing that the primary element of all nozzle, venturi and
orifice fuel flowmeters be visually inspected at the frequency
recommended by the manufacturer or once every three years, whichever is
the more frequent. The Agency believes that up to three years between
visual inspections is a technically sound period of time that will
assure the quality of fuel flow rate data, while providing regulatory
relief from the current annual requirement.
The Agency also has reconsidered the consequences of failure of a
visual inspection. The May 17, 1995 direct final rule added a
requirement to test a flowmeter's transmitters in the calendar quarter
following a failed inspection, but the rule does not explicitly require
that the primary element be repaired or replaced, nor does it
explicitly require data from the fuel flowmeter to be invalidated.
Today's proposed rule would require the primary element to be
removed following a failed visual inspection and would require the
problem to be corrected. The Agency believes that it is appropriate to
provide two options for correcting the problem: either replace the
element with a new one or repair it. This would provide flexibility to
facilities, while still assuring that the fuel flowmeter will be
repaired to give quality assured data.
Today's proposed rule would also change the timing of the
requirement for fuel flowmeter transmitter or transducer testing if a
primary element fails its visual inspection. The Agency believes that
it would be appropriate also to test the fuel flowmeter transmitters
before the fuel flowmeter is placed into service again. This would be a
more thorough quality assurance check of the entire fuel flowmeter than
simply addressing the problem with the primary element. Thus, when the
fuel flowmeter is placed into service again, its accuracy would be
tested as fully as possible. In addition, EPA proposes to remove the
requirement for a test on the flowmeter transmitters in the calendar
quarter following a failed visual inspection. This requirement might be
appropriate if it seemed that transmitter drift was likely to be a
problem or if the Agency had no other means of assuring the quality of
the data from the flowmeter after a problem with the primary element
was known to have occurred. However, the Agency believes that problems
with the primary element are separate from problems with drift in the
transmitters. Because today's proposal would require a check on the
fuel flowmeter transmitters after repair or replacement of the primary
element, requiring an additional test of the transmitters in the
following calendar quarter appears to be unnecessary.
The proposed rule gives procedures for data validation when a
primary element fails a visual inspection. The element would have to be
replaced or repaired, and the transmitters would have to be tested
before data would again be valid from the fuel flowmeter. During the
period in which the flowmeter data are considered invalid, the
appropriate missing data substitution procedures would be used. The
Agency has clarified that these data validation procedures would also
apply to failures of other fuel flowmeter quality assurance tests. EPA
believes that this will make facilities' obligations clearer. In
addition, the Agency believes that fuel flowmeter systems should be
treated as consistently as possible with CEMS. Consistent treatment
simplifies the part 75 requirements and is more equitable for sources
using different monitoring approaches.
(e) Orifice, Venturi, and Nozzle Flowmeter Transmitter Testing
Background
As discussed previously, once an orifice-, nozzle-, or venturi-type
flowmeter has been installed, one of the major causes of error in the
measured flow rates is drift in the transmitters or transducers that
determines the total pressure, differential pressure, and temperature.
The flow measurement error for these types of flowmeters is a
combination of the errors in these individual transmitters or
transducers and a constant error value associated with the physical
dimensions of the primary element. The May 17, 1995 direct final rule
added a requirement that flowmeter transmitters be tested at least
annually. The transmitters are also required to be retested in the next
calendar quarter if the overall flow rate error is greater than 1.0
percent of the upper range value of the flowmeter. For practical
purposes, this requires a facility to know the error from the physical
dimensions of the primary element in order to determine if the
flowmeter meets the overall accuracy requirement.
Some utilities asked the Agency how to determine the overall
flowmeter accuracy from individual transmitter values (see Docket A-97-
35, Item II-E-31). EPA addressed this issue in Policy Guidance (see
Docket A-97-35, Item II-I-9, Policy Manual, Question 10.17). This
guidance included a formula for calculating total flowmeter accuracy
from error in transmitter readings for differential pressure, static
pressure and temperature, and error from all other sources (i.e.
physical dimensions of the primary element). Some utilities indicated
that they do not always have information available on the constant
error from other portions of the primary element (see Docket A-97-35,
Item II-E-13). The policy guidance also indicated that a facility could
report test results electronically using the highest amount of error
from any of the three transmitters. Provided that the highest error
from an individual transmitter is 1.0 percent of the upper range value
of the transmitter or less, the overall flowmeter accuracy will be less
than 2.0 percent of the upper range value (see Docket A-97-35, Item II-
I-10).
EPA has also observed that transmitter test data reported for
orifice-, nozzle-, and venturi-type flowmeters have not been
consistent. Some facilities test each transmitter once at three
different levels, including a low, middle, and high value (see Docket
A-97-35, Item II-D-16). Others test each transmitter at five different
levels, including zero, full scale, and three intermediate levels (see
Docket A-97-35, Item II-D-17). The Agency had previously issued some
guidance on reporting test results, both for orifice flowmeters and
other flowmeters (see Docket A-97-35, Items II-I-4, p. 3-58, and II-I-
9, Policy Manual, Questions 10.17 and 12.27). However, this guidance
appears to have been insufficient, as utilities have continued to
request guidance in how to perform and report test results (see Docket
A-97-35, Item II-D-21). Questions have included the number of levels at
which transmitters should be tested, whether all of these levels must
be non-zero, the number of times the transmitter should be tested at a
particular level, if results may be reported in hardcopy or should be
reported electronically, and how data should be reported
electronically.
Discussion of Proposed Changes
Today's proposed rule would make the requirement to assess the
total accuracy of orifice-, nozzle-, and venturi-type fuel flowmeters
from the transmitter/transducer test results an option. As an
alternative, proposed section 2.1.6.5 in Appendix D would allow each of
the three transmitters (static pressure, differential pressure, and
temperature) individually to meet

[[Page 28098]]

an accuracy specification of 1.0 percent of the upper range value of
the transmitter.
Today's rulemaking also proposes a procedure in section 2.1.6.1 of
Appendix D for testing the accuracy of orifice-, nozzle-, and venturi-
type fuel flowmeters. Each transmitter would be calibrated against
NIST-traceable reference values at least once at the zero level and at
a minimum of two other levels across the range of values that the
transmitter reads during normal unit operation. Note that in many
instances this would be a portion of the full-scale range of the
transmitter, rather than the entire range. In addition, revised section
2.1.6.2 of today's proposed rule includes the new Equation D-1a to
clarify how to calculate the error from an individual transmitter.
Finally, today's proposal would clearly specify the consequences of
failure of an accuracy test on transmitters in section 2.1.6.5 of
Appendix D. Just as CEM data are considered invalid from the time that
a quality assurance test is failed until the test is subsequently
passed, data from a fuel flowmeter would be considered invalid from the
date and time of a failed transmitter accuracy test until the date and
time of a passed transmitter accuracy test.
Rationale
The Agency considered two main options for determining the accuracy
of a transmitter or transducer of an orifice-, nozzle-, or venturi-type
fuel flowmeter. In the first approach (which is consistent with current
policy guidance), these types of fuel flowmeters would be required to
meet an accuracy of 2.0 percent of the upper range value of the total
flow rate of the fuel flowmeter. The accuracy would be determined using
the square root of the sum of the squares of all sources of error in
the fuel flowmeter, according to the following equation:
[GRAPHIC] [TIFF OMITTED] TP21MY98.000

Where: dqv/qv = Error in the volumetric flow rate
due to transmitter drift at a given level;
K = Original error resulting from installation of orifice (including
all other variables);
dPf = Average difference between static pressure transmitter
reading(s) and reference static pressure reading(s) at a given level;
Pf = Average reference static pressure reading at a given
level;
dP = Average difference between differential pressure
transmitter reading(s) and reference differential pressure reading(s)
at a given level;
P = Average reference differential pressure reading at a given
level;
dTf = Average difference between temperature transmitter
reading(s) and reference temperature reading(s) at a given level; and
Tf = Average reference temperature reading at a given level.

If the error calculations for error from the primary element of the
fuel flowmeter were not available, then the facility could use a
default value of 1.0 percent of the upper range value error from all
parts of the fuel flowmeter except for the differential pressure,
static pressure, and temperature transmitters. (In other words, the
factor ``K'' in the equation above would be equal to 1.0 percent of the
upper range value.) However, this would almost certainly trigger the
requirement for recalibration or retesting of the accuracy of the
transmitters in the next calendar quarter because the fuel flowmeter
accuracy would exceed 1.0 percent of the upper range value. Based upon
statements from the American Gas Association, it is the Agency's
understanding that for an orifice-, nozzle-, or venturi-type fuel
flowmeter meeting AGA Report No. 3 or ASME MFC-3M, the maximum error
from portions of the meter other than the transmitters should be 1.0
percent of the upper range value (see Docket A-94-16, Item II-F-2, and
this Docket, A-97-35, Item II-E-18).
In the second approach to determining error for orifice-, nozzle-,
and venturi-type fuel flowmeters, each transmitter or transducer would
be tested separately for accuracy, and each transmitter or transducer
would be required to meet an accuracy specification of 1.0 percent of
the full scale range of the transmitter. Under this approach, it would
no longer be necessary to determine the total error in the flowrate
from the fuel flowmeter. Because this proposal would eliminate the
calculation of the total error in flowrate, there would no longer need
to be a requirement to retest the accuracy of the transmitters in the
next calendar quarter when the total fuel flowmeter accuracy exceeds
1.0 percent of the upper range value.
In today's rule, EPA proposes to allow both of the approaches
described above for calculating the total flowmeter accuracy. The
second approach (i.e., calculating individual transmitter accuracy) is
simpler than calculating the total error in the flow rate, although it
is less directly related to the accuracy of SO2 mass
emission rate and heat input measurements than the fuel flowrate. An
individual transmitter accuracy specification of 1.0 percent of the
full scale of each transmitter would be slightly stricter than a total
fuel flowmeter accuracy specification of 2.0 percent of the upper range
value of the fuel flowmeter, because one transmitter could potentially
have an error greater than 1.0 percent of its full scale range while
the entire error in the fuel flowrate would still be less than 2.0 of
the upper range value of the fuel flowmeter. Thus, the option of
calculating the total error in the fuel flowrate has been retained in
today's proposal. At least one industry representative suggested
allowing both approaches of calculating accuracy when testing
transmitters of an
orifice-, nozzle-, or venturi-type fuel flowmeter (see Docket A-97-35,
Item II-E-24).
The Agency considered two main methodologies for transmitter
testing on orifice-, nozzle-, and venturi-type flowmeters. The first
method would be to require a five-point test that checks the linearity
of the transmitter. The transmitter would be tested against an NIST
traceable method (e.g., testing a pressure transmitter against an NIST
traceable deadweight transmitter) at the following percentages of the
full scale range of the transmitter: 0.0 percent, 20.0 to 30.0 percent,
40.0 to 60.0 percent, 70.0 to 80.0 percent, and 100.0 percent. This is
the general approach that was taken by many utilities that provided
transmitter calibration results to EPA (see Docket A-97-35, Items II-D-
26 through 28).
The second method would be to require a comparison to an NIST
traceable transmitter at the zero level and at least two other levels
across the range of readings on the transmitter or transducer. This
would be different from the first method in that the transmitter would
only need to be tested across the range where the transmitter is

[[Page 28099]]

actually used. For example, if a fuel flowmeter transmitter's readings
never rise higher than 60.0 percent of the full scale range of the
transmitter, then the transmitter could be tested at 0.0 percent, 30.0
percent, and 60.0 percent of full scale. These procedures are reflected
in the proposed revised section 2.1.6.1 of Appendix D.
The Agency is proposing the second method in today's rule, i.e.,
that each individual transmitter must be tested at three or more points
across its normal range of readings. EPA realizes that it is standard
industry procedure to test a fuel flowmeter at five levels across its
entire range (see Docket A-97-35, Item II-E-24). However, the Agency is
aware of at least one case where a fuel flowmeter failed to meet an
accuracy specification of 2.0 percent of the upper range value when it
was tested at 100.0 percent of the upper range value. However, the fuel
flowmeter was never used to measure a rate greater than roughly 55.0
percent of the upper range value (see Docket A-97-35, Item II-D-15). If
this flowmeter had only been required to test across the range where
the fuel flowmeter actually measured fuel flow rates, it would have met
the accuracy specification. Section 2.1.5 requires fuel flowmeters that
are tested against a master fuel flowmeter to be tested across the
range of measured fuel flowrate only. Requiring testing of each
transmitter at three or more points across the range of all readings
would still ensure that the transmitter reads accurately across all
readings, while reducing the possibility that the transmitter might
fail an accuracy test because of a high error reading at the high end
of the transmitter's range where the transmitter is never used. At
least one utility has mentioned that this would be helpful (see Docket
A-97-35, Item II-E-24). The Agency solicits comment on the proposed
approach.
Today's proposed rule also includes Equation D-1a for calculating
error from an individual flowmeter transmitter. The Agency feels that
this would clarify the calculation. It also would prevent the possible
confusion that would occur if a facility attempted to use the existing
Equation D-1, which is designed for a fuel flowmeter that is compared
to another fuel flowmeter.
Finally, under today's proposal, when a transducer or transmitter
test is failed, a fuel flowmeter would be considered out-of-control,
and its data would be considered invalid until the date and time the
transmitter is retested and meets an accuracy of 1.0 percent of its
full scale.
(f) Reporting of Fuel Flowmeter Testing Data
Background
As mentioned above in Section III.P.5 of the preamble, utilities
have had questions about how to report the results of their fuel
flowmeter testing data. In certification applications and quality
assurance testing results, utilities have reported test data in a
variety of ways. In some cases, the Agency was unable to determine the
flowmeter accuracy from the testing information provided because data
were not labeled as reference flow rate data, flowmeter data, or
accuracy data. For example, for turbine flowmeters, data on the
reproducibility of the ``K-factor'' was often presented. However, these
are not flow rate data, nor is it clear what the accuracy of the flow
rate is (see Docket A-97-35, Item II-D-9). Sometimes data were
presented in tables. Other data were presented in graphs (see Docket A-
97-35, Item II-D-9). In many cases, Agency or state environmental
agency staff needed to request additional information from utilities to
determine if they had met the accuracy requirement for fuel flowmeters
(see Docket A-97-35, Items II-C-3, II-C-5).
To clarify the requirements for certification applications for fuel
flowmeters, the Agency issued policy guidance about the type of
information to provide (see Docket A-97-35, Item II-I-9, Policy Manual,
Question 12.27). This guidance included a sample table with an example
of how to submit information for a fuel flowmeter that is tested
against a master meter or flow prover reference value.
Discussion of Proposed Changes
EPA proposes to add a sample table to Appendix D (Table D-1) for
summarizing the results of accuracy tests of fuel flowmeters that are
calibrated by comparison against other fuel flowmeters or a prover. In
addition, EPA proposes to add a separate table for summarizing the
results of calibrations of the transmitters or transducers of an
orifice-, nozzle-, or venturi-type fuel flowmeter.
Rationale
In today's proposed rule, EPA would provide clarification in the
form of a table for summarizing the quality assurance test results of
fuel flowmeters that are compared against other fuel flowmeters or a
prover. A second table is provided for summarizing the results of
calibrations of transmitters or transducers of an orifice-, nozzle-, or
venturi-type fuel flowmeter. This second table accounts for differences
in the testing procedure for transmitters or transducers. In both
cases, EPA has tried to make clear what critical information would have
to be reported in order to demonstrate that the fuel flowmeter (or the
transmitter of an orifice-, nozzle-, or venturi-type fuel flowmeter)
meets the accuracy specification. In addition, EPA will design revised
electronic record types with this type of information so that test
results may be more easily reported electronically. The Agency is aware
that this has been difficult or confusing for some utilities (see
Docket A-97-35, Items II-D-23, and II-I-9, Policy Manual, Question
12.27). The Agency also considered adding a sample graph for reporting
accuracy data. However, EPA feels that it would be easier to compare
the data in tabular format and to enter it into the electronic data
format than to enter values from a graph. Most of the graphs provided
to EPA have been relatively easy to read, and there appears to be less
of a need for an example to be included in Appendix D (see Docket A-97-
35, Item II-D-9).
7. Use of Uncertified Commercial Gas Flowmeter
Background
Currently, a facility using Appendix D may either install its own
gas flowmeter or use a commercial gas flowmeter owned by a pipeline
natural gas supplier, provided that the meter meets the reporting and
accuracy requirements of Appendix D, including initial certification
and continuing quality assurance requirements. Some utilities have
suggested to EPA that they would like to be able to use data from the
commercial billing of pipeline natural gas without having to
demonstrate that the gas flowmeter meets initial certification and
continuing quality assurance requirements (see Docket A-97-35, Items
II-D-45, II-D-49). Those utilities assert that because the amount of
gas measured is already subject to market forces, the monitoring should
be sufficiently accurate for the Acid Rain Program. Utilities have
mentioned that gas companies often are already conducting meter
calibrations as quality assurance, but utility customers generally do
not have access to this information (see Docket A-97-35, Items II-D-49,
II-E-33). Facilities would find it advantageous to rely upon their
commercial billing charges for accounting for pipeline natural gas
usage because they would need to devote less time, effort, and money to
the maintenance of gas fuel flowmeters. This is particularly desirable
to facilities since the SO2 emissions from pipeline

[[Page 28100]]

natural gas are extremely low compared to the SO2 emissions
from other fuels.
Discussion of Proposed Rule Changes
Proposed section 2.1.4.2 of Appendix D would allow facilities to
record and report the gas flow rate, the heat input rate, and emission
values based on gas flowmeter readings from a flowmeter used for
commercial billing of pipeline natural gas without meeting the
certification requirements of section 2.1.5 of Appendix D or the
quality assurance requirements of section 2.1.6 of Appendix D under
specified conditions. Relief from the certification and quality
assurance requirements for gas flowmeters used for commercial billing
would be limited to flowmeters where the gas flowmeter is used for
commercial billing under a contract with another company having no
common owner with the unit(s) served by the flowmeter, which would
exclude any gas flowmeters used for transfers of gas between different
divisions, subsidiaries, or affiliates of the same company.
If the commercial billing gas flowmeter would be used without
undergoing certification or quality assurance under part 75
requirements, then the designated representative would need to report
hourly records of the gas flow rate, the heat input rate, and emissions
due to combustion of pipeline natural gas, as well as heat input rate
for each unit if the commercial billing gas flowmeter is on a common
pipe header. This would be similar to the reporting currently done for
a certified gas flowmeter, but no quality assurance records would be
required. The quarterly report would contain record types 303 for fuel
flow rate and heat input rate, record type 314 for the SO2
mass emission rate, either record type 320 or 323 for the
NO_X emission rate in lb/mmBtu, and either record type 330 or
331 for CO2 mass emissions. It also would be necessary for
the designated representative to identify the commercial billing gas
flowmeter in Table B (electronic record type 510) of the monitoring
plan for the unit.
So long as the records from the commercial billing gas flowmeter
are the values used for commercial billing, the designated
representative would report those values from the commercial billing
gas flowmeter without adjustment. If the records from the commercial
billing gas flowmeter are not consistent with the values used for
commercial billing because of some problem that needs to be reconciled
between the gas vendor and the facility customer, then the designated
representative would consider the readings from the commercial billing
gas flowmeter to be invalid for that billing period and would report
hourly records using the missing data procedures for fuel flowmeter
data found in section 2.4 of Appendix D for all hours of gas combustion
during that billing period. A facility would not be able to use the
commercial billing value in the quarterly report if the commercial
billing value was different from the value on the commercial billing
gas flowmeter.
Rationale
Utilities have suggested that the purchase of pipeline natural gas
from a vendor is subject to market forces that ensure accurate
monitoring (see Docket A-97-35, Item II-D-49). Utilities have stated
that gas vendors already have procedures for certification and meter
calibration and that the gas vendors have an even greater incentive
than utilities to maintain a high monitor ``uptime'' (i.e.,
availability) for gas fuel flowmeters. Typically, utilities will work
together with their gas vendors if they believe there is any sort of
discrepancy in their monthly billing for pipeline natural gas (see
Docket A-97-35, Items II-D-33, II-E-33).
The Agency believes that this argument is reasonable. However, EPA
also understands that some utilities require their gas vendor to
correct their billing values based upon the evidence of the utility's
own gas flowmeters. In addition, it is likely that utilities will be
combusting more pipeline natural gas in the future as they respond to
current and potential future environmental requirements for reducing
NO_X and CO2. Therefore, the Agency believes that
there must be conditions placed upon reporting emissions and heat input
for the Acid Rain Program from gas flowmeters used for commercial
billing if the gas flowmeters will not be required to meet the
certification and quality assurance requirements of part 75.
The Agency is proposing to limit the waiver from certification and
quality assurance requirements to commercial billing gas flowmeters
that are used in billing transactions between companies with entirely
different ownership (e.g., a pipeline natural gas vendor and a separate
electric utility company with no owners in common). Some utilities
requested the relief from quality assurance requirements based upon the
reasoning that a gas vendor would do its own quality assurance and
maintenance, and perhaps with better accuracy than a utility would be
able to maintain, but the utility would not necessarily have access to
the test results and would not have control over what quality assurance
might occur (see Docket A-97-35, Items II-D-49, II-E-33). This
reasoning is sound if the utility and the gas vendor have no common
owners, but it would not necessarily be sound if a gas supplier were
part of the same company as the electric utility. Also, utilities
suggested that a gas vendor may have an incentive to overstate the
amount of gas in order to bill more, rather than having an incentive to
underestimate or under-report (see Docket A-97-35, Item II-D-49). Once
again, this argument is reasonable if the gas vendor is a separate
entity, but may not be reasonable if the gas supplier has common owners
with the electric utility. Therefore, today's proposed rule includes a
limitation on the waiver from certification and quality assurance
requirements for commercial billing gas flowmeters to those gas
flowmeters used for commercial billing between companies with separate
ownership.
EPA solicits comment on the proposed approach of allowing the use
of uncertified fuel flowmeters for purposes of determining emissions
and heat input in the limited circumstances described above.
EPA has proposed in today's rule that a facility may only report
data from a commercial billing gas flowmeter if the data are used in a
commercial transaction. A group of utilities suggested that the Agency
allow facilities to report quarterly SO2 emissions based on
gas supplier data, including any reconciliation that has taken place
(see Docket A-97-35, Item II-D-45). Such a reconciliation between a gas
vendor and its customer may occur if the customer believes there is a
discrepancy in their monthly billing for pipeline natural gas (see
Docket A-97-35, Items II-D-33, II-E-33). If a facility and its gas
vendor determined that gas supply information from a fuel flowmeter
were not sufficiently accurate to purchase gas, then the Agency
presumes the gas supply information is also not sufficiently accurate
for emissions accounting.
The Agency also considered whether a facility should be able to use
the reconciled gas volumes agreed upon for billing if that value were
not from the commercial billing gas flowmeter. In general in the Acid
Rain Program, hand-typed corrections to emissions data are not
permitted (see Docket A-97-35, Item II-I-14), with the primary
exception of cases where sound engineering judgement indicates there is
an obvious error that cannot exist, such as a negative concentration
reading.

[[Page 28101]]

Allowing a facility to enter a commercial billing value by hand would
contradict this basic reporting policy of the Acid Rain Program.
Today's proposed rule also specifies the type and frequency of
information that would be required to be reported by a facility
concerning pipeline natural gas. Some utilities have requested the
ability to report only a quarterly cumulative SO2 mass
emission number for emissions from gas (see Docket A-97-35, Item II-D-
45). However, the Agency believes that there are several reasons for
maintaining hourly heat input rate and emissions data during combustion
of pipeline natural gas. First, hourly data is the most useful interval
of data for air quality modeling in order to see if progress is being
made in reducing emissions. Hourly data from combustion of pipeline
natural gas will become even more important as more units switch to
combusting pipeline natural gas in order to reduce their emissions. In
addition, hourly data are easier to check for anomalous values than
quarterly data. Further, hourly heat input rate data is necessary in
order to determine the NO_X emission rate when using the
NO_X-versus-heat input rate correlation of Appendix E to part
75. Also, since hourly data are already being recorded, reported, and
processed by automated computer data acquisition and handling systems,
a change to this requirement would require costly reprogramming for
industry and for EPA. For all of these reasons, EPA is proposing that
facilities continue to report hourly gas flow rates, heat input rates,
and emissions from commercial billing gas flowmeters that are not
required to meet the certification and quality assurance requirements
of part 75.

