United States
Environmental Protection Air and Radiation EPA/400/1 -91/006.C
Agency (ANR-445) April 1991 i
E PA Acid Rain Advisory
Committee Meeting:
January 28-29,1991
Monitoring
Issue Papers
-------
E-2
U.S. ENVIRONMENTAL PROTECTION AGENCY
ACID RAIN ADVISORY COMMITTEE
;- MINUTES
of -.-,
EMISSIONS MONITORING SUBCOMMITTEE MEETING
The first meeting of, the Emission Monitoring Subcommittee
was held on December 14, 1990,. at the Omni Shoreham Hotel, 2500
Calvert St., ,NW, Washington DC. The meeting convened'at-9:20
a.m..and concluded at about 12:20 p.m. It was agreed that
written materials for the next meeting.would be mailed by January
14;,the next meeting is scheduled for January 28, 1991.
PARTICIPANTS '
Subcommittee members attending included:
Mr. David Hawkins (Chairman)
Senior Attorney .
Natural Resources Defense Council
Washington, DC <
..'-,-'. . .
.,. Mr. Henry Beal . '
V vice President for Strategic Planning
.;'>., , Research-Cottre 11 Companies . . .
Branchburg, NJ . ..
Mr. Walter Canney
Administrator
Lincoln Electric System
Lincoln,. NE
Mr. Richard Riggs, for
,Mr. Jerry Golden . , .
-'Manager, Clean Air Program/ /
Tennessee Valley Authority . \ .
Chattanooga, TN "".''"
Mr. Robert McWhorter "~
Senior Vice President r -. -. '' -v
Ohio Edison ; '
Akron, OH ,
" " ' . : *
.''". Mr. ban Plumley ,
Director of Park Protection
The Adirondack Council
Elizabethtown, NY
* ^
. PRELIMINARY DRAFT FOR ARAC
DISCUSSION DOES NOT REPRESENT THE POSITION OF
U;S. ENVIRONMENTAL PROTECTION AGENCY
-------
Mr. Robert Bergstrom, Jr.
Chief Counsel
Iowa Southern
Centerville, IA
Mr. Richard Poirot
Air Quality Planner
Vermont Department of Environmental Conservation
Waterbury, VT
Ms. Nancy Wrona
Director, Office of Air Quality
Arizona Department of Environmental Quality
Phoenix, AZ
Also contributing was:
Mr. Roger Morris, for
Mr. Ted Williams
. U.S. Department of Energy
Washington, DC ;
The designated U.S. EPA representative on the Subcommittee '
was Mr. Larry Kertcher, Source Control Branch Chief, Acid Rain
Division (ARD), Office of Atmospheric and Indoor Air Programs
(OAIAP). He introduced the following staff from his branch as
key members on the EPA team responsible for promulgating the Acid
Rain Continuous Emissions Monitoring (CEM) regulation:
* *
Ms. Doris Price
Emissions Monitoring Section Chief
Mr. John Schakenbach
Senior Environmental Scientist
Ms. Margaret Sheppard
Environmental Scientist
Mr. Kertcher indicated that other staff from EPA Headquarters and
Regional Offices have been and would continue to be involved.
Other EPA- staff attending the Subcommittee meeting included: Mr.
Louis Paley of Stationary Source Compliance Division, Office of
Air Quality Planning and Standards (OAQPS); Mr. Anthony Wayne of
Technical Support Division, OAQPS; Mr. Paul Horwitz of Policy
Branch, ARD, OAIAP; and Mr. David Kee, Director, Air and
Radiation Division, Region V.
PRELIMINARY DRAFT FOR ARAC
DISCUSSION DOES NOT REPRESENT THE POSITION OF
U.S. ENVIRONMENTAL PROTECTION AGENCY
-------
Mr. Kertcher also introduced Mr. George Walsh, Director of
Research and Analysis Division, Entropy Environmentalists, Inc.,-
who is providing technical support to EPA in developing the Acid
Rain CEM regulation: Mr. Walsh was responsible for preparing the
Subcommittee meeting minutes. ' . . ,
* i .* t * ~
AGENDA AND MINUTES . . ' . ,
Prior to the formal opening of the meeting-by David Hawkins,
Subcommittee Chair, Larry Kertcher distributed.a proposed agenda.
Mr.-Hawkins-indicated that, he had no objection to the:agenda and
that a draft, issue paper prepared by EPA on "missing data
.periods1* was scheduled for discussion. .Mr. Kertcher commented
that discussion of the issue paper was not critical to1 the first
meeting, of,.the Subcommittee. Rather, the primary objectives were
to develop a list of all issues (technical, legal, economic, and ,
other) associated with the CEM regulation, and to prioritize
.these issues for consideration by the Subcommittee and the full .
Acid Rain Advisory Committee (ARAC). ~
- ' i--*/ . ' ' - - . :
A Subcommittee member asked, if minutes of the meeting would,
be made available. The reply was affirmative. A question was \
raised concerning the need, for.Subcommittee meetings between
scheduled meetings of the ARAC. It was decided that any '
"intermediate meetings" would be conference calls arranged*
through the Subcommittee Chair. '
'';'''' ..'''-' ' ' ' ' '--" .
LIST OF ISSUES - " - ' ' '-. \ ''*-,. - '
Mr'. Hawkins asked the-Subcommittee to generate a list of
major discussion items. - About ten to twelve items were listed x
rapidly. Additional items were formulated and added to the list
over the course of the meeting. The items, in the order listed
by the Subcommittee, are: /
o .Accuracy of flow determinations. *
o CEMS accuracy. < .
o Definition and criteria to evaluate "alternative"
.' systems. ....' . ._-.''- ...
o Averaging times, including frequency of measurement and
basic units for recording and reporting data.
o . 'Mature and type of emissions tracking system.
o- How to define and handle "missing".data. .
o How to handle emissions from multiple, boilers.exhausted
to a common stack.
o Audits of reported data. .
o Cost-effective improvements or changes to existing
' CEMS. . ,.; -- '-' . ' .
r I > ' '
PRELIMINARY DRAFT'FOR ARAC
DISCUSSION DOES NOT REPRESENT THE POSITION OF
U.S. ENVIRONMENTAL PROTECTION AGENCY
-------
o Format for reporting data. .
o Timely public disclosure of CEMS data.,
o QA/QC procedures for measurement systems.
o *. Uses of reported Acid Rain CEMS data in ambient sulfur
dioxide program. ...
o Availability of early guidance on hardware requirements
for Phase I units.
o Reliability of CEMS.
o Monitoring guidance for ^opt-in" facilities prior to
; . issuance of. regulations.
o v Range of performance levels and cost for CEMS.
o Information "acceptable to the Administrator."(
o Range of permissible carbon dioxide monitors.
o Should EPA define a close-ended list of substitute
. . methods and what are the associated legal aspects?
o -.- What control device parameters should be monitored to
determine if the controls are operation during CEMS
downtime?
o Are there different issues for monitoring nitrogen
oxides compared to sulfur dioxide?
-o Precision of alternative methods.
o Heed to revise CEMS requirements in 40 CFR Part 60 and|
the extent of anticipated changes.
o Definition of acceptable alternatives to flow rate
: monitors. .
After some discussion, Mr. Hawkins requested a consensus on
the highest priority issues. The Subcommittee agreed that the
seven items listed below are high priority issues:
.... y *
o Accuracy of flow determinations.
o Early guidance on hardware requirements for Phase I
units.
o Heed for and documentation of changes to existing CEMS
rules.
o Information on CEMS accuracy and availability, and
treatment of resulting data.
o Alternatives to flow rate monitors.
o CEMS and flow monitor source population data.
o Use of Acid Rain CEMS data for ambient sulfur dioxide
To ensure rapid progress on these issues, the Subcommittee
made the following assignments:
o Mr. McWhorter - Provide report from Utility Air
Regulatory Group (UARG) on precision
and availability of CEMS and another
PRELIMINARY DRAFT FOR ARAC
DISCUSSION DOES NOT REPRESENT THE POSITION OF
U.S. ENVIRONMENTAL PROTECTION AGENCY
-------
o ** Mr. Walsh
* / : *
o Mr. Beal
report oil alternatives to flow rate
'. monitors. ,-'...'
- Provide paper by John Richards on
; alternatives to flow rate; monitors.
- - Write paper on parameters that could be
' : monitored to demonstrate continued
, . performance of pollution control .system
. when CEMS is inoperative.
o Mr. Kertcher ''- Develop list of Subpart Da CEMS
elements that are candidates for
"*"' ' . ' - changes under Title IV. Also discuss
. ..: \ nitrogen oxides monitoring issues that
">'." may be different from sulfur dioxide
, . . monitoring. --. '
,. o Ms. Wrona - Compile information from state
. . . - programs on precision, availability and
cost of installed CEMS.
o Mr. Bergs trom - Write paper on whether EPA can define a
close-ended list of substitute methods;
. . '# '-'-.' - . include a discussion of legal aspects.
-'.. . ,: . . .' _' -.-. !- i _ . , ..'.. . .
DISCUSSION OP ISSUES . r. - . ^ ' .'[':... ." / ' -,.;. -''I
.-' " :. , :'- ' ''' ." '*- '"' . ' : ' ' *"'.; s'
' , During, the. meeting, a number of .topics and issues were ' \
discussed and rediscussed. This section of the minutes
summarizes .the discussions by topic, without regard for sequence.
; :- - * " '' -"''.'..',' :.:-~ - ' .-i . . :
'Accuracy of GEMS and Alternative Measurement Methods
- "' ' " '''' .'.*- ' '»*'-
.Many subcommittee members, expressed concerns on the subject
CEMS. of measurement, inaccuracy. The concerns included: < .
potential conflicts between desired, available, and;
cost-effective accuracies of existing measurement systems; the
'problem of not excluding yet-to-be-developed technologies .that
might offer greater accuracy at lower cost; and the apparent need
to understand CEMS accuracy in order to establish acceptability
criteria for alternative measurement methods. One member argued
that the marketplace is the proper place to resolve the issue.
Another argued that the seller bears the risk (or benefit)
derived from inaccurate data, since the buyer retains the \v, .
purchased allowances while the seller is responsible for the
"unknown": balance. This argument .was countered .with the comment
that contractual -relationships will probably tie .buyer and seller
in a long-term relationship. , '"-''> ' ..... ;.-/.:
Discussions of CEMS accuracy resurfaced during discussions
of procedures for estimating .emissions when the primary
measurement-system is inoperative. v. - ' .
,- ' , f. . . * * s. ^ . N ',''''".''
. ': ; 'P^tlMIHARY DRAFT TOR 'ARAC .
DISCUSSION 'DOES NOT REPRESENT THE POSITION OF
U.S. ENVIRONMENTAL PROTECTION AGENCY
-------
6
On at least two occasions, it was noted that data produced
by the measurement system are to be used without an error band,
and that the question of accuracy is only relevant as a criterion
to judge the adequacy of a measurement system.
Definition of Terns
4
Mr. Hawkins recommended that the Subcommittee define
specific terms in order to clarify the discussion. He proposed
the tern "designated method" to refer to the GEMS method;
"alternative method* to a site-specific substitute developed by
a source owner; and, "missing data method!* to procedures for
estimating emissions when the designated or alternative methods
are inoperative. Consensus not reached, however, on the
proposed definitions. .
Applicability of SybPart Da CEM Regulation '
It was suggested that existing CEM provisions in Subpart Da
40 CFR Part 60 could be used as a starting point for defining >
Acid Rain CEM requirements. Some may be appropriate as is f
whereas others may need modification to accommodate the goals of*
Title IV. One provision that appears inappropriate, according £o
Larry Kertcher, is Subpart Da's acceptance of 75 percent CEM data
capture. He indicated that the Acid Rain emissions tracking
system would need to account for 100 percent of sulfur dioxide
emissions from affected units and that clearly defined procedures
are needed to estimate emissions when measurement systems are not
operative. Several subcommittee members suggested that these
procedures should be complete and unambiguous, so as to eliminate
the need for case-by-case approval. Other members, however,
argued -that a provision for "approval by the Administrator" would
always be heeded in the regulation. .
