Proposed Rule: Air Pollution Standards of Performance For New Stationary Sources: Predictive Emission Monitoring Systems Performance Specifications Testing and Monitoring Provisions Amendments

45608 Federal Register / Vol. 70, No.
151 / Monday, August 8, 2005 / Proposed Rules
(EPA), Region 8, Mailcode 8P-AR, 999 ACTION: Proposed rule. through Friday, (except for Federal
18th Street, Suite 300, Denver, Colorado holidays), at the U.S. Environmental
80202–2466. Such deliveries are only SUMMARY: The Environmental Protection
Protection Agency, EPA Docket Center,
accepted Monday through Friday, 8 a.m. Agency is proposing performance
EPA West, Room 108, 1301 Constitution
to 4:55 p.m., excluding Federal specifications (PS) that evaluate the
Ave., Washington, DC 20004; telephone
holidays. Special arrangements should acceptability of predictive emission
(202) 566–1742. The docket office may
be made for deliveries of boxed monitoring systems (PEMS) when used
on stationary sources. This PS is needed charge a reasonable fee for copying.
information.
Please see the direct final rule which to provide sources and regulatory FOR FURTHER INFORMATION CONTACT:
is located in the Rules Section of this agencies with performance criteria for Foston Curtis, Emission Measurement
Federal Register for detailed evaluating this new technology. The Center, Mail Code D205–02, Emissions,
instructions on how to submit intended effect of this action is to Monitoring, and Analysis Division, U.S.
comments. establish standardized performance Environmental Protection Agency,
requirements that will be used to Research Triangle Park, North Carolina
FOR FURTHER INFORMATION CONTACT:
evaluate candidate PEMS uniformly. 27711; telephone (919) 541–1063;
Amy Platt, Environmental Protection
Agency, Region 8, 999 18th Street, Suite The affected industries and their facsimile number (919) 541–0516;
300, Denver, CO 80202–2466, (303) Standard Industrial Classification codes electronic mail address
312–6449, platt.amy@epa.gov. are listed under SUPPLEMENTARY
curtis.foston@epa.gov.
INFORMATION. In addition, we are
SUPPLEMENTARY INFORMATION: See the
proposing to make minor amendments SUPPLEMENTARY INFORMATION:
information provided in the Direct Final
to various testing provisions in the New
action of the same title which is located General Information
Source Performance Standards (NSPS)
in the Rules and Regulations section of
and National Emission Standards for A. Affected Entities
this Federal Register.
Hazardous Air Pollutants for Source
Authority: 42 U.S.C. 7401 et seq. Categories (MACT) to correct Predictive emission monitoring
Dated: July 14, 2005. inadvertent errors, make needed systems are not currently required in
Max H. Dodson, updates, and add flexibility. any Federal rule. However, they may be
Acting Regional Administrator, Region 8. DATES: Comments: Submit comments on used under the NSPS to predict nitrogen
[FR Doc. 05–15608 Filed 8–5–05; 8:45 am] or before October 7, 2005. oxides emissions from small industrial,
BILLING CODE 6560–50–P Public Hearing: If anyone contacts us commercial, and institutional steam
requesting to speak at a public hearing generating units. In some cases, PEMS
by August 23, 2005, we will hold a have been approved as alternatives to
ENVIRONMENTAL PROTECTION public hearing on September 7, 2005. CEMS for the initial 30-day compliance
AGENCY ADDRESSES: Comments. Comments may test at these facilities. Various State and
be submitted electronically, by mail, by Local regulations are incorporating
40 CFR Parts 60 and 63 facsimile, or through hand delivery/ PEMS as an emission monitoring tool.
courier. Follow the detailed instructions The major entities that are potentially
OAR–2003–0074 as provided in Unit IB of the affected by Proposed Performance
[FRL–7947–5] SUPPLEMENTARY INFORMATION section. Specification 16 and amendments to the
Public Hearing. If a public hearing is
RIN 2060–AG21 subparts are included in the following
held, it will be held at 10 a.m. in the
tables.
Performance Specification 16 for EPA Auditorium, Research Triangle
Predictive Emission Monitoring Park, North Carolina, or at an alternate
Systems and Amendments to Testing site nearby.
Docket. Docket No. OAR–2003–0074,
and Monitoring Provisions
contains information relevant to this
AGENCY: Environmental Protection rule. You can read and copy it between
Agency (EPA). 8:30 a.m. and 5:30 p.m., Monday
TABLE 1.—MAJOR ENTITIES POTENTIALLY AFFECTED BY THIS ACTION FOR PROPOSED PERFORMANCE SPECIFICATION 16
AND FOR PETROLEUM REFINERY NSPS, KRAFT PULP MILLS NSPS, MUNICIPAL SOLID WASTE LANDFILL NSPS
Examples of regulated entities SIC codes NAICS codes
Industrial, Commercial, Institutional Steam Generating Units ................................................................................. 3569 332410

Stationary Gas Turbines .......................................................................................................................................... 3511 333611
Petroleum Refineries ............................................................................................................................................... 2911 324110
Kraft Pulp Mills ......................................................................................................................................................... 2621 322110
Municipal Solid Waste Landfills ............................................................................................................................... 4953 562213
Surface Coatings ..................................................................................................................................................... 3479 336111,
336112
Coke Ovens ............................................................................................................................................................. 3312 33111111
TABLE 2.—MAJOR ENTITIES POTENTIALLY AFFECTED BY THIS ACTION FOR AMENDMENTS TO PERFORMANCE
SPECIFICATION 11 AND PROCEDURE 2, APPENDIX F, PART 60
Portland Cement Manufacturing .............................................................................................................................. 3559 333298
VerDate jul<14>2003 20:26 Aug 05, 2005 Jkt 205001 PO 00000 Frm 00038 Fmt 4702 Sfmt 4702 E:\FR\FM\08AUP1.SGM 08AUP1
Federal Register / Vol. 70, No. 151 / Monday, August 8, 2005 / Proposed Rules 45609
SPECIFICATION 11 AND PROCEDURE 2, APPENDIX F, PART 60—Continued
Hazardous Waste Incinerators ................................................................................................................................ 4953 562211
SPECIFICATION 2, APPENDIX B, PART 60
Fossil Fuel Steam Generators ................................................................................................................................. 3569 332410

Electric Generating Units ......................................................................................................................................... 3569 332410
Industrial/Commercial/Institutional Steam Generating Units ................................................................................... 3569 332410
Small Industrial/Commercial/Institutional Steam Generating Units ......................................................................... 3569 332410
Municipal Waste Combustors .................................................................................................................................. 4953 562213
Nitric Acid Plants ..................................................................................................................................................... 2873 525311
Sulfuric Acid Plants .................................................................................................................................................. 2819 325188
Petroleum Refineries ............................................................................................................................................... 2911 324110
Primary Copper Smelters ........................................................................................................................................ 3331 331411
Primary Zinc Smelters ............................................................................................................................................. 3339 331419
Primary Lead Smelters ............................................................................................................................................ 3339 331419
TABLE 4.—MAJOR ENTITIES POTENTIALLY AFFECTED BY THIS ACTION FOR AMENDMENTS TO METHOD 24, APPENDIX A,
PART 60
Rubber Tire Manufacturing ...................................................................................................................................... 3011 326211

Flexible Vinyl and Urethane Coating and Printing .................................................................................................. 2754 323111
Magnetic Tape Coating Facilities ............................................................................................................................ 3695 334613
Surface Coating of Plastic Parts for Business Machines ........................................................................................ 3479 326199
Polymetric Coating of Supporting Substrates Facilities .......................................................................................... 2824 332812
Surface Coating of Metal Furniture ......................................................................................................................... 2514 337124
Automobile and Light Duty Truck Surface Coating ................................................................................................. 5012 336111
Graphic Arts Industry: Publication Rotogravure Printing ......................................................................................... 2754 323111
Pressure Sensitive Tape and Label Surface Coating Operations .......................................................................... 2672 322222
Indusrial Surface Coating: Large Appliances .......................................................................................................... 5064 421620
Metal Coil Surface Coating ...................................................................................................................................... 3479 335931
Beverage Can Surface Coating ............................................................................................................................... 3411 332812
Aerospace ................................................................................................................................................................ 3721 33641
Boat and Ship Manufacturing and Repair Surface Coating .................................................................................... 3731, 3732 ........................