Q. Appendix G

1. Use of ASTM D5373-93 for Determining the Carbon Content of Coal
Background
Appendix G to part 75 provides procedures for determining
CO2 emissions from fuel sampling and analysis instead of
from a CO2 CEMS and a flow monitor. Section 2.1 of Appendix
G includes a mass-balance equation for determining CO2 (see
Equation G-1), the frequency for sampling fuel, and the specific
methods for analyzing fuel for carbon content. Section 2.3 of Appendix
G provides a method for determining CO2 mass emissions from
a gas-fired unit from its heat input using Equation G-4. Some
facilities use Appendix G procedures to determine CO2 mass
emissions every day for their units. Other facilities might use the
procedures of section 2.1 of Appendix G only to provide CO2
mass emissions during extended periods when CO2 data are
missing from their CO2 CEMS, under the provisions of
Sec. 75.36.
A utility and its fuel analysis laboratory contacted EPA concerning
use of an additional ASTM method for analysis of carbon content. The
industry staff felt that the new infrared analysis method, ASTM D5373-
93, was the most up-to-date method and that this method should be at
least as accurate as the methods specified in Appendix G to part 75
(see Docket A-97-35, Item II-D-25). Based upon the precision and bias
information in the method, EPA approved its use under Sec. 75.66 (see
Docket A-97-35, Item II-C-16).
Discussion of Proposed Changes
Today's proposed rule would allow the use of ASTM D5373-93,
``Standard Methods for Instrumental Determination of Carbon, Hydrogen,
and Nitrogen in Laboratory Samples of Coal and Coke,'' for Section 2.1
of Appendix G to part 75. This method is for determining the carbon
content of coal. ASTM D5373-93 would also be incorporated by reference
in Sec. 75.6. Facilities would also continue to have the option to use
ASTM D3178-89 to analyze coal for carbon content.
Rationale
EPA has previously approved the use of ASTM D5373-93 for analyzing
the carbon content of coal (see Docket A-97-35, Item II-C-16). The
Agency believes this method is of sufficient accuracy for use in the
Acid Rain Program. In addition, EPA historically has accepted
analytical methods from standard-setting organizations such as the
American Society for Testing and Materials (ASTM). The Agency solicits
comment on the use of ASTM D5373-93 for analyzing the carbon content of
coal.
2. Changes to Fuel Sampling Frequency
Background
Section 2.1 of Appendix G (as revised by the May 17, 1995 direct
file rule) specifies that fuel sampling should be done weekly for gas
or oil for each shipment for diesel fuel and at least once per month
for gaseous fuel. The sampling frequencies for diesel fuel and for
gaseous fuel are consistent with the frequency for sampling under
Appendix D to part 75.
Most gas-fired and oil-fired units that perform fuel sampling for
sulfur content under Appendix D also perform fuel sampling for carbon
content. Today's proposed rule would reduce the frequency with which
facilities need to sample oil or gas under Appendix D.
Discussion of Proposed Changes
The fuel sampling frequency specified in section 2.1 of Appendix G
would be made consistent with the proposed requirements for Appendix D
oil and gas sampling. Thus, all oil samples could be taken upon
delivery, either from the delivery vessel itself or from the storage
tank after a delivery is transferred. Gas samples would be taken
monthly (for pipeline natural gas), for each shipment (for gases
delivered in lots), or daily (for fuels that are analyzed daily for
sulfur). Coal samples would continue to be taken weekly.
Rationale
Appendix D of today's proposed rule would reduce the required
sampling frequency of oil and gaseous fuels delivered in lots. Based
upon information provided by one utility, the variability of carbon
content in oil is less than the variability of sulfur content (see
Docket A-97-35, Item II-D-18). Some utilities have stated that they
would prefer the procedures for sulfur and GCV to be similar (see
Docket A-97-35, Item II-D-24). Based upon this statement, the Agency
believes that facilities would also prefer to have consistent fuel
sampling procedures for Appendices D and G. Therefore, the Agency
believes it is appropriate to make the fuel sampling frequency for
carbon analysis under Appendix G consistent with the fuel sampling
frequency for sulfur content under Appendix D. Similarly, section 5.5
of Appendix F would be revised to make the gas sampling frequency
consistent with Appendix D. The Agency solicits comment on the proposed
changes to the fuel sampling frequency.
3. Addition of Missing Data Procedures for Fuel Analytical Data
Background
Appendix D provides procedures for substituting missing fuel
analytical information, either for sulfur or GCV. However, Appendix G
to part 75 does not specify what should be done if carbon content data
are missing.
Some software programmers asked EPA what missing data procedures
should be used for carbon content data (see Docket A-97-35, Item II-E-
5). The Agency responded to this question at a public conference and in
policy guidance (see Docket A-97-35, Items II-E-5, and II-I-9, Policy
Manual, Question 6.3). In its policy guidance, EPA stated that
facilities should ``[f]ill in the most recent carbon content . . .
available for that fuel type (gas, oil or

[[Page 28102]]

coal) of the same grade (for oil) or rank (for coal). If at all
possible, use a carbon content value from the same fuel supply.''
Discussion of Proposed Changes
Today's proposed rule would allow facilities to substitute for
missing carbon content prior to January 1, 2000, using either the most
recent carbon content for that fuel type, grade and rank, or procedures
parallel to those of Appendix D. Beginning January 1, 2000, facilities
would substitute for missing carbon content data using procedures
consistent with Appendix D. For gaseous fuels and for oil sampled
manually, these procedures would provide for a conservative maximum
carbon content value. Specifically, the permissible conservative carbon
content values would be either the maximum carbon content measured in
the previous calendar year or, if this information were not available,
a default value based upon handbook fuel characteristics. For weekly
coal samples or composite oil samples, CO2 mass emissions
would be calculated using the highest carbon content from the previous
four carbon samples available.
Rationale
Software programmers have already indicated that it is useful to
have a procedure for filling in missing carbon content data for
purposes of programming (see Docket A-97-35, Item II-E-5). Some
utilities have stated that they would prefer the missing data
procedures to be similar for both sulfur and GCV, even if both values
are conservative (see Docket A-97-35, Item II-E-24). Therefore, the
Agency believes that facilities would also prefer to have Appendix G
missing data procedures for carbon content that are parallel with those
for sulfur content and GCV in Appendix D. Thus, today's proposal would
allow for missing data for manual oil samples or for gaseous fuel using
the maximum carbon content measured in the previous calendar year or,
if this information were not available, a default value based upon
handbook fuel characteristics.
In determining the conservative default carbon content values that
would be used for missing data substitution in the event that no
previous carbon content samples are available, the Agency consulted
several handbook reference tables on fuel characteristics.
Specifically, the Agency reviewed handbook values for the carbon
content of coal (of various ranks), oil (of various grades), and gas
(of different types). (see Docket A-97-35, Items II-I-18, II-I-19, II-
I-20). In the case of coal, there was a fairly wide range of carbon
content values for different ranks of coal. Therefore, today's rule
would propose separate default carbon content values for Anthracite,
Bituminous, and Subbituminous/Lignite. In contrast, the carbon content
values for different grades of residual oil were fairly consistent. For
this reason, today's rule proposes a single default carbon content
value for all grades of oil. Finally, for gaseous fuels, the handbooks
which were reviewed presented a fairly narrow range of values for
natural gas but a much wider range of values for other types of gaseous
fuels. Therefore, today's rule proposes a value for natural gas and a
separate, conservative value for all other types of gaseous fuels.
The Agency solicits comment on the proposed revisions to the
missing data procedures under Appendix D.

R. Reporting Issues

1. Partial Unit Operating Hours and Emission and Fuel Flow Rates
Background
For affected units that use CEMS to account for emissions under
part 75, hourly emission rates of SO2 (in lb/hr),
NO_X (in lb/mmBtu), and CO2 (in tons/hr), and
hourly heat input rates (in mmBtu/hr) are calculated using the
applicable equations in Appendix F. For affected units that use fuel
flow meters and fuel analysis (or default emission rates) rather than
CEMS, the applicable equations in Appendices D, F and G (for certain
gas-fired units) are used to determine the hourly SO2 and
CO2 mass emission rates and heat input rates. For oil and
gas-fired peaking units that use Appendix E to account for
NO_X emissions, the hourly NO_X emission rates in
lb/mmBtu are derived from a graph of NO_X emission rate
versus heat input rate, the hourly heat input rates being derived from
the applicable equation in Appendix F. Under Sec. 75.54(b)(2), unit
operating time is reported by rounding the actual operating time up to
the nearest 15 minutes.
The equations in Appendices D through G assume that each unit
operating hour consists of a full 60 minutes of unit operation (or, for
common stacks, that emissions are discharged through the stack for 60
minutes in each hour); the equations do not attempt to account for
partial unit operating hours. This is a shortcoming in the current
rule, because partial unit operating hours sometimes occur during
periods of unit startup, shutdown, and malfunction. Therefore, to
ensure accurate accounting of SO2 and CO2 mass
emissions and unit heat input, part 75 should address the issue of
partial unit operating hours. Note, that because NO_X
emission rates are measured with respect to heat input (lb/mmBtu),
rather than with respect to time (lb/hr), this discussion is not
relevant for NO_X emission rate. Many vendors and utilities
have asked EPA for guidance on how to calculate mass emission rates
during partial unit operating hours (see, e.g., Docket A-97-35, Item
II-D-4).
The crux of the partial unit operating hour issue is when to adjust
the emission data for unit operating time, before the reporting of
hourly values or at the quarterly summation. For many units, there are
very few hours of partial operation, and adjusting the data for
operating time merely involves multiplying by 1, a seemingly
inconsequential issue. For other units, such as peaking and cycling
units, which start up and shut down often, the issue of how the data is
reported is relevant because there can be a significant amount of
partial unit operating hours. Definitive and standardized reporting
requirements allow facilities and/or vendors to program their software
such that their calculated result equals the result calculated by EPA.
For SO2 and CO2, the question is whether to
report hourly emissions on a mass basis (i.e., lb or tons) or on a mass
emission rate basis (i.e., lb/hr or tons/hr). For heat input, the
question is whether to report the total hourly heat input (in mmBtu) or
the hourly heat input rate (in mmBtu/hr). For example, suppose that a
unit emits for a full 60 minutes in a particular clock hour at an
SO2 concentration of 602.5 parts per million (ppm), a
CO2 concentration of 10.0 percent, a volumetric flow rate of
4,000,000 standard cubic feet per hour (scfh), and a heat input rate of
300 mmBtu/hr. Suppose further that the same unit operates for only 15
minutes in the next hour and all of the parameters (i.e.,
SO2 and CO2 concentration, flow rate, and heat
input rate) remain unchanged. If unit operating time is disregarded,
the SO2 mass emission rate (calculated from Equation F-1 in
Appendix F) would be the same (400 lb/hr) for both the partial
operating hour and the full unit operating hour. Similarly, the
CO2 mass emission rate would be the same (22.8 tons/hr) and
the heat input rate would be the same (300 mmBtu/hr) for both the full
and partial operating hours. The mass emission rates and heat input
rate for the partial unit operating hour are the same as the full-hour
values because they are based solely upon data recorded during unit
operation, i.e., in

[[Page 28103]]

the first 15 minutes of the hour. The hourly average rates for the
partial hour do not include ``zero'' values for the three 15-minute
periods of unit non-operation during the clock hour (e.g., an
SO2 emission rate of (400 lb/hr + 0 + 0 + 0)/4 = 100 lb/hr
would not be appropriate). If the emission and heat input rates are
adjusted by multiplying them by the operating time, then, for the full
operating hour (i.e., operating time = 1.0), the SO2 and
CO2 mass emissions and heat input would be, respectively,
400 lb SO2, 22.8 tons CO2, and 300 mmBtu. For the
partial hour (operating time = 0.25), the corresponding values would
all be divided by four, i.e., 100 lb SO2, 5.7 tons
CO2, and 75 mmBtu, respectively.
Software vendors and utilities have requested clarification as to
whether hourly SO2 mass emission values should be reported
as totals, in lb, or as rates, in lb/hr. As early as November of 1993,
EPA stated that hourly SO2 mass emission values should be
reported as rates in lb/hr. Then, when determining quarterly cumulative
SO2 mass emissions, each hourly emission rate would be
converted to a mass basis by multiplying it by the unit operating time
(expressed as a fraction of an hour) for the same hour. Similarly,
hourly heat input values would be expressed as rates, in mmBtu/hr, and
hourly CO2 mass emissions would be expressed as rates, in
tons/hr. Parallel issues were also addressed by the Agency's policy,
for units that determine SO2 and CO2 mass
emissions and heat input from fuel flow rates and fuel analyses under
Appendix D to part 75 (see Docket A-97-35, Item II-I-9, Policy Manual,
Questions 14.14, 14.36 and 14.37).
Some utilities have requested that the Agency change its policy and
allow reporting of hourly total SO2 and CO2 mass
emissions and heat input instead of mass emission rates and heat input
rates (see Docket A-97-35, Item II-E-14). The utilities argued that
this would simplify determination of the total year-to-date
SO2 mass emissions, in order to estimate the number of
allowances needed to cover a unit's emissions or to prepare a report on
mass emissions for a state environmental agency, because the reported
values would already be multiplied by the hourly operating time. Thus,
by performing the multiplication by operating time before reporting the
hourly value rather than waiting until calculating the quarterly value,
it might save a calculation step if a facility wanted to use the data
for another purpose. For these reasons, reporting of totals is a
preferred approach for some facilities. However, other utilities that
have incorporated the correct rate approach into their software have
indicated that they would prefer not to have to revise their software
to report in totals.
Partial unit operating hours must also be considered in the
recording and reporting of hourly unit load. The standard missing data
procedures in Sec. 75.33 require historical flow rate data to be placed
in load ``bins'' (ranges) based upon the maximum operating electrical
generation (or steam flow rate) of the unit. However, the recorded
hourly volumetric flow rate value in scfh applies only to the fraction
of the hour in which the unit operates. Therefore, the reported load
for the hour should be based upon the average electrical generation
during the period when the unit operates. Thus, the electrical
generation should be recorded as a rate for the period when the unit
operates, rather than an integrated total for the entire hour. The
units for reporting hourly load should, therefore, be MWe or 1000 lb/hr
of steam, and not MW-hr or 1000 lb of steam.
Discussion of Proposed Changes
In today's rulemaking, EPA is proposing to amend part 75 to clarify
that heat input, fuel flow, SO2 mass emissions, and
CO2 mass emissions are all to be reported on an hourly basis
as rates. Today's proposal also would clarify that the hourly emission
rates are to be based only upon data collected during periods of unit
operation (i.e., for partial unit operating hours, emission rates or
heat input rates of zero that are recorded during periods of non-
operation are not to be included in the hourly average emission rates).
These clarifications are found in proposed Sec. 75.57, and Appendices
D, E and F to part 75. Today's proposed rule would also clarify that
the proper units of reporting for load are MWe and lb/hr of steam.
Today's proposal would also provide new options for reporting unit
operating time. While the current requirement to report operating time
rounded to the nearest 15 minutes would be retained as an option, the
proposal would allow more flexibility by specifying that, for reporting
purposes, unit operating time be rounded up to the nearest fraction of
an hour (in equal increments that can range from one hundredth to one
quarter of an hour, at the option of the owner or operator).
Consistent with the requirement to report hourly SO2 and
CO2 mass emissions and hourly heat input as rates, today's
rulemaking proposes to revise the quarterly summation formulas for
SO2 and CO2 and to add summation formulas for
heat input in Appendix F to part 75. The proposed formulas show that
hourly mass emission rates or heat input rates would be multiplied by
unit operating time before summing to get total mass emissions. Today's
proposal also includes new formulas in Appendix D for summing hourly
SO2 mass emission rates and hourly heat input values from
fuel flowmeter systems in order to determine quarterly and annual total
SO2 mass emissions and total heat input. The Appendix D and
F equations revised or added to address summations include Equations D-
6, D-7, D-8, D-9, F-3, F-12, F-24, and F-25.
In addition, EPA is proposing optional recordkeeping provisions for
determining total heat input, total SO2 mass emissions or
total CO2 mass emissions for the hour. In addition to
reporting the required emission and heat input rates, owners or
operators could choose to report the total hourly heat input and mass
emissions under this option.
Rationale
As stated above, some utilities have expressed a preference for
reporting hourly total values for SO2 and CO2
mass emissions and heat input, rather than rates (see Docket A-97-35,
Item II-E-14). They have stated that this is easier to understand and
that reporting hourly total values, instead of or in addition to rates,
would make it easier to determine the cumulative total mass emissions
at any time during the year.
One representative requested that EPA consider allowing either
method of calculation (i.e., hourly rates or totals), so long as the
annual mass emissions and heat inputs are correctly determined and
reported. EPA notes that, although this approach may appear
advantageous because it would not require some facilities to reprogram
their DAHS software, it would require other facilities to reprogram
their software and it would make it difficult for EPA to verify
emissions calculations from reported hourly data. Because EPA considers
it essential to the Acid Rain Program to be able to recalculate annual
compliance values based upon hourly emission information reported by
facilities, the Agency is not revising the rule to take the
representative's suggestion. EPA considered using the total mass
emissions (or total heat input) approach instead of the mass emission
rate (or heat input rate) approach currently stated in Agency policy
(see Docket A-97-35, Item II-I-9, Policy Manual, Questions 14.14 and
14.36). In fact, as discussed in section III.H. of this preamble, the
Agency is proposing, under subpart H of part 75, model

[[Page 28104]]

reporting requirements for NO_X mass emissions that would (if
adopted by an applicable state or federal authority) require hourly
NO_X mass emissions to be reported as a total value (in lb)
rather than an hourly mass emission rate (in lb/hr). However, using
hourly mass emission totals for values currently reported to the Agency
would have the distinct disadvantage of requiring both EPA and the
utilities who correctly implemented the mass emission rate approach to
reprogram software to perform the new calculations, whereas retaining
the use of SO2 and CO2 emission and heat input
hourly rates offers several advantages.
First, using hourly mass emission rates and heat input rates
instead of totals is consistent with the units of measure in which flow
rate is recorded. Volumetric flow monitors measure flow rate during a
given time in standard cubic feet per hour scfh, rather than total flow
in standard cubic feet (scf). When SO2 concentration is
multiplied by volumetric flow rate, one calculates a mass emission rate
rather than a total mass of SO2. Similarly, multiplying a
volumetric flow rate by a diluent gas concentration yields a heat input
rate in mmBtu/hr, rather than a total heat input in mmBtu.
Second, the current missing data procedures for volumetric flow
rate, which are based upon the assumption that flow is a rate that is
comparable from one hour to another, rather than a total volumetric
flow that will vary depending upon the unit operating time, would no
longer be appropriate if volumetric flow rate were changed to a total
volumetric flow. Third, for Appendix E gas-fired or oil-fired peaking
units, it is critical that heat input rate, and not total heat input,
be used to determine the NO_X emission rate. The Appendix E
correlation curve formulas are based upon heat input rate rather than
total heat input. Appendix E allows a facility to create a correlation
of the NO_X emission rate measured in the stack during stack
testing and heat input combusted during that same period of time,
rather than installing CEMS on gas-fired or oil-fired peaking units. If
a facility were mistakenly to use the total heat input from an hour
rather than the heat input rate, it would correlate to the wrong
portion of the NO_X to heat input rate correlation curve and
would incorrectly estimate NO_X emission rate. For example,
if heat input totals were used to determine NO_X emission
rate from the Appendix E curve, the unit would have a different
NO_X emission rate if it combusted 25,000 mmBtu in half an
hour than if it combusted 25,000 mmBtu during a full hour. This would
apply both under the current provisions of Appendix E and today's
revised provisions to Appendix E.
In view of the above considerations, today's proposed rule would
affirm that facilities are to report SO2 and CO2
emissions and heat input as rates on an hourly basis. However,
facilities would also be allowed, at their discretion, to report
SO2 and CO2 emissions and heat input as hourly
totals, in addition to reporting them as rates. This approach would not
require reprogramming of computerized reporting software for those
utilities that are following EPA's current policy, and would provide
consistent reporting that allows EPA to recalculate emissions and heat
input values. Those utilities that find recording and reporting of
hourly total SO2 and CO2 mass emissions and heat
input to be desirable would be able to do so. EPA will provide the
necessary electronic record types to support this optional reporting.
Although today's proposed rule would affirm that emissions and heat
input are to be reported as rates, rather than totals, EPA has become
concerned that for partial unit operating hours, some utilities are
incorrectly calculating hourly average flow rates by including flow
rates of zero in the hourly average to represent periods of non-
operation, rather than basing the average flow rate solely on the
minutes of operation of the affected unit during the clock hour. In one
example, it appears that the software is designed to calculate the
average flow rate by including data from all minutes during those
fifteen-minute quadrants of an hour when the unit operates, thus
including some minutes when the unit is not operating, rather than
creating an average flow rate just from merely those minutes when the
unit is operating and emitting (see Docket A-97-35, Item II-C-17). EPA
suspects that still other utilities may be calculating an average
hourly flow rate that includes flow rates of zero for whole quadrants
of an hour when a unit does not operate. This can result in the flow
rate values for partial operating hours being under-reported to EPA and
a lowering of the average flow rates in the load ranges used to provide
substitute flow rate data, both of which can cause underestimation of
SO2 mass emissions.
The Agency is also concerned that this same kind of improper data
averaging may be occurring when hourly gas concentrations are
determined during partial operating hours. EPA would, therefore,
require in today's proposal that facilities base all of their reported
hourly average concentrations, flow rates, emission rates, and heat
input rates solely upon data that are recorded during unit operation
(that is, when the unit is combusting fuel and emitting).
Some utilities have indicated that the approach of averaging in
readings of zero from periods of non-operation has been incorporated to
compensate for having to report operating time rounded up to the
nearest fifteen minutes (Note, this is not an acceptable approach). A
utility representative indicated that reporting operating time to less
precision can cause overestimation of emissions because the operating
time is multiplied by the mass emission rate. Thus, a mass emission
rate of 400 lb/hr measured over a period of 20 minutes, during an hour
when the unit shut down, would be multiplied by an operating time of .5
hr (i.e., 20 minutes rounded up to the nearest fifteen minutes) and
would result in 200 lb of SO2 being reported rather than the
132 lb of SO2 that was actually emitted. The utility
suggested that a solution would be to allow operating time to be
reported to more precision than is currently allowed. Therefore,
today's proposal would allow flexibility for reporting unit operating
time to greater precision. While the current requirement to report
operating time rounded up to the nearest 15 minutes would be retained
as an option, the proposal would allow more flexibility by specifying
that unit operating time be rounded up to the nearest fraction of an
hour (in equal increments that can range from one hundredth to one
quarter of an hour, at the option of the owner or operator). Thus, a
facility could decide whether it had enough partial operating hours
(e.g., unit start-ups and shutdowns) to merit changing their software
to report operating time to more precision.
2. Use of Bias-Adjusted Flow Rates in Heat Input Calculations.
In late 1995, the first year of the Phase I SO2
allowance program, EPA conducted an audit of the Phase I-affected
units. Data from the second quarter of 1995 were retrieved from the
Emission Tracking System (ETS) in order to determine whether the
SO2 emission rates and heat input values were being properly
reported. The results of the audit showed that a number of sources were
not reporting heat input correctly. The problem in most instances was
that the unadjusted flow rate was being used in the heat input
equation, rather than the bias-adjusted value. EPA believes that this
is attributable to the fact that part 75 does not explicitly state that
the bias-adjusted flow rate is to be used in heat input

[[Page 28105]]

calculations. The Agency has attempted to clarify this through policy
guidance (see Docket A-97-35, Item II-I-9, Policy Manual, Question
14.81). To correct the situation, the necessary language would be added
to section 7.6.5 of Appendix A in today's proposed rule.
3. Removing the Restriction on Using the Diluent Cap Only for Start-Up
Background:
Based on the May 17, 1995 direct final rule, sections 3.3.4, 4.1,
4.4.1, 5.1, 5.2.1, 5.2.2, 5.2.3, and 5.2.4 of Appendix F currently
provide for the substitution of a constant CO2 or
O2 value for a measured value from a CO2 or
O2 monitor during unit start-up. This provision was
originally created in response to concerns from some utilities that
their NO_X emission rate in lb/mmBtu was being overestimated
during unit start-up (see Docket A-90-51, Item IV-D-220, Letter from
English, Mark G., Deputy General Counsel, Kansas City Power & Light
Company on EPA's Proposed Part 75 regulations; see also Docket A-94-16,
Item II-F-2). During unit start-up or other periods when the unit is at
a low load level, CO2 concentrations are lower than during
normal operation and O2 concentrations are higher than
during normal operation. The NO_X emission rate equation,
however, is not designed to be used in these situations because it
assumes complete combustion and normal operating conditions. As a
result, the NO_X emission rate equation overestimates the
NO_X emission rate when the CO2 concentration is
very low or the O2 concentration is very high, such as
during start-up. The equations for calculating emission rates in lb/
mmBtu use measured CO2 concentration or the difference
between ambient air's O2 concentration and the measured
O2 concentration in the denominator. For example,
NO_X emission rate is calculated using a NO_X
pollutant concentration monitor and a CO2 diluent monitor
using the following equation:
[GRAPHIC] [TIFF OMITTED] TP21MY98.001