Alternatives to-CEMS. ' '
. _; *-"'-*''. ' ... '
The possible use of cheaper, more reliable alternatives was
a general point of -repeated discussion and debate. Alternative
means of measuring flow rate was noted as a major concern when
several teller exhausts are ducted to a common stack. At issue
is the strong possibility that locations suitable for
representative measurements will not be found. As a result,
alternative ways to measure flow rate will be .needed.
PROCEDURES FOR ESTIMATING EMISSIONS . ~ . .
- ' * - - .
It was generally agreed that measurement systems would be
PRELIMINARY DRAFT FOR ARAC
DISCUSSION DOES NOT REPRESENT THE POSITION OF
U.S. ENVIRONMENTAL PROTECTION AGENCY
-------
inoperative, for planned and unplanned reasons, and that the
period of inoperat ion could vary from an hour or less to a month
or more. It was also noted that only one of several measurements
might be missing, or that the entire system could be off-line.
At issue vas the nature and intent of the calculation procedure
to be prescribed in the regulation promulgated by EPA.
Subcommittee members pretty much agreed that emissions
estimates should tend to overestimate (rather than underestimate)
actual emissions in order to provide an incentive for sources to
operate and. maintain reliable emissions measurement systems.
There was considerable discussion, however, about. the appropriate
magnitude and form of this incentive. Some persons indicated
that the calculated emissions should represent "maximum reported"
values; others argued for. values that represented operation of
the source "in an uncontrolled/* manner.".. The use .of statistical
inference based on historical data was > suggested.. It was noted
that plant operating data might 'be used to demonstrate that
emissions were not significantly changed during periods of
inoperative measurement systems. .
Consistency with Existing Monitoring Systems *
Several Subcommittee members stated that the Title IV
regulations should be consistent with established CEM
requirements and installed monitoring systems. It was suggested
that EPA provide early guidance on hardware requirements for
Phase I units .in order to provide lead time for equipment
acquisition. The potential "tension1* between early guidance and
the results of public comments was noted.
The Subcommittee Chair commented that monitoring
requirements for sources that elect "opt-in" might also be needed
sooner than anticipated. :
PRELIMINARY DRAFT FOR ARAC
DISCUSSION DOES NOT REPRESENT THE POSITION OF
U.S. ENVIRONMENTAL PROTECTION AGENCY
-------
E3
Issue Paper - Missing' Data Periods
Continuous Emissions Monitoring
DRAFT
ISSUE -' .- . ' . '"'"". ' : -
... How should uncontrolled' emissions be calculated when primary
or substitute emissions and/or, flow data cannot be provided by an
affected unit? -.;* / V ;: -.
DISCUSSION "_.' , '"""' .
The explicit and precise language in.Section 412 of Title IV
restricts the options for'calculating emissions during periods when
data from a continuous 'emissions monitoring .system (CEMS) or an
approved alternative.monitoring system is unavailable. The Title
requires EPA to ''deem the unit,to be operating in an uncontrolled
manner during the entire period for which the data was not
available," unless the owner or operator can provide satisfactory
information on emissions during that period. Further, EPA must
prescribe a standard, method to calculate emissions for missing data.
periods in a continuous emissions monitoring (CEM) regulation to be
promulgated no later ..than .May 15, 1992. t' '
"* ' ' t ' *
There are two/inextricable subissues associated .with the issue
.of missing data periods':.. ' " , ; '.' ' - ''..''.'*/.'
. . (1) How should we define "uncontrolled emissions" for various
types of affected units under different operating
circumstances?- . _ . . , .;"'_' .'
: (2) What constitutes satisfactory information on emissions
'.:';.. during, periods when primary CEM (or approved alternative
monitoring) data is unavailable? ;
The primary; intent of this initial issue paper is to set forth
options for the first subissue which, when resolved, would lead to
a ' standard method to calculate emissions during missing data
periods. An adequate treatment of this subissue, however, would
appear :to require some (perhaps considerable) discussion of
"alternative monitoring methods. ' ' ':.'' ' ' .
One . important factor for defining "uncontrolled emissions" is
whether the unit normally operates1 with add-on pollution control
equipment for sulfur dioxide (SO,) and/ or . nitrogen oxides i (NO^)
emissions. (Units with add-on -pollution control equipment can emit
either "controlled" or '? uncontrolled" pollutant concentrations in
their flue gases, depending on the operational status of their
control device (s).) Another factor . ; is the required units of
measure for emissions estimates: S02 emissions must be calculated
'.-.- .PRELIMINARY DRAFT FOR
DISCUSSION DOES NOT REPRESENT THE POSITION OF
- : U.S. ENVIRONMENTAL PROTECTION AGENCY
-------
as mass per unit time (i.e., tons/year) for comparison with the
allowances the unit holds; whereas NO emissions should be
expressed in Ibs/MMBtu for comparison with the maximum allowable
emission rates mandated in Section 407.
Accordingly, separate options may need to be developed for the
definition of "uncontrolled emissions" (as well as acceptable
substitute emissions data) for the following combinations of unit
emissions and pollution control equipment;:
e SOZ emissions *
Units with flue gas desulfurization (FGD) and/or
other control devices (SO2)
,.,-. - Units without control devices (SO2)
* * ,
NOX emissions .
Units with selective catalytic reduction (SCR),
, urea injection, and/or other add-on (post-
combustion) control devices (NOX)
Units without add-on control devices (NOX).
*
Units employing low NOX burner'control technologies would normally
be included in the last category since - this . form of pollution
control is integral to the boiler and cannot be bypassed.
OPTIONS FOR SO. EMISSIONS
Our goal is to develop rules for inferring a unit's potential
SO, emissions during periods when primary CEM data is unavailable
that will: .
(1) Minimize the amount of missing and/or substitute data in
the unit's cumulative annual total of SO2 emissions for
comparison with the allowances it holds;
(2) Account for all S02 emissions throughout the year; and
(3) Provide conservative methods for estimating SO2 emissions
' during missing data periods to assure the integrity of
the allowance trading system and to provide incentives
for proper operation and maintenance of CEMSs.
Units with Control Devices. As a starting point, estimates will be
needed for the unit's Maximum Uncontrolled Emissions Rate (MUER)
for SO2. Calculating a realistic, yet conservative, MUER is not
PRELIMINARY DRAFT FOR ARAC
DISCUSSION DOES NOT REPRESENT THE POSITION OF
U.S. ENVIRONMENTAL PROTECTION AGENCY
-------
necessarily a straightforward exercise, particularly if one tries
.to use source-specific test data instead of average emissions
factors. What constitutes acceptable, test data? What capacity
'factor should be used when calculating the uncontrolled emissions
rates? . . '.."'' ' ; '' ' : '
One .method for calculating a unit's MUER during an outage is to
assume 100% capacity and use a combination of the specific.unit's
S02 coal quality valuation from the mine in conjunction with an
appropriate SO2 content variability factor to account for the
.maximum emission rate as shown below:
. MUERS02 (tons/outage hrsj - (Coal quality measurement in Ibs
. . SO2/ton coal burned) x (coal
, . variability factor) x (tons coal
'.,,'.'. burned/outage hrs) /(2000 Ibs/ton)
' * ' '' ' .''..' ' . '*..'.
(Outage hours could be converted into 'fraction of a year for
comparison with allowances.) Using some ^combination of good
source-specific , inlet test data or bypass monitoring values for
;maximum uncontrolled S02 emissions/ when 'available, may also be
possible. '. '"..'-" '.- *'. . !'.v. '_ ', '. '. ":.v'f- * :
If the source can, demonstrate':that the unit's ^"pollution
control device was functioning properly, it might be reasonable to
use the unit's Maximum Recorded Emissions Rate (MRERJ for SO2
emissions during the missing data' period.' -Since CEM equipment
.should be in operation for at least a "year before the SO2 emissions
allowances begin, there appears to be ample time for establishing
a reasonable . baseline for determining MRERs. A problem would
arise, however, if the unit's normal operating capacity changes
significantly or if modifications .are.made to the unit's pollution
control equipment between the baseline period and the actual
emissions reporting period.
* "* ' " '," s-
MRER could be defined as the maximum hourly .emissions rate
recorded by the CEMS during either.the preceding calendar year or
the preceding monitored quarter. The idea is to define a long
enough time period to encompass the highest emission rates likely
to be emitted .from the unit in question.
Another' potential option would be to allow the use of
substitute data, for SO2 emissions from backup systems, probably on
a limited basis, when valid primary CEM '. (or approved alternative)
monitoring' system data is unavailable. Backup systems could
include: ' duplicate CEMS; statistically valid, as-fired or as-
bunkered fuel sampling and analyses (sulfur and heat content) for
Ibs/MMBtu determination (no sulfur retention credit should be
allowed); FGD system parameter monitoring for percent SO2 removal
efficiency;, 'and* boiler steam flow output monitoring for indirect
PRELIMINARY DRAFT FOR ARAC
DISCUSSION DOES NOT REPRESENT THE POSITION OF
U.S. ENVIRONMENTAL PROTECTION AGENCY
-------
calculation of hourly heat input. ' *
Should backup emissions monitoring methods be subject to as
stringent quality control (QC) and quality assurance (QA)
requirements as primary monitoring systems? If not, limiting the
number of hours of backup emissions.data seems appropriate. We
suggest that an affected unit only be allowed to provide up to 500
hours of backup monitoring system data per year, after that, the
unit's uncontrolled emissions rate or MRER must be used. A 500-
hour period would accommodate normal periods of daily calibration,
quarterly audits, and limited primary monitoring system
malfunctions. Obviously, such limitations would not apply where
the backup is a duplicate GEMS.
Units Without Control Devices. * The issues of defining
"uncontrolled emissions" and acceptable substitute data for units
without SO2 control equipment are similar, though not identical, to
those equipped with add-on controls. MRER values could be used for
"uncontrolled emissions" for these units since, presumably, they
would be equivalent to HUERs at the actual units' operating rates.
On the other hand, mandating a MUER value based on 100% capacity
would provide greater market incentive for units to operate and
maintain their CEMSs effectively.
i .
Backup systems for units without F6D or other SO? emissions
control equipment would include the alternate monitoring methods
listed in the previous subsection. The concerns articulated in
that discussion about limiting use of substitute data and/or
providing appropriate QA/QC for backup systems would also apply to
these units. . ' . . . . ,
OPTIONS FOR. NQX EMISSIONS.
The rules for inferring a unit's potential NOX emissions
during missing data periods have essentially the same basic goal as
the rules for SO2 emissionsthat is, to: '
(1) Minimize the amount of missing and/or substitute data in
the record of a unit's annual NOX emissions;
(2) Account for all NOX emissions throughout the year; and
(3) Provide conservative methods for estimating NOX emissions
during missing data periods to assure validity of annual
emissions averaging and to provide incentives for proper
operation and maintenance of CEMSs.
Clearly, the application of these rules may differ (perhaps be less
rigid) from those applying to SO% emissions because there is no
explicit requirement for a cumulative annual total of NOX emissions
PRELIMINARY DRAFT FOR ARAC
DISCUSSION DOES NOT REPRESENT THE POSITION OF
U.S. ENVIRONMENTAL PROTECTION AGENCY
-------
from each affected unit. Since the statutory limitation, is
expressed .as an annual average, it appears that some consideration
may be given to the impact of discrete (small) data "pieces" on the
overall '.annual average. How the rule would apply to units
participating in the annual emissions averaging pools allowed under
Section 401(e) is also of concern. ' . .
*> *
Units With! 'Control Devices; It is anticipated that very few Phase
I -units will fall into this category since, as mentioned
previously, most low NO burner technologies' are integral to the
combustion process in. the boiler and; thus, cannot be bypassed.