Fabric Printing, Coating and Dyeing ....................................................................................................................... 2759 ........................
Leather Finishing ..................................................................................................................................................... 3111 ........................
Miscellaneous Coating Manufacturing ..................................................................................................................... 3479 ........................
Miscellaneous Metal Parts and Products ................................................................................................................ 3479 ........................
Paper and other Web Surface Coating ................................................................................................................... 2741
Plastic Parts Surface Coating .................................................................................................................................. 3479
Printing and Publishing Surface Coating ................................................................................................................. 2741 ........................
Wood Building Products .......................................................................................................................................... 2499 ........................
Wood Furniture ........................................................................................................................................................ 2511, 2521 ........................
These tables are not intended to be documents specifically referenced in OAR–2003–0074, EPA Docket Center,
exhaustive, but rather provides an this action, any public comments (EPA/DC) EPA West, Room B102, 1301
example of entities that may be affected received, and other information related Constitution Ave., NW., Washington,
by this action. If you have any questions to this action. Although a part of the DC 20460; telephone (202) 566–1742.
regarding the applicability of this action official docket, the public docket does The docket facility is open from 8:30
to a particular entity, consult the person not include Confidential Business a.m. to 4:30 p.m., Monday through
listed in the preceding FOR FURTHER Information (CBI) or other information Friday, excluding legal holidays. The
INFORMATION CONTACT section. whose disclosure is restricted by statute. telephone number for the Public
Documents in the official public docket Reading Room is (202) 566–1744.
B. How Can I Get Copies of This
are listed in the index list in EPA’s 2. Electronic Access. You may access
Document and Other Related
Information? electronic public docket and comment this Federal Register document
system, EDOCKET. Documents may be electronically through the EPA Internet
1. Docket. EPA has established an available either electronically or in hard under the ‘‘Federal Register’’ listings at
official public docket for this action copy. Electronic documents may be http://www.epa.gov/fedrgstr/ or you can
under Docket ID No. OAR–2003–0074. viewed through EDOCKET. Hard copy go to the federal wide eRulemaking site
The official public docket consists of the documents may be viewed at Docket at http://www.regulations.gov.
45610 Federal Register / Vol. 70, No. 151 / Monday, August 8, 2005 / Proposed Rules
An electronic version of the public be placed in EPA’s electronic public or other contact information unless you
docket is available through EDOCKET. docket along with a brief description provide it in the body of your comment.
You may use EPA Dockets at http:// written by the docket staff. ii. http://www.regulations.gov.
www.epa.gov/edocket/ to submit or For additional information about Electronic comments may also be sent
view public comments, access the index EPA’s electronic public docket, visit through the federal wide eRulemaking
listing of the contents of the official EDOCKET online or see 67 FR 38102, web site at http://www.regulations.gov.
public docket, and to access those May 31, 2002. iii. E-mail. Comments may be sent by
documents in the public docket that are electronic mail (e-mail) to a-and-r-
C. How and to Whom Do I Submit docket@epamail.gov, Attention: Docket
available electronically. Once in the Comments?
system, select ‘‘search,’’ then key in the ID No. OAR–2003–0074. In contrast to
appropriate docket identification You may submit comments EPA’s electronic public docket, EPA’s e-
number. electronically, by mail, by facsimile, or mail system is not an ‘‘anonymous
Certain types of information will not through hand delivery/courier. To access’’ system. If you send an e-mail
be placed in the EPA Dockets. ensure proper receipt by EPA, identify comment directly to the Docket without
Information claimed as CBI and other the appropriate docket identification going through EPA’s electronic public
information whose disclosure is number in the subject line on the first docket, EPA’s e-mail system
restricted by statute, which is not page of your comment. Please ensure automatically captures your e-mail
included in the official public docket, that your comments are submitted address. E-mail addresses that are
will not be available for public viewing within the specified comment period. automatically captured by EPA’s e-mail
in EPA’s electronic public docket. The Comments received after the close of the system are included as part of the
EPA’s policy is that copyrighted comment period will be marked ‘‘late.’’ comment that is placed in the official
material will not be placed in EPA’s The EPA is not required to consider public docket and made available in
electronic public docket but will be these late comments. However, late EPA’s electronic public docket.
available only in printed, paper form in comments may be considered if time iv. Disk or CD ROM. You may submit
the official public docket. To the extent permits. comments on a disk or CD ROM that
feasible, publicly available docket 1. Electronically. If you submit an you mail to the mailing address
materials will be made available in electronic comment as prescribed identified in Unit I.C.2. These electronic
EPA’s electronic public docket. When a below, EPA recommends that you submissions will be accepted in
document is selected from the index list include your name, mailing address, WordPerfect or ASCII file format. Avoid
in EDOCKET, the system will identify and an e-mail address or other contact the use of special characters and any
whether the document is available for information in the body of your form of encryption.
viewing in EPA’s electronic public comment. Also include this contact 2. By Mail. Send duplicate copies of
docket. Publicly available docket information on the outside of any disk your comments to: ‘‘Performance
materials that are not available or CD ROM you submit and in any cover Specification 16 for Predictive Emission
electronically may be viewed at the letter accompanying the disk or CD Monitoring Systems,’’ Environmental
docket facility identified in Unit I.B. ROM. This ensures that you can be Protection Agency, Mail Code 6102T,
The EPA intends to work towards identified as the submitter of the 1200 Pennsylvania Ave., NW.,
providing electronic access to all of the comment and allows EPA to contact you Washington, DC, 20460, Attention
publicly available docket materials in case EPA cannot read your comment Docket ID No. OAR–2003–0074.
through EPA’s electronic public docket. due to technical difficulties or needs 3. By Hand Delivery or Courier.