When a small CO2 concentration is entered into this
equation, the calculated NO_X emission rate will be very high
and will overestimate the actual emissions.
The idea of capping CO2 or O2 concentration
was implemented in part 75 for determination of NO_X emission
rate, CO2 mass emissions, and heat input during unit start-
up. The cap concentration was set at a minimum CO2
concentration of 5.0 percent CO2 and a maximum O2
concentration of 14.0 percent O2, based upon some
information provided by utilities for boilers (see Docket A-94-16, Item
II-D-34).
Some utilities asked EPA to consider extending this cap on diluent
gas concentrations to other situations when a unit is operating at a
low level (see, e.g., Docket A-97-35, Items II-D-20 and 30, and Docket
A-97-35, Items II-E-13 and II-E-14). In addition to unit start-up, this
might include periods of unit shutdown or unit ``banking,'' where a
unit is combusting a very small amount of fuel to keep the boiler warm,
but little or no electricity is generated. During these other
situations where a unit operates at a low level, the CO2
concentration will be very low and the O2 concentration will
be very high, resulting in high calculated NO_X emission rate
values like those during unit start-up. One software vendor
specifically mentioned that it would be easiest to implement the
diluent cap if it could be used any time the CO2
concentration would fall below or the O2 concentration would
rise above the cap value (see Docket A-97-35, Item II-E-7). This could
be implemented mathematically in the software, rather than having to
examine the unit operation or the number of hours since the unit
started operating in order to trigger use of the diluent cap.
During the process of implementing the May 17, 1995 direct final
rule, EPA issued guidance that explained that facilities may use the
diluent cap values for calculating NO_X emission rate during
unit start-up whenever the CO2 concentration is below 5.0
percent or the O2 concentration is above 14.0 percent, and
also may use the actual measured CO2 or O2
concentration values at all times for calculating CO2 mass
emissions or heat input (see Docket A-97-35, Item II-I-9, Policy
Manual, Question 14.39). In Question 14.39, EPA recommended that even
if the diluent cap is used to calculate NO_X emission rate,
the actual diluent measurement should be used for the purpose of
calculating CO2 mass emissions or heat input, because the
purpose of the diluent cap was ``to avoid using an extreme diluent
concentration in the denominator of the equation to calculate emission
rate in lb/mmBtu.'' The formulas for calculating hourly CO2
mass emission rate or hourly heat input rate do not use the
CO2 or O2 concentrations in the denominator of
the equation. Thus, use of the diluent cap would tend to overestimate
both CO2 mass emission rate and hourly heat input.
Discussion of Proposed Changes
Today's proposed rule would allow facilities to use diluent cap
values of 14.0 percent O2 or 5.0 percent CO2 for
boilers and 19.0 percent O2 or 1.0 percent CO2
for turbines. For the purpose of calculating NO_X emission
rates in lb/mmBtu, the diluent cap would be allowed to be used for any
hour in which the average measured CO2 concentration is
below the cap value or the average measured O2 concentration
is above the cap value. Diluent cap values would still be allowed to be
used to calculate CO2 mass emissions or heat input, as well
as NO_X (or SO2) emission rate in lb/mmBtu.
Rationale
EPA acknowledges that there are periods of low unit operation or
low load in addition to unit start-up where the calculated
NO_X emission rate would be overestimated if it were based
upon measured diluent concentrations. Therefore, the Agency believes
that extending use of the diluent cap is appropriate. The Agency
believes that allowing use of the diluent cap anytime when the actual
measured value is above the cap (for O2) or below the cap
(for CO2) is easier to program and to implement than
limiting the use of the diluent cap based upon unit load, another
option that EPA considered. The Agency believes that it is unlikely
that a unit would ever be able to operate at a high load and still have
an O2 or CO2 concentration beyond the diluent cap
value. Therefore, it is not necessary to limit the use of the diluent
cap value based on unit load.
The Agency is also proposing new diluent cap values for turbines.
Turbines tend to operate with much higher levels of excess
O2 than boilers. For example, Method 20 of Appendix A, 40
CFR part 60, the procedure for testing SO2, NO_X
and diluent gas from stationary gas turbines subject to the NSPS,
requires testers to correct data to a typical concentration of 15.0
percent O2. Emissions data reported to EPA confirms that for
turbines, hourly concentrations of O2 are typically between
14.0 and 16.0 percent and hourly concentrations of CO2 are
typically between 3.0 and 4.0 percent. Thus, a turbine's diluent gas
concentration is likely to consistently exceed the diluent cap value of
14.0 percent O2 and to be consistently below the cap value
of 5.0 percent CO2 promulgated in the May 17, 1995 direct
final rule. If these values were allowed to be used by turbines at all
times rather than just during unit start-up, a turbine

[[Page 28106]]

could conceivably report its NO_X emission rate using only
the diluent cap value and never report the actual monitored diluent
concentrations, thereby consistently underestimating the NO_X
emission rate. Therefore, today's proposal provides diluent cap values
of 19.0 percent O2 or 1.0 percent CO2 that are
clearly beyond the typical O2 or CO2
concentrations measured at turbines, while still providing some relief
at extreme diluent concentrations. It is EPA's observation that
turbines with NO_X CEMS have not reported emissions using the
diluent cap thus far. Thus, no turbines should need to reprogram
software in order to report the use of the new diluent cap value for
turbines with a new method of determination code.
EPA considered removing the option for facilities to use the
diluent cap for heat input rate and CO2 concentration, as
well as for NO_X (and SO2) emission rate in lb/
mmBtu, but is not proposing to do so in today's proposal. As explained
previously, the diluent cap was created in order to calculate more
representative NO_X emission rate data during certain unusual
circumstances. However, when a diluent cap value is used to calculate
the hourly CO2 mass emission rate or the heat input rate,
the final calculation would often be less representative of actual
emissions or heat input during those hours. The Agency also found that
allowing some facilities to use the diluent cap only for NO_X
emission rate and others to use the diluent cap also for hourly
CO2 mass emission rate and heat input rate makes it
difficult to check emissions and heat input rate data to verify that
calculations are performed correctly. This is because a data
acquisition and handling system could use either the actual reported
diluent gas concentration or the diluent cap value to calculate
NO_X emission rate, CO2 mass emission rate, or
heat input rate, but there is currently no provision in the electronic
data reporting format for a facility to indicate which value was used
to calculate the heat input. However, some utilities have indicated
that making a change to discontinue using the diluent cap for
calculations of heat input rate and CO2 mass emission rate
would require a significant change in their software calculations (see
Docket A-97-35, Item II-E-25). Therefore, today's proposed rule would
allow facilities the options of (1) not using the diluent cap at all,
(2) using the diluent cap only for calculating NO_X (or
SO2) emission rate in lb/mmBtu, or (3) using the diluent cap
for calculating NO_X (or SO2) emission rate in lb/
mmBtu, heat input rate, and CO2 emissions. In addition, EPA
is proposing to add a minor additional reporting requirement to
indicate whether the diluent cap is used in calculating CO2
and heat input in the electronic data reporting format. This would
allow EPA to verify facilities' calculations, while requiring less
reprogramming than changing the calculations for heat input and
CO2 emissions.
The Agency solicits comment on the proposed revisions relating to
the diluent cap.
4. Complex Stacks--General Issues
Background
Many power plants regulated under part 75 have relatively simple
stack and monitoring configurations. Many utilities have one stack for
each affected unit and have CEMS installed on the stack. Other plants
have more than one unit discharging to the atmosphere through a common
stack, with CEMS installed on the common stack. Still others have
individual units that exhaust into multiple stacks and have CEMS
installed on each stack. The monitoring requirements for these various
configurations are addressed in Secs. 75.13, 75.16, 75.17, and 75.18.
EPA has issued guidance to assist utilities in preparing quarterly
reports for these unit and stack configurations (see Docket A-97-35,
Items II-I-4 and II-I-9, Policy Manual, Section 17).
For the configurations described above, the process of accounting
for emissions and heat input from the units and stacks will follow
simple mathematical rules. For example, for single unit-single stack
configurations, the emissions and heat input for the unit are directly
determined from the stack CEMS (or from an excepted methodology, where
applicable). For units discharging through a common stack with CEMS on
the common stack, the combined emissions and heat input are determined
from the CEMS, and the heat input to each individual unit is determined
by apportionment of the combined heat input, using a ratio of the unit
load to the combined load of all units utilizing the common stack. For
a single unit exhausting through multiple stacks, the sum of the
SO2 and CO2 mass emissions and heat input for the
different stacks equals the total SO2 and CO2
mass emissions and heat input for the unit.
However, in implementing part 75, EPA has become aware of a number
of affected units that have stack exhaust configurations which are more
complex than the configurations described above. For example, one
utility has a configuration in which two units can emit through two
different stacks at the same time, combining their emissions in both
stacks (see Docket A-97-35, Items, II-C-1, II-D-12). In this case, the
stack configuration is both a common stack and a multiple stack
configuration. EPA has had significant problems in determining the
emissions and heat input from these units, and in one case, EPA
rejected the quarterly reports for the units (see Docket A-97-35, Item
II-C-8). The utility worked closely with EPA to resolve the reporting
issues resulting from this unusual situation (see Docket A-97-35, Item
II-D-21). Other utilities with similar situations have contacted the
Agency to ensure there would not be problems with their reporting (see,
e.g. Docket A-97-35, Item II-D-5).
There have been other cases in which a unit that is accountable for
holding SO2 allowances shares a common stack with a unit
that does not hold SO2 allowances (e.g., where an affected
unit and a non-affected unit share a common stack or, prior to 1/1/
2000, where a Phase I unit and a Phase II unit share a common stack).
These are termed ``subtractive stack'' situations in the following
discussion. Utilities with subtractive stack situations have generally
used the provisions of Sec. 75.16(a)(2)(ii)(C) or
Sec. 75.16(b)(2)(ii)(B). These provisions allow a facility to monitor
separately the common stack and the unit with no allowance requirement
and to subtract the emissions from the non-affected or Phase II unit
from the common stack emissions. In some cases, it has not been clear
in the electronic quarterly reports whether a utility is reporting
combined emissions from all of the units using the common stack or
whether the emissions from the non-affected unit(s) have already been
subtracted out of the reported emissions (see Docket A-97-35, Item II-
C-18). This confusion in interpreting the quarterly emissions reports
has made compliance determination difficult.
The Agency found that there is a potential problem with the
underestimation of emissions using this subtractive approach. In some
cases, the error in the monitors' measurements might be such that a
larger emissions value is subtracted from a smaller value, resulting in
the reporting of false negative emissions (see Docket A-97-35, Item A-
94-16-IV-D-18, Comments from Monitor Labs). In other cases, there may
be an incentive for making inaccurate measurements with the monitoring
systems installed on a unit with no allowance requirement. For

[[Page 28107]]

example, if the SO2 pollutant concentration monitor on a
unit with no allowance requirement did not operate properly and had a
significant amount of missing data, the facility would calculate
SO2 emissions from the unit using a conservative, high
concentration value. Therefore, emissions reported for the units with
allowance requirements would, as a result of the subtraction, be less
than the actual emissions. Thus, a facility might have a disincentive
for good monitor performance and accuracy, because it could lower the
emissions reported for the units with allowance requirements. Though
allowed under the current wording of Appendix A to part 75 and subpart
D of part 75, this is contrary to the intent of the missing data
substitution procedures, which is to encourage good monitor performance
while preventing any systematic underestimation of emissions. (See
Docket A-97-35, Items II-B-13, II-E-4, and II-I-12.)
Discussion of Proposed Changes
Today's proposed rulemaking would add a general regulatory
requirement to Secs. 75.16 and 75.17 for facilities with complex stack
configurations (i.e., subtractive stack situations or configurations
involving combinations of common stacks and multiple stacks) to receive
approval from EPA's Administrator for a method of calculating and
reporting emissions from the units and stacks in the configuration. The
facility would be required to reach agreement with the Agency on issues
such as: identification of the stack in its quarterly report,
representation of the configuration in its monitoring plan, groups of
units for which cumulative emissions must be reported, testing
procedures, use of the bias test, and use of the missing data
substitution procedures. This would apply both to sources that already
have certified monitoring equipment and are submitting quarterly
reports and to units that do not yet have certified monitoring systems
(e.g. new units).
Rationale
The Agency evaluated two basic approaches to resolving issues in
these complex stack monitoring configurations. First, EPA considered
resolving the issues through policy guidance and through instructions
for submitting quarterly reports. Second, the Agency considered putting
detailed instructions in part 75 for reporting from and testing of
monitoring systems installed in these complex stack configurations.
These rule provisions would have explicitly addressed missing data
substitution to ensure that when emissions are reported, they are not
underestimated from units with an allowance requirement or a
NO_X emission limitation. For example, EPA could have
required, for the subtracted unit(s), that the facility only use those
provisions of the standard missing data procedures that are not
intended to be conservative estimates, such as the average
SO2 concentration during the hour before and the hour after
a missing data period. Another approach for missing data substitution
could have been to count zero emissions for the unit with no allowance
requirement during any missing data periods. Or perhaps creation of a
site-specific missing data procedure could have been required (see
Docket A-97-35, Items II-E-4 and II-I-12). To prevent a potential
underestimation of emissions and a disincentive for more accurate
monitoring due to application of a bias adjustment on a monitor on a
unit with no allowance requirement where its emissions are subtracted
from a common stack, EPA could have required that the bias calculation
be based upon both the monitors on the common stack and the monitors on
units with no allowance requirement, resulting in a single bias
adjustment factor for the subtractive stack situation.
However, EPA's experience thus far in implementing the program
indicates that each complex monitoring configuration tends to be
unique. Thus, the Agency has rejected the two approaches discussed
above and has decided instead to make General regulatory revisions that
allow for case-by-case resolution of issues in individual plant
situations, rather than making extensive, detailed revisions to part 75
to address each unique situation.
The Agency prefers to make regulatory revisions rather than
addressing issues solely through policy and guidance. In some cases,
the Agency has given advice to utilities on how to report emissions,
and the utility involved has not followed the Agency guidance (see
Docket A-97-35, Items II-C-7, II-C-24, and II-D-8). In another case,
the current provisions of part 75 for missing data substitution and for
the bias test appeared to be in conflict with guidance that the Agency
wanted to issue in order to ensure that emissions are not
underestimated in a subtractive stack situation (see Docket A-97-35,
Item II-B-13). Therefore, today's proposed rule would require owners or
operators of facilities with complex stack configurations to apply for
approval of their monitoring plans and reporting methodologies from
EPA's Administrator on a case-by-case basis. The Agency believes that
the General regulatory provisions requiring approval of a complex
monitoring situation by EPA's Administrator will give both facilities
and the Agency flexibility to deal with site-specific cases, while also
giving the Agency regulatory authority to resolve any case-specific
problems.
It is possible that any final rule resulting from today's proposal
may not be promulgated until 1999. Thus, EPA is proposing to require
the Administrator's approval of the monitoring plans and reporting
methodologies only for those situations that will exist on and after
January 1, 2000. Any subtractive stack situations that exist only
during the duration of Phase I would not fall under this requirement.
However, complex stack situations that exist where affected and non-
affected units share a common stack would need to meet today's proposed
requirement. Similarly, in situations where coal-fired units sharing a
common stack have different NO_X emission limitations under
part 76, or situations where some units sharing a common stack have a
NO_X emission limitation under part 76 and others have no
NO_X emission limitations under part 76, any complex
monitoring configuration would need to be approved by EPA's
Administrator.
5. Complex Stacks--Heat Input at Common Stacks
Background
For a unit that utilizes a flow monitor to determine SO2
mass emissions, section 5 of Appendix F to part 75 requires heat input
to be calculated using the installed flow monitor and a diluent gas
(O2 or CO2) monitor. The January 11, 1993 final
rule indicated that units with common stacks, multiple stacks, or
bypass stacks should follow the same General procedures for monitoring
heat input as are used for monitoring SO2 under Sec. 75.16.
As written, those procedures allowed facilities to monitor their heat
input either by placing individual monitors on each unit that serves a
common stack or by placing monitors only on the common stack and
measuring a combined heat input from all of the units sharing the
common stack. The May 17, 1995 rule required the combined heat input
measured by monitors on the common stack to be apportioned to the
individual units, in two specific provisions. First, unit level heat
input was required under Sec. 75.16(e)(2) for cases in which a
knowledge of the heat input for each unit is critical to compliance
determination (i.e., for situations where any units using the common
stack have

[[Page 28108]]

a NO_X emission limit). Second, Sec. 75.16(e)(3) required
unit level heat input to be determined for all other common stacks, but
only until the year 2000. The November 20, 1996 rule outlined the
acceptable methodology for apportioning heat input, i.e., by using the
ratio of the unit load in MWe or lb of steam per hour to the combined
load of all units utilizing the common stack (provided that all of the
units utilizing the common stack are combusting fuel with the same F-
factor).
Discussion of Proposed Changes
Today's proposed rule would revise the existing requirements found
in Sec. 75.54(b) and two specific provisions of Sec. 75.16(e) for
accounting of heat input for units serving a common stack, a by-pass
stack, or multiple stacks. First, EPA would require determination and
reporting of the unit level heat input to be continued after the year
2000 for all affected units, rather than restricting it to certain
situations after 2000. Second, EPA would clarify that the proper units
of measure for load to be used in an apportionment of common stack heat
input to determine unit level heat input are totals of MWe-hr and 1000
lb of steam, rather than rates of MWe and 1000 lb/hr of steam.
Rationale
EPA considered leaving the current provisions of Sec. 75.16(e) and
Sec. 75.54(b) from the May 17, 1995 and November 20, 1996 rules
unchanged. However, this would have the serious drawback of requiring
the facilities to reprogram their computer software for certain units
and not for others. Corresponding monitoring plan changes would also be
required. Additionally, EPA would have to reprogram its emission
tracking software to accommodate two different heat input reporting
methodologies for common stacks. In view of these considerations, EPA
is proposing to continue to receive individual heat input data from all
affected units. This information is useful for developing inventories
of total NO_X mass emissions in tons in support of other
Agency rulemakings. Without such information, the inventories would be
based on assumptions about how units operate, rather than being based
on unit level heat input as reported from the facility.
The Agency believes that a relatively small number of sources would
be affected by this proposed change. This is because (1) most coal-
fired units would still need to report unit level heat input under the
current provisions of Sec. 75.16(e)(2), even after the year 2000; and
(2) gas-fired and oil-fired units using fuel flowmeters to determine
heat input and to implement the procedures of Appendix D or Appendix E
would still be required to monitor heat input for each unit under
section 2.1 of Appendix D. Because of the usefulness of having heat
input data for individual units, because of the burden of reprogramming
software to remove the heat input apportionment by the year 2000, and
because of the small number of sources that would benefit from
retaining the current provisions of Sec. 75.16(e)(3), EPA believes it
is reasonable to require all units that measure combined heat input at
a common stack to continue to apportion heat input to the individual
units. The Agency solicits comment on the number of sources that would
be affected by this revision.
6. Start-Up Reporting--Units Shutdown Over the Compliance Deadline
Background
As currently written, part 75 requires that units which are
shutdown over an applicable compliance date specified in Sec. 75.4 must
submit a notice of the planned and (if different) actual shutdown date.
In addition, Sec. 75.4(d) provides an extended certification deadline
for such units of ``the earlier of 45 unit operating days or 180
calendar days after the date that the unit recommences commercial
operation of the affected unit.'' If an owner or operator subsequently
recommences commercial operation of the unit, a notice related to the
planned and (if different) actual date of recommencement of commercial
operation is required. In addition to these notices, Sec. 75.64
requires that after the applicable compliance date passes, the owner or
operator must submit quarterly reports for such units. If the unit
remains shut down and does not operate during the quarter, the
quarterly report must show zero emissions. Utility commenters (see,
e.g., Docket A-97-35, Items II-D-20, II-D-30) have recommended that
this quarterly report requirement for shutdown units be deleted because
it is unnecessary and burdensome.
Discussion of Proposed Changes
Section 75.64(a) would be modified so that quarterly reporting is
not required until the first quarter in which a previously shutdown
unit recommences commercial operation. In this case, the first
quarterly report would contain data beginning with the hour in which
the unit recommences commercial operation.
Rationale
Units that are shutdown over their applicable certification
deadlines are required to submit notice, pursuant to Sec. 75.61(a)(3),
of the planned date of recommencement of commercial operation and also
must submit a follow-up notice if the actual date of recommencement of
commercial operation is different from the planned date. As a result of
these notice provisions, EPA will know whenever the status of a
shutdown unit changes. Because shutdown units have no emissions, the
Agency believes that quarterly reporting in addition to the notice
provisions is unnecessary to fulfill the emission reporting objectives
of the Act.
The Agency notes, however, that the proposed revision differs from
that suggested by certain utilities (see Docket A-97-35, Item II-D-30).
The utilities proposed tying the reporting requirement to the
certification deadline in Sec. 75.4(d). However, under Sec. 75.4(d),
facilities are required to report emissions data using special
provisions in that section prior to the extended certification deadline
in Sec. 75.4(d). Thus, the proposed revisions would tie the obligation
for quarterly reporting to the quarter in which commercial operation is
recommenced.
7. Start-Up Reporting--New Units
Background
As currently written, Sec. 75.64(a) requires the first quarterly
report for new units to be submitted for the quarter corresponding to
the compliance date in Sec. 75.4. However, the current provision is
unclear about which hourly emissions data need to be included in the
first quarterly report if the compliance deadline does not correspond
to the first hour in the quarter.
Discussion of Proposed Changes
Section 75.64(a) would be modified to clarify that a new unit must
start reporting data beginning with the earlier of the date and time of
provisional certification or the compliance deadline in Sec. 75.4(b).
Rationale
These proposed revisions are generally consistent with existing
implementation of the new unit reporting requirements, and primarily
would serve to clarify ambiguous elements of the current rule.

[[Page 28109]]

8. Recordkeeping and Reporting Provisions
Background
Subpart F and subpart G of the existing part 75 regulation set
forth the recordkeeping and reporting requirements that accompany the
monitoring provisions of part 75. Specifically, in subpart F,
Sec. 75.53 contains the monitoring plan requirements, Sec. 75.54
contains the general recordkeeping provisions, Sec. 75.55 lists the
general recordkeeping provisions for specific situations, and
Sec. 75.56 consists of the certification, quality assurance and quality
control record provisions. In subpart G, Sec. 75.62 lists the
monitoring plan reporting provisions, Sec. 75.62 contains the reporting
requirements for initial certification and recertification
applications, and Sec. 75.64 discusses the provisions for quarterly
reports. Quarterly reports are electronic data files containing
emissions and operating data from affected units, as well as monitoring
plan information and the results of certification and quality assurance
tests. Under Sec. 75.64, these electronic data reports are required to
be submitted to the Agency each calendar quarter. This electronic
information is used by the Agency for many different purposes,
including implementation of the SO2 allowance trading
program, determination of compliance with emission limits, development
of reports on utility emissions, and modeling of air quality to assess
the effectiveness of the Act.
In order to effectively use the electronic quarterly report
information, EPA created a standardized reporting format, the
electronic data reporting (EDR) format. The electronic file formats and
record structures of the EDR provide the vehicle by which required
information is submitted to the Agency every calendar quarter. The EDR
primarily defines the order, length, and placement of information
within the electronic report or file. The individual tables of the EDR
define the record type, type code, start column, data element
description, units, range, length, and FORTRAN format for each data
element in the electronic report. The information in the EDR fields
mirrors the required information set forth in subparts F and G of part
75. Considering both the volume of information contained in each
quarterly report (e.g, operating and emissions data for each of the
hours in the quarter) and the number of reports submitted to the Agency
(i.e., currently, 1765 reports are received each quarter for the 2055
affected units; some reports contain information for more than one unit
if several units are interrelated, as in a common stack configuration),
a standard format is critical in order for the Agency to review,
verify, and use the information reported. A standard format allows the
Agency to develop software to receive and verify the files and to
correlate and separate out specific information for compliance
determinations. A standard format also allows software vendors to
create standard software which can be utilized by many affected units.
This is more cost effective than developing site-specific software and
thus reduces the software cost to industry.
Today's rulemaking proposes a number of revisions to subparts F and
G of part 75 (the reporting and recordkeeping sections of the rule).
The majority of these changes are necessary to implement the proposed
substantive revisions to the sections of the rule and appendices
discussed elsewhere in this notice. In addition, EPA is
proposingrevisions to these subparts in order to streamline
implementation of the program and to coordinate reporting under the
Acid Rain Program with other programs.
To support the changes to the recordkeeping provisions, new
Secs. 75.57, 75.58, and 75.59 would be added. These sections would
replace existing Secs. 75.54, 75.55, and 75.56. The addition of new
sections is necessary because the proposed revisions would not be
mandatory until January 1, 2000, and to have the proposed revisions
listed throughout existing effective sections could lead to confusion.
However, an owner or operator would be free to follow the provisions of
Secs. 75.57, 75.58, and 75.59 before January 1, 2000, if he chooses to
do so. In addition, the owner or operator would be required to satisfy,
prior to January 1, 2000, the elements in these sections that support a
regulatory option proposed in other sections of part 75 if the owner or
operator elects to implement that option prior to January 1, 2000.
Because, as discussed above, the Acid Rain Program relies on a
standardized electronic data reporting format, EPA has also developed
draft revisions to the EDR formats and instructions (draft EDR version
2.1). The following discussion refers to both the rule sections and EDR
record types (RTs) that would be affected by the proposed revisions.
Discussion of Proposed Changes
There are a number of proposed rule changes to the recordkeeping
and reporting requirements of part 75 and corresponding draft EDR
revisions that would be necessary to implement the substantive
revisions proposed by EPA and discussed elsewhere in this preamble.
These include the following requirements:
(1) Changes to support new CO2 missing data requirements
(see Sec. 75.57 and RT 202, 210, and 211);
(2) Changes to support new reporting, QA and missing data
requirements for moisture monitoring (see Secs. 75.53, 75.57, and
75.59, and RT 211, 212, 220, and 618);
(3) Changes to support optional Appendix I (flow methodology for
gas and oil units) (see Secs. 75.57 and 75.58, and RT 220, 302, 303,
608, and 609);
(4) Changes to support more flexibility for units that have
multiple range analyzers (see Secs. 75.53 and 75.59, and RT 230, 530,
600, 601, and 602);
(5) Changes to support the use of the diluent cap during all hours
(see Sec. 75.57 and RT 300 and 330);
(6) Changes to support test exemptions and extensions for units
that operate infrequently (see Secs. 75.59 and 75.64, and RT 301, 697,
and 698);
(7) Changes to support increased flexibility in fuel sampling (see
Sec. 75.58 and RT 302, 303, 313, and 314);
(8) Changes to allow reporting of hourly total values in addition
to hourly rates (see Sec. 75.57 and RT 300, 310, and 330);
(9) Changes to support the proposed re-definition of unit operating
loads (see Secs. 75.53 and 75.59, and RT 535 and 611);
(10) Changes to support reporting of conditional data during
recertification events (see Sec. 75.59, and RT 556);
(11) Changes to support a new quarterly flow-to-load QA check for
flow monitors (see Sec. 75.59, and RT 605 and 606);
(12) Changes to allow QA test grace periods (see Sec. 75.59, and RT
699);
(13) Changes to support simplified reporting for low mass emissions
units (see Secs. 75.53, 75.58, and 75.63, and RT 360, 508, and 531);
(14) Changes to support fuel flow-to-load QA checks for fuel flow
meters (see Sec. 75.59, and RT 628 and 629); and
(15) Changes to support expanded reporting of RATA supporting
information (see Sec. 75.59, and RT 614, 615, 616, 617, and 618).
In addition, since the EDR version 1.3 was released, EPA has
developed additional record types to aid in the implementation of the
program, by allowing the designated representative to certify the
validity of quarterly reports using an electronic certification
statement. The proposed revisions would adopt the necessary rule
language to implement these miscellaneous record types (see Sec. 75.64,
and RT 900, 901, 910, and 920).