Like the SO2 emissions, estimates' will be needed for the unit's
MUER for NOx:'. The capacity factor is also an issue for ,NOX
emissions 'as is . the use of average uncontrolled emissions
estimates.
One method for calculating a unit's MUER is to assume 100%
capacity and use the AP-42 maximum emissions factor for
uncontrolled NOX emissions for the appropriate .coal and boiler
category as shown below:
(Ibs/MMBtu) = (AP-42 NO factor in Ibs NOx/ton
burned)/(MMbtu/ton coal burned)
coal
The NOX MUER is expressed in Ibs/MMBtu for comparison with the
maximum allowable emission rates mandated in Section 407.
Also, if the source can demonstrate that the unit's SCR or
other add-on pollution control device was functioning properly, it
might be reasonable to substitute the unit's MRER emissions during
the missing data period. Using the MRER may not provide sufficient
incentive, however, for units to operate and maintain their CEMSs
properly.
No alternative monitoring systems have been identified for NOX
emissions so backup systems would be limited to duplicate CEMSs.
It may be argued that a small amount of missing data, say, 5%
or less could be tolerated for NOX emissions from, units not
averaging their emissions with others. A cumulative annual total
for NOX emissions is not required for these units. On- the other
hand, some estimating rules will be needed to account for all NO
emissions and to gauge progress towards achievement Of the mandated
2-2.5 million ton annual NOX reduction.
Units Without Control Devices. The potential methods discussed
previously, for calculating SO2 emissions during missing data
periods for these units should also apply to NOX emissions. Valid
substitute data would be restricted, however, to data from a
duplicate CEMS since no acceptable'alternative monitoring systems
PRELIMINARY DRAFT FOR ARAC
DISCUSSION DOES NOT'REPRESENT THE POSITION OF
U.S. ENVIRONMENTAL PROTECTION AGENCY
-------
have been devised for NOX emissions. The rationale given above for
allowing a limited, amount of missing data for NOX emissions from
units not averaging their annual emissions would also apply.
We hope that this initial issue paper on missing data periods
provides grist for a productive discussion with the Continuous
Emissions Monitoring Subcommittee of the Acid Rain Advisory
Committee (ARAC). A later issue paper will elaborate upon the
related issue of alternative monitoring methods. We look forward
to working with members of the Subcommittee in developing an
effective CEM regulation consistent with the statutory intent of
Section 412 of Title IV and the Acid Rain Program Vision Statement.
* I
PRELIMINARY DRAFT FOR ARAC
DISCUSSION DOES NOT REPRESENT THE POSITION OF
U.S. ENVIRONMENTAL PROTECTION AGENCY
-------
DRAFT
E4
Applicability of Continuous Emissions Monitoring (OEM)
Provisions in K8PS Subpart Da Regulation to.
... Acid Rain CEM Regulation
- *A , .
BACKGROUND AND OBJECTIVES
. The New Source Performance Standards (NSPS) Subpart Da
regulation, promulgated by EPA in 1979, launched a new era for
continuous emissions monitoring (CEM) by specifying that CEM data
would be used to .assess affected sources' compliance with
applicable S02 -and NOX emissions standards. (Prior to this rule,
EPA used CEM only as a relative indicator of the proper operation
and maintenance of pollution -control equipment.) Subpart Da
represents a natural starting point for the development of the Acid
Rain CEM regulation since it has become the standard reference for
all subsequent federal and state CEM regulations.
In most cases, industry now looks to the CEM provisions in
NSPS Subpart Da for quality assuring their CEM devices, whether
monitoring new or existing sources, and uses them to. shape their
internal quality assurance (QA) programs. Thus; these standards..
and .procedures (contained in Appendices B and.F.of 40 CFR Part 60)
provide a logical reference for . defining the performance
specifications, certifying tests, QA and audit procedures in the
Acid'Rain CEM regulation. One of our goals, in developing this
regulation is to minimize the need for replacing existing CEM
systems (CEMSs). Utilizing NSPS Subpart Da and-,, its related
procedures will accomplish this. . .
**" -
While adopting the CEM provisions in NSPS Subpart Da as our
initial framework is a major step forward, some changes need .to be
considered. .The goals of the acid rain control program are not
entirely consistent with those of the new sources program. The
Acid' Rain CEM regulation will need to account .for all SO2 mass
emissions.(in tons/year) rather.than .relating, SO2 concentrations
back to fuel usage (in Ibs/MMBtu). Another objective is to
identify operation and maintenance procedures likely to improve the
performance and reliability of today's CEMSs to assure consistent
quality of the measurements used as the basis for trading in the
emissions market.
We have conducted a close review of Subpart Da-and EPA's
knowledge-of CEM operating, history. From this effort, we. have
identified areas in Subpart Da that may not serve the objectives of
the acid rain control program. These areas of difference generally
stem from the different program and statutory requirements of
Section ill of the Clean Air Act (for .New Source Performance
PRELIMINARY DRAFT FOR ARAC
DISCUSSION DOES NOT REPRESENT THE, POSITION .OF
U.S. ENVIRONMENTAL PROTECTION"AGENCY
-------
Standards) and Title IV, which establishes the Acid Rain (AR)
Program; .
SUMMARY OF FINDINGS ....
t' ' - *'
Majpr areas of difference and similarity between Subpart Da
and the AR Program are summarized below. We also present
suggestions for modifications of specific provisions in Subpart Da
in .-'light of .the (different) goals of the AR Program. ;
*' *, * '
. / - ', Subpart Da applies only to hew electric utility units
. with a specific (large) capacity constructed on or after
September 19, 1978. No exemptions are allowed except for
/a governor's emergency. The'hew acid rain rule would
; apply to both new and existing utility'units, including
:. - ' some smaller units (i.e., down to 25 MW capacity), as
well as nonutility units that opt into the program. No
-exemptions are anticipated for the acid rain rule.
1 4 ' "***.
, .;' Under Subpart Da, certification that CEMSs meet the
performance specifications, is obtained on non-
standardized forms from several sources (e.g., States,
local agencies, EPA Regions). Under the acid rain
regulation, this certification needs to;be obtained on a
\ * standardized form from a single, central source (i.e.,
. Federal EPA).. ' , .
1 ' * - * ' - "f '
. , Pollutant emissions limits in Subpart .Da are expressed as
. mass per unitfof heat input (i.e., Ibs/MMBtu). For.the
' ,acid> rain irul'e, S02 emissions must be measured and
expressed as mass per unit of time (i.e., tons/year) for
. .' comparison with allowances a unit holds. . NOX limits
t. .,, under tlie acid rain rule would be expressed in the same
..mass per unit of heat input, but would be an annual
average rather than, a 30-day rolling-average..
'' " -. '"-'.- ' ; '**'' ''''.'-,-' . '."'.. ,-.
^ *^fc.~ _-. »*-'H. -,
-. Subpart. .Da requires theI. use"* of CEMSs for SO2V NOX,
. .' opacity, and O2 or CO2. The CEMSs for the acid rain rule
' , ', .will need to add flow monitors for., computing SO2
,- emissions in mass per unit time. Monitoring of CO2
emissions also will be required under the provisions of
'" ' Title VIII.' .Under both Subpart Da and acid rain
regulations, CEM data is used to determine compliance.
.'*'**' - '"-
.-.*- ' » - ' .*' <
, -\, . Unlike Subpart Da, no compliance. determination for
".'",' . opacity, will be made under the 'acid rain rule; however,
like Subpart :Da, opacity will be monitored.
Subpart Da allows'for a specified amount (i.e., up to
t.",,i~' . '-"; '* ' ** ,
* \ ' ' .« * '
. , :.- PRELIMINARY DRAFT FOR ARAC .
DISCUSSION DOES NOT REPRESENT THE POSITION OF
' U.S. ENVIRONMENTALPROTECTION AGENCY
-------
25%) of missing data. The acid rain rule will probably
require 100% accounting of SO2 emissions.
The acid rain regulation may need to require 4 equally
spaced data points for quality assurance of the hourly
average emissions; Subpart Da requires a minimum of 2
data points.
* *
Subpart Da allows the use of commercial grade test gases
for quarterly CEM accuracy checks. To assure less bias
in emissions measurements (and thus, allowances
consumed), the acid rain rule could require the use of
certified reference materials (CRN), standard reference
materials (SRM), or Protocol 1 gases. A new SRM for
higher concentration levels may need to be developed to
accommodate the potentially wide range of pollutant
concentrations (including uncontrolled emissions). Also/
since it would be impractical to adjust for bias
retroactively, particularly over long periods, certified
gases may be recommended for daily tests as well as for
periodic audits.
Subpart Da allows the use of either manual wet-chemistry
(e.g., RM6 and RM7) or instrumental (e.g., RM6C and BM7E)
methods for quarterly audits of CEM relative accuracy.
To ensure greater consistency in CEM precision (which
will help foster certainty in the allowance trading
market), the acid rain rule might consider limiting the
options to instrumental methods. Also, instrumental
reference methods could be more cost-effective.
The acid rain regulation might also consider tightening
the relative accuracy specification for CEMS from 20% to,
say, 10%. (This revision . would also help foster
certainty and robustness in the allowance trading
market.) Historical data suggest that most of today's
CEMSs routinely perform better than the Subpart Da
specification. Also, the Subpart Da alternate
specification of using 10% of the applicable emissions
standard is not appropriate for the acid rain rule
because SO2 Ibs/MMBtu emission rates can vary under Title
IV. . .'.;.
Unlike Subpart Da, Title IV CEMSs will need to be
accurate over the entire possible range of pollutant
concentrations, not just at low concentrations.
Therefore, a four-point linearity check is being
considered. This check would ensure monitor accuracy
over the entire range of the instrument.
PRELIMINARY DRAFT FOR ARAC
DISCUSSION DOES NOT REPRESENT THE POSITION OF
U.S. ENVIRONMENTAL PROTECTION AGENCY
-------
DETAILED ANALYSIS
The following tables present a detailed comparison of the
potential components of an Acid .Rain GEM regulation with NSPS
Subpart Da and 40 CFR Part 60, .Appendices B and F (addressing
initial CEMS performance requirements and ongoing quality assurance
requirements, respectively). ,
PRELIMINARY DRAFT FOR ARAC
DISCUSSION DOES NOT REPRESENT THE POSITION OF
U.S. ENVIRONMENTAL PROTECTION AGENCY
-------
DRAFT
JANf7f
-------
PURPOSE -
To Review Provisions of Subpart Da-' '
And Performance Specifications
To Identify Areas That May Require Revisions
in Development of Proposed Acid Rain Regulations
-------
Citation
(a)
ro
(a)(2)
(b)
(c)
(d)
PRELIMINARY DRAFT FOR ARAC .
DISCUSSION DOES NOT REPRESENT THE POSITION OF THE
U.S. ENVIRONMENTAL PROTECTION AGENCY
40 CFR 60.40a Applicability and Designation of Affected Facility
Subpart Da Acid Rain Regulation
Electric utility steam generating units
i . .
. Capable of combusting more than
73 Megawatts (MM) heat input
(either, alone or in combination
. with other fuel),. '
Constructed or modified after
September 18, 1978.
Electric utility combined cycle gas
turbines capable of combusting over
73 MM that input of fossil fuel in
steam generator (includes only
.emissions from combustion of fuels
in steam generating.unit; emissions
from gas turbines not included).
Change to an existing unit to'
accomodate the use of combustion
materials other than fossil fuels
does not bring unit under rule.
-* ?
f. .
Change to an existing unit*,
originally designed to fire
gaseous or liquid fossil fuels,
to accomodate the use of any other
fuel does not bring unit under rule.
Applicability will need to be broader than Subpart
Da; would need, to include:
Existing coal-fired .utility units over 100 MM
and 2.5 Ibs/mmBtu subject to Phase I of acid
rain control program. -
Existing utility units over 25 MM and 1.2
Ib/mmBtu subject to Phase II requirements.
New units upon commencement of operations.