For public commenters, it is further information on the substance of Deliver your comments to: EPA Docket
important to note that EPA’s policy is your comment. The EPA’s policy is that Center, EPA West, Room 108, 1301
that public comments, whether EPA will not edit your comment, and Constitution Ave., NW., Washington,
submitted electronically or on paper, any identifying or contact information DC 20460, Attention: Docket ID No.
will be made available for public provided in the body of a comment will OAR–2003–0074. Such deliveries are
viewing in EPA’s electronic public be included as part of the comment that only accepted during the Docket’s
docket as EPA receives them and is placed in the official public docket normal hours of operation as identified
without change, unless the comment and made available in EPA’s electronic in Unit I.B.1.
contains copyrighted material, CBI, or public docket. If EPA cannot read your 4. By Facsimile. Fax your comments
other information whose disclosure is comment due to technical difficulties to: 202–566–1741, Attention: Docket ID.
restricted by statute. When EPA and cannot contact you for clarification, No. OAR–2003–0074.
identifies a comment containing EPA may not be able to consider your
comment. D. How Should I Submit CBI to the
copyrighted material, EPA will provide
i. EDOCKET. Your use of EPA’s Agency?
a reference to that material in the
version of the comment that is placed in electronic public docket to submit Do not submit information that you
EPA’s electronic public docket. The comments to EPA electronically is consider to be CBI electronically
entire printed comment, including the EPA’s preferred method for receiving through EPA’s electronic public docket
copyrighted material, will be available comments. Go directly to EDOCKET at or by e-mail. Send or deliver
in the public docket. http://www.epa.gov/edocket, and follow information identified as CBI only to the
Public comments submitted on the online instructions for submitting docket address to the attention of
computer disks that are mailed or comments. To access EPA’s electronic Docket ID No. OAR–2003–0074. You
delivered to the docket will be public docket from the EPA Internet may claim information that you submit
transferred to EPA’s electronic public Home Page, select ‘‘Information to EPA as CBI by marking any part or
docket. Public comments that are Sources,’’ ‘‘Dockets,’’ and ‘‘EDOCKET.’’ all of that information as CBI (if you
mailed or delivered to the Docket will Once in the system, select ‘‘search,’’ and submit CBI on disk or CD ROM, mark
be scanned and placed in EPA’s then key in Docket ID No. OAR–2003– the outside of the disk or CD ROM as
electronic public docket. Where 0074. The system is an ‘‘anonymous CBI and then identify electronically
practical, physical objects will be access’’ system, which means EPA will within the disk or CD ROM the specific
photographed, and the photograph will not know your identity, e-mail address, information that is CBI). Information so
marked will not be disclosed except in F. Performance Specification 11 of also be required to comply with PS–16
accordance with procedures set forth in Appendix B of Part 60 at the discretion of the applicable
40 CFR Part 2. G. Method 303 of Appendix A of Part 63 regulatory agency or permit writer.
IV. Statutory and Executive Order Reviews
In addition to one complete version of
A. Executive Order 12866: Regulatory C. What Are the Basic Requirements of
the comment that includes any Planning and Review PS–16?
information claimed as CBI, a copy of B. Paperwork Reduction Act
the comment that does not contain the C. Regulatory Flexibility Act The PS–16 requires owners and
information claimed as CBI must be D. Unfunded Mandates Reform Act operators of affected PEMS to: (1) Select
submitted for inclusion in the public E. Executive Order 13132: Federalism a PEMS that satisfies basic design
docket and EPA’s electronic public F. Executive Order 13175: Consultation criteria; (2) verify and document their
docket. If you submit the copy that does and Coordination With Indian Tribal PEMS; (3) validate their PEMS against a
Governments reference method using prescribed
not contain CBI on disk or CD ROM, G. Executive Order 13045: Protection of
mark the outside of the disk or CD ROM statistical procedures prior to placing it
Children From Environmental Health
clearly that it does not contain CBI. Risks and Safety Risks
into operation; and (4) periodically
Information not marked as CBI will be H. Executive Order 13211: Action reassess their PEMS’s performance. The
included in the public docket and EPA’s Concerning Regulations That performance requirements for PS–16
electronic public docket without prior Significantly Affect Energy Supply, follow the general performance
notice. If you have any questions about Distribution, or Use requirements for continuous emission
I. NTTAA: National Technology Transfer monitoring systems (CEMS) in
CBI or the procedures for claiming CBI, and Advancement Act
please consult the person identified in Appendix B of Part 60. A relative
the FOR FURTHER INFORMATION CONTACT I. Background accuracy (RA) test of the PEMS against
section. a reference method is the primary
Today we are proposing Performance assessment of accuracy. The number of
E. What Should I Consider as I Prepare Specification 16 for Predictive Emission runs prescribed for the RA test will
My Comments for EPA? Monitoring Systems to Appendix B, Part depend upon the underlying regulation.
60. Predictive emission monitoring
You may find the following systems are a new and innovative tool D. What Is the Rationale for the
suggestions helpful for preparing your for monitoring pollutant emissions Performance Criteria in PS–16?
comments: without the traditional hardware
1. Explain your views as clearly as The Agency is allowing, but not
analyzers. The PEMS predicts a unit’s requiring, PEMS use in a number of
possible. emissions indirectly using process
2. Describe any assumptions that you recently-promulgated rules, and a
parameters that have a known number of facilities regulated by State
used. relationship to pollutant concentration.
3. Provide any technical information and Local agencies are considering their
Their principle of operation can range use. Past EPA approvals of PEMS were
and/or data you used that support your from a relatively simple relationship
views. based on criteria provided in the draft
based on combustion principles to the performance specifications on the
4. If you estimate potential burden or
more complex computer models that are Agency’s Emission Measurement Center
costs, explain how you arrived at your
trained to predict emissions using website. In other cases, performance
estimate.
neural networks technology. They have specifications developed by State or
5. Provide specific examples to
been used for monitoring purposes at Local Agencies were used to evaluate
illustrate your concerns.
industrial, commercial, and institutional the PEMS. We are proposing PS–16 to
6. Offer alternatives.
7. Make sure to submit your steam-generating units, gas turbines, provide regulatory agencies a uniform
comments by the comment period internal combustion engines, and other procedure for assessing the capabilities
deadline identified. combustion processes where process of this new monitoring tool.
8. To ensure proper receipt by EPA, parameters have a predictable
relationship to emissions. We are also III. Summary of Other Amendments
identify the appropriate docket
identification number in the subject line proposing to make amendments to the A. Petroleum Refinery (Subpart J) NSPS
on the first page of your response. It testing and monitoring provisions of
In the petroleum refinery NSPS in
would also be helpful if you provided various NSPS and MACT rules.
§ 60.106(b)(3) the equation for
the name, date, and Federal Register II. Summary of Proposed Performance determining the coke burnoff rate is
citation related to your comments. Specification 16 being corrected.
Outline. The information presented in
this preamble is organized as follows: A. What Is the Purpose of PS–16? B. Kraft Pulp Mill (Subpart BB) NSPS
I. Background
The purpose of PS–16 is to establish In the monitoring provisions of the
II. Summary of Proposed Performance the initial installation and performance kraft pulp mills NSPS in § 60.284, a
Specification 16 procedures that candidate PEMS must paragraph requiring continuous
A. What Is the Purpose of PS–16? meet to be acceptable for use. The emission monitors be subject to the
B. Who Must Comply With PS–16? specification stipulates equipment quality assurance provisions of
C. What Are the Basic Requirements of PS– design and documentation, location, Appendix F that was added by mistake
16? and addresses initial and periodic in an October 17, 2000 amendment is
D. What Is the Rationale for the performance tests of the PEMS. being deleted.
Performance Criteria in PS–16?