[[Page 28110]]

The proposed revisions would also set forth optional requirements
for reporting of NO_X mass emissions that states or EPA could
adopt as part of a NO_X mass trading program, such as the OTC
NO_X Budget Program. In this situation both a rule change and
an EDR change would be needed (see Secs. 75.57 and 75.64 and RT 301,
307, and 328).
The proposed rule revisions also include a number of changes that
EPA believes will facilitate implementation of the program. These
include:
(1) Reporting of test numbers, reasons for tests and indicators of
aborted tests (see Sec. 75.59, and RT 560, 600, 601, 602, 603, 610, and
611);
(2) Changing the deadlines for reporting the RATA supporting
information that was originally required on January 1, 1998 (see
Sec. 75.59, and RT 614, 615, 616, 617, and 618);
(3) Reporting of an optional record type that will allow facilities
to provide contact person information that many facilities currently
provide in quarterly report cover letters (see Sec. 75.59, and RT 999);
(4) Based on comments received, the rule would be revised so that
reporting the reasons for missing data as part of the quarterly report
would become optional, but would still need to be maintained on-site
(see Secs. 75.56 and 75.59, and RT 550);
(5) Reporting of facility location, identification, and EDR version
numbers to support the transition from EDR 1.3 to EDR 2.1 (see
Sec. 75.64, and RT 100 and 102);
(6) Reporting of information documenting the calculation of heat
input (see Sec. 75.57, and RT 300);
(7) Reporting of reference method backup QA data (see
Sec. 75.59(a)(11), and RTs 260, 261, and 262);
(8) Expanded reporting of unit definition information (see
Secs. 75.53, and RTs 504, 585, 586, and 587);
(9) Reporting of Appendix E segment ID information (see Sec. 75.58,
and RT 323, 324, and 560);
(10) Reporting of qualification data for peaking units or gas-fired
units (see Sec. 75.53, and RT 507);
(11) Reporting of the qualifying test for off-line calibrations
(see Sec. 75.59, and RT 623);
(12) Reporting of Appendix E emission rate test data (see
Secs. 75.59, and RT 650-653);
(13) Reporting of span effective date information and flow rate
span values (see Sec. 75.53, and RT 530); and
(14) Removal of the recordkeeping provisions of Secs. 75.50, 75.51,
and 75.52 that are no longer effective.
Rationale
The majority of the proposed changes to subparts F and G are needed
to support proposed substantive changes elsewhere in part 75. EPA is
also proposing certain minor revisions to the order and wording of
provisions in these subparts so that the records required by the rule
match up consistently with the record type descriptions in the EDR.
Certain utility groups previously had objected that EPA had not made
the EDR format available for formal public notice and comment. The
Agency maintains that it is not required to provide notice and comment
for the EDR. The data included in (or proposed to be included in) the
EDR are also listed in the rule (or the proposed rule revisions) as
requirements under the recordkeeping and/or reporting provisions of
Secs. 75.53 through 75.64, which have already undergone (or are
undergoing) public notice and comment. Since the EDR simply shows how
to present electronically the data whose submission is (or will be)
required by the rule, it is the rule, not the EDR, that imposes the
data requirements. Notice and comment on the contents of the EDR would
therefore be unnecessary and duplicative. Moreover, the requirement to
present the rule's data requirements in a specified format is
authorized by Sec. 75.64(d), which requires a quarterly report to be
submitted in the format specified by the Administrator. Like the data
requirements, this format requirement in part 75 was adopted after
public notice and comment.
In today's rulemaking, EPA has developed draft EDR revisions
simultaneously with the proposed rule revisions and is therefore
including the draft EDR revisions in the docket for comment at the same
time as the proposed rule revisions (see Docket A-97-35, Item II-A-12).
EPA is also posting the draft EDR v2.1 revisions and draft EDR v2.1
reporting instructions on the Acid Rain Homepage (www.epa.gov/
acidrain). However, the Agency maintains that notice and comment are
not necessary for revisions to the EDR so long as the data included in
the EDR is the same as the data required by rule provisions that have
undergone or are undergoing notice and comment. Thus, future EDR
revisions may be made without prior notice and comment on the EDR in
order to implement rule revisions for which notice and opportunity for
comment are provided. However, the Agency will continue its informal
procedures for involving the affected stakeholders in any such EDR
revisions.
There are a number of other proposed changes to Secs. 75.54-75.64
that have been included to implement existing provisions in other
sections of part 75. First, information on test numbers and reasons for
tests would be required so that quality-assurance test data can be more
easily correlated and interpreted. Second, the reporting of various
run-specific and point-specific RATA support information would be
required (e.g., point velocity head readings, gas reference method
quality-assurance data, moisture reference method data, etc.). The
Agency believes that most testing companies currently either collect
these data electronically or enter the data into computer programs
manually to determine RATA results. By requiring the reporting of these
data elements in a standard electronic format, the Agency believes that
both facilities and regulatory personnel would be able to more easily
interpret data that are currently provided by test contractors in many
different hardcopy formats.
The Agency is proposing not to require the electronic reporting of
RATA support information prior to the year 2000. Sections 75.56
(a)(5)(iii)(F) and (a)(7) and Sec. 75.64(a)(1) of part 75 currently
require RATA supporting information to be reported in the electronic
quarterly report. EPA believes, however, that it would be more cost
effective to require the more detailed RATA support records to be
electronically reported beginning in the year 2000, rather than having
a two-stage implementation. The Agency has notified all designated
representatives that this RATA supporting information will not be
required to be reported electronically, in RT612 and 613 of the
quarterly report, prior to January 1, 2000.
The Agency notes that certain data elements (e.g., yaw angle, pitch
angle, axial velocity, wall effect point identifier, etc.) have been
included in anticipation of future revisions to EPA Reference Method 2.
EPA is presently evaluating a number of alternative flow rate
measurement methodologies, such as the use of a 3-dimensional probe.
Depending on the outcome of the Agency's evaluation, one or more of
these alternative flow measurement techniques may be allowed beginning
in the year 2000. Therefore, EPA believes it is appropriate to include
data elements to support these anticipated Method 2 revisions in draft
EDR version 2.1.
Finally, by changing the requirements for reporting the results of
the most recent RATA from requiring it to be reported in the quarter in
which it was

[[Page 28111]]

performed, to requiring it to be reported in the quarter in which it
was performed and each subsequent quarter in which a BAF that was
calculated using the results of that RATA are used, EPA would make the
individual quarterly reports more self contained and make it easier for
people who are using the reported data to understand how the BAFs
reported in those reports were applied. EPA considered adding a field
to the hourly emissions data record for each pollutant to indicate the
BAF applied in that hour. However, the Agency received requests from
utilities on an early draft of the EDR revisions that the hourly
emissions data record types not be revised to add a field for BAF. The
Agency believes that reporting the results of the most recent RATA,
including the BAF, in each quarterly report would accommodate the
utilities' requests not to add the BAF to each hourly record type and
would achieve the objective of making the quarterly reports easier to
interpret because the BAF being applied will be found in each quarterly
report. In addition, since electronic RATA results involve a relatively
small amount of information that can be copied into subsequent reports
and does not have to be recreated, it should not be a significant
burden to reporting facilities.
The proposed revisions would also remove the requirement to report
the reasons for missing data and make it optional. However, even if the
information is not reported, the reasons for missing data would have to
be maintained on site in a manner suitable for inspection. Based on the
high data availability achieved during initial implementation of the
program, the Agency believes that this type of information is not
needed in the review of most quarterly reports. For those situations in
which the Agency may wish to review this information, the records would
still be on-site for audit purposes or for submittal to the Agency.
The EPA is also proposing to incorporate additions which would
allow the reporting of electronic signatures and certification
statements so that no hardcopy reporting of any kind (e.g., cover
letters) would be necessary to meet the quarterly report requirements.
Finally, the removal of recordkeeping Secs. 75.50, 75.51, and 75.52
(and the corresponding explanatory text included in Appendix J to the
existing rule) is necessary because those sections were scheduled for
replacement during the May 17, 1995 rule revisions. At that time,
Secs. 75.54, 75.55, and 75.56 were added as replacements for
Secs. 75.50, 75.51, and 75.52, effective January 1, 1996. Because the
effective date is now past, the old sections and Appendix J will be
removed and reserved in order to prevent any confusion.
9. Electronic Transfer of Quarterly Reports
Background
Sections 75.64(a) and (d) of the original January 11, 1993 Acid
Rain rule requires emissions, monitoring, and quality assurance data to
be electronically reported to the Administrator on a quarterly basis in
a format to be specified by the Administrator. Version 1.3 of the
Electronic Data Reporting (EDR) format (see Docket A-97-35, Item II-I-
5) further specifies the record structures to be used to report the
required data elements. Page 3-3 of the May 1995 Acid Rain Program CEMS
Submission Instructions (see Docket A-97-35, Item II-I-4) further
specifies the mode of transmission of the electronic data file to the
Agency. Three modes of transfer are listed as options: (a) by mail on
diskette, (b) by mail on magnetic tape, or (c) through direct
electronic transfer.
Since the beginning of the program, the Agency has received
quarterly reports by mail on diskette and through direct electronic
transfer. To date, the magnetic tape option has never been utilized.
Based on the first four years of implementation of part 75, the Agency
believes that the use of the direct electronic transfer mode of
transmission has many advantages to the Agency and to the affected
sources. In fact, more than seventy percent of the reports for sources
currently affected by part 75 were submitted directly to the EPA
mainframe with EPA-provided software in second quarter 1997, and the
number of sources using this option has steadily increased over time
(see Docket A-97-35, Item II-I-8).
Discussion of Proposed Changes
Today's proposal would require quarterly reports to be submitted
via direct electronic transfer unless otherwise approved by the
Administrator. This would remove the option of sending files through
the mail on interceding media except for hardship cases where a modem
is not available or where technical difficulties prevent the successful
transmission of files via modem.
An additional revision to section 4 of Appendix A to part 75 would
require data acquisition and handling systems (DAHS) to be capable of
transmitting a record of measurements and other required information by
direct computer-to-computer electronic transfer via modem and EPA-
provided software.
Rationale
For each quarterly report submitted, the Agency performs an
assessment which results in a feedback report for the submitting
designated representative. This feedback report provides information to
the facility that may be used in making trading decisions, that may
indicate that a change is needed to the facility software, and/or that
may indicate that the file needs to be corrected and resubmitted. A
major advantage of submission through direct electronic transfer with a
modem and EPA-provided software is that the designated representative
submitting the file receives the EPA assessment of the submitted data
much more quickly than for a file that is transmitted through the mail
on diskette. Currently, for a file that is submitted to the Agency by
electronic transfer via modem and EPA-provided software, the EPA
assessment is received by the designated representative, via modem and
EPA-provided software, immediately (typically within ten minutes) after
the transmission of the quarterly report file. However, for files
submitted on diskette that must travel through the mail system and be
processed by Agency personnel, a letter containing the EPA assessment
is currently sent to the designated representative through the mail and
arrives 45 days or later from when the submission was originally
received by the Agency. Therefore, with direct electronic transfer,
potential errors get corrected and resolved more quickly and trading
decisions can be made with assurance that submitted data meets the
minimum quality standards acceptable to the Agency. Additionally, the
source may electronically submit the quarterly report, via modem and
EPA software, prior to the deadline, immediately receive the EPA
assessment, fix any errors, and resubmit the file by the deadline. Many
utilities have indicated that this is an important advantage over
submission of the quarterly report by diskette.
Another benefit of direct electronic transfer is the reduced risk
of error in transmission to the Agency or handling at the Agency.
Throughout the implementation of the program, many files submitted on
diskette through the mail have been lost, returned to the sender,
damaged in transit, or contained viruses (see Docket A-97-35, Item II-
I-8). When a file is submitted using direct electronic transfer of a
quarterly report, the designated representative submitting the file(s)
receives an immediate

[[Page 28112]]

confirmation that the file was received by the Agency.
Further, immediate feedback from the agency on quarterly report
submissions may also contribute to cost savings for facilities if a
file submitted via direct electronic transfer is rejected and required
to be amended and resubmitted. Utilities have indicated that submitting
the report to EPA, receiving feedback, and making the necessary
corrections to the file in a single work session significantly reduces
the cost of reworks, particularly for facilities that retain their
master file at the individual plant locations.
An additional advantage to direct electronic transfer is the
reduced cost to the Agency resulting from the minimized EPA labor hours
required to process a diskette. For instance, a diskette transmitted
through the mail must be catalogued, scanned for readability and
viruses, uploaded to the EPA mainframe Emissions Tracking System, and
renamed. On the other hand, transmission of a file by direct computer-
to-computer electronic transfer using EPA software eliminates all of
those manual steps because they are performed automatically by the EPA
software used for transmission of the report.
A possible concern about a requirement to submit the quarterly
report via modem is the possibility that source may not be equipped
with a modem and electronic transfer capability. Although the Agency
believes that most sources currently have a modem or will have a modem
by the year 2000, the Agency understands that a very small percentage
might not. Therefore, the Agency would accept petitions from sources
unable to transmit files via modem in order to allow transmission via
diskette for hardship cases.
Additionally, a utility group representative raised a concern about
the possibility of a computer at either the facility source or at the
EPA being inoperative at the time of the deadline for transmission,
preventing a source from successfully transferring the quarterly report
to the Agency. In order to minimize the risk of this type of problem,
there is a wide window, currently thirty days, during which EPA will
accept quarterly report transmissions each quarter. Additionally, EPA
has instituted preventative measures to minimize the possibility that
the EPA computer would be inoperative for an extended length of time,
preventing quarterly report transmission. Nevertheless, the Agency
accepts that it is conceivable that a technical difficulty could
prevent the successful electronic submission of a quarterly report and,
therefore, would also approve diskette submission on an as-needed basis
for sources unable to transfer a file via modem and EPA-provided
software due to technical difficulties. Furthermore, EPA solicits
comment on whether it should allow a grace period for late submissions
due to a technical difficulty with the EPA computer.
Finally, section 4 of Appendix A to part 75 would be amended to
require the DAHS to be capable of transmitting the required information
by direct electronic transfer via modem and EPA-provided software, for
consistency with the proposed Sec. 75.64(f). In addition, section 4 of
Appendix A to part 75 would retain the requirement for the DAHS to be
capable of transmitting a record of measurements and other required
information via an IBM-compatible personal computer diskette so that an
on-site inspector could collect electronic data on a diskette for
review.

S. Revised Traceability Protocol for Calibration Gases

Background
Currently, Appendix H to part 75 requires affected units to follow
a 1987 version of EPA Protocol procedures for developing calibration
gases. This protocol document has been superseded by a later version,
the ``EPA Traceability Protocol for Assay and Certification of Gaseous
Calibration Standards,'' September 1997, EPA 600/R-97/121. The 1997
document is actually five protocols. Two of these protocols (formerly
known as Protocols 1 and 2) have been combined to allow both CEMS and
ambient air analyzers to be calibrated from gases produced either
without dilution (Procedure G1) or with dilution (Procedure G2). The
remaining three protocols (Procedures P1, P2, and P3) describe
procedures that are mandatory for ambient air quality analyzers (not
continuous emission monitoring systems).
The 1997 Protocol document, described above, is required by other
parts of the CFR, such as the NSPS provisions in part 60. Because the
old and new protocols specify different certification periods (i.e.,
useful shelf lives) for most calibration gases, some affected units
that must comply with both part 60 and part 75 have been forced to
replace calibration gas cylinders more frequently because of the
shorter certification period in the 1987 Protocol procedures required
by part 75.
Under the 1987 Protocol document, affected units with low
SO2 emission rates occasionally had difficulty finding
calibration gases that were within the concentration ranges required by
Appendix A to part 75. The 1997 Protocol document allows calibration
gases to be developed over a wider range of concentrations than was
previously allowed.
Under the current part 75 rule, ``Protocol 1 gases must be vendor-
certified to be within 2.0 percent of the concentration specified on
the cylinder label (tag value).'' However, no method is specified to
determine the uncertainty value. The overall uncertainty in the
concentration estimated for a calibration gas comes from many different
sources, including uncertainty in the reference standards, uncertainty
in the analyzer multi-point calibration, uncertainty in the zero/span
correction factors, and measurement imprecision.
Discussion of Proposed Changes and Rationale
Today's rule proposes to remove Appendix H and revise parts 72 and
75 to be consistent with the 1997 Protocol document. The following
sections of part 75 would be revised: Secs. 72.2 and 72.3; sections
5.1.1 through 5.1.6, 6.2, and 6.3.1 of Appendix A; and all of Appendix
H.
The final rule would incorporate by reference the 1997 Protocol
document. This is the preferred option for the following reasons: (a)
calibration gas certification periods would be identical under parts 60
and 75, thereby allowing affected units to reduce expenditures on
calibration gas without sacrificing accuracy or performance; (b) lower
emitting affected units would more easily be able to comply with the
required range of calibration gas concentrations; (c) improved assaying
procedures and accuracy determinations would be allowed; and (d) a
wider selection of calibration gases would be allowed.
While today's proposal would retain the requirement for EPA
protocol gases to be within 2.0 percent of the tag value, section 5.1.3
in Appendix A would be revised to specify the use of the uncertainty
calculation procedure in section 2.1.8 of the 1997 Protocol document
for estimating the analytical uncertainty associated with the assay of
the calibration gas. This uncertainty estimate includes the uncertainty
of the reference standard and any gas manufacturer's intermediate
standard (GMIS) and interference correction equation that may be used
in developing the calibration gas.
EPA proposes to change the term ``Protocol 1 gas'' to ``EPA
protocol gas'' because the 1997 Protocol document combines the Protocol
1 and Protocol 2

[[Page 28113]]

procedures; therefore, the term ``Protocol 1 gas'' would no longer be
used.
Today's proposal would also continue to allow a ``research gas
mixture'' to be used as a calibration gas. However, an RGM would need
to meet the same 2.0 percent uncertainty requirement that a protocol
gas would meet.
The proposed rule would explicitly allow GMISs to be used as
calibration gas for two reasons. First, an EPA protocol gas may be made
from a GMIS. Therefore, GMISs are at least as accurate as EPA protocol
gases. Second, GMISs are more readily available and less expensive than
standard reference material or National Institute of Standards and
Technology (NIST) traceable reference material, both of which are
allowable as calibration gas under part 75.
Today's proposal clarifies that NIST/EPA-approved certified
reference materials (CRMs) would be acceptable as calibration gas by
adding those CRMs to the definition of ``calibration gas'' in
Sec. 72.2.
The 1997 Protocol document accepts primary reference standards from
the Netherlands Measurement Institute as being equivalent to standard
reference materials from the NIST. As a result, today's proposal adds
``standard reference material-equivalent compressed gas primary
reference material'' to the ``calibration gas'' definition in Sec. 72.2
and to section 5.1.2 of Appendix A.
Finally, the definition of ``zero air material'' would be revised
to accommodate other acceptable procedures.
Major differences between the 1987 Protocol procedures and the 1997
Protocol procedures are explained on pages 1-1 through 1-3 of the 1993
Protocol document and on pages 1-1 through 1-2 of the 1997 Protocol
document (see Docket A-97-35, Items II-I-23 and 24).

T. Appendix I--New Optional Stack Flow Monitoring Methodology

Background
Section 412 of the Act requires that units subject to title IV
install SO2 concentration monitors and volumetric flow
monitors for the purpose of determining SO2 emissions. The
purpose of the volumetric flow requirement is to enable a unit to
convert SO2 concentrations into mass emission rates of
pounds per hour (lbs/hr). Volumetric flow is also used to determine
heat input rate in mmBtu/hr and CO2 mass emission rate in
ton/hr.
In December 1991, 56 FR 63002 (December 3, 1991), EPA proposed an
exception to the requirement to install SO2 concentration
monitors and volumetric flow monitors at oil- and gas-fired units in
Appendix D to part 75. The exception relies on fuel flowmeters and fuel
sampling and analysis to determine SO2 emissions from oil-
and gas-fired units. In comments on the December 1991 proposed rule,
some industry commenters also advocated allowing oil- and gas-fired
units to use a diluent monitor, an F-factor, and a fuel flowmeter as an
alternative to a volumetric flow monitor. An F-factor is a fuel-
specific constant that relates the heat content of a fuel and the
volume of gases given off upon combustion. It is used to convert
pollutant concentrations into units of pounds of pollutant per million
British thermal units of heat input (lb/mmBtu). EPA already allows the
use of F-factors in emissions monitoring under part 75 and under 40 CFR
part 60, subparts Da and Db. Method 19 of Appendix A to part 60 uses F-
factors as the reference methods for calculating SO2 and
NO_X emissions in terms of lb/mmBtu for subpart Da and Db
units. F-factors also are used in the performance tests for certain
pollutants required under Sec. 60.8 to determine if a source is in
compliance with a particular emission standard in lb/mmBtu. Part 75
also uses F-factors in conjunction with diluent gas and volumetric flow
data to determine heat input under section 5 of Appendix F to part 75.
Table 19-1 of Method 19 in Appendix A to part 60 and Table 1 in section
3.3.5 of Appendix F to part 75 list the appropriate F-factors for
different types of fuel, including oil and natural gas.
Although the commenters supported the two exceptions included in
Appendix D, some commenters did not believe the exceptions would be
economical at all oil- and gas-fired units. According to one commenter,
fuel sampling protocols have an inherently high bias because they
assume a 100 percent conversion of fuel sulfur into SO2,
which results in higher emissions reporting from fuel sampling
protocols than from CEMS. The commenter claimed that the high bias
appears to be in the range of 5 to 10 percent. According to the
commenter, the higher emissions reporting ``penalty'' that is inherent
in fuel sampling protocols would justify installing SO2 CEMS
at some oil- and gas-fired units, particularly large, base-loaded oil-
fired units. In addition, the commenter claimed that, for oil- and gas-
fired units which install SO2 CEMS, use of the ``F-factor/
fuel flow method''--which includes use of an F-factor, a fuel
flowmeter, fuel sampling data, and a diluent (CO2 or O2)
concentration monitor--would provide much more accurate and precise
information than volumetric flow monitors (see Docket A-90-51, Item IV-
D-184).
In a four-day experiment performed in 1991 by one commenter,
measurements from the F-factor/fuel flow method were compared to those
generated by a combined SO2 CEMS and a volumetric flow
monitor. However, EPA did not believe that four consecutive days of
data were sufficient to support a conclusive equivalency determination.
Instead, in the January 11, 1993 final rule (58 FR 3590, 3643), EPA
reserved Appendix I to part 75 for the F-factor/fuel flow method and
stated that, to be approved, the method would have to meet the criteria
for alternative methods as required by section 412 of the Act and the
provisions of Sec. 75.40 in a 30-day (720 hour) trial.
Section 412 of the Act requires that an alternative monitoring
system provide information with ``the same precision, reliability,
accessibility, and timeliness as that provided by CEMS . . .'' 42
U.S.C. 7651k. To be approved, the alternative monitoring system must
meet the criteria for alternative methods in a 720 hour trial as
required by the provisions of subpart E of part 75. The rule designates
a certified CEMS or a reference method according to Appendix A to part
60 as the reference for evaluating the alternative monitoring system's
performance.
In order to meet the precision and reliability criteria, an
alternative monitoring system must achieve performance specifications
and quality assurance requirements equivalent to those for CEMS. In
addition, to demonstrate precision, an alternative monitoring system
must pass three statistical tests evaluating the flow CEMS and
alternative method in terms of their respective systematic error,
random error, and correlation. Additionally, to meet the reliability
criterion, the alternative monitoring system is required to match a
certified CEMS in terms of annual availability. Finally, to meet the
accessibility and timeliness criteria, an alternative monitoring system
must match the CEMS' ability to record requisite emissions data on an
hourly basis and report results within 24 hours.
In 1995, Long Island Lighting Company (LILCO) sponsored an
``alternative flow monitor demonstration project'' to demonstrate the
equivalency of fuel flow measurements and F-factor calculations to
stack instrument flue gas measurements for the determination of
volumetric flow. The project was