Utility or nonutility units that opt into
Phase I or Phase II program.
-------
, 40 CFR 60.41a Definitions
Citation Subpart Da Acid Rain Regulation
New definitions in Title IV such as:
"Affected unit" ,
"Existing unit"
"New unit"
New definition for continuous emission
monitoring system with flow monitor.
(Consider using or modifying definition
in Performance Specification 6).
-------
40 CFR 60.42a Standards for Particulate Natter
Citation Subpart Da
(a) ' * Participates not in excess of:
(a)(l) 13 ng/J from the combustion of
solid, liquid, or gaseous fuel,
(a)(2) 1 percent of potential combustion
/ concentration (99 percent reduction)
. ' when combusting solid fuel, and
(a)(3) 30 percent of potential combustion
concentration (70 percent reduction)
when combusting liquid fuel.
(b) Opacity not to exceed 20 percent
(6-minute average) except for one
6-minute period per hour of not more .
than 27 percent opacity.
Acid Rain Regulation
Regulation will not contain particulate matter
standards. Rule needs only to require opacity
monitoring, recordkeeping, and reporting, but is
likely to require compliance with existing opacity
provisions in NSPS.or SIPs.
If no Federal rule or' SIP applies, consider
requiring 20 percent opacity as threshold for
reporting data.
Consider use of 6-minute average as a default
requirement.
-------
Citation Subpart Da
(a) (2)
Ul
(c)
(d)
40 CFR 60.43a SOa Limits
Acid Rain Regulation
S02 emissions from solid-derived fuel
not to exceed 520 ng/J heat input and
10 percent of potential combustion
concentration or 30 percent of.potential
combustion concentration (90 percent
reduction) when emissions are less
than 260 ng/J heat input.
'*. - ' '
S02 emissions from liquid or gaseous
fuels not to exceed 340 ng/J heat input
and 10 percent of potential combustion
cpncentration (90 percent reduction)
or iOO percent of potential combustion
concentration (zero percent reduction)
when emissions are less than .86 ng/J.
S02 emissions from solid solvent-refined
coal not ,to exceed 520 ng/J heat input and
.15 percent of potential combustion, concentration
(85 percent reduction); compliance with
emission limitation based on 30 day average
and compliance with percent reduction standard
is on a 24-hour basis.
S02 emissions not to exceed 520 ng/J heat
input from an affected facility that:
' *
Combusts 100 percent anthracite.
Is a resource recovery facility.
Is located in a noncontinental area
and combusts solid fuel,or solid-
derived fuel. "
Numerical limits for S02 emission rates not
applicable; emission limitations covered in permits
rulemaking. Sources must comply with S02
allowances (tons/yr) on a calendar year basis.
-------
Citation Subpart Da
(e)
(f)
(9)
(h)
40 CFR 60.43a S02 Limits (Con't)
Acid Rain Regulation
S02 emissions hot to exceed 340 ng/J
heat input from an affected facility
located in a noncontinental area and
combusting liquid or gaseous fuels
(except solid-derived fuels).
Less stringent standards apply to
facility under S02 commercial
demonstration permit;
Compliance with emission limitation
and percent reduction requirements based
on 30-day rolling average.
Gives formulas for determining standard
when different fuels are combusted
simultaneously.
Not applicable. Permits covered under
another rulemaking.
30-day rolling average not applicable.
Compliance will be based on continuous
emission monitoring system.
Not applicable for acid rain rule.
-------
Citation Subpart Da -
.-I,' , " . ' ' ' -
(a) NOX limits based on a 30-day
rolling average..
(a)(l) NOX emissions from gaseous fuels
not to exceed 210 ng/J for coal-derived
fuels and 86 ng/J.for all others.
(a)(l) NOX emissions from liquid fuel not
to exceed 210 ng/J for coal derived
fuels and shale oil, and 130-ng/J for
all others. , ..; ^
>v' ~ * *. ' -1 *"v *
(a)(1) NOX emissions from solid fuels'not to;.
exceed 210 ng/J for coal-derived fuels
and different provisions for use of
fuel containing over 25 % lignite.*
(a)(l) Fuels containing over 25 % by weight
of coal refuse exempt from'1 limits.
(b) NOX limits do not apply, to an affected
facility combusting coalrderlyed liquid
fuel operating under commercial
. r: demonstration' permit. :'
(c) Gives formulas for determining-standard
by proratipn when two or-.more fuels are
combusted simultaneously."
40 CFR 60.44a NQX Limits .; .
Acid Rain Regulation
30-day rolling average not applicable. Emission
limitations covered in,NOX and permits rulemaking.
Sources must comply with annual average NOX
Ib/mmBtu emission rates: , * .
0.45 Ib/mmBtu for tangentially fired boilers.
0.50 Ib/mmBtu for dry bottom wall-fired
boilers (other than units applying cell
burner technology).
Not applicable. Permits covered under another
rulemaking.
Not applicable for acid rain rule.
-------
40 CFR 60.45a Commercial Demonstration Permit
Citation Subpart Da
(a) Owner or operatory may apply for
commercial demonstration permit
to demonstrate emerging "'
technology. "''. .
(b)-(e) Different S02 and NOX emission
limits for facilities using one
of four listed technologies and
that have a commercial demonstration
permit.
00
Acid Rain Regulation
Not applicable. Permits covered under
permits rulemaking.'
Not applicable. Allowances and emission
limits to be covered in allowance and permits
rulemakings, respectively.
Clean coal technology projects and repowering
projects will be covered in an Interpretative
rule by OAQPS, and in the acid rain permits
rulemaking.
-------
Citation
(a)-(b)
(c)
(d)
(e)
Subpart Da
40 CFR 60.46a Compliance Provisions
; f Acid Rain Regulation
(f)
Compliance with the emission limits -
for participates and NOX constitutes
compliance with the percent reduction
requirements.
Standards, for participates and NOX apply
all times except during startup,/shutdown,'
or malfunction (startup, shutdown, or
emergency conditions for S02);'
Malfunctioning flue gas desulfurization
system may be operated if S02 emissions
are minimimi zed by following certain ..-.
procedures;
Compliance with 502 emission, limits/
percentage reduction requirements and
NOx emission limits based on arithmetic
average for 30 successive days of boiler
operation (data for 30 day average
generated by continuous emission
monitoring system). '."...
-".*' . , t _ 1
For initial performance.test; compliance
with S02 emission limits/percent;reduction
requirements and NOX emission.limits .based
on average emission rates for 30 days;
new average calculated for-each 30-day
period thereafter. . .
Not applicable. No percent reduction requirement
in acid rain program.
Acid rain emissions monitoring regulation may or
.may not have similar exemptions for NOX.
Covered in permits'rulemaking.
Not applicable. Compliance for acid rain
.emissions.monitoring regulation needs to .
be based on use of the continuous emission
monitoring system (or acceptable alternatives)
in tons/yr for the S02 limit and annual
average Ib/mmBtu for NOX.
Not applicable. Same as above.
-------
40 CFR 60.46a Compliance-Provisions (Con't)
Citation Subpart Da Acid Rain Regulation
(g) Data from startup, shutdown, malfunction Needs revision to account for all emissions,
(NOX only), or emergency conditions (S02
, only) excluded. ,
(h) If minimum quantity of emission data not Needs to account for use of MUER, and
; obtained, missing data may be determined substitute data acceptable to the
according to procedures in Method 19 Administrator.
(i.e. fuel sampling).
-------
Citation
(a)
(b)
Subpart Da
40 CFR 60.47a Emission Monitoring
' . ... '. Acid Rain Regulation
Install, calibrate, operate, and
maintain a continuous emission
monitoring system and. record the
output of the system for opacity
(except where gaseous fuel is the
only fuel combusted). Includes
provisions .for monitoring where
interferences exist.
Install, calibrate, operate, and
maintain a continuous emission
monitoring system and record the
output of the system for S02
(except where natural gas is the
only, fuel combusted);.
Monitor at inlet and outlet
. of control device.
If burning solid fuel or
solid-derived fuel, monitor
as discharged to the
atmosphere.
As-fired system may be used
to determine potential S02
emissions at inlet to S02
control device rather than
continuous emission monitoring
system.
Every affected unit needs a continuous
emission monitoring system,.(or acceptable
alternative). .
Revisions are needed to account for flow
monitor and data recorder, and for multiple
units utilizing a single stack.
C02 monitoring also would be required under
Title VIII.
Same as above.
-------
Citation Subpart Da
(c)
(d)
(e)
.(f)
(9)
40 CFR 60.47a Emission Monitoring (Con't)
Acid Rain Regulation
Install, calibrate, operate, and
maintain a continuous emission
monitoring system and record the
output for NOX discharged to the
atmosphere.
Install, calibrate, operate, and
maintain a continuous emission
monitoring system and record the
output.of the system for measuring
the 02 or C02 content of the
flue gases at each location
where S02 or NOX is monitored.
Systems and data recorder to be
operated during all periods of
operation including startup,
shutdowns, malfunctions, and
emergency conditions, except
during breakdowns, repairs,
calibration checks, and zero/
span adjustments.
Minimum-data requirement of 18 hours
in at least 22 of 30 succesive boiler
operating days; owner or operator can
supplement data with other approved
monitoring systems or with manual
tests/procedures cited in rule.
One-hour averages used to calculate
average emission rates. At least
two data points needed.
Same as above.
Same as above.
Needs revision to account for all emissions
throughout year, including those during calibration
periods and zero/span adjustments.
Needs revision to account for all emissions.
Alternative methods must be as.precise,
reliable, accessible, and timely as the
continuous emission monitoring system.
Four data points may be required for average
hourly emissions.
-------
Citation
(h) '
(0(3)
(0(4)
Subpart Da
40 CFR 60.47a Emission Monitoring (Con.'t)
; .Acid Rain Regulation
Specific manual or instrumental, methods
(e.g., Methods 6, 7, 3B, and 19) must
be used when supplementing data to meet
minimum data; requirements.
*
Manual or instrumental reference methods
for 02, S02, and NOx) roust be used for'
performance evaluations and calibration
checks. '-I;? '' ' ' ; ';.'
SQ2 or NOX must be used for,preparing
calibration gas mixtures.
For facilities burning only-fossil
fuel, span,value for continuous
opacity .monitoring system is 60-80
percent; span value for system measuring
NOx varies depending on type of fuel.
Span values for burning combinations
of fossil fuels are rounded'to the
nearest 500 ppni.
For facilities burning fossil- fuel
alone or in combination with.ribnfossil
fuel, span value for system measuring
S02 at the inlet to the control device
is 125 percent of the maximum estimated
hourly potential emissions of the fuel
fired and the outlet of the device is
50 percent of maximum estimated hourly ,
potential emissions of the fuel.fired.
May be restricted to instrumental methods
for greater consistency.
Same as above.
May need revision to use standard reference
material or certified reference material
gases. Also, acid rain'regulation may require
at least a minimum quality of calibration gas.
Span values need revision to better
account for all potential emissions;
wider range of concentration and
flows, expected over the .one-year acid
rain,emission measurements time period. -
Same as above. , .
'Same as above.
-------
40 CFR 60.47a Emission Monitoring (Con't)
, i ""
Citation Subpart Da '-.' Acid Rain Regulation
. Phase I units must install, operate, and
: quality assure data (and perhaps have EPA
.'.--. ' receive certification reports) by November 15,
. 1993. Phase II units must install, operate, and
quality assure data (and perhaps have EPA receive
certification reports) by January 1, 1995.
Certifying agents include States, local . EPA would be the sole certifying agent.
agencies, and/or EPA Regions. .
No requirement for EPA to approve EPA would turn around cetification report
certification report within a certain within a definite time period.
time period.
No standardized certification form. Acid rain regulation would have standardized
- . certification form. . - '
Acid rain regulation would specify minimum
; . - contents of certification reports, perhaps
with Regional review prior to submission.