III. Summary of Other Amendments B. Who Must Comply With PS–16? C. Municipal Solid Waste Landfill
A. Petroleum Refinery (Subpart J) NSPS If adopted as a final rule, all PEMS (Subpart WWW) NSPS
B. Kraft Pulp Mill (Subpart BB) NSPS
C. Municipal Solid Waste Landfills
that will be used to comply with 40 CFR Under the municipal solid waste
(Subpart WWW) NSPS Parts 60, 61, and 63 will be required to landfill NSPS in § 60.752, the
D. Method 24 of Appendix A of Part 60 comply with PS–16. In addition to new requirement to test open flares for heat
E. Performance Specification 2 of PEMS that are installed after the content and flare exit velocity using
Appendix B of Part 60 effective date of PS–16, other PEMS may Methods 18 and ASTM D1946 is being
changed to require Method 3C. These and power correlation models, and monitoring tool that may be used in
open flares must comply with the adding a note following paragraph (5)(v) some cases in place of current source
general flare provisions of 40 CFR 60.18, concerning the application of monitoring requirements. These criteria
which require that flare gas heat content correlation equations to calculate PM do not add information collection
and flare exit velocity be within concentrations using the response data requirements beyond those currently
prescribed limits. The heat content of from an operating PM CEMS. We are required under the applicable
flare gas is determined from an analysis also renumbering some equations and regulation. The additional amendments
of its organic compound and hydrogen references for clarification, consistency, being made to the testing requirements
content using Method 18 and ASTM and accuracy. in 40 CFR part 60 do no add information
D1946, respectively. Methane is the collection requirements but make minor
only significant organic compound in G. Method 303 of Appendix A of Part 63
corrections to existing testing
landfill gas and hydrogen is not likely In Method 303 of Appendix A of Part methodology.
to be present. Therefore, Method 18 and 63, we are proposing to add a statement Burden means the total time, effort, or
ASTM D1946 are not practical methods on varying the time of day runs are financial resources expended by persons
for landfill applications. Method 3C is taken that was deleted by mistake in a to generate, maintain, retain, or disclose
less labor-intensive than Method 18 and recent amendment of the method. or provide information to or for a
has the preferred measuring range for Federal agency. This includes the time
methane levels encountered at landfills. IV. Statutory and Executive Order
Reviews needed to review instructions; develop,
In addition, Method 3C determines acquire, install, and utilize technology
oxygen and nitrogen which are A. Executive Order 12866: Regulatory and systems for the purposes of
currently determined by an additional Planning and Reviews collecting, validating, and verifying
method and are needed to calculate the information, processing and
Under Executive Order 12866 (58 FR
flare gas exit velocity. We are proposing maintaining information, and disclosing
51735 October 4, 1993), we must
that Method 3C be required as the test and providing information; adjust the
determine whether this regulatory
method for methane in place of Method existing ways to comply with any
action is ‘‘significant’’ and therefore
18 and ASTM D1946 for organics and previously applicable instructions and
subject to Office of Management and
hydrogen. requirements; train personnel to be able
Budget (OMB) review and the
D. Method 24 of Appendix A of Part 60 requirements of this Executive Order. to respond to a collection of
The Order defines ‘‘significant information; search data sources;
Method 24, Part 60, Appendix A is
regulatory action’’ as one that is likely complete and review the collection of
used to determine the contents and
to result in a rule that may: (1) Have an information; and transmit or otherwise
properties of surface coatings under
annual effect on the economy of $100 disclose the information.
NSPS applications. Method 24 currently
references ASTM D2369 as the method million or more or adversely affects in An agency may not conduct or
for determining volatiles content. The a material way the economy, a sector of sponsor, and a person is not required to
American Society for Testing and the economy, productivity, competition, respond to a collection of information
Materials has recommended that ASTM jobs, the environment, public health or unless it displays a currently valid OMB
D6419 be allowed as an alternative to safety, or State, Local, or Tribal control number. The OMB control
D2369 in this case. We are proposing to governments or communities; (2) create numbers for EPA’s regulations in 40
amend Method 24 to allow this option. a serious inconsistency or otherwise CFR are listed in 40 CFR part 9.
interferes with an action taken or C. Regulatory Flexibility Act
E. Performance Specification 2, Part 60, planned by another agency; (3)
Appendix B materially alter the budgetary impact of The RFA generally requires an agency
In Performance Specification 2, Part entitlements, grants, user fees, or loan to prepare a regulatory flexibility
60, Appendix B, an inadvertent programs, or the rights and obligations analysis of any rule subject to notice
omission in an October 17, 2000 of recipients thereof; or (4) raise novel and comment rulemaking requirements
amendment removed an allowance for legal or policy issues arising out of legal under the Administrative Procedure Act
relative accuracy relief for low-emitters. mandates, the President’s priorities, or or any other statute unless the agency
We are proposing to reinstate the the principles set forth in the Executive certifies that the rule will not have a
allowance. Order. significant economic impact on a
We have determined that this rule is substantial number of small entities.
F. Performance Specification 11 of not a ‘‘significant regulatory action’’ Small entities include small businesses,
Appendix of Appendix B of Part 60 under the terms of Executive Order small organizations, and small
The publication on January 12, 2004 12866 and is therefore not subject to governmental jurisdictions.
of Performance Specification 11 for OMB review. We have determined that For purposes of assessing the impacts
Appendix B and Procedure 2 for Part 60, this regulation would result in none of of today’s rule on small entities, small
Appendix F contained technical and the economic effects set forth in Section entity is defined as: (1) A small business
typographical errors and unclear 1 of the Order because it does not as defined by the Small Business
instructions. We are revising the impose emission measurement Administration’s regulations at 13 CFR
definition of confidence interval half requirements beyond those specified in 121.201; (2) a small governmental
range to clarify the language, replacing the current regulations, nor does it jurisdiction that is a government of a
the word ‘‘pairs’’ with ‘‘sets’’ to avoid change any emission standard. city, county, town, school district or
possible confusion regarding the use of special district with a population of less
paired sampling trains, correcting errors B. Paperwork Reduction Act than 50,000; and (3) a small
in Equations 11–22, 11–27, and 11–37, This action does not impose an organization that is any not-for-profit
correcting the procedures in paragraphs information collection burden under the enterprise which is independently
(4) and (5) of section 12.3 for provisions of the Paperwork Reduction owned and operated and is not
determining confidence and tolerance Act, 44 U.S.C. 3501 et seq. This actions dominant in its field. Entities
interval half ranges for the exponential provides performance criteria for a new potentially affected by this action
include those listed in Table 1 of officials of affected small governments 67249, November 6, 2000), requires EPA
SUPPLEMENTARY INFORMATION. to have meaningful and timely input in to develop an accountable process to
After considering the economic the development of EPA regulatory ensure ‘‘meaningful and timely input by
impacts of today’s proposed rule on proposals with significant Federal tribal officials in the development of
small entities, I certify that this action intergovernmental mandates, and regulatory policies that have tribal
will not have a significant economic informing, educating, and advising implications.’’ ‘‘Policies that have tribal
impact on a substantial number of small small governments on compliance with implications’’ is defined in the
entities. We are allowing, but not the regulatory requirements. Executive Order to include regulations
requiring, PEMS use in a number of Today’s rule contains no Federal that have ‘‘substantial direct effects on
recently-promulgated rules, and a mandates (under the regulatory one or more Indian tribes, on the
number of facilities regulated by State provisions of Title II of the UMRA) for relationship between the Federal
and Local agencies are considering their State, Local, or Tribal governments or government and the Indian tribes, or on
use. The intended effect of this action is the private sector. The rule imposes no the distribution of power and
to facilitate the use of PEMS by enforceable duty on any State, Local, or responsibilities between the Federal
establishing levels of acceptability for Tribal governments or the private sector. government and Indian tribes.’’