[[Page 28114]]

performed by Entropy at LILCO's Port Jefferson Unit 4, a 180 MW oil-
fired unit that burns residual oil with a maximum sulfur content of one
percent. The components of the alternative method consisted of a fuel
flowmeter and a CO2 CEMS. The alternative F-factor/fuel flow
method was compared to a flue gas volumetric flow CEMS.
Testing of the F-factor/fuel flow method took place in April-May
1995, and 739 hours of data were collected over a wide range of
operating loads (40 MW--190 MW). Fuel oil samples were taken daily and
analyzed for density and carbon content. The alternative method
successfully passed statistical tests but showed statistically
significant bias (see Docket A-97-35, Item II-D-14). Due to the bias
uncovered during the test, EPA concluded that the alternative flow
monitor demonstration project did not meet the requirements of subpart
E of part 75 for an alternative monitoring system. However, EPA is
proposing that a default multiplier, derived from the demonstration
data, be incorporated into the equations used under Appendix I to
compensate for the detected systematic bias and thereby help to ensure
that emissions are not underestimated when using the F-factor/fuel flow
method. With these provisions, EPA proposes to include the F-factor/
fuel flow method as an excepted method for determining flow in Appendix
I to part 75. The proposed default multiplier, 1.12, is based on the
data and results of the LILCO demonstration and is supported by EPA and
the Class of `85 Regulatory Response Group. The default multiplier
would be incorporated into the equations used under Appendix I whenever
a relative accuracy test audit is performed on a component-by-component
basis as was proposed in the LILCO demonstration.
Discussion of Proposed Changes
EPA proposes to include the F-factor/fuel flow method in Appendix I
as an excepted method for use in place of a volumetric flow monitor for
oil- and gas-fired units that burn only natural gas and/or fuel oil.
The F-factor/fuel flow method uses fuel flow measurement, fuel sampling
data, CO2 (or O2) CEMS data and F-factors to
determine the flow rate of the stack gas. EPA proposes limiting use of
the F-factor/fuel flow method to oil- and gas-fired units that burn
only natural gas and/or fuel oil because of the greater fuel
consistency of oil and natural gas and because the fuel flow rates of
oil and natural gas can be monitored accurately with a fuel flowmeter,
unlike the feed rate of coal.
Appendix I flow monitoring would be done using any of the following
combinations of components: a CO2 monitor and a volumetric
oil flowmeter, a CO2 monitor and a mass oil flowmeter, a
CO2 monitor and a volumetric gas flowmeter, an O2
monitor and a volumetric oil flowmeter, an O2 monitor and a
mass oil flowmeter, or an O2 monitor and a volumetric gas
flowmeter.
Today's proposal would amend Sec. 75.20, ``Certification and
Recertification Procedures,'' to add certification and recertification
procedures for units using Appendix I flow monitoring systems. Initial
certification of the components of the F-factor/fuel flow method would
be performed either component by component or on a system basis. If
each component is tested separately, then the fuel flowmeter would be
tested in accordance with section 2.1.5 of Appendix D, and the
CO2 or O2 monitor would have to pass a 7-day
calibration test, a linearity check, a cycle time test and a relative
accuracy test audit (RATA) using Method 3A from Appendix A to part 60.
A bias test would also have to be conducted. If the excepted Appendix I
flow monitoring system is tested as an entire system, then the
following tests would be performed: a 7-day calibration error test, a
linearity check, and a cycle time test on the CO2 or
O2 monitor, and a relative accuracy test audit on the entire
excepted flow monitoring system using Method 2 from Appendix A to part
60, and a bias test. The owner or operator would also test the data
acquisition and handling system. Upon successful completion of all
certification tests, the Appendix I system would be considered
provisionally certified.
Today's proposal would amend Sec. 75.21, ``Quality Assurance and
Quality Control Requirements,'' to include Appendix I flow monitoring
systems. A unit utilizing the optional F-factor/fuel flow method would
have to meet ongoing quality assurance testing requirements. First, the
daily and quarterly assessment requirements for a CO2 or
O2 monitor in sections 2.1 and 2.2 of Appendix B would have
to be followed. Second, one of the following would have to be met,
depending on whether the owner or operator chooses to test the method
on a component-by-component basis or on a system level: (1) the fuel
flow meter quality assurance requirements and a separate RATA on the
CO2 (or O2) monitor; or (2) a system level flow
RATA. If the components are tested separately, the applicable
procedures in section 2.1.6 of Appendix D would have to be followed for
the fuel flowmeter quality assurance (i.e., a flow meter accuracy test,
a transmitter accuracy test and primary element inspection, and/or the
supplemental quarterly fuel flow-to-load quality assurance testing) and
the applicable RATA procedures in sections 6.5 through 6.5.2.2 of
Appendix A for the CO2 (or O2) monitor would be
followed. In addition, the bias test would have to be performed on the
CO2 (or O2) monitor and, if the bias test is
failed, a bias adjustment factor (BAF) would have to be calculated and
applied to hourly data.
If the entire system is tested, the applicable procedures in
sections 6.5 through 6.5.2.2 of Appendix A would have to be used to
meet the performance specifications for flow relative accuracy in
section 3.3.4 of Appendix A. The bias test would have to be performed
on the volumetric flow data and, if the bias test is failed, a BAF
would have to be calculated using the procedures in section 7.6 of
Appendix A.
Several other sections of the rule would be modified or added in
order to incorporate the new excepted method described in Appendix I,
including Secs. 75.30, 75.57, 75.58, and 75.59. Section 75.30,
``General Provisions'' (for missing data substitution procedures),
would be modified by adding quality assured data from a certified
excepted flow monitoring system under Appendix I to the list of
monitoring systems that measure flow rate data, for which the missing
data substitution procedures of subpart D are required. If fuel
sampling data, fuel flow rate data, and diluent gas data are missing,
then the data acquisition and handling system would have to substitute
for missing volumetric flow data. In addition, Sec. 75.57, would
include additional information that Appendix I flow monitoring systems
must record. This includes fuel flow rate data and data from component
monitors. Section 75.58(g) would be added to address specific
volumetric flow rate record provisions for units using the optional
protocol in Appendix I. Section 75.59, ``Certification, Quality
Assurance and Quality Control Record Provisions,'' would also include
certification and quality assurance information that facilities must
record for Appendix I flow monitoring system tests.
Finally, the new proposed Appendix I would describe the
applicability, procedures, calculations, missing data, and
recordkeeping and reporting requirements for units using Appendix I to
determine flow.
The Appendix I formulas are more complex if an O2
monitor is used. EPA proposes to allow the use of an O2
monitor for Appendix I; however, the

[[Page 28115]]

initial programming of the formulas and monitoring plan development may
take longer for Appendix I flow monitoring systems that use an
O2 monitor.
Volumetric stack flow rate during oil combustion would be
calculated from (1) a bias adjustment factor from the applicable bias
test results; (2) the fuel flow rate (in gal/hr); (3) the fuel density
(in lb/gal); (4) the percent carbon by weight; (5) the CO2
(or O2) concentration percent by volume; and (6) the
appropriate conversion factor. The carbon content of the fuel would
have to be determined according to the procedures in section 2.1 of
Appendix G and the density of the oil would have to be determined
according to the procedures in section 2.2 of Appendix D.
Rationale: EPA is proposing an F-factor/fuel flow method in
Appendix I to part 75 as an excepted method to measure volumetric flow
directly with a flow monitor because this method would allow fuel flow
measurement with a gas or oil flowmeter, fuel sampling data,
CO2 (or O2) CEMS data, and F-factors to determine
the flow rate of the stack gas rather than a volumetric flow monitor.
The F-factor/fuel flow method would be available for use by oil-fired
and gas-fired units, as defined under Sec. 72.2, provided that they
only burn natural gas and/or fuel oil. For these units, EPA believes
that the proposed method would provide acceptably accurate measurements
of volumetric flow, while affording cost savings that some industry
representatives estimate could be substantial. The Agency solicits
comment on the proposed Appendix I and associated changes to part 75.
Appendix I may offer cost savings to some oil and gas fired units.
Representatives from oil- and gas-fired units have estimated that the
costs of operating, maintaining and testing volumetric flow monitors
range from approximately $15,000 to $25,000 per year. In contrast,
using the F-factor/fuel flow method is estimated to result in costs of
only approximately $5,000 to $7,000 per year due to elimination of the
operating, maintenance, testing and fuel costs associated with the
volumetric flow monitor.

U. The Use of Predictive Emissions Modeling Systems (PEMS)

A number of parties have submitted preliminary field test data
designed to demonstrate that EPA should set forth specific requirements
for alternative monitoring methodologies that predict NO_X
emission rates at gas-fired units. These ``predictive emissions
modeling systems'' (PEMS) use mathematical models to predict
NO_X emission rates based on sensor readings of key operating
parameters. The agency is evaluating the submitted data and will
consider taking further action under a future rulemaking if additional
study demonstrates the equivalency of PEMS to CEMS for well defined
classes of units.

IV. Administrative Requirements

A. Public Hearing

If requested as specified in the DATES section of this preamble, a
public hearing will be held to discuss the proposed regulations.
Persons wishing to make oral presentations at the public hearing should
contact EPA at the address given in the ADDRESSES section of this
preamble. If necessary, oral presentations will be limited to 15
minutes each. Any member of the public may file a written statement
with EPA before, during, or within 30 days of the hearing. Written
statements should be addressed to the Air Docket address given in the
ADDRESSES section of this preamble.
A verbatim transcript of the public hearing, if held, and all
written statements will be available for public inspection and copying
during normal working hours at EPA's Air Docket in Washington, DC (see
the ADDRESSES section of this preamble).

B. Public Docket

The Docket for this regulatory action is A-97-35. The docket is an
organized and complete file of all the information submitted to or
otherwise considered by EPA in the development of this proposed
rulemaking. The principal purposes of the docket are: (1) to allow
interested parties a means to identify and locate documents so that
they can effectively participate in the rulemaking process, and (2) to
serve as the record in case of judicial review. The docket is available
for public inspection at EPA's Air Docket, which is listed under the
ADDRESSES section of this preamble.

C. Executive Order 12866

Under Executive Order 12866 (58 FR 51735, October 4, 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.

This proposed rule is not expected to have an annual effect on the
economy of $100 million or more. However, pursuant to the terms of
Executive Order 12866, it has been determined that this proposed rule
is a significant action because it raises novel policy issues. As such,
the proposed rule has been submitted for OMB review. Any written
comments from OMB and any EPA response to OMB comments are in the
public docket for this proposal.

D. Unfunded Mandates Reform Act

Title II of the Unfunded Mandates Reform Act of 1995 (UMRA), P.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, for proposed and final rules with ``Federal mandates'' that
may result in expenditures to State, local, and tribal governments, in
the aggregate, or to the private sector, of $100 million or more in any
one year. Before promulgating an EPA rule for which a written statement
is needed, section 205 of the UMRA generally requires EPA to 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 publishes with the final rule an explanation why that
alternative was not adopted. Before EPA establishes any regulatory
requirements that may significantly or uniquely affect small
governments, including tribal governments, it must have developed under
section 203 of the UMRA a small government agency plan. The plan must
provide for notifying potentially affected small governments, enabling
officials of affected small governments

[[Page 28116]]

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.
This proposed rule is not expected to result in expenditures of
more than $100 million in any one year and, as such, is not subject to
section 202 of the UMRA. Although the proposed rule is not expected to
significantly or uniquely affect small governments, the Agency has
notified all potentially affected small governments that own or operate
units potentially affected by the proposal in order to assure that they
have the opportunity to have meaningful and timely input on the
proposed rule. EPA will continue to use its outreach efforts related to
part 75 implementation, including a policy manual that is generally
updated on a quarterly basis, to inform, educate, and advise all
potentially impacted small governments about compliance with part 75.

E. Paperwork Reduction Act

The information collection requirements in this proposal have been
submitted for approval to the OMB under the Paperwork Reduction Act, 44
U.S.C. 3501, et seq. An Information Collection Request (ICR) document
has been prepared by EPA (ICR No. 1835.01), and a copy may be obtained
from Sandy Farmer, OPPE Regulatory Information Division; U.S.
Environmental Protection Agency (2137); 401 M Street, SW, Washington,
DC 20460, by calling (202) 260-2740, or via the Internet at www.gov/
icr.
Currently, all affected utilities are required to keep records and
submit electronic quarterly reports under the provisions of part 75.
The proposed rule includes several new options for compliance with part
75 which have been requested by affected utilities. To implement these
options, EPA would have to modify the existing recordkeeping and
reporting requirements. In some circumstances, these changes would
result in significant reductions in the reporting and recordkeeping
burdens or costs for some units (such as low mass emissions units).
However, these changes would require modifications to the software used
to generate electronic reports. In addition, there would be some
increased burden or costs for certain units to fulfill the new quality
assurance procedures proposed in these proposed revisions. Finally,
several other technical revisions to the existing reporting and
recordkeeping requirements have been proposed to clarify existing
provisions or to facilitate reporting for other regulatory programs in
the context of Acid Rain Program reporting. Although these one-time
software changes would tend to increase the short-term burdens
allocated to the Acid Rain Program, such changes should reduce a
source's overall long-term burden by streamlining the source's
reporting obligations under both the Acid Rain Program and the Act.
The average annual projected hour burden is 2,608,836, which is
based on an estimated 835 likely respondents (on a per utility basis).
The projected cost burden resulting from the collection of information
is $47,555,000, which includes a total projected capital and start-up
cost of $1,436,000 (for monitoring equipment/software), and a total
projected operation and maintenance cost (which includes purchase of
testing contractor services and total projected fuel sampling and
analysis cost of $716,000) of $46,119,000. Burden means the total time,
effort, or financial resources expended by persons to generate,
maintain, retain, 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 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.
An agency may not conduct or sponsor and a person is not required
to respond to a collection of information, unless it displays a
currently valid OMB control number. The OMB control numbers for EPA's
regulations are listed in 40 CFR part 9 and 48 CFR chapter 15.
Comments are requested on the Agency's need for this information,
the accuracy of the provided burden estimates, and any suggested
methods for minimizing respondent burden, including through the use of
automated collection techniques. Send comments on the ICR to the
Director, OPPE Regulatory Information Division; U.S. Environmental
Protection Agency (2137); 401 M Street, SW, Washington, DC 20460; and
to the Office of Information and Regulatory Affairs, Office of
Management and Budget, 725 17th Street, NW, Washington, DC 20503,
marked ``Attention: Desk Officer for EPA.'' Include the ICR number in
any correspondence. Since OMB is required to make a decision concerning
the ICR between 30 and 60 days after May 21, 1998, a comment to OMB is
best assured of having its full effect if OMB receives it by June 22,
1998. The final rule will respond to any OMB or public comments on the
information collection requirements contained in this proposal.

F. Regulatory Flexibility Act

The Regulatory Flexibility Act (RFA), 5 U.S.C. 601, et seq.,
generally requires an agency to conduct a regulatory flexibility
analysis of any rule subject to notice and comment rulemaking
requirements unless the agency certifies that the 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 governmental jurisdictions. This proposed rule would
not have a significant impact on a substantial number of small
entities.
Today's proposed revisions to part 75 result in a net cost
reduction to utilities affected by the Acid Rain Program, including
small entities. Most importantly, the proposed changes to Appendix D
and the addition of an optional calculation procedure instead of actual
monitoring for oil- and gas-fired units with low mass emissions would
significantly reduce the cost of complying with part 75 for oil-and
gas-fired units, many of which are owned or operated by small entities.
Therefore, I certify this action will not have a significant economic
impact on a substantial number of small entities.

G. National Technology Transfer and Advancement Act

Section 12(d) of the National Technology Transfer and Advancement
Act of 1995 (``ANTTAA''), Pub L. No. 104-113 15 USC 272 note, directs
EPA to use voluntary consensus standards in its regulatory activities
unless to do so would be inconsistent with applicable law or otherwise
impractical. Voluntary consensus standards are technical standards
(e.g., materials specifications, test methods, sampling procedures,
business practices, etc.) that are developed or adopted by voluntary
consensus standards bodies. The NTTAA requires EPA to provide Congress,
through OMB, explanations when the Agency decides not to use available
and applicable voluntary consensus standards.
This regulatory action proposes to incorporate by reference
voluntary consensus standards pursuant to Sec. 12(d) of the NTTAA. The
EPA has adopted the general policy of using voluntary

[[Page 28117]]

consensus standards from technically knowledgeable groups such as the
Organization for International Standards (ISO), the American Society
for Testing and Materials (ASTM), the American Society of Mechanical
Engineers (ASME), the American Gas Association (AGA), the Gas
Processors Association (GPA), and the American Petroleum Institute
(API).
EPA invites public comment on the voluntary consensus standards
which are proposed to be incorporated by reference for use in part 75.
EPA has not identified any additional voluntary consensus standards
which might be applicable to this rulemaking. This does not indicate
that other applicable standards do not exist or that any other
standards should not be allowed. Therefore, EPA also invites public
comment on any other voluntary consensus standards which may be
appropriate for the proposed regulatory action. Further, if additional
applicable voluntary consensus standards are identified in the future,
the designated representative may petition under Sec. 75.66(c) to use
an alternative to any standard incorporated by reference and prescribed
in this part.
EPA proposes to incorporate by reference the following voluntary
consensus standards for use under part 75:
a. ASTM D5373-93 ``Standard Methods for Instrumental Determination
of Carbon, Hydrogen and Nitrogen in laboratory samples of Coal and
Coke.'' This standard is proposed to be incorporated by reference for
use under section 2.1 of Appendix G to part 75 and is discussed further
in section III.Q.1 of this preamble.
b. API Section 2 ``Conventional Pipe Provers'' from Chapter 4 of
the Manual of Petroleum Measurement Standards, October 1988 edition.
This standard is proposed to be incorporated by reference for use under
paragraph (g)(1)(i) of Sec. 75.20 and under section 2.1.5.1 of Appendix
D to part 75. The proposal to incorporate this standard by reference is
discussed further in section III.P.6.(b) of this preamble.

List of Subjects in 40 CFR Parts 72 and 75

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

Dated: April 27, 1998.
Carol M. Browner,
Administrator, U.S. Environmental Protection Agency.

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

PART 72--PERMITS REGULATION

1. The authority for part 72 continues to read as follows:

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

2. Section 72.2 is amended by revising the definitions of
``calibration gas,'' ``excepted monitoring system,'' ``gas-fired,''
``pipeline natural gas,'' ``span,'' ``stationary gas turbine,'' and
``zero air material''; by revising paragraph (2) of ``oil-fired'' and
paragraph (2) of the ``peaking unit''; by adding paragraph (3) to the
definition of ``peaking unit''; by adding new definitions for
``conditionally valid data,'' ``EPA protocol gas,'' ``gas
manufacturer's intermediate standard,'' ``low mass emissions unit,''
``maximum rated hourly heat input,'' ``ozone season,'' ``probationary
calibration error test,'' ``research gas mixture (RGM)'', and
``standard reference material-equivalent compressed gas primary
reference material''; and by removing the definition of ``protocol 1
gas,'' to read as follows:

Sec. 72.2 Definitions.

* * * * *
Calibration gas means:
(1) A standard reference material;
(2) A standard reference material-equivalent compressed gas primary
reference material;
(3) A NIST traceable reference material;
(4) NIST/EPA-approved certified reference materials;
(5) A gas manufacturer's intermediate standard;
(6) An EPA protocol gas;
(7) Zero air material; or
(8) A research gas mixture.
* * * * *
Conditionally valid data means data from a continuous monitoring
system that are not quality assured, but which may become quality
assured if certain conditions are met. Examples of data that may
qualify as conditionally valid are: data recorded by an uncertified
monitoring system prior to its initial certification; or data recorded
by a certified monitoring system following a significant change to the
system that may affect its ability to accurately measure and record
emissions. A monitoring system must pass a probationary calibration
error test, in accordance with section 2.1.1 of appendix B of part 75
of this chapter, to initiate the conditionally valid data status. In
order for conditionally valid emission data to become quality assured,
one or more quality assurance tests or diagnostic tests must be passed
within a specified time period.
* * * * *
EPA protocol gas means a calibration gas mixture prepared and
analyzed according to section 2 of the ``EPA Traceability Protocol for
Assay and Certification of Gaseous Calibration Standards,'' September
1997, EPA-600/R-97/121 or such revised procedure as approved by the
Administrator.
* * * * *
Excepted monitoring system means a monitoring system that follows
the procedures and requirements of Sec. 75.19 of this chapter or of
appendix D or E to part 75 for approved exceptions to the use of
continuous emission monitoring systems.
* * * * *
Gas-fired means:
(1) For all purposes under the Acid Rain Program, except for part
75 of this chapter, the combustion of:
(i) Natural gas or other gaseous fuel (including coal-derived
gaseous fuel), for at least 90.0 percent of the unit's average annual
heat input during the previous three calendar years and for at least
85.0 percent of the annual heat input in each of those calendar years;
and
(ii) Any fuel, except coal or solid or liquid coal-derived fuel for
the remaining heat input, if any.
(2) For purposes of part 75 of this chapter, the combustion of:
(i) Natural gas or other gaseous fuel with a total sulfur content
no greater than the total sulfur content of natural gas (including
coal-derived gaseous fuel) for at least 90.0 percent of the unit's
average annual heat input during the previous calendar years and for at
least 85.0 percent of the annual heat input in each of those calendar
years; and
(ii) Fuel oil, for the remaining heat input, if any.
(3) For purposes of part 75 of this chapter, a unit may initially
qualify as gas-fired if the designated representative demonstrates to
the satisfaction of the Administrator that the requirements of
paragraph (2) of this definition are met, or will in the future be met,
through one of the following submissions:
(i) For a unit for which a monitoring plan has not been submitted
under Sec. 75.62 of this chapter,
(A) The designated representative submits fuel usage data for the
unit for