-------
Citation
(a)
* -
(b)
un
*"» "
(c)
(d)
(e)
(f)
40 CFR 60.48a Compliance Determination Test Methods and Procedures
'"'.' ' r ' ..w' ' . . - ";'."''.'
Subpar't Da ' . - " Acid Rain Regulation
/ . . ,l
.Must use methods and procedures -
in Appendix A or specified alternative
methods in rule; '
Gives methods, and procedures for
determining compliance with particulate
standards by calculation of emission rate
and for opacity. Method 9 (visible
emissions observations), must.be used
to determine compliance with opacity
standards.
Gives methods and procedures for
determining compliance with percent
reduction limit for S02I concentrations
of S02 and C02 or 02 from continuous
monitoring system shall be used.
K
Gives'methods and procedures for
determining compliance with percent
reduction limit for NOX; concentrations
of NOX and C02 or 02 from continous
monitoring system shall be used.
Gives alternative methods and procedures
for facilities with or without wet FGD
systems. . ; t
> '"*!. . *',
Gives provisions for testing data obtained
from performance testing of gas turbines
under other NSPS rules for electric
utility combined cycle gas turbines..
Acid rain regulation also would require
compliance with specifications (whether
specifications are contained in.a separate
Appendix or in the regulation itself).
Not applicable. 'Continuous .emission monitoring
system (or; acceptable alternative) isythe .
compliance method .for acid rain. No compliance
determination for opacity will be made; under
acid rain rule. .
Needs revision to account .for compliance with S02
allowances (tons/yr) on a calendar year basis,.
-\
Needs revision to account for compliance with
annual average NOX Jb/mmBtu emission rates.
Alternative methods must be as precise, reliable,
accessible, and timely as continuous emission
monitoring system. ~ '". ~ . , . .
Not applicable. All affected units must install,
operate, and quality assure (and perhaps have EPA
receive certification reports) for continuous
emission monitoring systems.
-------
Citation Subpart Da
40 CFR 60.49a Reporting Requirements
Acid Rain Regulation
(a)
(b)
(c).
(d)
(e)
(f)
(9)
Results from initial performance
test required under General Provisions
and from performance evaluation of
continuous monitoring systems.
Daily records of average S02 and
NOx emission rates for each 30 boiler
operating days and associated infor-
mation on missing data, type of fuel,
whether compliance achieved, etc.
Information pertaining to minimum
data requirements for each 30-day
period.
Information on exceeding standard
during emergency conditions because
of control system malfunctions.
Information on the use of fuel
pretreatment credit toward S02
emission limit.
Signed statement by owner or operator
indicating whether any changes to the
control system were made during any
period of data unavailability.
Signed statement by owner or operator
indicating whether all calibration, span,
and drift checks and other periodic
audits have been made as required.
General Provisions initial performance test
not applicable. Rule would require reporting
of results of certification tests for monitoring
system (to ensure systems meet- performance
specifications).
Needs revision on type of information and format of
data to be reported for acid rain S02 allowances
and NOx emission limitations. ' ,
Acid rain regulation would need comparable
requirement for use of HRER and MUER.
A similar provision may be needed, but would be
covered under permits regulation.
Not applicable. The S02 allowances will be
included in the allowance rulemaking*.
A similar provision is unlikely.
Under consideration. Also could require the
results of these checks to be.reported.
-------
Citation Subpart Da
(i)
40 CFR 60.49a Reporting Requirements (Con't)
5
Acid Rain Regulation
Periods of excess emissions of opacity
defined as all 6-minute periods when
average opacity exceeds standards.
40 CFR 60.7 requires quarterly reports
of emissions in excess of opacity
standards.
Quarterly reports of 30-day averages
and other information (except opacity).
Under consideration.
Frequency of reports under review. Report
contents will need to include actual SO?
emissions and other data.
Reports sent to State."local, or Regions. Reports should be sent to EPA.
No standardized format for reports in
rule. . , s ..-.;;,' .-
Other recordkeeping and reporting
requirements included in various
performance specifications*.'"1 ;
NSPS General Provisions also require
records of startups, shutdowns, mal-
functions, and periods when continuous
monitoring system inoperative.. Records
of all measurements must be maintained
for at least 2 years. ' .
.Standardized format for reports likely
to be needed.
Recordkeeping and reporting requirements in
performance specifications may be retained
but may be integrated in recordkeeping/reporting
requirements in rule.
Likely to be retained in acid rain rule.
-------
PERFORMANCE SPECIFICATIONS (40 CFR 60)
Performance Specification 2 - Specifications and Test Procedures for S02 and NOX
Continuous Emission Monitoring Systems in Stationary Sources. (Appendix B)
Performance Specification 3 - Specifications and Test Procedures for 02 and C02
Continuous Emission Monitoring Systems in Stationary Sources. (Appendix 8)
Performance Specification 6 - Specifications and Test Procedures for Continuous
Emission Rate Monitoring Systems in Stationary Sources. (Appendix B)
^ Quality Assurance Procedures. (Appendix F)
00 . .
-------
Performance Specifications (Including Nos. 2, 3,
Citation Requirement
PS 2 Installation and measurement location
Sec. 3.1 of continuous'emission monitoring system
PS 3
PS 2
Sec. 3:2
.PS 2
Sec. 4.1
For. diluent monitor (02 and
(Same .as. PS2/ Sec. 3)
Reference method location and traverse
points v '(* V.
Equipment/Performance Specifications
Data recorder
Data recorder scale;must include
zero'and a high-level value - '
chosen by owner or operator. !
For CEM measuring uncontrolled
emissions/high-level value must be-
between 1.25 arid 2 times the average
potential emission level .(unless
otherwise specified in regulation).
and 6 in Appendix B and Appendix F of 40 CFR 60)
Acid Rain Regulation
Need to add specifications for flow rate sensor
location, including determination of represen-
. tativeness. . . .
, May only need for NOx because S02 is a mass/time
measurement. .: ..
Instrumental analyzer procedures in 40 CFR 60,
Appendix A are suggested. May need to measure
moisture to correct for use of extractive monitor
in a wet stack (extractive monitor analyzes sample
on dry basis;...flow would be on wet basis.. Flow and
concentration measurements need to be on the same
basis): v v
Needs to reflect'data recording system that
produces a continuous permanent record of emissions
and that has an adequate range to record all
expected measurements.
Scale must be .high enough to cover uncontrolled
emissions, if control equipment goes down.
Under consideration to cover periods when control
equipment is down. _ ;
-------
Performance Specifications (Including Nos. 2, 3, and 6 in Appendix B and Appendix F of 40 CFR 60)
Citation Requirement . Acid Rain Regulation
ro
o
For CEM measuring controlled
emissions or emissions in compliance
with an applicable regulation, high-
level value must be between 1.5 times
the pollutant concentration corresponding
to the emission standard level and the
span value.. .
If lower level value used/ source
must be capable of measuring emissions
which exceed full scale limits of the
CEMS in accordance with requirements
of the applicable regulation.
Data recorder output must have high*
level output value read between 90 and
100 percent of data recorder full scale.
Calibration gas, optical filter, or
cell values used to establish data
recorder scale should produce the zero
and high-level values. Alternatively,
a calibration gas, optical filter, or
cell value between 50 and 100-percent '
of the high-level value may be used in
place of the high-level value provided
the data recorder full-scale
requirements are.met. . " '
Under consideration. Range should be high enough
to cover all expected pollutant concentrations
(including uncontrolled emissions). . .
Same as above.
Same as above.
Same as above.
-------
Performance Specifications (Including Nos. 2, 3, and 6 in Appendix B and Appendix F of 40 CFR 60)
Citation
Sec. 4.2
PS3
Requirement
Design must also allow determination
of calibration drift at the zero and
high-level values. If this not possible
or practical, design must allow determi-
nations to be conducted at a low-level
value (zero to 20 percent of the high-
level value) and at a value between
50 and 100 percent of the high-level
value.
In special cases, if not already
approved, Administrator.may approve
: a single-point calibration drift
determination. : .
Over a 7rday period,, CEM calibration
drift shall not deviate from
reference value of-the gas cylinder
. gas cell,;or optical filter by more
than 2.5 percent of the span value.
If system,includes pollutant and
diluent monitors, drift must be
measured separately in terms of
concentrations. , .3
~.v, ..-.,.,'
Di luent monitor, cal ibration must
not drift by more than 0.5 percent
02 or C02 from^the reference value
of the gas, gas cell; or optical
filter.
Acid Rain Regulation
No change-necessary.
Under review.
No change needed.
No change needed.
No change needed.
Specifications would be added for measuring
flow monitor calibration drift.
-------
Performance Specifications (Including Nos. 2, 3, and 6 in Appendix B and Appendix F of 40 CFR 60)
Citation Requirement . Acid Rain Regulation
Sec. 4.3
(PS 2)
(NJ
ro
PS3
CEM relative accuracy may not
exceed more than.20 percent of.
the mean value of the reference
method test data in terms of the
'units of the emission standard
or 10 percent of the applicable
standard, whichever is greater.
For S02 emission standards
between 0.3 and 0.2 Ib/million Btu,
use*15 percent of the applicable
standard; below 0.2 .Ib/million Btu,
use 20 percent of the applicable
standard.
For diluent monitor, the relative
accuracy of the continuous emission
monitoring system must be no greater
than 20 percent of the mean value of
the reference method test data or 1.0
percent 02 or C02, whichever is greater.
Historical CEMS data being reviewed to determine if
relative accuracy specification should be
tightened. -At a potential 10 percent relative
accuracy, clean sources may not be able to qualify.
Unlike Da, the relative accuracy should be met over
the entire range of possible concentrations, not
just at low concentrations. Perhaps should tighten
relative accuracy for high emitting units. In
addition, the alternative use of 10 percent of the
standard would not be appropriate for acid rain
because S02 Ib/mmBtu emission rates can vary under
Title IV.
Not applicable to acid rain. SO? emissions are
monitored on mass/time for comparison with
allowances at the end of the calendar year. NOX
emissions would be recorded in Ib/mmBtu for
comparison with annual average emission rates.
Relative accuracy is being reviewed to see if
improvements can be made. ' . -
Relative accuracy is being reviewed to see if
'improvements can be made.
New .specifications would be needed for measuring
combined relative accuracy for continuous
emission monitoring system and flow monitor.
New specifications would be needed for measuring
the relative accuracy for individual flow monitor.
New specifications would be needed for measuring
cycle time/ response time.
-------
Performance Specifications'(including Nos. 2, 3,
.Citation Requirement .
Sec. 5 Performance Specification Test Procedures
Sec. 5.1 Pretest preparation
Sec. 5.2 Calibration drift test period
Sec. 5.3 Relative accuracy test period
" . « . . ' T :
Sec. 6 Calibration Drift Test Procedure
Sec. 7 Relative Accuracy Test Procedure
Alternative procedures' are allowed
under the General'Provisions.
and 6 in Appendix B and Appendix F of 40 CFR 60)
Acid Rain Regulation .'-; .'...-;'
No change necessary. .-,,-:
' ; ' "*. '
No change necessary.
No change necessary.
Changes would be needed to indicate that different
high and low reference points are used. New
procedures would be needed to account for flow
rate sensor drift test.
Could include linearity test. If so, new
procedures would be needed.
Could include .cycle time/response specification.
If so, new procedures would be added for tests.
Revisions may be made to specify use of
instrumental reference methods (RM 3A, 6C, and
7E) rather than wet chemistry.methods.
* " .
If alternative procedures are allowed, calibration
gases should:be standard reference material or
certified reference material (same as for required
daily calibration drift checks). A new standard
reference material (SRM) for higher, concentration
level may need to be developed;
Sec. 8
Equations
No changes necessary.
-------
Performance Specifications (Including Nos. 2, 3, and 6 in Appendix B and Appendix F of 40 CFR 60)
Citation Requirement Acid Rain Regulation
Sec. 9 Reporting .
Check with Region, State, or local
office for additional requirements.