candidate PEMS. In addition, we are In any event, EPA has determined that This proposed rule does not have
proposing to make minor amendments this rule does not contain a Federal tribal implications. It will not have
to various testing provisions in the New mandate that may result in expenditures substantial direct effects on tribal
Source Performance Standards (NSPS) of $100 million or more for State, Local, governments, on the relationship
and National Emission Standards for and Tribal governments, in the between the Federal government and
Hazardous Air Pollutants for Source aggregate, or the private sector in any Indian tribes, or on the distribution of
Categories (MACT) to correct one year. Thus, today’s rule is not power and responsibilities between the
inadvertent errors, make needed subject to the requirements of Sections Federal government and Indian tribes,
updates, and add flexibility. We invite 202 and 205 of the UMRA. as specified in Executive Order 13175.
comments on all aspects of the proposal In this proposed rule, we are simply
and its impacts on small entities. E. Executive Order 13132: Federalism
allowing an alternative emission
Executive Order 13132, entitled
D. Unfunded Mandates Reform Act monitoring tool that applicable facilities
‘‘Federalism’’ (64 FR 43255, August 10,
Title II of the Unfunded Mandates may use. Thus, Executive Order 13175
1999), requires EPA to develop an
Reform Act of 1995 (UMRA), Public does not apply to this rule.
accountable process to ensure
Law 104–4, establishes requirements for ‘‘meaningful and timely input by State G. Executive Order 13045: Protection of
Federal agencies to assess the effects of and Local officials in the development Children From Environmental Health
their regulatory actions on State, Local, of regulatory policies that have Risks and Safety Risks
and Tribal governments and the private federalism implications.’’ ‘‘Policies that
sector. Under section 202 of the UMRA, Executive Order 13045 applies to any
have federalism implications’’ are
EPA generally must prepare a written rule that EPA determines (1) is
defined in the Executive Order to
statement, including a cost-benefit ‘‘economically significant’’ as defined
include regulations that have
analysis, for proposed and final rules under Executive Order 12866, and (2)
‘‘substantial direct effects on the States,
with ‘‘Federal mandates’’ that may the environmental health or safety risk
on the relationship between the national
result in expenditures to State, Local, addressed by the rule has a
government and the States, or on the
and Tribal governments, in the disproportionate effect on children. If
distribution of power and
aggregate, or to the private sector, of the regulatory action meets both criteria,
responsibilities among the various
$100 million or more in any one year. the Agency must evaluate the
levels of government.’’
Before promulgating an EPA rule for This rule does not have federalism environmental health or safety effects of
which a written statement is needed, implications. It will not have substantial the planned rule on children and
section 205 of the UMRA generally direct effects on the States, on the explain why the planned regulation is
requires EPA to identify and consider a relationship between the national preferable to other potentially effective
reasonable number of regulatory government and the States, or on the and reasonably feasible alternatives
alternatives and adopt the least costly, distribution of power and considered by the Agency.
most cost-effective or least burdensome responsibilities among the various The EPA interprets Executive Order
alternative that achieves the objectives levels of government, as specified in 13045 as applying only to regulatory
of the rule. The provisions of section Executive Order 13132. Thus, the actions that are based on health or safety
205 do not apply when they are requirements of Section 6 of the risks, such that the analysis required
inconsistent with applicable law. Executive Order do not apply to this under section 5–501 of the Executive
Moreover, section 205 allows EPA to rule. Order has the potential to influence the
adopt an alternative other than the least In the spirit of Executive Order 13132, regulation. This proposed rule is not
costly, most cost-effective or least and consistent with EPA policy to subject to Executive Order 13045
burdensome alternative if the promote communications between EPA because it is not based on health or
Administrator publishes with the final and State and Local governments, EPA safety risks.
rule an explanation why that alternative specifically solicits comment on this H. Executive Order 13211: Actions
was not adopted. Before EPA establishes proposed rule from State and Local Concerning Regulations That
any regulatory requirements that may officials. Significantly Affect Energy Supply,
significantly or uniquely affect small
F. Executive Order 13175: Consultation Distribution, or Use
governments, including tribal
governments, it must have developed and Coordination With Tribal This action is not subject to Executive
under section 203 of the UMRA a small Governments Order 13211, ‘‘Actions Concerning
government agency plan. The plan must Executive Order 13175, entitled Regulations that Significantly Affect
provide for notifying potentially ‘‘Consultation and Coordination with Energy Supply, Distribution, or Use’’ (66
affected small governments, enabling Indian Tribal Governments’’ (65 FR FR 28355, May 22, 2001) because it is
not a significant regulatory action under § 60.284 Monitoring of emissions and 7. In Appendix B, Performance
Executive Order 12866. operations. Specification 11:
* * * * * A. By revising Sections 3.4 and 8.6;
I. NTTAA: National Technology B. By revising paragraphs (1)(ii), (2),
(f) The procedures under § 60.13 shall
Transfer and Advancement Act (4), and (5) of Section 12.3;
be followed for installation, evaluation,
Section 12(d) of the National and operation of the continuous C. By revising paragraph (3)(ii) of
Technology Transfer and Advancement monitoring systems required under this Section 12.4;
Act of 1995 (NTTAA), Public Law 104– section. All continuous monitoring D. By revising (2) and (3) of Section
113 (15 U.S.C. 272), directs us to use systems shall be operated in accordance 13.2;
voluntary consensus standards (VCSs) E. By adding references 16.8 and 16.9
with the applicable procedures under
in our regulatory activities unless to do to Section 16.0; and
Performance Specifications 1, 3, and 5 F. By revising Table 1 in Section 17.0.
so would be inconsistent with of appendix B of this part.
applicable law or otherwise impractical. The revisions and addition read as
* * * * * follows:
Voluntary consensus standards are
technical standards (e.g., materials § 60.752 [Amended] Performance Specification 11—
specifications, test methods, sampling 4. By revising § 60.752(b)(2)(iii)(A) to Specifications and Test Procedures for
procedures, business practices, etc.) that read as follows: Particulate Matter Continuous Emission
are developed or adopted by VCS Monitoring Systems at Stationary
bodies. The NTTAA requires us to § 60.752 Standards for air emissions from Sources
provide Congress, through OMB, municipal solid waste landfills
* * * * *
explanations when we decide not to use * * * * * 3.4 ‘‘Confidence Interval Half Range
available and applicable VCSs. We are (b) * * * (CI)’’ is a statistical term and means one-
not proposing new test methods in this (2) * * * half of the width of the 95 percent
rulemaking but are adding performance (iii) * * * confidence interval around the
requirements for a new monitoring tool (A) An open flare designed and predicted mean PM concentration (y
that can be used as an alternative to operated in accordance with § 60.18, value) calculated at the PM CEMS
what has already been mandated. except that the net heating value of the response value (x value) where the
Therefore, NTTAA does not apply. combusted landfill gas is calculated confidence interval is narrowest.
from the concentration of methane in Procedures for calculating CI are
List of Subjects in 40 CFR Parts 60 and
the landfill gas as measured by Method specified in section 12.3. The CI as a
63
3C. Other organic components, percent of the emission limit value
Environmental protection, Air hydrogen, and carbon monoxide are not (CI%) is calculated at the appropriate
pollution control, New sources, Test measured; PM CEMS response value and must
methods and procedures, Performance * * * * * satisfy the criteria specified in Section
specifications, and Continuous emission
Appendix A [Amended] 13.2 (2).
monitors.