[[Page 28118]]

the three calendar years immediately preceding the date of initial
submission of the monitoring plan for the unit under Sec. 75.62; or
(B) For a unit that does not have fuel usage data for one or more
of the three calendar years immediately preceding the date of initial
submission of the monitoring plan for the unit under Sec. 75.62, if the
designated representative submits: the unit's designated fuel usage;
all available fuel usage data (including the percentage of the unit's
heat input derived from the combustion of gaseous fuels), beginning
with the date on which the unit commenced commercial operation; and the
unit's projected fuel usage.
(ii) For a unit for which a monitoring plan has already been
submitted under Sec. 75.62, that has not qualified as gas-fired under
paragraph (3)(i) of this definition, and whose fuel usage changes, the
designated representative submits either:
(A) Three calendar years of data following the change in the unit's
fuel usage, showing that no less than 90.0 percent of the unit's
average annual heat input during the previous three calendar years, and
no less than 85.0 percent of the unit's annual heat input during any
one of the previous three calendar years is from the combustion of
gaseous fuels with a total sulfur content no greater than the total
sulfur content of natural gas and the remaining heat input is from the
combustion of fuel oil; or
(B) A minimum of 720 hours of unit operating data following the
change in the unit's fuel usage, showing that no less than 90.0 percent
of the unit's heat input is from the combustion of gaseous fuels with a
total sulfur content no greater than the total sulfur content of
natural gas and the remaining heat input is from the combustion of fuel
oil, and a statement that this changed pattern of fuel usage is
considered permanent and is projected to continue for the foreseeable
future.
(iii) If a unit qualifies as gas-fired under paragraph (2)(i) or
(ii) of this definition, the unit is classified as gas-fired as of the
date of the submission under such paragraph.
(4) For purposes of part 75 of this chapter, a unit that initially
qualifies as gas-fired must meet the criteria in paragraph (2) of this
definition each year in order to continue to qualify as gas-fired. If
such a unit fails to meet such criteria for a given year, the unit no
longer qualifies as gas-fired starting January 1 of the year after the
first year for which the criteria are not met. If a unit failing to
meet the criteria in paragraph (2) of this definition initially
qualified as a gas-fired unit under paragraph (3)(ii) of this
definition, the unit may qualify as a gas-fired unit for a subsequent
year only under paragraph (3)(i) of this definition.
* * * * *
Gas manufacturer's intermediate standard (GMIS) means a compressed
gas calibration standard that has been assayed and certified by direct
comparison to a standard reference material (SRM), an SRM-equivalent
PRM, a NIST/EPA-approved certified reference material (CRM), or a NIST
traceable reference material (NTRM), in accordance with section 2.1.2.1
of the ``EPA Traceability Protocol for Assay and Certification of
Gaseous Calibration Standards,'' September 1997, EPA-600/R-97/121.
* * * * *
Low mass emissions unit means a gas-fired or oil-fired unit that
burns only natural gas and/or fuel oil and that qualifies under
Secs. 75.19(a) and (b) of this chapter.
* * * * *
Maximum rated hourly heat input means a unit-specific maximum
hourly heat input (mmBtu) which is the higher of the manufacturer's
maximum rated hourly heat input or the highest observed hourly heat
input.
Oil-fired means:
* * * * *
(2) For purposes of part 75 of this chapter, a unit may qualify as
oil-fired if the unit burns only fuel oil and gaseous fuels with a
total sulfur content no greater than the total sulfur content of
natural gas and if the unit does not meet the definition of gas-fired.
* * * * *
Ozone season means the period of time from May 1st to September
30th, inclusive.
* * * * *
Peaking unit means:
* * * * *
(2) For purposes of part 75 of this chapter, a unit may initially
qualify as a peaking unit if the designated representative demonstrates
to the satisfaction of the Administrator that the requirements of
paragraph (1) of this definition are met, or will in the future be met,
through one of the following submissions:
(i) For a unit for which a monitoring plan has not been submitted
under Sec. 75.62,
(A) The designated representative submits capacity factor data for
the unit for the three calendar years immediately preceding the date of
initial submission of the monitoring plan for the unit under
Sec. 75.62; or
(B) For a unit that does not have capacity factor data for one or
more of the three calendar years immediately preceding the date of
initial submission of the monitoring plan for the unit under
Sec. 75.62, the designated representative submits: all available
capacity factor data, beginning with the date on which the unit
commenced commercial operation; and projected capacity factor.
(ii) For a unit for which a monitoring plan has already been
submitted under Sec. 75.62, that has not qualified as a peaking unit
under paragraph (2)(i) of this definition, and where capacity factor
changes, the designated representative submits either:
(A) Three calendar years of data following the change in the unit's
capacity factor showing an average capacity factor of no more than 10.0
percent during the three previous calendar years and a capacity factor
of no more than 20.0 percent in each of those calendar years; or
(B) One calendar year of data following the change in the unit's
capacity factor showing a capacity factor of no more than 10.0 percent
and a statement that this changed pattern of operation resulting in a
capacity factor less than 10.0 percent is considered permanent and is
projected to continue for the foreseeable future.
(3) For purposes of part 75 of this chapter, a unit that initially
qualifies as a peaking unit must meet the criteria in paragraph (1) of
this definition each year in order to continue to qualify as a peaking
unit. If such a unit fails to meet such criteria for a given year, the
unit no longer qualifies as a peaking unit starting January 1 of the
year after the year for which the criteria are not met. If a unit
failing to meet the criteria in paragraph (1) of this definition
initially qualified as a gas-fired unit under paragraph (2)(ii) of this
definition, the unit may qualify as a peaking unit for a subsequent
year only under paragraph (2)(i) of this definition.
* * * * *
Pipeline natural gas means natural gas that is provided by a
supplier through a pipeline and that contains 0.3 grains or less of
hydrogen sulfide per 100 standard cubic feet. The hydrogen sulfide
content of the natural gas must be documented either through quality
characteristics specified by a purchase contract or pipeline
transportation contract, through certification of the gas vendor, based
on routine vendor sampling and analysis, or through at least one year's
worth of analytical data on the fuel hydrogen sulfide content from
samples taken at least monthly, demonstrating that all samples contain

[[Page 28119]]

0.3 grains or less of hydrogen sulfide per 100 standard cubic feet.
* * * * *
Probationary calibration error test means an on-line calibration
error test performed in accordance with section 2.1.1 of appendix B of
part 75 of this chapter that is used to initiate a conditionally valid
data period.
* * * * *
Research gas mixture (RGM) means a calibration gas mixture
developed by agreement of a requestor and NIST that NIST analyzes and
certifies as ``NIST traceable.'' RGMs may have concentrations different
from those of standard reference materials.
* * * * *
Span means the highest pollutant or diluent concentration or flow
rate that a monitor component is required to be capable of measuring
under part 75 of this chapter.
* * * * *
Standard reference material-equivalent compressed gas primary
reference material (SRM-equivalent PRM) means those gas mixtures listed
in a declaration of equivalence in accordance with section 2.1.2 of the
``EPA Traceability Protocol for Assay and Certification of Gaseous
Calibration Standards,'' September 1997, EPA-600/R-97/121.
* * * * *
Stationary gas turbine means a turbine that is not self-propelled
and that combusts natural gas, other gaseous fuel with a total sulfur
content no greater than the total sulfur content of natural gas, or
fuel oil in order to heat inlet combustion air and thereby turn a
turbine, in addition to or instead of producing steam or heating water.
* * * * *
Zero air material means either:
(1) A calibration gas certified by the gas vendor not to contain
concentrations of SO2, NO_X, or total hydrocarbons
above 0.1 parts per million (ppm), a concentration of CO above 1 ppm, a
concentration of CO2 above 400 ppm; or
(2) Ambient air conditioned and purified by a CEMS for which the
CEMS manufacturer or vendor certifies that the particular CEMS model
produces conditioned gas that does not contain concentrations of
SO2, NO_X, or total hydrocarbons above 0.1 ppm, a
concentration of CO above 1 ppm, or a concentration of CO2
above 400 ppm; or
(3) For dilution-type CEMS, conditioned and purified ambient air
provided by a conditioning system concurrently supplying dilution air
to the CEMS; or
(4) A multicomponent mixture certified by the supplier of the
mixture that the concentration of the component being zeroed is less
than or equal to the applicable concentration specified in paragraph
(1) of this definition, and that the mixture's other components do not
interfere with the specific CEM readings or cause the CEM being zeroed
to read concentrations of the gas being zeroed.
3. Section 72.3 is amended by adding in alphabetical order, new
acronyms for kacfm, kscfh, and NIST to read as follows:

Sec. 72.3 Measurements, abbreviations, and acronyms.

* * * * *
kacfm--thousands of cubic feet per minute at actual conditions.
kscfh--thousands of cubic feet per hour at standard conditions.
NIST--National Institute of Standards and Technology.
* * * * *

Sec. 72.6 [Amended]

4. Section 72.6 is amended by removing from paragraph (b)(1) the
word ``operation'' and adding, in its place, the words ``commercial
operation.''
5. Section 72.90 is amended by revising paragraph (c)(3) to read as
follows:

Sec. 72.90 Annual compliance certification report.

* * * * *
(c) * * *
(3) Whether all the emissions from the unit, or a group of units
(including the unit) using a common stack, were monitored or accounted
for through the missing data procedures and reported in the quarterly
monitoring reports, including whether conditional data were reported in
the quarterly report. If conditional data were reported, the owner or
operator shall indicate whether the status of all conditional data has
been resolved and all necessary quarterly report resubmissions have
been made.
* * * * *

PART 75--CONTINUOUS EMISSION MONITORING

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

Authority: 42 U.S.C. 7601 and 7651k.

7. Section 75.1 is amended by revising paragraph (a) to read as
follows:

Sec. 75.1 Purpose and scope.

(a) Purpose. The purpose of this part is to establish requirements
for the monitoring, recordkeeping, and reporting of sulfur dioxide,
nitrogen oxides, and carbon dioxide emissions, volumetric flow, and
opacity data from affected units under the Acid Rain Program pursuant
to Sections 412 and 821 of the Clean Air Act, 42 U.S.C. 7401-7671q as
amended by Public Law 101-549 (November 15, 1990) (the Act). In
addition, this part sets forth provisions for the monitoring,
recordkeeping, and reporting of NO_X mass emissions with
which EPA, individual States, or groups of States may require sources
to comply in order to demonstrate compliance with a NO_X mass
emission reduction program, if these provisions are adopted as
requirements under such a program.
* * * * *
8. Section 75.2 is amended by revising paragraph (a) and adding a
new paragraph (c) to read as follows:

Sec. 75.2 Applicability.

(a) Except as provided in paragraphs (b) and (c) of this section,
the provisions of this part apply to each affected unit subject to Acid
Rain emission limitations or reduction requirements for SO2
or NO_X.
* * * * *
(c) The provisions of this part may apply to sources subject to a
State or federal NO_X mass emission reduction program, if
these provisions are adopted as requirements under such a program.
9. Section 75.4 is amended by revising paragraphs (a) introductory
text and (d)(1) and adding a new paragraph (i) to read as follows:

Sec. 75.4 Compliance dates.

(a) The provisions of this part apply to each existing Phase I and
Phase II unit on February 10, 1993. For substitution or compensating
units that are so designated under the Acid Rain permit which governs
that unit and contains the approved substitution or reduced utilization
plan, pursuant to Sec. 72.41 or Sec. 72.43 of this chapter, the
provisions of this part become applicable upon the issuance date of the
Acid Rain permit. For combustion sources seeking to enter the Opt-in
Program in accordance with part 74 of this chapter, the provisions of
this part become applicable upon the submission of an Opt-in permit
application in accordance with Sec. 74.14 of this chapter. The
provisions of this part for the monitoring, recording, and reporting of
NO_X mass emissions become applicable on the deadlines
specified in the applicable State or federal NO_X mass
emission reduction program, if these provisions are adopted as
requirements under such a program. In accordance with Sec. 75.20, the
owner or operator of each existing affected unit shall ensure that all
monitoring systems required by

[[Page 28120]]

this part for monitoring SO2, NO_X,
CO2, opacity, and volumetric flow are installed and that all
certification tests are completed no later than the following dates
(except as provided in paragraphs (d) through (h) of this section):
* * * * *
(d) * * *
(1) The maximum potential concentration of SO2, the
maximum potential NO_X emission rate, the maximum potential
flow rate, as defined in section 2.1 of appendix A to this part, or the
maximum potential CO2 concentration, as defined in section
2.1.3.1 of appendix A to this part.
* * * * *
(i) In accordance with Sec. 75.20, the owner or operator of each
affected unit at which SO2 concentration is measured on a
dry basis or at which moisture corrections are required to account for
CO2 emissions, NO_X emission rate in lb/mmBtu, or
heat input, shall ensure that the continuous moisture monitoring system
required by this part is installed and that all applicable initial
certification tests required under Sec. 75.20(c)(5), (c)(6), or (c)(7)
for the continuous moisture monitoring system are completed no later
than the following dates:
(1) January 1, 2000, for a unit that is existing and has commenced
commercial operation by October 3, 1999; or
(2) For a new affected unit which has not commenced commercial
operation by October 4, 1999, not later than 90 days after the date the
unit commences commercial operation; or
(3) For an existing unit that is shutdown and is not yet operating
by January 1, 2000, not later than the earlier of 45 unit operating
days or 180 calendar days after the date that the unit recommences
commercial operation.
10. Section 75.5 is amended by revising paragraph (f)(2) to read as
follows:

Sec. 75.5 Prohibitions.

* * * * *
(f) * * *
(2) The owner or operator is monitoring emissions from the unit
with another certified monitoring system or an excepted methodology
approved by the Administrator for use at that unit that provides
emission data for the same pollutant or parameter as the retired or
discontinued monitoring system; or
* * * * *
11. Section 75.6 is amended by redesignating paragraph (a)(40) as
paragraph (a)(41) and by adding new paragraphs (a)(40) and (f) to read
as follows:

Sec. 75.6 Incorporation by reference.

* * * * *
(a) * * *
(40) ASTM D5373-93, ``Standard Methods for Instrumental
Determination of Carbon, Hydrogen, and Nitrogen in Laboratory Samples
of Coal and Coke,'' for appendix G to this part.
* * * * *
(f) The following materials are available for purchase from the
following address: American Petroleum Institute, Publications
Department, 1220 L Street NW, Washington, DC 20005-4070: American
Petroleum Institute (API) Section 2, ``Conventional Pipe Provers,''
from Chapter 4 of the Manual of Petroleum Measurement Standards,
October 1988 (Reaffirmed 1993), for Sec. 75.20 and appendix D to this
part.
12. Section 75.10 is amended by revising paragraphs (d)(3) and (f)
to read as follows:

Sec. 75.10 General operating requirements.

* * * * *
(d) * * *
(3) Failure of an SO2, CO2, or O2
pollutant concentration monitor, flow monitor, or NO_X
continuous emission monitoring system to acquire the minimum number of
data points for calculation of an hourly average in paragraph (d)(1) of
this section, shall result in the failure to obtain a valid hour of
data and the loss of such component data for the entire hour. An hourly
average NO_X or SO2 emission rate in lb/mmBtu is
valid only if the minimum number of data points is acquired by both the
pollutant concentration monitor (NO_X or SO2) and
the diluent monitor (O2 or CO2). For a moisture
monitoring system consisting of one or more oxygen analyzers capable of
measuring O2 on a wet-basis and a dry-basis, an hourly
average percent moisture value is valid only if the minimum number of
data points is acquired for both the wet-and dry-basis measurements.
Except for SO2 emission rate data in lb/mmBtu, if a valid
hour of data is not obtained, the owner or operator shall estimate and
record emission, moisture, or flow data for the missing hour by means
of the automated data acquisition and handling system, in accordance
with the applicable procedure for missing data substitution in subpart
D of this part.
* * * * *
(f) Minimum measurement capability requirement. The owner or
operator shall ensure that each continuous emission monitoring system
and component thereof is capable of accurately measuring, recording,
and reporting data, and shall not incur a full scale exceedance, except
as provided in sections 2.1.1.5, 2.1.2.5, and 2.1.4.3 of appendix A to
this part.
* * * * *
13. Section 75.11 is amended by revising paragraphs (a), (b),
(d)(1), (d)(2), (e)(2), (e)(3) introductory text, (e)(3)(ii),
(e)(3)(iv), and (e)(4) and by adding paragraph (d)(3), to read as
follows:

Sec. 75.11 Specific provisions for monitoring SO2 emissions
(SO2 and flow monitors).

(a) Coal-fired units. The owner or operator shall meet the general
operating requirements in Sec. 75.10 for an SO2 continuous
emission monitoring system and a flow monitoring system for each
affected coal-fired unit while the unit is combusting coal and/or any
other fuel, except as provided in paragraph (e) of this section, in
Sec. 75.16, and in subpart E of this part. During hours in which only
natural gas or gaseous fuel with a total sulfur content no greater than
the total sulfur content of natural gas (i.e., 20 grains
per 100 standard cubic feet (gr/100 scf)) is combusted in the unit, the
owner or operator shall comply with the applicable provisions of
paragraph (e)(1), (e)(2), or (e)(3) of this section.
(b) Moisture correction. Where SO2 concentration is
measured on a dry basis, the owner or operator shall install, operate,
maintain, and quality assure a continuous moisture monitoring system
for measuring and recording the moisture content of the flue gases, in
order to correct the measured hourly volumetric flow rates for moisture
when calculating SO2 mass emissions (in lb/hr) using the
procedures in appendix F to this part. The following continuous
moisture monitoring systems are acceptable: a continuous moisture
sensor; an oxygen analyzer (or analyzers) capable of measuring
O2 both on a wet basis and on a dry basis; or a stack
temperature sensor and a moisture look-up table, i.e., a psychrometric
chart (for saturated gas streams following wet scrubbers, only). The
moisture monitoring system shall include as a component the automated
data acquisition and handling system (DAHS) for recording and reporting
both the raw data (e.g., hourly average wet and dry-basis O2
values) and the hourly average values of the stack gas moisture content
derived from those data. When a moisture look-up table is used, the
moisture monitoring system shall be represented as a single component,
the certified DAHS, in the monitoring plan for the unit or common
stack.
* * * * *
(d) * * *
(1) By meeting the general operating requirements in Sec. 75.10 for
an SO2 continuous emission monitoring system

[[Page 28121]]

and flow monitoring system. If this option is selected, the owner or
operator shall comply with the applicable provisions in paragraph
(e)(1), (e)(2), or (e)(3) of this section during hours in which the
unit combusts only natural gas (or gaseous fuel with a total sulfur
content no greater than the total sulfur content of natural gas);
(2) By providing other information satisfactory to the
Administrator using the applicable procedures specified in appendix D
to this part for estimating hourly SO2 mass emissions.
Appendix D shall not, however, be used when the unit combusts gaseous
fuel with a total sulfur content greater than the total sulfur content
of natural gas (i.e., > 20 gr/100 scf); when such fuel is burned, the
owner or operator shall comply with the provisions of paragraph (e)(4)
of this section; or
(3) By using the low mass emissions excepted methodology in
Sec. 75.19(c) for estimating hourly SO2 mass emissions if
the affected unit qualifies as a low mass emissions unit under
Sec. 75.19(a) and (b).
(e) * * *
(2) When gaseous fuel with a total sulfur content no greater than
the total sulfur content of natural gas (i.e., 20 gr/100
scf) is combusted in the unit, the owner or operator may, in lieu of
operating and recording data from the SO2 monitoring system,
determine SO2 emissions by certifying an excepted monitoring
system in accordance with Sec. 75.20 and with appendix D to this part,
by following the fuel sampling and analysis procedures in section 2.3.1
of appendix D to this part, by meeting the recordkeeping requirements
of Sec. 75.55 or Sec. 75.58, as applicable, and by meeting all quality
control and quality assurance requirements for fuel flowmeters in
appendix D to this part. If this compliance option is selected, the
hourly unit heat input reported under Sec. 75.54(b)(5) or
Sec. 75.57(b)(5), as applicable, shall be determined using a certified
flow monitoring system and a certified diluent monitor, in accordance
with the procedures in section 5.2 of appendix F of this part. The flow
monitor and diluent monitor shall meet all of the applicable quality
control and quality assurance requirements of appendix B of this part.
(3) When gaseous fuel with a total sulfur content no greater than
the total sulfur content of natural gas (i.e., 20 gr/100
scf) is burned in the unit, the owner or operator may determine
SO2 mass emissions by using a certified SO2
continuous monitoring system, in conjunction with a certified flow rate
monitoring system. However, on and after January 1, 2000, the
SO2 monitoring system shall be subject to the following
provisions; prior to January 1, 2000, the owner or operator may comply
with these provisions:
* * * * *
(ii) The calibration response of the SO2 monitoring
system shall be adjusted, either automatically or manually, in
accordance with the procedures for routine calibration adjustments in
section 2.1.3 of appendix B to this part, whenever the zero-level
calibration response during a required daily calibration error test
exceeds the applicable performance specification of the instrument in
section 3.1 of appendix A to this part (i.e., 2.5 percent
of the span value or 5 ppm, whichever is less
restrictive). This calibration adjustment is optional if gaseous fuel
is burned in the affected unit only during unit startup.
* * * * *
(iv) In accordance with the requirements of section 2.1.1.2 of
appendix A to this part, for units that sometimes burn natural gas (or
gaseous fuel with a total sulfur content no greater than the total
sulfur content of natural gas) and at other times burn higher-sulfur
fuel(s) such as coal or oil, a second low-scale SO2
measurement range is not required when natural gas (or gaseous fuel
with a total sulfur content no greater than the total sulfur content of
natural gas) is combusted. For units that burn only natural gas (or
gaseous fuel with a total sulfur content no greater than the total
sulfur content of natural gas) and burn no other type(s) of fuel(s),
the owner or operator shall set the span of the SO2
monitoring system to a value no greater than 200 ppm.
(4) During any hours in which a unit combusts only gaseous fuel(s)
with a total sulfur content no greater than the total sulfur content of
natural gas (i.e., 20 gr/100 scf), the owner or operator
shall meet the general operating requirements in Sec. 75.10 for an
SO2 continuous emission monitoring system and a flow
monitoring system.
* * * * *
14. Section 75.12 is amended by revising the title; by
redesignating existing paragraphs (b), (c), and (d) as paragraphs (c),
(d), and (f), respectively; by adding new paragraphs (b) and (e); and
by revising the newly designated paragraph (c), to read as follows:

Sec. 75.12 Specific provisions for monitoring NO_X emission
rate (NO_X and diluent gas monitors).

* * * * *
(b) Moisture correction. If a correction for the stack gas moisture
content is needed to properly calculate the NO_X emission
rate in lb/mmBtu, i.e., if the NO_X pollutant concentration
monitor measures on a different moisture basis from the diluent
monitor, the owner or operator shall install, operate, maintain, and
quality assure a continuous moisture monitoring system, as defined in
Sec. 75.11(b).
(c) Determination of NO_X emission rate. The owner or
operator shall calculate hourly, quarterly, and annual NO_X
emission rates (in lb/mmBtu) by combining the NO_X
concentration (in ppm), diluent concentration (in percent O2
or CO2), and percent moisture (if applicable) measurements
according to the procedures in appendix F to this part.
* * * * *
(e) Low mass emissions units. Notwithstanding the requirements of
Secs. 75.12(a) and (c), the owner or operator of an affected unit that
qualifies as a low mass emissions unit under Sec. 75.19(a) and (b)
shall comply with one of the following:
(1) Meet the general operating requirements in Sec. 75.10 for a
NO_X continuous emission monitoring system;
(2) Meet the requirements specified in paragraph (d)(2) of this
section for using the excepted monitoring procedures in appendix E to
this part, if applicable; or
(3) Use the low mass emissions excepted methodology in
Sec. 75.19(c) for estimating hourly NO_X emission rate and
hourly NO_X mass emissions.
* * * * *
15. Section 75.13 is amended by revising paragraphs (a) and (c) and
by adding paragraph (d) to read as follows:

Sec. 75.13 Specific provisions for monitoring CO2
emissions.

(a) CO2 continuous emission monitoring system. If the
owner or operator chooses to use the continuous emission monitoring
method, then the owner or operator shall meet the general operating
requirements in Sec. 75.10 for a CO2 continuous emission
monitoring system and flow monitoring system for each affected unit.
The owner or operator shall comply with the applicable provisions
specified in Secs. 75.11(a) through (e) or Sec. 75.16, except that the
phrase ``SO2 continuous emission monitoring system'' is
replaced with ``CO2 continuous emission monitoring system,''
the phrase ``SO2 concentration'' is replaced with
``CO2 concentration,'' the term ``maximum potential
concentration of SO2'' is replaced with ``maximum potential
concentration of CO2,'' and the phrase ``SO2 mass
emissions'' is replaced with ``CO2 mass emissions.''
* * * * *
(c) Determination of CO2 mass emissions using an O2
monitor

[[Page 28122]]

according to appendix F. If the owner or operator chooses to use the
appendix F method, then the owner or operator may determine hourly
CO2 concentration and mass emissions with a flow monitoring
system; a continuous O2 concentration monitor; fuel F and
Fc factors; and, where O2 concentration is
measured on a dry basis, a continuous moisture monitoring system, as
defined in Sec. 75.11(b), using the methods and procedures specified in
appendix F to this part. For units using a common stack, multiple
stack, or bypass stack, the owner or operator may use the provisions of
Sec. 75.16, except that the phrase ``SO2 continuous emission
monitoring system'' is replaced with ``CO2 continuous
emission monitoring system,'' the term ``maximum potential
concentration of SO2'' is replaced with ``maximum potential
concentration of CO2,'' and the phrase ``SO2 mass
emissions'' is replaced with ``CO2 mass emissions.''
(d) Determination of CO2 mass emissions from low mass
emissions units. The owner or operator of a unit that qualifies as a
low mass emissions unit under Secs. 75.19(a) and (b) shall comply with
one of the following:
(1) Meet the general operating requirements in Sec. 75.10 for a
CO2 continuous emission monitoring system and flow
monitoring system;
(2) Meet the requirements specified in paragraph (b) or (c) of this
section for use of the methods in appendix G or F to this part,
respectively; or
(3) Use the low mass emissions excepted methodology in
Sec. 75.19(c) for estimating hourly CO2 mass emissions.
16. Section 75.16 is amended by:
a. Revising paragraphs (b)(2)(ii)(B), (b)(2)(ii)(D), (d)(2), and
(e)(1);
b. Removing paragraphs (e)(2) and (e)(3);
c. Redesignating existing paragraphs (e)(4) and (e)(5) as
paragraphs (e)(2) and (e)(3), respectively;
d. Revising the last sentence and adding a new sentence to the end
of the newly designated paragraph (e)(3); and
e. Adding a new paragraph (e)(4), to read as follows:

Sec. 75.16 Special provisions for monitoring emissions from common,
bypass, and multiple stacks for SO2 emissions and heat input
determinations.