Summarize-in tabular form the results
of the calibration drift tests and
relative accuracy tests or alternative
procedures. Include all data sheets
necessary to substantiate that perfor-
mance of CEMS meets specifications.
Includes standardized forms for recording
calibration drift and relative accuracy
tests.
App. F Quality Assurance Procedures
Sec. 3 Quality Control Requirements
-. Written procedures for calibrations/
adjustments, preventive maintenance,
: data recording, relative accuracy audits,
and corrective actions.
Sec. 4 Daily check of calibration drift and
adjust if drift exceeds two times the
drift specification.
Reporting procedures would be revised
to support the determination of compliance
with allowances and emission limitations
and to correlate with the allowance tracking
system.
Standardized forms likely to be used in acid
rain regulation.
Same procedures would likely be used.
Daily check the same. Checks would be made
using standard reference materials or certified
reference materials or Protocol 1 gases. Action
criteria and procedures are being reviewed to
determine if they are adequate to. meet
specifications and requirements of new rule.
Under new rule, units may be considered to be
out-of-control if calibration drift specifications
are exceeded.
-------
Performance Specifications (Including Nos. 2,,3, and 6 in Appendix B and Appendix F of 40 CFR 60)
Citation Requirement *'' s '.- ." * «'.'-. Acid Rain Regulation
Ul
If either the zero,(or low level)-or.
high level calibration drift exceeds
twice the applicable drift specifica-
tion for ,5 consecutive.daily periods,
the'continuous emission monitoring,
system is out-of-control.. > .,. .
If either the zero .(or low level) or
high level.calibration drift exceeds
four,times the applicable drift
specification during any.check,-the,
continuous emission monitoring system
is out-of-control..;.».
-'. ' - :" * :,. ;
Out-of-control period begins with
completion.of fifth, consecutive.
daily check with calibration drift
in excess, of 2'timeV the limit (or,.
.the time corresponding to the
comp 1 etion of the dai ly
-------
Performance Specifications (Including Nos. 2, 3, and
. i
Citation Requirement
Sec. 5.1
ro
en
Quarterly Relative Accuracy Test Audit
(RATA)
Quarterly Cylinder Gas Audit (CGA)
Quarterly Relative Accuracy Audit .(RAA)
Alternative audits "
If relative accuracy, using the
RATA, exceeds 20 percent or 10
percent of the applicable standard,
whichever is greater, the continuous ,
emission monitoring system is out-
of-control. ,
For S02 emission standards between
0.3 and 0.2 Ib/million Btu, use 15
percent of the applicable standard;
below 0.2 Ib/million Btu, use 20
percent of the applicable standard.
If inaccuracy exceeds > 15 percent
using the CGA or the RAA, or for the
RAA, 7.5 percent of the applicable
standard, whichever is greater, the
continuous emission monitoring system
is out-bf-control.
6 in Appendix B and Appendix F of 40 CFR 60)
Acid Rain Regulation . ' - .
Quarterly requirement the. same. Audit
provisions may need to account for flow monitors
and combined systems.
f
Quarterly requirement the same. Only standard
reference material or certified reference material
gases would be allowed. A new SRM for a higher
concentration level may be required.
Quarterly requirement the same.. Audit
provisions may need to account for flow monitors
and combined systems.
No change necessary.
Criteria for excessive inaccuracy is being reviewed
in light of existing CEM data analysis to determine
if changes are appropriate. Alternative.using a
percent of the standard is not applicable to acid
rain.
Same as above.
Same as above.
-------
* Performance Specifications (Including Nos. 2, 3, and 6 in Appendix B arid Appendix F of 40 CFR 60)
i ' ' " "' i ' *" ' '
Citation Requirement ; \ ,\ Acid Rain Regulation
.If excessive inaccuracies occur^for No change necessary.
"./. . two consecutive,quarters, owner. V
. operator must revise QC procedures ' . : ,
or modify or replace system. : :
Sec. 6 Calculations for Data Accuracy No change necessary.
Sec. 7 .. Daily assessment report containing . Reports would likely be sent to EPA rather
daily drift checks/quarterly audit than to State/local agencies. The
. information and -information on data . ; frequency of reports is under review.
accuracy must be reported quarterly.
(NJ
-J
-------
DRAFT
UTILITY BOILER PARAMETER MONITORING
FOR DETERMINATION OF FLUE GAS FLOW RATES
Prepared for:
Entropy Environmentalists, Inc.
Research Triangle Park, N.C. 27709
Prepared by:
. - >..
Richards Engineering
Durham, N.C. 2770S
December 4, 1989
-------
UTILITY BOILER PARAMETER MONITORING
. .OF FLUE1-GAS'"FLOWRATES ' .'
Possible amendments to the Clean .Air Act of 1977 may require sulfur .
dioxide..and nitrogen oxides.mass flow continuous monitoring for .large, fossil-
fuel fired utility boilers. Flue gas flowrate data will be necessary in order
to upgrade presently available gaseous pollutant concentration data to the
desired mass flow data.:vThe two general approaches to flue,.gas flowrate deter-
mination include (1) the use of direct-reading* continuous gas, flow analyzers,
and (2) the use boiler,parameter monitoring. The latter approach uses one or
more routinely .measured operating conditions to indirectly calculate flue gas
flowrate:'' -. . . ' ' ;'.;'.. . '.. ''-.'
_, * ^ - . - -
i ' '.-.- _« __.* ~ f t
>*' This preliminary study examines the feasibility of using routinely meas-
ured utility boiler operating .parameters to continuously monitor flue.gas flow
rate. The specific, issues evaluated include the potential accuracy of the
method, the practicality of the method,, and possible costs. Also, the
applicability of the method for different averaging times has-been.considered.
o ,"..,. , ' . ... ' " '."(.'
1.0 Background Information .
'' '"'' '":" :' '' .'" ,'.' .'.. '<,-'' - : . -.-j.
There are three principal options available for parameter monitoring of .
flue gas flowrate at fossil fuel fired steam boilers. These include the
following: . .. .. ..,; ;. i- '-...-... ,: ' ., ' '
1. Use of'.induced draft fan operating .data and fan .curves
2. Use of feedwater (or steam) flowrate - flue gas flowrate
' . correlations, or use of.electrical power output - flue gas
flowrate correlations > > .
3. Use of combustion stoichiometry calculations
. '»* " . " "- ' ..'.-*..
Procedures based on induced draft fan operating use either static
pressure .data or fan motor horsepower data. This is the only approach which
does not 'require an oxygen,monitor in the,stack to correct for changes in
excess'-air/ The feedwater. steam, or electrical output flowrate correlations.
all depend on standard .plant monitors. Some, enhancement of the: steam rate
meters would be necessary,at -some facilities to account for variations in
temperature. Alsbv some preliminary testing would be necessary to establish
the necessary correlations...- The, combustion stoichiometry approach would
require accurate, coal feedrate measurements at either the main coal conveyor
or at the individual pulverizer feeders. Ultimate analyzes of .these samples
would have to manually, entered into system calculating gas flowrate.
-------
2. Use of Induced Draft Fan Data ,
*
. The function of .the induced draft fan. is to move the flue gas produced in
the furnace area through the various flow resistances downstream of the boiler.
Since it is the "prime mover" of the flue gas, it is logical to assume that the
fan operating parameters have a useful relationship -with gas flowrate which can
be exploited. Unfortunately, there are several practical complications vhich
make this general approach for flue gas flow monitoring difficult at best.
Some background information concerning ID fans at utility boilers is necessary
to support subsequent discussions of these fundamental monitoring problems.
; ' . -
- * i- »
Centrifugal fans are generally used for induced draft service. Due
partially to the advances in flyash control efficiency, there has been a rapid ,
progression away from relatively simple radial blade designs to more energy
efficient designs such as radial tip fans and backward curved fans. -The per-
formance of centrifugal fans is described using fan curves. These, are determined
by the fan manufacturers under idealized test block conditions, and the curves
are presented in standard pressures and temperatures. A typical fan curve for
a radial tip fan is shown in Figure I. The top curve, labelled "a", is the fan
characteristic curve. It describes the possible combinations of static pressure
rise and gas f lowrate with the fan operating at the specific rotational speed.
Over most of the operating range, it is apparent that centrifugal fans can
develop progressively less static pressure rise as gas flow increases. The
second curve, labelled "b" included in Figure 1, is the the fan motor brake
horsepower. This is the actual input power to the motor and it generally
increases as gas flowrate increases. The third curve, labelled "c," is the
system resistance curve which is a function of the square of the gas flowrate.. '
The fan will operate at the intersection between the fan characteristic curve
and the system resistance curve. . -.
Theoretically, the flue gas flowrate through the fan can be determined
simply by measuring the actual fan static pressure rise and/or the fan motor
horsepower and then applying this data to the fan curves. This is illustrated
in Figure 2. However, this simplistic approach is subject to major errors and
numerous practical problems. These limitations are described in the following
subsections.
2.1 Accuracy of test block fan curves
The fan curves provided by the manufacturer are for idealized conditions -.
which are not representative of actual installations. There are a variety of
nonuniform gas flow conditions at the fan inlet and outlet which, significantly
influence actual performance. These are collectively described as system
effects factors (References 2 and 3). These can routinely affect expected gas
flow by 1SZ and expected static.pressure rise by approximately 302. In some
cases, the deviations from test block conditions could be much more extreme.
Accordingly, it would not be advisable to simply incorporate the fan manufac-
turers' test block fan curves into a flue gas flowrate monitoring program.
-------
o 14
»
W . , 0
« 12
o
5 10
«
b
o
Vt
CLi
en
8
6
4
2
Curve c
Curve b
140
120
100
80
60
40
20
10 20 30, 40 50 60 70 80 90 100 110 120,;
Gas Flow, Thousands-of SCFM' * .
W:
« '
.1"
CO
w
£
2
Figure 1. Typical fan characteristics curve and brake horsepower curve
for a radial tip centrifugal fan '....;,
en
10- 20 3f -40 50 =60 70 3u .90 lOO 110 120
,. - , :, ,v Gas > Flow, Thousands of.SCFM- -
Figure 2. /Presumed technique for determining gas flowrate from fan
operating data / - > .'
I
0.
-------
A series of field performance tests could be conducted to derive a revised
fan curve which takes into account the system effects. Such tests would have
to be performed using procedures similar to those specified in ASME Power Test
Code FTC 11-1984 or in the Air Moving and Conditioning, Association (AMCA)
Procedure 203. There is increasing application of these procedures for deter-
mining conformance with fan performance guarantees. The tests are expensive
and time consuming. Furthermore, they are not entirely suitable for developing
an'"actual conditions* fan curve. The procedures are designed for evaluating
performance at the maximum rating and do not address conditions at partial load.
The operating point for fans with inlet dampers varies substantially as indicted
in Figure 3. The operating point moves downward along the system resistance
curve, labelled "c", as the inlet damper opening is changed.
Row. % tost block
Figure 3. Changes in fan operating point due to changes in inlet damper
position and pitch (based on Figure 9, Reference 2)
The present fan field test procedures are limited to steady state fan
operation at maximum rating. and they are not subject to the changes shown in
Figure 3. Accordingly, these procedures do not have to take into account
the measurement errors involved with varying fan inlet/outlet static pressures
and temperatures during load variations. It would be necessary to revise and
expand these already rigorous and difficult procedures so-that.they would be
appropriate for developing accurate fan curves.
-------
2.2 Fan speed variation
.The "actual conditions* fan curves discussed above would also have to
consider fan speed variations.. There is increasing use of variable speed
drives such as two speed motors and hydraulic drive couplings. A change in fan
speed has a strong influence on,the'position of the fan curves as indicated in
Figure 4. Accordingly, any field tests would have to be conducted over the
entire operating range of the variable speed drives. This means that a large
family of curves must be developed, not just a single curve.
50 75
Ftow, %tMt Mock
Figure 4. Changes in fan characteristic curves at different fan speeds
(Figure 10v Reference 4) .