* * * * *
Dated: July 26, 2005. 5. In Appendix A, by adding Section 8.6 How do I conduct my PM CEMS
Stephen L. Johnson, 6.7 to Method 24 to read as follows: correlation test? You must conduct the
Administrator. Method 24—Determination of Volatile correlation test according to the
For the reasons stated in the Matter Content, Water Content, Density, procedure given in paragraphs (1)
preamble, the Environmental Protection Volume Solids, and Weight Solids of through (5) of this section. If you need
Agency proposes to amend title 40, Surface Coatings multiple correlations, you must conduct
chapter I of the Code of Federal testing and collect at least 15 sets of
Regulations as follows: * * * * * reference method and PM CEMS data
6.7 ASTM D 6419–00, Test Method for calculating each separate correlation.
PART 60—STANDARDS OF for Volatile Content of Sheet-Fed and
* * * * *
PERFORMANCE FOR NEW Coldset Web Offset Printing Inks. 12.3 How do I determine my PM
STATIONARY SOURCES * * * * * CEMS correlation?
1. The authority citation for Part 60 Appendix B [Amended] * * * * *
continues to read as follows: (1) How do I evaluate a linear
6. In Appendix B, by adding a correlation for my correlation test data?
Authority: 42 U.S.C. 7401, 7411, 7413, sentence to Section 13.2 of Performance
7414, 7416, 7601, and 7602. * * * * *
Specification 2 to read as follows:
(ii) Calculate the half range of the 95
§ 60.106 [Amended] Performance Specification 2— percent confidence interval (CI) for the
2. By revising the equation in Specifications and Test Procedures for predicted PM concentration (ŷ) at the
§ 60.106(b)(3) to read as follows: SO2 and NOX Continuous Emission mean value of x, using Equation 11–8:
Monitoring Systems in Stationary
§ 60.106 Test methods and procedures.
Sources 1
* * * * * CI = t df , 1− a / 2 ⋅ S L ( Eq. 11-8)
(b) * * * * * * * * n
(3) * * * 13.2 * * * For SO2 emission Where:
RC = K1Qr (%CO2 + %CO) + (K2Qa ¥ standards of 130 to and including 86 CI = the half range of the 95 percent
K3Qr)((%CO / 2) + (%CO2 + %O2)) ng/J (0.30 and 0.20 lb/million Btu), confidence interval for the
inclusive, use 15 percent of the predicted PM concentration at the
* * * * *
applicable standard; below 86 ng/J (0.20 mean x value,
§ 60.284 [Amended] lb/million Btu), use 20 percent of the tdf,1-a/2 = the value for the t statistic
3. By revising § 60.284(f) to read as emission standard. provided in Table 1 for df = (n–2),
EP08AU05.045</MATH>
follows: * * * * * and
SL = the scatter or deviation of ŷ values predicted PM concentration at the SL = as calculated using Equation 11–
about the correlation curve, which mean x value, and 9:
is determined using Equation 11–9: EL = PM emission limit, as described
in section 13.2. k T = u n ’ ⋅ v df ( Eq. 11-12)
n
1
∑ (yˆ i − y i ) Where:
2 (iii) Calculate the half range of the
SL = ( Eq. 11-9)
n − 2 i =1 tolerance interval (TI) for the predicted n’ = the number of test runs (n),
PM concentration (ŷ) at the mean x un, = the tolerance factor for 75 percent
Calculate the confidence interval half
value using Equation 11–11: coverage at 95 percent confidence
range for the predicted PM
concentration (ŷ) at the mean x value as TI = k T ⋅ SL ( Eq. 11-11) provided in Table 1 for df = (n–2),
a percentage of the emission limit (CI%) and
using Equation 11–10: Where: vdf = the value from Table 1 for df =
TI = the half range of the tolerance (n—2).
CI
CI% = ⋅ 100% ( Eq. 11-10) interval for the predicted PM Calculate the half range of the
EL concentration (ŷ) at the mean x tolerance interval for the predicted PM
Where: value, concentration (y) at the mean x value as
CI = the half range of the 95 percent kT = as calculated using Equation 11– a percentage of the emission limit (TI%)
confidence interval for the 12, and using Equation 11–13:
EP08AU05.056</MATH>
TI
TI% = ⋅100% ( Eq. 11-13)
EL
EP08AU05.055</MATH>
Where: To evaluate a polynomial correlation, yˆ = b 0 + b1x + b 2 x 2 ( Eq. 11-16)
TI = the half range of the tolerance follow the procedures described in
interval for the predicted PM paragraphs (2)(i) through (iv) of this Where:
concentration (ŷ) at the mean x section. ŷ = the PM CEMS concentration
EP08AU05.054</MATH>
value, and predicted by the polynomial
(i) Calculate the polynomial
EL = PM emission limit, as described correlation equation, and
correlation equation, which is indicated b0, b1, b2 = the coefficients determined
in section 13.2.
by Equation 11–16, using Equations 11– from the solution to the matrix
* * * * * 17 through 11–22:
(2) How do I evaluate a polynomial equation Ab=B
EP08AU05.053</MATH>
correlation for my correlation test data? Where:
n S1 S0  b 0  S 5  n n n n
A =  S1 S2 ( ) ( )
S3  , b =  b1  , B = S6  ⋅ S 1 = ∑ ( x i ) , S2 = ∑ x 2i , S3 = ∑ x 3i , S 4 = ∑ x i4 ( ) (Eq. 11-17)
17)
EP08AU05.052</MATH>
S2 S3 S 4   b 2  S 7  i =1 i =1 i =1 i =1
  
n n n
S5 = ∑ ( y i ) , S6 = ∑ (x i y i ) , S 7 = ∑ x 2i y i ⋅( )
EP08AU05.051</MATH>
( Eq. 11-18)
i =1 i =1 i =1
Where: n = the number of test runs. Equations 11–19 through 11–21,
EP08AU05.050</MATH>
xi = the PM CEMS response for run i, respectively:
Calculate the polynomial correlation
yi = the reference method PM
curve coefficients (b0, b1, and b2 ) using
concentration for run i, and
EP08AU05.049</MATH>
b0 =
(S5 ⋅ S2 ⋅ S4 + S1 ⋅ S3 ⋅ S7 + S2 ⋅ S6 ⋅ S3 − S7 ⋅ S2 ⋅ S2 − S3 ⋅ S3 ⋅ S5 − S4 ⋅ S6 ⋅ S1 ) ( Eq. 11-19)
det A
EP08AU05.048</MATH>
b1 =
(n ⋅ S6 ⋅ S4 + S5 ⋅ S3 ⋅ S2 + S2 ⋅ S1 ⋅ S7 − S2 ⋅ S6 ⋅ S2 − S7 ⋅ S3 ⋅ n − S4 ⋅ S1 ⋅ S5 ) (Eq. 11- 20)
det A
(n ⋅ S2 ⋅ S7 + S1 ⋅ S6 ⋅ S2 + S5 ⋅ S1 ⋅ S3 − S2 ⋅ S2 ⋅ S5 − S3 ⋅ S6 ⋅ n − S7 ⋅ S1 ⋅ S1 )
EP08AU05.047</MATH>
b2 = (Eq. 11- 21)

det A
EP08AU05.046</MATH>
Where:
EP08AU05.071</MATH>
det A = n ⋅ S2 ⋅ S 4 − S2 ⋅ S2 ⋅ S2 + S1 ⋅ S3 ⋅ S2 − S3 ⋅ S3 ⋅ n + S2 ⋅ S1 ⋅ S3 − S 4 ⋅ S1 ⋅ S1 (Eq.11-22)
(ii) Calculate the 95 percent first calculating the C coefficients (C0 to
EP08AU05.070</MATH>
confidence interval half range (CI) by C5) using Equations 11–23 and 11–24:
C0 =
(S 2 ⋅ S4 − S32 ) , C = (S3 ⋅ S2 − S1 ⋅ S4 )
, C2 =
(S ⋅ S
1 3 − S22 ) , C = (nS 4 − S22 ) , C = (S ⋅ S
1 2 − nS3 )
, C5 =
(nS 2 − S12 ) (Eq. 11-23)
1 3 4
EP08AU05.069</MATH>
D D D D D D
Where:
EP08AU05.068</MATH>
( )
D = n S2 ⋅ S 4 − S32 + S1 (S3 ⋅ S2 − S1 ⋅ S 4 ) + S2 S1 ⋅ S3 − S22 ( ) (Eq. 11-24)
Calculate D using Equation 11–25 for
EP08AU05.067</MATH>
each x value:
∆ = C 0 + 2 C1 x + ( 2 C 2 + C 3 )x 2 + 2 C 4 x 3 + C 5 x 3 (Eq. 11-25)
EP08AU05.066</MATH>
Determine the x value that Equation 11–29 for the polynomial SP = as calculated using Equation 11–
corresponds to the minimum value D correlation, using Equations 11–30 and 26, and
(Dmin). Determine the scatter or deviation 11–31: Sy = as calculated using Equation 11–
EP08AU05.065</MATH>
of ŷ values about the polynomial 15.