* * * * *
(b) * * *
(2) * * *
(ii) * * *
(B) Install, certify, operate, and maintain an SO2
continuous emission monitoring system and flow monitoring system in the
duct from each nonaffected unit; determine SO2 mass
emissions from the affected units as the difference between
SO2 mass emissions measured in the common stack and
SO2 mass emissions measured in the ducts of the nonaffected
units, not to be reported as an hourly average value less than zero;
combine emissions for the Phase I and Phase II affected units for
recordkeeping and compliance purposes; calculate and report
SO2 mass emissions from the Phase I and Phase II affected
units, pursuant to an approach approved by the Administrator, such that
these emissions are not underestimated; or
* * * * *
(D) Petition through the designated representative and provide
information satisfactory to the Administrator on methods for
apportioning SO2 mass emissions measured in the common stack
to each of the units using the common stack and on reporting the
SO2 mass emissions. The Administrator may approve such
demonstrated substitute methods for apportioning and reporting
SO2 mass emissions measured in a common stack whenever the
demonstration ensures that there is a complete and accurate accounting
of all emissions regulated under this part and, in particular, that the
emissions from any affected unit are not underestimated.
* * * * *
(d) * * *
(2) Install, certify, operate, and maintain an SO2
continuous emission monitoring system and flow monitoring system in
each stack. Determine SO2 mass emissions from each affected
unit as the sum of the SO2 mass emissions recorded for each
stack. Notwithstanding the prior sentence, if another unit also
exhausts flue gases to one or more of the stacks, the owner or operator
shall also comply with the applicable common stack requirements of this
section to determine and record SO2 mass emissions from the
units using that stack and shall calculate and report SO2
mass emissions from the affected units and stacks, pursuant to an
approach approved by the Administrator, such that these emissions are
not underestimated.
(e) * * *
(1) The owner or operator of an affected unit using a common stack,
bypass stack, or multiple stack with a diluent monitor and a flow
monitor on each stack may choose to install monitors to determine the
heat input for the affected unit, wherever flow and diluent monitor
measurements are used to determine the heat input, using the procedures
specified in paragraphs (a) through (d) of this section, except that
the terms ``SO2 mass emissions'' and ``emissions'' are
replaced with the term ``heat input'' and the phrase ``SO2
continuous emission monitoring system and flow monitoring system'' is
replaced with the phrase ``a diluent monitor and a flow monitor.'' The
applicable equation in appendix F to this part shall be used to
calculate the heat input from the hourly flow rate, diluent monitor
measurements, and (if the equation in appendix F requires a correction
for the stack gas moisture content) hourly moisture measurements.
Notwithstanding the options for combining heat input in paragraphs
(a)(1)(ii), (a)(2)(ii), (b)(1)(ii), and (b)(2)(ii) of this section, the
owner or operator of an affected unit with a diluent monitor and a flow
monitor installed on a common stack to determine the combined heat
input at the common stack shall also determine and report heat input to
each individual unit.
* * * * *
(3) * * * The heat input may be apportioned either by using the
ratio of load (in MWe-hr) for each individual unit to the total load
for all units utilizing the common stack or by using the ratio of steam
flow (in 1000 lb) for each individual unit to the total steam flow for
all units utilizing the common stack. The heat input should be
apportioned according to the procedures in appendix F to this part.
(4) Notwithstanding paragraph (e)(1) of this section, any affected
unit that is using the procedures in this part to meet the monitoring
and reporting requirements of a State or federal NO_X mass
emission reduction program must also meet the requirements for
monitoring heat input in Secs. 75.71 and 75.72.
17. Section 75.17 is amended by adding introductory text before
paragraph (a) and by revising paragraph (a)(2)(i)(C) to read as
follows:

Sec. 75.17 Specific provisions for monitoring emissions from common,
by-pass, and multiple stacks for NO_X emission rate.

Notwithstanding the provisions of paragraphs (a), (b), and (c) of
this section, the owner or operator of an affected unit that is using
the procedures in this part to meet the monitoring and reporting
requirements of a State or federal NO_X mass emission
reduction program must also meet the provisions for monitoring
NO_X emission rate in Secs. 75.71 and 75.72.
(a) * * *
(2) * * *
(i) * * *
(C) Each unit's compliance with the applicable NO_X
emission limit will be determined by a method satisfactory to

[[Page 28123]]

the Administrator for apportioning to each of the units the combined
NO_X emission rate (in lb/mmBtu) measured in the common stack
and for reporting the NO_X emission rate, as provided in a
petition submitted by the designated representative. The Administrator
may approve such demonstrated substitute methods for apportioning and
reporting NO_X emission rate measured in a common stack
whenever the demonstration ensures that there is a complete and
accurate estimation of all emissions regulated under this part and, in
particular, that the emissions from any unit with a NO_X
emission limitation are not underestimated.
* * * * *
18. Section 75.19 is added to subpart B to read as follows:

Sec. 75.19 Optional SO2, NO_X, and CO2
emissions calculation for low mass emissions units.

(a) Applicability. (1) Consistent with the requirements of
paragraphs (a)(2) and (b) of this section, the low mass emissions
excepted methodology in paragraph (c) of this section may be used in
lieu of continuous emission monitoring systems or, if applicable, in
lieu of excepted methods under appendix D or E to this part, for the
purpose of determining hourly heat input, hourly NO_X
emission rate, and hourly NO_X, SO2, and
CO2 mass emissions from a low mass emissions unit. A low
mass emissions unit is a gas-fired or oil-fired unit that burns only
natural gas and/or fuel oil and that:
(i) Emits no more than 25 tons of SO2 annually and no
more than 25 tons of NO_X annually; and
(ii) Has calculated emissions of no more than 25 tons of
SO2 annually and no more than 25 tons of NO_X
annually based on the maximum rated hourly heat input, the actual
operating time for each fuel burned, and the low mass emissions
excepted methodology, calculations, and values in paragraph (c) of this
section.
(2) A unit may initially qualify as a low mass emissions unit only
under the following circumstances:
(i) The designated representative provides historical actual and
calculated emissions data from the previous three calendar years
immediately prior to the submission of an application to use the low
mass emissions excepted methodology, and the data demonstrates to the
satisfaction of the Administrator that the unit meets the criteria in
paragraphs (a)(1)(i) and (ii) of this section; or
(ii) If a unit does not have the historical data required in
paragraph (a)(2)(i) of this section for any one or more of the previous
three calendar years, the designated representative submits:
(A) Any historical annual emissions and operating data, as required
in paragraphs (a)(1)(i) and (a)(1)(ii) of this section, beginning with
the unit's first calendar year of commercial operation, and the data
demonstrates to the satisfaction of the Administrator that the unit
meets the criteria in paragraphs (a)(1)(i) and (a)(1)(ii) of this
section; and
(B) A demonstration satisfactory to the Administrator that the unit
will continue to qualify as a low mass emissions unit under the
requirements of this paragraph (a). The demonstration shall include any
historical emissions and operating data for less than a calendar year
for the unit and projected emissions information for the unit, as
determined using projected operating hours and fuel usage, and the low
mass emissions excepted methodology, calculations, and values in
paragraph (c) of this section.
(b) Disqualification. If a unit that initially qualifies as a low
mass emissions units under this section changes the fuel that is burned
in the unit such that a fuel other than natural gas or fuel oil is
combusted in the unit, the unit is disqualified from using the low mass
emissions excepted methodology as of the first hour that the new fuel
is combusted in the unit. In addition, if a unit that initially
qualifies as a low mass emissions unit under this section emits more
than 25 tons of SO2 or 25 tons of NO_X in any
calendar year or has calculated emissions greater than 25 tons of
SO2 or 25 tons of NO_X in any calendar year, as
determined using the low mass emission equations in paragraph (c) of
this section, the owner or operator of the unit shall have two quarters
from the end of the quarter in which the exceedance occurs to install,
certify, and report SO2, NO_X, and CO2
from monitoring systems that meet the requirements of Secs. 75.11,
75.12, and 75.13, respectively. The unit shall be disqualified as a low
mass emissions unit as of the end of the second quarter following the
quarter in which either of the 25 ton limits was exceeded. A unit that
has been disqualified from using the low mass emissions excepted
methodology may subsequently qualify again as a low mass emissions unit
under paragraph (a)(2) of this section, provided that if such unit
qualified under paragraph (a)(2)(ii) of this section, the unit may
subsequently qualify again if the unit meets the requirements of
paragraph (a)(2)(i) of this section.
(c) Low mass emissions excepted methodology, calculations, and
values.--(1) Operating time. (i) Report an hourly record if the unit
operated for any portion of the hour or if records are missing, as to
whether or not the unit operated for any portion of that hour.
(ii) Quarterly operating time (hr) is equal to the sum of all of
the reported operating hours in the quarter, such that any hour in
which the unit combusted fuel for any portion of the hour is considered
a full hour.
(iii) Year-to-date cumulative operating time (hr) is equal to the
sum of all of the reported operating hours in the year to date, such
that any hour in which the unit combusted fuel for any portion of the
hour is considered a full hour.
(2) Heat input. (i) Hourly heat input (mmBtu) is equal to the
maximum rated hourly heat input, as defined in Sec. 72.2 of this
chapter. However, the owner or operator of an affected unit may
petition the Administrator under Sec. 75.66 for a lower value for
maximum rated hourly heat input than that defined in Sec. 72.2 of this
chapter. The Administrator may approve such lower value if the owner or
operator demonstrates that either the maximum hourly heat input
specified by the manufacturer or the highest observed hourly heat
input, or both, are not representative of the unit's current
capabilities because modifications have been made to the unit, limiting
its capacity permanently.
(ii) Calculate the quarterly total heat input (mmBtu) using
Equation 7a as follows:

HIqtr = Tqtr x HIhr

(Eq. 7a)
where:

Tqtr = Actual number of operating hours in the quarter, in
hr.
HIhr = Hourly heat input under paragraph (c)(2)(i) of this
section, in mmBtu.

(iii) Calculate the year-to-date cumulative heat input (mmBtu) as
the sum of all of the hourly heat input values in the year to date.
(3) SO2. (i) Calculate the hourly total SO2
mass emissions (lbs) using Equation 7b and the appropriate fuel-based
SO2 emission factor from Table 1a for the fuel being burned
in that hour. If more than one fuel is burned in the hour, use the
highest emission factor for all of the fuels burned in the hour. If
records are missing as to which fuel was burned in the hour, use the
highest emission factor for all of the fuels capable of being burned in
that unit.

Table 1a.--SO2 Emission Factors (lb/mmBtu) for Various Fuel Types
------------------------------------------------------------------------
Fuel type SO2 Emission factors
------------------------------------------------------------------------
Pipeline Natural Gas...................... 0.0006 lb/mmBtu.

[[Page 28124]]

Natural Gas............................... 0.06 lb/mmBtu.
Residual Oil.............................. 2.1 lb/mmBtu.
Diesel Fuel............................... 0.5 lb/mmBtu.
------------------------------------------------------------------------

WSO2 = EFSO2 x HIhr

(Eq. 7b)

Where:

WSO2 = SO2 mass emissions, in lbs.
EFSO2 = Fuel-based SO2 emission factor
from Table 1a of this section, in lb/mmBtu.
HIhr = Hourly heat input under paragraph (c)(2)(i) of this
section, in mmBtu.

(ii) Calculate the quarterly total SO2 mass emissions
(tons) by summing all of the hourly SO2 mass emissions under
paragraph (c)(3)(i) of this section in the quarter and dividing by 2000
lb/ton.
(iii) Calculate the year-to-date cumulative SO2 mass
emissions (tons) by summing all of the SO2 mass emissions
under paragraph (c)(3)(i) of this section in the year to date.
(4) NO_X. (i) Determine the hourly NO_X
emission rate (lb/mmBtu) by using the appropriate fuel and boiler type
default NO_X emission rate in Table 1b for the fuel being
burned in that hour. If more than one fuel is burned in the hour, use
the highest emission rate for all of the fuels burned in the hour. If
records are missing as to which fuel was burned in the hour, use the
highest emission factor for all of the fuels capable of being burned in
that unit.

Table 1b.--NOX Emission Rates (lb/mmBtu) for Various Boiler/Fuel Types
------------------------------------------------------------------------
NOX
Boiler type Fuel type Emission
rate
------------------------------------------------------------------------
Tangentially fired................. Oil................... 0.366
Tangentially fired................. Gas................... 0.290
Dry Bottom Wall fired.............. Oil................... 0.490
Dry Bottom Wall fired.............. Gas................... 0.400
Combustion Turbine................. Oil................... 0.258
Combustion Turbine................. Gas................... 0.172
Combined Cycle..................... Oil................... 0.273
Combined Cycle..................... Gas................... 0.273
------------------------------------------------------------------------

(ii) Calculate the hourly total NO_X mass emissions (lbs)
as the product of the NO_X emission rate (lb/mmBtu) and
hourly heat input (mmBtu), using Equation 7c as follows:

WNOX = EFNOX x HIhr

(Eq. 7c)
where:

WNOX = NO_X mass emissions, in lbs.
EFNOX = Boiler-type and fuel-type NO_X emission
factor from Table 1b of this section, in lb/mmBtu.
HIhr = Hourly heat input under paragraph (c)(2)(i) of this
section, in mmBtu.

(iii) Calculate the quarterly average NO_X emission rate
(lb/mmBtu) by summing all of the hourly NO_X emission rates
for the quarter and dividing the total by the number of reported
operating hours under paragraph (c)(1)(i) of this section in the
quarter.
(iv) Calculate the quarterly total NO_X mass emissions
(tons) by summing all of the hourly NO_X mass emissions under
paragraph (c)(4)(ii) of this section in the quarter and dividing the
total by 2000 lb/ton.
(v) Calculate the year-to-date cumulative average NO_X
emission rate (lb/mmBtu) by summing all of the hourly NO_X
emission rates for all of the hours in the year to date and dividing
the total by the number of reported operating hours under paragraph
(c)(1)(i) of this section in the year to date.
(vi) Calculate the year-to-date cumulative NO_X mass
emissions total (tons) by summing all of the hourly NO_X mass
emissions under paragraph (c)(4)(ii) of this section in the year to
date.
(5) CO2. (i) Calculate the hourly total CO2
mass emissions (tons) using Equation 7d and the appropriate fuel-based
CO2 emission factor from Table 1c for the fuel being burned
in that hour. If more than one fuel is burned in the hour, use the
highest emission factor for all of the fuels burned in the hour. If
records are missing as to which fuel was burned in the hour, use the
highest emission factor for all of the fuels capable of being burned in
that unit.

Table 1c.--CO2 Emission Factors (ton/mmBtu) for Gas and Oil
------------------------------------------------------------------------
Fuel type CO2 emission factors
------------------------------------------------------------------------
Natural Gas............................... 0.059 ton/mmBtu.
Oil....................................... 0.081 ton/mmBtu.
------------------------------------------------------------------------

WCO2=EFCO2 x HIhr

(Eq. 7d)

Where:

WCO2 = CO2 mass emissions, in tons.
EFCO2 = Fuel-based CO2 emission factor from Table
1c, in ton/mmBtu.
HIhr = Hourly heat input under paragraph (c)(2)(i) of this
section, in mmBtu.

(ii) Calculate the quarterly total CO2 mass emissions
(tons) by summing all of the hourly CO2 mass emissions under
paragraph (c)(5)(i) of this section in the quarter.
(iii) Calculate the year-to-date cumulative CO2 mass
emissions (tons) by summing all of the hourly CO2 mass
emissions under paragraph (c)(5)(i) of this section in the year to
date.
(d) The quality control and quality assurance requirements in
Sec. 75.21 are not required for a low mass emissions unit for which the
optional low mass emissions excepted methodology in paragraph (c) of
this section is being used in lieu of a continuous emission monitoring
system or an excepted monitoring system under appendix D or E to this
part.

Subpart C--[Amended]

19. Section 75.20 is amended by:
a. Revising the title of the section;
b. Revising the titles of paragraphs (a)(3), (a)(4), (c), (d), (g),
(g)(1), (g)(2), (g)(4), and (g)(5);
c. Revising paragraphs (a) introductory text, (a)(1), (a)(3),
(a)(4) introductory text, (a)(4)(i), (a)(4)(ii), (a)(4)(iii),
(a)(5)(i), (b), (c) introductory text, (c)(1)(iii), (d)(1), (d)(2), (g)
introductory text, (g)(1) introductory text, (g)(1)(i), (g)(2), (g)(4),
and (g)(5);
d. Removing existing paragraph (c)(3);
e. Revising and redesignating existing paragraphs (c)(4), (c)(5),
(c)(6), (c)(7), and (c)(8) as paragraphs (c)(3), (c)(4), (c)(8),
(c)(9), and (c)(10), respectively; and revising newly designated
paragraphs (c)(4) introductory text, (c)(8) introductory text,
(c)(8)(i),

[[Page 28125]]

(c)(9)(ii), and (c)(10) introductory text; and
f. Adding new paragraphs (c)(5), (c)(6), (c)(7), (g)(6), (g)(7),
(h), and (i), to read as follows:

Sec. 75.20 Initial certification and recertification procedures.

(a) Initial certification approval process. The owner or operator
shall ensure that each continuous emission or opacity monitoring system
required by this part, which includes the automated data acquisition
and handling system, and, where applicable, the CO2
continuous emission monitoring system, meets the initial certification
requirements of this section and shall ensure that all applicable
initial certification tests under paragraph (c) of this section are
completed by the deadlines specified in Sec. 75.4 and prior to use in
the Acid Rain Program. In addition, whenever the owner or operator
installs a continuous emission or opacity monitoring system in order to
meet the requirements of Secs. 75.13 through 75.18, where no continuous
emission or opacity monitoring system was previously installed, initial
certification is required.
(1) Notification of initial certification test dates. The owner or
operator or designated representative shall submit a written notice of
the dates of initial certification testing at the unit as specified in
Sec. 75.61(a)(1).
* * * * *
(3) Provisional approval of certification (or recertification)
applications. Upon the successful completion of the required
certification (or recertification) procedures of this section for each
continuous emission or opacity monitoring system or component thereof,
each continuous emission or opacity monitoring system or component
thereof shall be deemed provisionally certified (or recertified) for
use under the Acid Rain Program for a period not to exceed 120 days
following receipt by the Administrator of the complete certification
(or recertification) application under paragraph (a)(4) of this
section, provided that no continuous emission or opacity monitor
systems for a combustion source seeking to enter the Opt-in Program in
accordance with part 74 of this chapter shall be deemed provisionally
certified (or recertified) for use under the Acid Rain Program. Data
measured and recorded by a provisionally certified (or recertified)
continuous emission or opacity monitoring system or component thereof,
in accordance with the requirements of appendix B to this part, will be
considered valid quality-assured data (retroactive to the date and time
of provisional certification or recertification)), provided that the
Administrator does not invalidate the provisional certification (or
recertification) by issuing a notice of disapproval within 120 days of
receipt by the Administrator of the complete certification (or
recertification) application. Note that if the data validation
procedures of paragraph (b)(3) of this section are applied to the
initial certification (or recertification) of a continuous emissions
monitoring system, it is possible for data recorded by the CEMS during
the certification (or recertification) test period to be quality
assured retrospectively, upon completion of all of the certification
(or recertification) tests. Therefore, in certain instances, the date
and time of provisional certification (or recertification) of the CEMS
may be earlier than the date and time of completion of the required
certification (or recertification) tests.
(4) Certification (or recertification) application formal approval
process. The Administrator will issue a notice of approval or
disapproval of the certification (or recertification) application to
the owner or operator within 120 days of receipt of the complete
certification (or recertification) application. In the event the
Administrator does not issue such a written notice within 120 days of
receipt, each continuous emission or opacity monitoring system which
meets the performance requirements of this part and is included in the
certification (or recertification) application will be deemed certified
(or recertified) for use under the Acid Rain Program.
(i) Approval notice. If the certification (or recertification)
application is complete and shows that each continuous emission or
opacity monitoring system meets the performance requirements of this
part, then the Administrator will issue a written notice of approval of
the certification (or recertification) application within 120 days of
receipt.
(ii) Incomplete application notice. A certification (or
recertification) application will be considered complete when all of
the applicable information required to be submitted in Sec. 75.63 has
been received by the Administrator, the EPA Regional Office, and the
appropriate State and/or local air pollution control agency. If the
certification (or recertification) application is not complete, then
the Administrator will issue a written notice of incompleteness that
provides a reasonable timeframe for the designated representative to
submit the additional information required to complete the
certification (or recertification) application. If the designated
representative has not complied with the notice of incompleteness by a
specified due date, then the Administrator may issue a notice of
disapproval specified under paragraph (a)(4)(iii) of this section. The
120-day review period shall not begin prior to receipt of a complete
application.
(iii) Disapproval notice. If the certification (or recertification)
application shows that any continuous emission or opacity monitoring
system or component thereof does not meet the performance requirements
of this part, or if the certification (or recertification) application
is incomplete and the requirement for disapproval under paragraph
(a)(4)(ii) of this section has been met, the Administrator shall issue
a written notice of disapproval of the certification (or
recertification) application within 120 days of receipt. By issuing the
notice of disapproval, the provisional certification (or
recertification) is invalidated by the Administrator, and the data
measured and recorded by each uncertified continuous emission or
opacity monitoring system or component thereof shall not be considered
valid quality-assured data beginning with the following time: from the
hour of the probationary calibration error test that began the initial
certification (or recertification) test period, if the data validation
procedures of paragraph (b)(3) of this section were used to
retrospectively validate data; or from the date and time of completion
of the invalid certification tests until the date and time that the
owner or operator completes subsequently approved initial certification
tests, if the data validation procedures of paragraph (b)(3) of this
section were not used. The owner or operator shall follow the
procedures for loss of initial certification in paragraph (a)(5) of
this section for each continuous emission or opacity monitoring system
or component thereof which is disapproved for initial certification.
For each disapproved recertification, the owner or operator shall
follow the procedures of paragraph (b)(5) of this section.
* * * * *
(5) * * *
(i) Until such time, date, and hour as the continuous emission
monitoring system or component thereof can be adjusted, repaired, or
replaced and certification tests successfully completed, the owner or
operator shall substitute the following values, as applicable, for each
hour of unit operation during the period of invalid

[[Page 28126]]

data specified in paragraph (a)(4)(iii) of this section or in
Sec. 75.21: the maximum potential concentration of SO2 as
defined in section 2.1.1.1 of appendix A to this part to report
SO2 concentration; the maximum potential NO_X
emission rate, as defined in Sec. 72.2 of this chapter to report
NO_X emissions; the maximum potential flow rate, as defined
in section 2.1.4.1 of appendix A to this part to report volumetric
flow; or the maximum potential concentration of CO2, as
defined in section 2.1.3.1 of appendix A to this part to report
CO2 concentration data; and
* * * * *
(b) Recertification approval process. Whenever the owner or
operator makes a replacement, modification, or change in a certified
continuous emission monitoring system or continuous opacity monitoring
system that is determined by the Administrator to significantly affect
the ability of the system to accurately measure or record the
SO2 or CO2 concentration, stack gas volumetric
flow rate, NO_X emission rate, or opacity, or to meet the
requirements of Sec. 75.21 or appendix B to this part, the owner or
operator shall recertify the continuous emission monitoring system or
continuous opacity monitoring system, according to the procedures in
this paragraph. Furthermore, whenever the owner or operator makes a
replacement, modification, or change to the flue gas handling system or
the unit operation that is determined by the Administrator to
significantly change the flow or concentration profile, the owner or
operator shall recertify the monitoring system according to the
procedures in this paragraph. Examples of changes which require
recertification include: replacement of the analyzer; change in
location or orientation of the sampling probe or site; changing of flow
rate monitor polynomial coefficients; and complete replacement of an
existing continuous emission monitoring system or continuous opacity
monitoring system. The owner or operator shall recertify a continuous
opacity monitoring system whenever the monitor path length changes or
as required by an applicable State or local regulation or permit. Any
change to a stack flow rate or gas monitoring system for which the
Administrator determines that a RATA is not necessary shall not be
considered a recertification event. In such cases, any other tests that
the Administrator determines to be necessary (linearity checks,
calibration error tests, DAHS verifications, etc.) shall be performed
as diagnostic tests, rather than recertification tests. The data
validation procedures in paragraph (b)(3) of this section shall be
applied to linearity checks, 7-day calibration error tests, and cycle
time tests when these are required as diagnostic tests. When the data
validation procedures of paragraph (b)(3) of this section are applied
in this manner, replace the word ``recertification'' with the word
``diagnostic.''
(1) Tests required. For recertification testing after changing the
flow rate monitor polynomial coefficients, the owner or operator shall
complete a 3-level RATA. For all other recertification testing, the
owner or operator shall complete all initial certification tests in
paragraph (c) of this section that are applicable to the monitoring
system, except as otherwise approved by the Administrator.
(2) Notification of recertification test dates. The owner,
operator, or designated representative shall submit notice of testing
dates for recertification under this paragraph as specified in
Sec. 75.61(a)(1)(ii), unless all of the tests in paragraph (c) of this
section are required for recertification, in which case the owner or
operator shall provide notice in accordance with the notice provisions
for initial certification testing in Sec. 75.61(a)(1)(i).
(3) Recertification test period requirements and data validation.
(i) In the period extending from the hour of the replacement,
modification or change made to a monitoring system that triggers the
need to perform recertification test(s) of the CEMS to the hour of
successful completion of a probationary calibration error test
(according to paragraph (b)(3)(ii) of this section) following the
replacement, modification, or change to the CEMS, the owner or operator
shall either substitute for missing data, according to the standard
missing data procedures in Secs. 75.33 through 75.37, or report
emission data using a reference method or another monitoring system
that has been certified or approved for use under this part.
(ii) Once the modification or change to the CEMS has been completed
and all of the associated repairs, component replacements, adjustments,
linearization, and reprogramming of the CEMS have been completed, a
probationary calibration error test is required to establish the
beginning point of the recertification test period. In this instance,
the first successful calibration error test of the monitoring system
following completion of all necessary repairs, component replacements,
adjustments, reprogramming, and any preliminary tests (e.g., trial RATA
runs or a challenge of the monitor with calibration gases other than
those used to perform the daily calibration error test) shall be the
probationary calibration error test. The probationary calibration error
test must be passed before any of the required recertification tests
are commenced.
(iii) Beginning with the hour of commencement of a recertification
test period, emission data recorded by the CEMS are considered to be
conditionally valid, contingent upon the results of the subsequent
recertification tests.
(iv) Each required recertification test shall be completed no later
than the following number of unit operating hours after the
probationary calibration error test that initiates the test period:
(A) For a linearity test and/or cycle time test, 168 consecutive
unit operating hours;
(B) For a RATA (whether normal-load or multiple-load), 720
consecutive unit operating hours; and
(C) For a 7-day calibration error test, 21 consecutive unit
operating days.
(v) All recertification tests shall be performed hands-off, as
follows. No adjustments to the calibration of the CEMS, other than the
adjustments described in section 2.1.3 of appendix B to this part, are
permitted prior to or during the recertification test period. Routine
daily calibration error tests shall be performed throughout the
recertification test period, in accordance with section 2.1.1 of
appendix B to this part. The additional calibration error test
requirements in section 2.1.3 of appendix B to this part shall also
apply during the recertification test period.
(vi) If all of the required recertification tests and required
daily calibration error tests are successfully completed in succession
with no failures, and if each recertification test is completed within
the time period specified in paragraph (b)(3)(iv)(A), (B), or (C) of
this section, then all of the conditionally valid emission data
recorded by the CEMS shall be considered quality assured, from the hour
of commencement of the recertification test period until the hour of
completion of the required test(s).
(vii) If a required recertification test is failed or aborted due
to a problem with the CEMS, or if a calibration error test is failed
during a recertification test period, data validation shall be done as
follows:
(A) If any required recertification test is failed, it shall be
repeated. If any recertification test other than a 7-day calibration
error test is failed or aborted due to a problem with the CEMS, the
original recertification test period is ended, and a new
recertification test period must be commenced with a