During routine use of the field curve, it.would be necessary to monitor
the variations in fan speed so that' the proper curve could be used by the gas
flowrate calculation algorithm. There are numerous techniques for monitoring
fan speed..'However; the accuracy of the instruments during long term-, routine
use is unknown. Since the fan speed is controlled by the furnace draft
controller; it may also be possible to intercept and record the signal from
this "control loop.. The advantage of this approach is that it would not be
necessary to maintain the fan tachometer. However, the accuracy of the flue
|;asflow rate procedure would be limited by the accuracy of the relationship
between this signal and the actual fan speed. The possible errors are large.
-------
2.3 Fan static pressures and brake horsepower . .
There are significant problems in applying fan curves even if the curves
are known and accurate. In order to determine the instantaneous gas flovrate,
either the static pressure rise or the brake horsepower oust be measured. It
is difficult to measure either one on a long tern, continuous oasis.
2.3.1 Brake horsepower - Only the fan motor current is routinely monitored,
and this can not be used as a direct measurement of the brake horsepower. As
indicated in the equation below, the relationship between current and horse-
power also includes the power factor.
Horsepower -yJ x V x I x cosjl
where: Hp - Brake Horsepower
V - Operating Voltage, Volts A.C.
I - Operating Current, Amps. A.C.
tf - Power Factor, Dimensionless
The power factor is a measure of the difference in the voltage and current
phase angles. It is normally in the range of 0.90 to 0.95 at high boiler loads,
and it drops slightly as the fan inlet dampers are closed. To avoid energy
losses, some plants install additional capacitance on the fan motor controller
in order to minimize decreases in power factor at low load. Unfortunately, it.
is difficult to routinely monitor power factor. Therefore, the extent of unit-
to-unit variation as a function of load has not been extensively studied. An
assumed power factor would probably introduce an error of at least 52, and it
could be much higher.
2.3.2 Static pressure - Measurement of static pressure rise across the ID fan
is complicated by the extreme localized pressure variations at the fan bound-
aries. Testing grid arrangements upstream and downstream of the fan similar
to those discussed in ASME PTC 11-1984 would have to be used to accurately
determine the average static pressure change. A study would be necessary to
determine the most appropriate locations of the static pressure sensor grid.
The use of a single point sensor or multiple wall-mounted sampling ports could
results in measurement errors of at least several inches of water. This error
would translate into a 10 to 20Z error in the calculated gas flowrate.
Any array of static pressure sensors near the ID fan would require
blowback capability in order to cleanse the probes of accumulated flyash. The
probes would have to be removed occasionally to check for any erosion of the
probe tips which would affect the calibration. Also, there would have to be a
microprocessor for averaging the static pressure signals. The static pressure
monitoring instrument system mounted near the fan would be at least as compli-
cated as any direct reading gas flow rate instrument mounted in the more
friendly environment of the stack. . .
-------
2.4 Gas density
.Any fan curve based gas flovrate calculations would have to take into
account changes in. gas'density due to gas temperature changes. Ave'rage gas
temperatures can vary 20 to 40 degrees Fahrenheit between full load and 60X
load. Also, air infiltration upstream of the ID fan can lead to gradual
decreases in both full load and low load gas temperatures. A_change of 40 *
degrees Fahrenheit is roughly, equivalent (pressure differences not included)
'to a gas density change of 51. This would translate into a 5X error in the
fflue.gas flovrate calculated from the fan curve. Accordingly,' it would be
necessary to compensate for gas density changes. '.''
,""**' ;
2.5 Summary - .... . , "
,» , > -, \"
Any flue gas flow calculations based on fan operating data would be more
complicated -and less accurate than direct measurements using stack mounted
instruments.. An extensive initial test program would be necessary to select .
the appropriate'test.locations.and to acquire the fan curves applicable to
.actual conditions. - Continuous,, reliable monitoring systems would have to be
'developed for determining either static pressure ;or fan motor .power, fact or. ..',
Microprocessors would have to be installed to average the data from multiple
sensors and to perform any gas density corrections. The computer system would
have to record and'average the data, over the averaging time period of the
future standard(s). : Furthermore, it would be very difficult to conduct
independent audits of the fan monitoring instruments and flowrate calculation
procedures^because of the complexity of .the monitoring site flow conditions
and the normal variability,of boiler operation. For all of these reasons; the
use of stack mounted direct gas flowrate monitors would -be more accurate .and
practical.. ." .' ;..;.-. ", f -.','..,-.. '.->,' '- ' "* '""" ' .
** *' I '*.r, > t
' '-
3. Feedwater Flowrate, Steam flowrate, and Electrical Power Output Correlations
There are several boiler operating rate parameters-which can be used as
independent .variables in flue,gas flowrate correlations. All of these approaches
use existing and well: understbond monitoring instruments. .
3.1 Feedwater,or-steam flovrate correlations* ' " \ i - ' - :
"- ' ' : ' - ,; . '.'.'-i-' '-.- '.'.-'.* - '' .
Boiler steam,rates and/or feedwater .rates could-theoretically be used 'in a
site-specific correlation to calculate flue gas flowrate. It would be necessary
to conduct a set of baseline, flue'gas velocity traverses in the vicinity of the
CEMs in order to derive the.steam/feedwater rates versus flue gas .flowrates.
It would also be necessary to monitor the stack oxygen and correct for changes ,
in the excess,air,rates since the baseline period. The advantage of this
approach is that the steam'and f eedwater flowrate meters and the oxygen analyzer
can provide short -term data. Furthermore, .signals from both, instruments can go
into an computerized system for calculation and storage of the calculated flue
gas flow data. .'."*'..'-.
-------
The most accurate boiler operating rate parameter is the feedwater flow
rate. This can be measured by a variety of differential pressure techniques
such as orifice meters and flow nozzles. Accuracy of measurement could be
maintained at plus or minus IX. Steam flow meters are measured by similar
techniques. However, they are vulnerable to errors due to steam temperature
variability (Reference 1). For reasonable accuracy, steam flow meters would
have -to be equipped with temperature and pressure compensator microprocessors.
Also, it would be necessary to monitor both the main steam flow to, the high
pressure turbines and the extraction steam flow to the intermediate pressure
turbines. It is less complicated and potentially more accurate to use the ;
.boiler feedwater rates as the basis of the correlation. .
. A series of Method 2 gas flow tests would have to be conducted at various
loads in order to establish the baseline curve relating gas flow to feedwater
.flowrate.- Obviously, the accuracy of this method is limited to the accuracy
of these tests. It would also be necessary to factor in long term changes in
the unit heat rate (efficiency) , intentional changes in the boiler excess air
levels, and unintentional air infiltration related- changes in the observed
stack excess air levels. The adequacy of the relationship would have to be
occasional y reverif led due to these conditions . .
3.2 Electrical power output correlations
The procedures involved in developing a electrical power output correlation
are essentially identical to those for the feedwater flowrate correlation. It
would necessary to conduct preliminary Method 2 type velocity traverses to
establish .the correlation. These tests would have to be repeated following any
significant reductions in unit heat efficiency. An oxygen monitor in the stack
would be necessary to account for boiler excess air changes and for air infil-
tration downstream of the boiler.
4. Combustion Stoichiometry
Given the ultimate analysis of the coal, It is possible to perform basic
combustion Stoichiometry calculations and determine the necessary combustion
air requirements. This can be combined with the measured excess air levels,
coal average heating value data, and coal feed rate data to determine the .
resulting flue gas flowrate. The method is basically a continuous exercise
of "weight method combustion calculations* using a computerized program. The
basic equations are described in References 5 and 6 and are listed on the ASME
Abbreviated Efficiency Test Forms included in the appendix.
For'maximum accuracy,- the coal samples should closely approximate actual
"as fired" coal characteristics. .The sample should be obtained using the
cyclonic collectors .mounted on each of the individual coal burner tubes from
the pulverizers. , The samples collected over the specified time period would
be physically integrated over time. :
The. main disadvantage of this approach is that the accuracy of the calcu-
lated flue gas flowrate is a strong function of the coal feed rate monitors
-------
and of the coal laboratory analyzes. Errors in;coal feed rate alone could
range from 5 to 20X. 'For short averaging time determinations, thecoal ;feed
rate must be determined between the various coal bunkers;and.pulverizers. .
The coal bunkers introduce a mass capacitance effect which could affect the
accuracy of the short term data during periods of varying load. At the present
time, most utilities do not use the expensive and difficult to maintain gravi-
metric feeders between the bunkers and'the'pulverizers. Instead, total coal
flow is-measured using belt scales along .the main coal'conveyor. The adoption.
'ofshort term averaging times would.demand.extensive"arid expensive modifica- .
tions .to existing facilities. Maintenance requirements would increase. ' ' t
For -long term averaging times; the main coal'belt scales can. conceivably
provide adequate data. These can be calibrated to within IX accuracy. In the
.past, the need for this level of accuracy has'been restricted to compliance
tests and-equipment performance tests. Long term performance of these scales
is 'not known.-."" ''..>." - -.'' "-' , '_ ..".-''''- . " -.-/.'
. The laboratory tests for coal heating value are generally within plus or
minus IX (ASME Power Test Code 4.1-1954(R1985)).' However, greater error is ~
undoubtedly introduced by coal .sampling problems-. Also, the. ultimate and
proximate analyses are relatively, expensive and, time consuming.. Each coal
sample tested independently would cost between $200 and $400. If the averaging
time, is on the order of 1 hour or even four hours,'the daily "cost for the coal
samples can reach very .high levels. If the averaging time is 24 hours or
longer, the analytical-costs become,less significant. ~However, long averaging
times create the need for'analytical techniques to blend the collected coal .
samples .prior to the tests. It.should be remembered that long averaging- times
are probably necessary simply because of the need to use the main coal belt
scale rather than.installing 4 to 8 separate gravimetric feeders for .each
boiler system'. .. - ,. ., :/ . ;
The accuracy of the method is'also dependent on the quality of the flue
gas oxygen measurements. Multiple point sampling points may be. needed in
ducts and stacks subject to flow stratification due either to localized air .
infiltration or persistent-flow nonuniformities originating at the burners.
It should be noted'that -both-conditions are very common. For this reason,
the ASME Power Test Codes for on-site fan testing (ASME'PTC 11-1984) require
a sampling grid for gas samples. - \
- "' *-*' *^ ' * "
The excess air rate could be determined using the standard equation
shown below, which assumes that carbon monoxide concentrations are negligible.
Excess Air - [ 2i/(21 r Oxygen)] x 100 . :
The oxygen.should be measured in the stack at the same location as the
continuous gas monitors in order to take into account 'air infiltration across
the particulate control systems, the air preheaters, and the induced draft fans.
The oxygen concentration routinely increases 0.10 to 0.50X across each of the
system components. . For example, boiler outlet oxygen concentrations of 3.OX
-------
could increase to 3.3 to 4.SZ at the stack GEM location. Control device
corrosion-and expansion joint deterioration can cause much greater shifts in
the stack oxygen concentrations. . : ^
The combustion stoichiouetry approach does not take into account the
combustibles content of the flyash; Generally, combustion in large fossil fue]
boilers is very complete and the combustibles content of the flyash is in the
range of 3 to 5X by weight. However, pulverizer wear problems and poor coal
grindability conditions can result in flyash combustibles in the range of 8 to
15X by weight. At these undesirable levels, the calculated flue gas flowrate
would be biased approximately 0.5 to l.OX to higher than actual flow rates.
'>"_ " * -
. f _ _ f
5. General Comments and Summary . , .
This preliminary review of possible parameter monitoring options suggests
that none are especially accurate or practical. Under the best conditions,
the accuracies would be limited to plus or minus 10X, and for most facilities
the- calculated flue gas flowrates would probably be in error 20. to 30Z.