correlation curve (SP) using Equation TI = k T SP ( Eq. 11-29 * * * * *
11–26: Where: (4) How do I evaluate an exponential
correlation for my correlation test data?
1 n k T = u n ’ ⋅ v df ( Eq. 11-30) To evaluate an exponential correlation,
∑ (yˆ i − y i )
2
SP =
EP08AU05.064</MATH>
( Eq. 11-26) which has the form indicated by
n − 3 i =1
Equation 11–37, follow the procedures
Calculate the half range of the 95 1 described in paragraphs (4)(i) through
percent confidence interval (CI) for the n’= ( Eq. 11-31) (v) of this section:
predicted PM concentration (ŷ) at the x ∆ min
EP08AU05.063</MATH>
value that corresponds to Dmin using un, = the value indicated in Table 1 for yˆ = b 0 e b1
x
( Eq. 11-37)
Equation 11–27: df = (n’ ¥ 3), and
vdf = the value indicated in Table 1 for (i) Perform a logarithmic
CI = t df ⋅ Sp ∆ min ( Eq. 11-27) df = (n’ ¥ 3). transformation of each PM
concentration measurement (y values)
EP08AU05.062</MATH>
Where: Calculate the tolerance interval half using Equation 11–38:
df = (n ¥ 3), and range for the predicted PM
concentration at the x value that
tdf = as listed in Table 1 (see section
corresponds to Dmin as a percentage of y ’i = Ln( y i ) ( Eq. 11-38)
17).
the emission limit (TI%) using Equation Where:
Calculate the half range of the 95
EP08AU05.061</MATH>
11–32: y’i = is the transformed value of yi, and
percent confidence interval for the
predicted PM concentration at the x Ln(yi) = the natural logarithm of the
TI
value that corresponds to Dmin as a TI% = ⋅100% ( Eq. 11-32) PM concentration measurement for
percentage of the emission limit (CI%) EL run i.
Where:
EP08AU05.060</MATH>
using Equation 11–28: (ii) Using the values for y’i in place of
TI = the tolerance interval half range the values for yi, perform the same
CI for the predicted PM concentration procedures used to develop the linear
CI% = ⋅ 100% ( Eq. 11-28) correlation equation described in
EL at the x value that corresponds to
Dmin, and paragraph (1)(i) of this section. The
Where:
EP08AU05.059</MATH>
EL = PM emission limit, as described resulting equation will have the form

CI = the half range of the 95 percent indicated by Equation 11–39.
in section 13.2.
confidence interval for the
predicted PM concentration at the x (iv) Calculate the polynomial
correlation coefficient (r) using Equation yˆ ’= b ’0 + b1x ( Eq. 11-39)
value that corresponds to Dmin, and
EL = PM emission limit, as described 11–33: Where:
EP08AU05.058</MATH>
in section 13.2. ŷ’ = the predicted log PM concentration

(iii) Calculate the tolerance interval S2P value,
half range (TI) for the predicted PM r = 1− ( Eq. 11-33) b’0 = the natural logarithm of b0, and the
S2y
concentration at the x value that variables b0, b1, and x are as defined
EP08AU05.057</MATH>
corresponds to Dmin, as indicated in Where: in paragraph (1)(i) of this section.
(iii) Using the values for y’i in place UTL’ = y’+ TI’ ( Eq. 11- 44) paragraph (1)(i) of this section. The
of the values for yi, calculate the half Where: resulting equation will have the form
range of the 95 percent confidence indicated by Equation 11–47:
LTL’ = the lower 95 percent tolerance
interval (CI’), as described in paragraph
limit for the mean value ȳ’,
(1)(ii) of this section for CI. Note that CI’ UTL’ = the upper 95 percent tolerance yˆ ’= b ’0 + b1x ’ ( Eq. 11- 47)
is on the log scale. Next, calculate the limit for the mean value ȳ’,
upper and lower 95 percent confidence Where:
ȳ’ = the mean value of the log-
limits for the mean value ȳ’ using transformed PM concentrations, ŷ’ = the predicted log PM concentration
Equations 11–40 and 11–41: and value, and
TI’ = the half range of the 95 percent x’ = the natural logarithm of the PM
LCL’= y’ − CI’ (Eq. 11-40) tolerance interval for the predicted CEMS response values,
PM concentration (ŷ’), as calculated b’0 = the natural logarithm of b0, and the
in Equation 11–11. variables b0, b1, and x are as defined
UCL’= y’+ CI’ (Eq. 11-41) Calculate the half range tolerance in paragraph (1)(i) of this section.
Where: interval (TI) on the original PM
concentration scale using Equation 11– (iii) Using the same procedure
LCL’ = the lower 95 percent confidence described for exponential models in
limit for the mean value ȳ’, 45:
paragraph (4)(iii) of this section,
UCL = the upper 95 percent confidence calculate the half range of the 95 percent
limit for the mean value ȳ’, e UTL’ - e LTL’
TI = ( Eq. 11- 45) confidence interval for the predicted PM
ȳ’ = the mean value of the log- 2 concentration corresponding to the
transformed PM concentrations, TI = the half range of the 95 percent mean value of x’ as a percentage of the
and tolerance interval on the original emission limit.
CI’ = the half range of the 95 percent PM scale, and UTL’ and LTL’ are as (iv) Using the same procedure
confidence interval for the defined previously. described for exponential models in
predicted PM concentration (ŷ’), as Calculate the tolerance interval half paragraph (4)(iv) of this section,
calculated in Equation 11–8. range for the predicted PM calculate the tolerance interval half
Calculate the half range of the 95 concentration corresponding to the range for the predicted PM
EP08AU05.080</MATH>
percent confidence interval (CI) on the mean value of x as a percentage of the concentration corresponding to the
original PM concentration scale using emission limit (TI%) using Equation 11– mean value of x’ as a percentage of the
Equation 11–42: 13. emission limit.