[[Page 28127]]

probationary calibration error test. The tests that are required in
this new recertification test period will include any tests that were
required for the initial recertification event which were not
successfully completed and any recertification or diagnostic tests that
are required as a result of changes made to the monitoring system to
correct the problems that caused the failure of the recertification
test. The new recertification test sequence shall not be commenced
until all necessary maintenance activities, adjustments,
linearizations, and reprogramming of the CEMS have been completed;
(B) If a linearity test, RATA, or cycle time test is failed or
aborted due to a problem with the CEMS, all conditionally valid
emission data recorded by the CEMS are invalidated, from the hour of
commencement of the recertification test period to the hour in which
the test is failed or aborted. Data from the CEMS remain invalid until
the hour in which a new recertification test period is commenced,
following corrective action, and a probationary calibration error test
is passed, at which time the conditionally valid status of emission
data from the CEMS begins;
(C) If a 7-day calibration error test is failed within the
recertification test period, previously-recorded conditionally valid
emission data from the CEMS are not invalidated, provided that the
calibration error on the day of the failed 7-day calibration error test
does not exceed twice the performance specification in section 3 of
appendix A to this part; and
(D) If a calibration error test is failed (i.e., the results of the
test exceed twice the performance specification in section 3 of
appendix A to this part) during a recertification test period, the CEMS
is out-of-control as of the hour in which the calibration error test is
failed. Emission data from the CEMS shall be invalidated prospectively
from the hour of the failed calibration error test until the hour of
completion of a subsequent successful calibration error test following
corrective action, at which time the conditionally valid status of data
from the monitoring system resumes. Failure to perform a required daily
calibration error test during a recertification test period shall also
cause data from the CEMS to be invalidated prospectively, from the hour
in which the calibration error test was due until the hour of
completion of a subsequent successful calibration error test.
Previously-passed recertification tests in the sequence and previously-
recorded conditionally valid data shall not be affected by a late
calibration error test. Whenever a calibration error test is failed or
missed during a recertification test period, no further recertification
tests shall be performed until the required subsequent calibration
error has been passed, re-establishing the conditionally valid status
of data from the monitoring system.
(viii) If any required recertification test is not completed within
its allotted time period, data validation shall be done as follows. For
a late linearity test, RATA, or cycle time test that is passed on the
first attempt, data from the monitoring system shall be invalidated
from the hour of expiration of the recertification test period until
the hour of completion of the late test. For a late 7-day calibration
error test, whether or not it is passed on the first attempt, data from
the monitoring system shall also be invalidated from the hour of
expiration of the recertification test period until the hour of
completion of the late test. For a late linearity test, RATA, or cycle
time test that is failed on the first attempt or aborted on the first
attempt due to a problem with the monitor, all conditionally valid data
from the monitoring system shall be considered invalid back to the hour
of the first probationary calibration error test which initiated the
recertification test period. Data from the monitoring system shall
remain invalid until the hour of successful completion of the late
recertification test and any additional recertification or diagnostic
tests that are required as a result of changes made to the monitoring
system to correct problems that caused failure of the late
recertification test.
(ix) If any required recertification test of a monitoring system
has not been completed by the end of a calendar quarter and if data
contained in the quarterly report is conditionally valid pending the
results of test(s) to be completed in a subsequent quarter, the owner
or operator shall indicate this by means of a suitable conditional data
flag in the electronic quarterly report for that quarter. The owner or
operator shall resubmit the report for that quarter if the required
recertification test is subsequently failed. In the resubmitted report,
the owner or operator shall use the appropriate missing data routine in
Sec. 75.31 or Sec. 75.33 to replace with substitute data each hour of
conditionally valid data that was invalidated by the failed
recertification test. In addition, if the owner or operator submits any
conditionally valid data (as defined in Sec. 72.2 of this chapter) in
any of the four quarterly reports for a given year, the owner or
operator shall indicate the status of the conditionally valid data
(i.e., resolved or unresolved) in the annual compliance certification
report required under Sec. 72.90 of this chapter for that year.
Alternatively, if any required recertification test is not completed by
the end of a particular calendar quarter but is completed no later than
30 days after the end of that quarter (i.e., prior to the deadline for
submitting the quarterly report under Sec. 75.64), the test data and
results may be submitted with the earlier quarterly report even though
the test date(s) are from the next calendar quarter. In such instances,
if the recertification test(s) are passed in accordance with the
provisions of paragraph (b)(3) of this section, conditionally valid
data may be reported as quality-assured, in lieu of reporting a
conditional data flag. If the recertification test(s) is failed and if
conditionally valid data are replaced, as appropriate, with substitute
data, then neither the reporting of a conditional data flag nor
resubmission is required.
(x) If the replacement, modification, or change requiring
recertification of the CEMS is such that the data collected by the
prior certified monitoring system are no longer representative, such as
after a change to the flue gas handling system or unit operation that
requires changing the span value to be consistent with section 2.1 of
appendix A to this part, the owner or operator shall substitute for
missing data as follows, in the period extending from the hour of
commencement of the replacement, modification, or change requiring
recertification of the CEMS to the hour of commencement of the
recertification test period:
(A) For a change that results in a significantly higher
concentration or flow rate, substitute maximum potential values
according to the procedures in paragraph (a)(5) of this section; or
(B) For a change that results in a significantly lower
concentration or flow rate, substitute data using the standard missing
data procedures.
(C) The owner or operator shall then use the initial missing data
procedures in Sec. 75.31, beginning with the first hour of quality
assured data obtained with the recertified monitoring system, unless
otherwise provided by Sec. 75.34 for units with add-on emission
controls.
(4) Recertification application. The designated representative
shall apply for recertification of each continuous emission or opacity
monitoring system used under the Acid Rain Program. The owner or
operator shall submit the recertification application in accordance
with Sec. 75.60, and each complete recertification application shall
include the information specified in Sec. 75.63.
(5) Approval or disapproval of request for recertification. The
procedures for

[[Page 28128]]

provisional certification in paragraph (a)(3) of this section shall
apply to recertification applications. The Administrator will issue a
written notice of approval or disapproval according to the procedures
in paragraph (a)(4) of this section. In the event that a
recertification application is disapproved, data from the monitoring
system are invalidated and the applicable missing data procedures in
Sec. 75.31 or Sec. 75.33 shall be used from the date and hour of
receipt of such notice back to the hour of the probationary calibration
error test that began the recertification test period. Data from the
monitoring system remain invalid until a subsequent probationary
calibration error test is passed, beginning a new recertification test
period. The owner or operator shall repeat all recertification tests or
other requirements, as indicated in the Administrator's notice of
disapproval, no later than 30 unit operating days after the date of
issuance of the notice of disapproval. The designated representative
shall submit a notification of the recertification retest dates, as
specified in Sec. 75.61(a)(1)(ii), and shall submit a new
recertification application according to the procedures in paragraph
(b)(4) of this section.
(c) Initial certification and recertification procedures. Prior to
the deadline in Sec. 75.4, the owner or operator shall conduct initial
certification tests and in accordance with Sec. 75.63, the designated
representative shall submit an application to demonstrate that the
continuous emission or opacity monitoring system and components thereof
meet the specifications in appendix A to this part. The owner or
operator shall compare reference method values with output from the
automated data acquisition and handling system that is part of the
continuous emission monitoring system being tested. Except as specified
in paragraphs (b)(1), (d), and (e) of this section, the owner or
operator shall perform the following tests for initial certification or
recertification of continuous emission or opacity monitoring systems or
components according to the requirements of appendix A to this part:
(1) * * *
(iii) A relative accuracy test audit. For the NO_X-
diluent system, the RATA shall be done on a system basis, in units of
lb/mmBtu.
* * * * *
(3) The initial certification test data from an O2-or a
CO2-diluent gas monitor certified for use in a
NO_X continuous emission monitoring system may be submitted
to meet the requirements of paragraph (c)(4) of this section. Also, for
a diluent monitor that is used both as a CO2 monitoring
system and to determine heat input, only one set of diluent monitor
certification data need be submitted (under the component and system
identification numbers of the CO2 monitoring system).
(4) For each CO2 pollutant concentration monitor, each
O2 monitor which is part of a CO2 continuous
emission monitoring system, each diluent monitor used to monitor heat
input and each SO2-diluent continuous emission monitoring
system:
* * * * *
(5) For each continuous moisture monitoring system consisting of
wet-and dry-basis O2 analyzers:
(i) A 7-day calibration error test of each O2 analyzer;
(ii) A cycle time test of each O2 analyzer;
(iii) A linearity test of each O2 analyzer; and
(iv) A RATA, directly comparing the percent moisture measured by
the monitor to a reference method.
(6) For each continuous moisture sensor:
(i) A 7-day calibration error test; and
(ii) A RATA, directly comparing the percent moisture measured by
the monitor sensor to a reference method.
(7) For a continuous moisture monitoring system consisting of a
temperature sensor and a data acquisition and handling system (DAHS)
software component programmed with a moisture lookup table:
(i) A demonstration that the correct moisture value for each hour
is being taken from the moisture lookup tables and applied to the
emission calculations. At a minimum, the demonstration shall be made at
three different temperatures covering the normal range of stack
temperatures.
(ii) [Reserved]
(8) The owner or operator shall ensure that initial certification
or recertification of a continuous opacity monitor for use under the
Acid Rain Program is conducted according to one of the following
procedures:
(i) Performance of the tests for initial certification or
recertification, according to the requirements of Performance
Specification 1 in appendix B to part 60 of this chapter; or
* * * * *
(9) * * *
(ii) Proper computation and application of the missing data
substitution procedures in subpart D of this part and the bias
adjustment factors in section 7 of appendix A to this part.
(10) The owner or operator shall provide, or cause to be provided,
adequate facilities for initial certification or recertification
testing that include:
* * * * *
(d) Initial certification and recertification and quality assurance
procedures for optional backup continuous emission monitoring systems.
(1) Redundant backups. The owner or operator of an optional
redundant backup continuous emission monitoring system shall comply
with all the requirements for initial certification and recertification
according to the procedures specified in paragraphs (a), (b), and (c)
of this section. The owner or operator shall operate the redundant
backup continuous emission monitoring system during all periods of unit
operation, except for periods of calibration, quality assurance,
maintenance, or repair. The owner or operator shall perform upon the
redundant backup continuous emission monitoring system all quality
assurance and quality control procedures specified in appendix B to
this part, except that the daily assessments in section 2.1 of appendix
B to this part are optional for days on which the redundant backup
monitoring system is not used to report emission data under this part.
For any day on which a redundant backup monitoring system is used to
report emission data, the system must meet all of the applicable daily
assessment criteria in appendix B to this part.
(2) Non-redundant backups. The owner or operator of an optional
non-redundant backup continuous emission monitoring system shall comply
with all of the following requirements for initial certification,
quality assurance, recertification, and data reporting:
(i) For a non-redundant backup gas monitoring system that has its
own separate probe, sample interface, and analyzer or for a non-
redundant backup flow monitor, all of the tests in paragraph (c) of
this section are required for initial certification of the system,
except for the 7-day calibration error test.
(ii) For a non-redundant backup gas monitoring system consisting of
one or more like-kind replacement analyzers that use the same probe and
sample interface as a primary monitoring system, no initial
certification of the non-redundant backup monitoring system is
required. Note that a non-redundant backup analyzer, connected to the
same probe and interface as a primary analyzer in order to satisfy the
dual span requirements of section

[[Page 28129]]

2.1.1.4 or 2.1.2.4 of appendix A to this part, shall be considered a
like-kind, non-redundant backup analyzer.
(iii) Each non-redundant backup monitoring system shall comply with
the daily and quarterly quality assurance and quality control
requirements in appendix B to this part for each day and quarter that
the non-redundant backup monitoring system is used to report data,
except that the requirements for when a linearity test must be
performed are superseded by the requirements of this section. The owner
or operator shall ensure that each non-redundant backup continuous
emission monitoring system passes a linearity check (for pollutant
concentration and diluent gas monitors) or a calibration error test
(for flow monitors) prior to each use for recording and reporting
emissions. For a non-redundant backup NO_X-diluent or
SO2-diluent monitoring system consisting of a primary
pollutant analyzer and a like-kind replacement diluent analyzer (or
vice-versa), provided that the primary analyzer is operating and is not
out-of-control with respect to any of its quality assurance
requirements, only the like-kind replacement analyzer must pass a
linearity check before the system is used for data reporting. When a
non-redundant backup monitoring system is brought into service prior to
conducting the linearity test, a probationary calibration error test
(as described in paragraph (b)(3)(ii) of this section), which will
begin a period of conditionally valid data, may be performed in order
to allow the use of data retrospectively, as follows. Conditionally
valid data from the CEMS are validated back to the hour of completion
of the probationary calibration error test if the following conditions
are met: if no adjustments are made to the monitor other than those
specified in section 2.1.3 of appendix B to this part between the
probationary calibration error test and the successful completion of
the linearity test, and if the linearity test is passed within 168 unit
operating hours of the probationary calibration error test. However, if
the linearity test is either failed, aborted due to a problem with the
CEMS, or not completed as required, then all of the conditionally valid
data are invalidated back to the hour of the probationary calibration
error test, and data from the CEMS remain invalid until the hour of
completion of a successful linearity test.
(iv) When data are reported from a non-redundant backup monitoring
system, the appropriate bias adjustment factor (BAF) shall be
determined as follows:
(A) Apply the BAF from the most recent RATA of the non-redundant
backup system (even if that RATA was done more than 12 months
previously); or
(B) If no RATA results are available for the non-redundant backup
system (e.g., for a non-redundant backup gas monitoring system that
uses the same probe and sample interface as the primary monitoring
system), apply the primary monitoring system BAF.
(v) A non-redundant backup system may not be used for reporting
data from a particular affected unit or common stack for more than 720
hours in any one calendar year, unless the monitoring system passes a
RATA at that same unit or stack.
(vi) For each non-redundant backup gas monitoring system that has
its own separate probe, sample interface, and analyzer and for each
non-redundant backup flow monitor, no more than eight successive
calendar quarters shall elapse following the quarter in which the last
RATA of the monitoring system was done at a particular unit or stack,
without performing a subsequent RATA. Otherwise, the monitoring system
may not be used to report data from that unit or stack until the hour
of completion of a successful RATA at that location.
* * * * *
(g) Initial certification and recertification procedures for
excepted monitoring systems under appendices D and E. The owner or
operator of a gas-fired unit, oil-fired unit, or diesel-fired unit
using the optional protocol under appendix D or E to this part shall
ensure that an excepted monitoring system under appendix D or E to this
part meets the applicable general operating requirements of Sec. 75.10,
the applicable requirements of appendices D and E to this part, and the
initial certification or recertification requirements of this
paragraph.
(1) Initial certification and recertification testing. The owner or
operator shall use the following procedures for initial certification
and recertification of an excepted monitoring system under appendix D
or E to this part.
(i) When the optional SO2 mass emissions estimation
procedure in appendix D to this part or the optional NO_X
emissions estimation protocol in appendix E to this part is used, the
owner or operator shall provide data from a flowmeter accuracy test (or
shall provide a statement of calibration if the flowmeter meets the
accuracy standard by design) for each fuel flowmeter, according to the
appropriate calibration procedures using one of the following standard
methods: ASME MFC-3M-1989 with September 1990 Errata, ``Measurement of
Fluid Flow in Pipes Using Orifice, Nozzle, and Venturi''; ASME MFC-4M-
1986 (Reaffirmed 1990) ``Measurement of Gas Flow by Turbine Meters'';
ASME MFC-5M-1985, ``Measurement of Liquid Flow in Closed Conduits Using
Transit-Time Ultrasonic Flowmeters''; ASME MFC-6M-1987 with June 1987
Errata, ``Measurement of Fluid Flow in Pipes Using Vortex Flow
Meters''; ASME MFC-7M-1987 (Reaffirmed 1992), ``Measurement of Gas Flow
by Means of Critical Flow Venturi Nozzles''; ASME MFC-9M-1988 with
December 1989 Errata, ``Measurement of Liquid Flow in Closed Conduits
by Weighing Method''; ISO 8316: 1987(E) ``Measurement of Liquid Flow in
Closed Conduits--Method by Collection of the Liquid in a Volumetric
Tank''; Section 8, Calibration from American Gas Association
Transmission Measurement Committee Report No. 7: Measurement of Gas by
Turbine Meters (1985 Edition); American Gas Association Report No. 3:
Orifice Metering of Natural Gas and Other Related Hydrocarbon Fluids
Part 1: General Equations and Uncertainty Guidelines (October 1990
Edition), Part 2: Specification and Installation Requirements (February
1991 Edition), and Part 3: Natural Gas Applications (August 1992
Edition), excluding the modified calculation procedures of Part 3; or
American Petroleum Institute (API) Section 2, ``Conventional Pipe
Provers,'' from Chapter 4 of the Manual of Petroleum Measurement
Standards, October 1988 (Reaffirmed 1993), as required by appendices D
and E to this part (all methods incorporated by reference under
Sec. 75.6).
* * * * *
(2) Initial certification and recertification testing notification.
The designated representative shall provide initial certification
testing notification and periodic retesting notification for an
excepted monitoring system under appendix E to this part as specified
in Sec. 75.61. The designated representative shall submit
recertification testing notification, as specified in Sec. 75.61, for
quality assurance related NO_X emission rate testing under
section 2.3 of appendix E to this part for an excepted monitoring
system under appendix E to this part. Initial certification testing
notification or periodic retesting notification is not required for
testing of a fuel flowmeter or for testing of an excepted monitoring
system under appendix D to this part.
* * * * *

[[Page 28130]]

(4) Initial certification or recertification application. The
designated representative shall submit an initial certification or
recertification application in accordance with Secs. 75.60 and 75.63.
(5) Provisional approval of initial certification and
recertification applications. Upon the successful completion of the
required initial certification or recertification procedures for each
excepted monitoring system under appendix D or E to this part, each
excepted monitoring system under appendix D or E to this part shall be
deemed provisionally certified for use under the Acid Rain Program
during the period for the Administrator's review. The provisions for
the initial certification or recertification application formal
approval process in paragraph (a)(4) of this section shall apply,
except that ``continuous emission or opacity monitoring system'' shall
be replaced with ``excepted monitoring system'' and except that ``shall
follow the procedures for loss of initial certification in paragraph
(a)(5)'' or ``shall follow the procedures of paragraph (b)(5)'' shall
be replaced with ``shall follow the procedures for loss of
certification in paragraph (g)(7)''. Data measured and recorded by a
provisionally certified excepted monitoring system under appendix D or
E to this part will be considered quality assured data from the date
and time of completion of the last initial certification or
recertification test, provided that the Administrator does not revoke
the provisional certification by issuing a notice of disapproval in
accordance with the provisions in paragraph (a)(4) or (b)(5) of this
section.
(6) Recertification requirements. Recertification of an excepted
monitoring system under appendix D or E to this part is required for
any modification to the system or change in operation that could
significantly affect the ability of the system to accurately account
for emissions and for which the Administrator determines that an
accuracy test of the fuel flowmeter or a retest under appendix E to
this part to re-establish the NO_X correlation curve is
required. Examples of such changes or modifications include fuel
flowmeter replacement, changes in unit configuration, or exceedance of
operating parameters.
(7) Procedures for loss of certification or recertification for
excepted monitoring systems under appendices D and E to this part. In
the event that a certification or recertification application is
disapproved for an excepted monitoring system, data from the monitoring
system are invalidated, and the applicable missing data procedures in
section 2.4 of appendix D or section 2.5 of appendix E to this part
shall be used from the date and hour of receipt of such notice back to
the hour of the provisional certification. Data from the excepted
monitoring system remain invalid until all required tests are repeated
and the excepted monitoring system is again provisionally certified.
The owner or operator shall repeat all certification or recertification
tests or other requirements, as indicated in the Administrator's notice
of disapproval, no later than 30 unit operating days after the date of
issuance of the notice of disapproval. The designated representative
shall submit a notification of the certification or recertification
retest dates if required under paragraph (g)(2) of this section and
shall submit a new certification or recertification application
according to the procedures in paragraph (g)(4) of this section.
(h) Initial certification and recertification procedures for low
mass emission units using the excepted methodologies under Sec. 75.19.
The owner or operator of a gas-fired, oil-fired, or diesel-fired unit
using the optional low mass emissions excepted methodologies under
Sec. 75.19 shall meet the applicable general operating requirements of
Sec. 75.10, the applicable requirements of Sec. 75.19, and the
applicable certification requirements of this paragraph (h).
(1) Monitoring plan. The designated representative shall submit a
monitoring plan in accordance with Secs. 75.53 and 75.62.
(2) Certification application. The designated representative shall
submit a certification application in accordance with
Sec. 75.63(a)(1)(iii).
(3) Approval of certification applications. Upon submission of the
required certification application for approval to use the low mass
emissions excepted methodology under Sec. 75.19, the excepted
methodology shall be deemed provisionally certified for use under the
Acid Rain Program during the period for the Administrator's review. The
provisions for the certification application formal approval process in
the introductory text of paragraph (a)(4) and in paragraphs (a)(4)(i),
(ii), and (iv) of this section shall apply, except that ``continuous
emission or opacity monitoring system'' shall be replaced with
``excepted methodology.''
(4) Disapproval of certification applications. If the Administrator
determines that the certification application does not demonstrate that
the unit meets the requirements of Secs. 75.19(a) and (b), the
Administrator shall issue a written notice of disapproval of the
certification application within 120 days of receipt. By issuing the
notice of disapproval, the provisional certification is invalidated by
the Administrator, and the data recorded under the excepted methodology
shall not be considered valid. The owner or operator shall follow the
procedures for loss of certification:
(i) The owner or operator shall substitute the following values, as
applicable, for each hour of unit operation during the period of
invalid data specified in paragraph (a)(4)(iii) of this section or in
Secs. 75.21(e) (introductory paragraph) and 75.21(e)(1): the maximum
potential concentration of SO2, as defined in section 2.1 of
appendix A to this part to report SO2 concentration; the
maximum potential NO_X emission rate, as defined in Sec. 72.2
of this chapter to report NO_X emissions; the maximum
potential flow rate, as defined in section 2.1 of appendix A to this
part to report volumetric flow; or the maximum CO2
concentration used to determine the maximum potential concentration of
SO2 in section 2.1.1.1 of appendix A to this part to report
CO2 concentration data until such time, date, and hour as a
continuous emission monitoring system or excepted monitoring system,
where applicable, is installed and provisionally certified;
(ii) The designated representative shall submit a notification of
certification test dates, as specified in Sec. 75.61(a)(1)(ii), and a
new certification application according to the procedures in paragraph
(a)(2) of this section; and
(iii) The owner or operator shall install and provisionally certify
continuous emission monitoring systems or excepted monitoring systems,
where applicable, no later than 180 unit operating days after the date
of issuance of the notice of disapproval.
(i) Initial certification and recertification procedures for
excepted flow monitoring systems under appendix I. The owner or
operator of a gas-fired unit, oil-fired unit, or diesel-fired unit
using the optional protocol under appendix I to this part shall ensure
that an excepted flow monitoring system under appendix I to this part
meets the applicable general operating requirements of Sec. 75.10, the
applicable requirements of appendix I to this part, and the initial
certification and recertification requirements of this paragraph.
(1) Initial certification and recertification testing. The owner or
operator shall, where applicable, use the

[[Continued on page 28131]]

Notices For
2009
2008
2007
2006
2005
2004
2003
2002
2001
2000
1999
1998
1997
1996
1995
1994

[[pp. 28081-28130]] Acid Rain Program; Continuous Emission Monitoring Rule Revisions

Local Navigation