Furthermore, all of these involve either extensive preliminary testing and/or
in-stack oxygen monitors for excess air correction. None of these adequately
account for combustibles content of the flyash. For these reasons, the direct
.measurement of flue gas flowrate using stack mounted instruments appears to
be more attractive. . .
There are several other considerations .that also favor direct measurement
over indirect boiler parameter monitoring at the present time. These are
described below.
t *
Very few new fossil fuel fired boilers have been installed during the
last five years. The aging population of relatively small 75 to 200 megawatt
fossil units is being subjected to greater cyclic load conditions due to base-
loading of nuclear units or more energy efficient fossil units. The accuracy
of the plant instruments during rapidly varying boiler operating conditions is
not well known. A direct reading;gas flow instrument would have less diffi-
culty following these load variations. .
» ^ i H * '
The accuracy of plant instrumentation during start-up and shutdown
conditions has not be assessed. It is unlikely that.the calibration procedures
are relevant to these transient conditions. " '
-The 'population of fossil fuel-fired boilers is very diverse. The '
application of a parameter monitoring approach of essentially any design would
be subject to numerous site specific application problems. The agency would be
besieged with numerous significant clarification questions and problems. Also,
it would be difficult at best to conduct on-site audits.
Plant personnel would probably be more receptive to a direct reading
instrument system rather than a parameter monitoring program. The latter is
more intrusive due to the comprehensive quality assurance procedures which
would accompany the monitoring requirements. It is also remotely possible that
the regulation-driven monitoring requirements would threaten the reliability of
10
-------
the - boiler system and lead to occasional boiler trips. A stack mounted' direct
gas flowrate monitor is far removed from the complex boiler system, and, there-
fore, it is less of a burdenand a threat..
It should be noted that advances in coal monitoring/analytical procedures,
microprocessor-based performance monitoring systems, and plant expert systems
may improve the viability of parameter monitoring in the next 5 to 10 years.
However, at the"present time, parameter monitoring approaches are definitely
premature unless the standards, allow long average times, and require only very
modest accuracies. ".'*>,
6. Recommendations
1. Developmental work on flue gas flowrate monitoring should emphasize
direct stack-mounted'instruments.
2. Parameter monitoring should.be retained only as an option which can
be chosen by plants not wishing to use the direct gas flowrate monitors.
3. The possible accuracy, attainable by .steam flow, feedwater flow, and
electrical output correlations parameter monitoring should be evaluated
in a more detailed study.
7. References
1. Hakansi, J. and J. Reason. "Monitoring Powerplant Performance,"
Power, September 1984, pages SI to S24.
2. Reason, J. "Fans, A Special Report," Power, September 1983,
pages SI to S24.
3. Rinek, L.P. and R. 0. Walter. "Avoid Costly Errors in Field
Tests of Big Fans," Power, October.1984, pages 51 to 54.
4. Aberbach, R. J. "Fans, A Special Report," Power, March 1968,
pages Si to S24.
5. Dukelo, S, The Control of Boilers. Instrument Society of
America., 1986.
6. Babcock & Wilcox Company, Steam. Its Generation and Use.
page 6-10, 1978.
11
-------
. " APPENDIX v
ASME Test.Forms for Abbreviated Efficiency Test
-------
CALCULATION SHtCT
ASMI TUT 'off.*
AIMIVIATEO MIOINCV TUT
PTC4.1-KI944)
11
M
'*
M
AS
M
47
41
AT
M
n
»
* OVNCt 0* n.Mtt " " , ' TtlT MO, . WILfa NO. DAT!
' lTt« U ITtM If
ua-vcoiiB.mmFuuuAiFi.1 FITMFVII
' '"".*» |l»l«» ltm» 1 IMOOLOil
CAItONtUMID I , . tIFAIATCL
FUfl , * L -»
OtT BAt Fft LI IICOi « W, TOtj COt .
AUtMID .*" * ew / \ t r- /
ITIMH ITtMIJ IITIHM ITCH Ml 1 iTIMM
:»»..... " "\ ' ...../ .
(iTmn iTiiiwX L
....*......!
/
ktyw
LutourrtAM ''
MFFt «ATt HIALLT'
i«Lteo«Tfirr, TWIT.
ItTIMATIO
r. iiiucnoNt.
*n. . . .
*'»
* 'I
«,J
1 i-
.. * * NIAf tail IFFieilMCT
NIATLOttMM L»»«T6»» ITMtt (ITIMIll.pTMItl
TOMTOM FI«L§« »c. »pi^-*-rt« «oj«- ""."«".
FIMOFUIL * tMi ....
.IfMTNALFV OF LNMttOAT T AI«J . 1Z!£2L a KtMTNALFT OF VAMM ,
KMT LOU Mil »0 MjB FROM COM. OF Mb ' "S « [itNTHALFT OF.VAFM AT 1 MIA » T OAI
' L*OI - If MTMALFT OF LIOUIO AT T AIM]
. t > *T*»** B Iff MTHALFT OF V*FM AT 1 FUA t T ITIH III . IIHTHALFT OF LIOUIO AT
. 1M . TITIBIII).
Mt AT LOU OM TO ' ITtMU ITtMtt-
CCMUtTIBLI W MFUIt » .
HIATLOUOUITO TOTAL »TU«ABIATio»« LOU PI »-
AOUTMH* UMFMfBFUCL - itim.
UHHIAUIMO lOUtl ~ .. ;
T0t*t ' . '
irncKHCr . IMI . iTtn ni
MFIMt
FUtL
'/;
'.
.
M «
i.«.
**.MI.
41
TT
iT1"*-
S---
a
.
' .-
, LOU
*
'
*
-------
SUMMARY SHEET
ASM! TEST FORM
FOR ABBREVIATED EMICIENCT TEST
PTC4.1..(I9«41
- ' 1EITMO. MILE* HO.
PATC
OM t» OP PLANT -, . ' ' .- LOCATION '
TIITCOMOUCTEB>n . - OEJtCTIVE OT Tf IT
OtMATNM
OILC*. HAM » TTPf . . BATtO CAPACITY
ITOCf t. TTPf A UH ..--, . ' .
«
PULVE 43ZEI. TTPE A UK .. '"' -.' OV*NK. TTPC A tltf
ruiLWEP mm t COUMTT IT ATI
titf AI PIMO
. p*titu*f s 4 Tf UPCIATUIIE:* . ruiu DATA -
i
. i
>
4
I
f
1
""a
,
10
II
M
JI
14
ITf AM rituuti M taut o*m '
ITIAM PMUUM AT t, M, OUTLET
tTlAM PkEUUn AT «. M. IMltT
ITf AM pnuum AT I H OUTLPT
ITf AM TtMPtRATimt AT 1. M. OUTLCT
tTf AM Tf MPffHATUfff AT P..N- MLP.T
tT|AM Tf MPf ftATUM AT P...M.. OUTLCT
WATE f TP.MP. f NTEMNG JCCON lltOILf R4
ITI 4MQUALITT«MOIlTtn» Ml P. P.M.
Al* TEMP. AmUMO OILt* tAMOIEHT)
TIMP.AI* POd COMBUSTION
fTH. It «.>->». ' | | | |1 T
TCMPf *ATU» Or ruEL
CAi TEMP, LEAVING C««iU.I !!<>.) IAI> »».l
6AI TCMP. EMTCHIMO Ml (II MnrifMMw M W
4 ^ ' ' '* ^
OMIT OUAMTITIII
11
M
,
II
It
M
II
a
11
14
*»
1*
I?
It
It
10
11
tNTMALPT OP IAT.LIOUIO ITOTAL HI ATI
ENTHALPY arilATU*ATEO)ISUPf *NIAT»>
fix.
tNTMALPT Or IAT. PEIO TO MOILED
neoN.i
ENTHALPY OP RfHtATtD ITf AM I.M. INLf T
(N TMAL PT OP MNI ATI P ITIAJI «. K.
OUTLET
HEAT A*VL*OP ITtAM (ITIM U.ITtH IT)
MIAT Alf/LI ».M. STEAK (ITEM It. ITEM 111
BIT REFUSE (AIM PIT PLT ABO PER L*
ft riHEO PUCL
I* LB W ItPUSI l«f IBMTEO AVIVACC)
CAIOON UIIHf P ft* L* At PIMP FUIL
OP.T CAI PCI LS At rwto rufL nwMto
ACTUAL «ATf« IVAPOMTfO
MMIA1 STIAM PLO*
ATE or PUEL rwwe »» PIMD i
HEAT OUTPUT IM «LO«.OO*N BATE*
IS If- (IM. M.IM. »MI»->7rti» lll«u_ 11
OUTPUT I0»
PLOT 0*1 AMAL- lOWLflllf CONI «*! MT»I
12
11
34
11
li
CO,
0,
CO
N, tor airrt*tncr.i
Ei:fn*»
pita
**
P»'«
!_
r'
r '
T
t
'
f
r
r
t
N>/n>
IWIk
ta^H.
h1"*
IM/U
UWI»
lM/lk
Ih/lk
(/Ik
Ik/Ik
Ik/Ik
IkA. .
kA.
'k*.
klA*
iBAt-
k«A>
^^^MM
-
-
.
OUTLCT
it VOL
*VOL
^ VOL
^^^^^H
- » 1
.
COAL At PWtO
P«OI. AMALVIII
n
*
v
n
41
41
Cl
4]
Jtt.
A|
17'
41
»
MOIITUM
VOL MATT(*
PltEO CA*WNI
AM
TOTAL
>. « ik At rimo
ASM IOPT TCMP.-
ASTM Mf THOO '
«.
>AL a* ML At riMO
ULTIMATt ANAL Till
CARBON
NTOBO^tfN | a
OITMN
MITMMSEN
UJLPMU*
AM
MOinun
TOTAL
^^^MMIIIM
COAL PULVMttATIOM
"
4t
st
M
CIIMOACILITT
IMDEI*
PMENEU «THW
WH>.
FIHEKEII « TMtU
n
n
..
4,
41
. . «>
PLAtM POINT f-
tp.BOTMrD»t.API>
':
VUCOWTT AT tSU*
TOTAL MTOMOtM
._>»
CAS
H
SS
Jt_
a
jt
v
M
t|
CO
CM. MCTMANt
C.M. ACETTLENt
C.M. 1THTLENE
C.H, ETHAHE
H.I
CO,
M, HTOmCIN
TOTAL
41
fj
41
TOTAL HTOMCf N
«M
HVOL
m^m^m^m^m
'
OEMSIT* 4* P
ATM. PMIS.
tM Pt» CU PT
10 Pf L» '
MPUT.OUTPUT ' ITIM1I
EPPICIEHCT OP UNIT m ITEM 1
- NIAT LOIS irpKllMCT
tl
M
a
M
w
»
n
Hf Al LOW DUE TO OUT CAI
MIAT LOIS PUf TO MOUTIMI Itt Put L
HEAT LOU Out TO H,O PROM COMC.OPMj
MIAT LOU DUt TO IABUTNM ' ;
UNMEAlMtO LOSSII
TOTAL '
UK
t
A.P. PUtL
-'
CPPKIIMCT » UN . I«H fll
PUtL
-------
January 17,1991
El
DRAFT
"*; - " yyptS GEMS SUBSET ANALYSIS:
SO, GEMS AVAILABILITY DATA (1988-19901
Category
Subpait Da and PSD
Utilities
All NSPS and PSD
Utilities
All SIP Utilities
All Utilities
All Repotted Quarters
No. Units
44
167
153
320
No. Quarters
253
1237
895
2132
% Availability
94.68
96.11
95.42
95.82
-------
DRAFT
SO2 GEMS PERCENT AVAILABILITY --
DISTRIBUTION OF QUARTERLY REPORTS FOR
ALL UTILITIES (1988-1990)
Percent of Reported '
Quarters Out of '2132
100
80
1669
60
40
20
22 i«- 77 s a 1 a a i a a i
"
>95 90 86 . 80 76 70 65 60 66 60 46 40 36 30 26 20 15
10
1
6
Percent Availability
------- |