(v) Using the values for y’i in place of
(v) Using the values for y’i in place of
e UCL’ − e LCL’ the values for yi, calculate the
CI = (Eq. 11-42) the values for yi, calculate the
EP08AU05.079</MATH>
correlation coefficient (r) using the
2 correlation coefficient (r) using the
procedure described in paragraph (1)(iv)
Where: procedure described in paragraph (1)(iv)
of this section.
CI = the half range of the 95 percent of this section.
(5) How do I evaluate a power
confidence interval on the original correlation for my correlation test data? Note: PS–11 does not address the
EP08AU05.078</MATH>
PM concentration scale, and UCL’ To evaluate a power correlation, which application of correlation equations to
and LCL’ are as defined previously. has the form indicated by Equation 11– calculate PM emission concentrations using
46, follow the procedures described in PM CEMS response data during normal
Calculate the half range of the 95 operations of a PM CEMS. However, we will
percent confidence interval for the paragraphs (5)(i) through (v) of this
provide guidance on the use of specific
predicted PM concentration section. correlation models (i.e., logarithmic,
EP08AU05.077</MATH>
corresponding to the mean value of x as exponential, and power models) to calculate
a percentage of the emission limit (CI%) yˆ = b 0 x b1 ( Eq. 11- 46) PM concentrations in an operating PM CEMS
using Equation 11–10. (i) Perform logarithmic in situations when the PM CEMS response
(iv) Using the values for y’i in place values are equal to or less than zero, and the
transformations of each PM CEMS
of the values for yi, calculate the half correlation model is undefined.
EP08AU05.076</MATH>
response (x values) and each PM
range tolerance interval (TI’), as concentration measurement (y values) * * * * *
described in paragraph (1)(iii) of this using Equations 11–35 and 11–38, 12.4 What correlation model should
section for TI. Note that TI’ is on the log respectively. I use?
scale. Next, calculate the half range (ii) Using the values for x’i in place of
tolerance limits for the mean value ȳ’ * * * * *
EP08AU05.075</MATH>
the values for xi, and the values for y’i
using Equations 11–43 and 11–44: in place of the values for yi, perform the (3) * * *
same procedures used to develop the (ii) Calculate the minimum value
LTL’ = y’ − TI’ ( Eq. 11-43) linear correlation equation described in using Equation 11–48.
EP08AU05.074</MATH>
b1
maximum or minimum = - ( Eq. 11- 48)
2b 2
EP08AU05.073</MATH>
* * * * * (2) The confidence interval half range (i) For linear or logarithmic
13.2 What performance criteria must must satisfy the applicable criterion correlations, the 95 percent confidence
my PM CEMS correlation satisfy? specified in paragraph (2)(i), (ii), or (iii) interval half range at the mean PM
of this section, based on the type of CEMS response value from the
* * * * *
correlation model. correlation test must be within 10
EP08AU05.072</MATH>
percent of the PM emission limit value specified in paragraph (3)(i), (ii), or (iii) 75 percent coverage at the mean of the
specified in the applicable regulation. of this section, based on the type of logarithm of the PM CEMS response
Therefore, the CI% calculated using correlation model. values from the correlation test must be
Equation 11–10 must be less than or (i) For linear or logarithmic within 25 percent of the PM emission
equal to 10 percent. correlations, the half range tolerance limit value specified in the applicable
(ii) For polynomial correlations, the interval with 95 percent confidence and regulation. Therefore, the TI%
95 percent confidence interval half 75 percent coverage at the mean PM calculated using Equation 11–13 must
range at the PM CEMS response value CEMS response value from the be less than or equal to 25 percent.
from the correlation test that correlation test must be within 25
* * * * *
corresponds to the minimum value for percent of the PM emission limit value
D must be within 10 percent of the PM specified in the applicable regulation. 16.0 Which references are relevant
emission limit value specified in the Therefore, the TI% calculated using to this performance specification?
applicable regulation. Therefore, the Equation 11–13 must be less than or * * * * *
CI% calculated using Equation 11–28 equal to 25 percent. 16.8 Snedecor, George W. and
must be less than or equal to 10 percent. (ii) For polynomial correlations, the
(iii) For exponential or power Cochran, William G. (1989), Statistical
half range tolerance interval with 95
correlations, the 95 percent confidence Methods, Eighth Edition, Iowa State
percent confidence and 75 percent
interval half range at the mean of the University Press.
coverage at the PM CEMS response
logarithm of the PM CEMS response value from the correlation test that 16.9 Wallis, W.A. (1951) ‘‘Tolerance
values from the correlation test must be corresponds to the minimum value for Intervals for Linear Regression,’’ in
within 10 percent of the PM emission D must be within 25 percent of the PM Second Berkeley Symposium on
limit value specified in the applicable emission limit value specified in the Mathematical Statistics and Probability,
regulation. Therefore, the CI% applicable regulation. Therefore, the ed. J. Neyman, Berkeley: University of
calculated using Equation 11–10 must TI% calculated using Equation 11–32 California Press, pp. 43–51.
be less than or equal to 10 percent. must be less than or equal to 25 percent. 17.0 What Reference Tables and
* * * * * (iii) For exponential or power
Validation Data Are Relevant to PS–11?
(3) The tolerance interval half range correlations, the half range tolerance
must satisfy the applicable criterion interval with 95 percent confidence and * * * * *
TABLE 1.—FACTORS FOR CALCULATION OF CONFIDENCE AND TOLERANCE INTERVAL HALF RANGES
Tolerance interval with 75% coverage and 95%
Student’s t, confidence level
df t df
V df (95%) u n, (75%) kT
3 ....................................................................................................................... 3.182 2.920 1.266 3.697

4 ....................................................................................................................... 2.776 2.372 1.247 2.958
5 ....................................................................................................................... 2.571 2.089 1.233 2.576
6 ....................................................................................................................... 2.447 1.915 1.223 2.342
7 ....................................................................................................................... 2.365 1.797 1.214 2.183
8 ....................................................................................................................... 2.306 1.711 1.208 2.067
9 ....................................................................................................................... 2.262 1.645 1.203 1.979
10 ..................................................................................................................... 2.228 1.593 1.198 1.909
11 ..................................................................................................................... 2.201 1.551 1.195 1.853
12 ..................................................................................................................... 2.179 1.515 1.192 1.806
13 ..................................................................................................................... 2.160 1.485 1.189 1.766
14 ..................................................................................................................... 2.145 1.460 1.186 1.732
15 ..................................................................................................................... 2.131 1.437 1.184 1.702
16 ..................................................................................................................... 2.120 1.418 1.182 1.676
17 ..................................................................................................................... 2.110 1.400 1.181 1.653
18 ..................................................................................................................... 2.101 1.384 1.179 1.633
19 ..................................................................................................................... 2.093 1.370 1.178 1.614
20 ..................................................................................................................... 2.086 1.358 1.177 1.597
21 ..................................................................................................................... 2.080 1.346 1.175 1.582
22 ..................................................................................................................... 2.074 1.335 1.174 1.568
23 ..................................................................................................................... 2.069 1.326 1.173 1.555
24 ..................................................................................................................... 2.064 1.316 1.172 1.544
25 ..................................................................................................................... 2.060 1.308 1.172 1.533
26 ..................................................................................................................... 2.056 1.300 1.171 1.522
27 ..................................................................................................................... 2.052 1.293 1.170 1.513
28 ..................................................................................................................... 2.048 1.286 1.170 1.504
29 ..................................................................................................................... 2.045 1.280 1.169 1.496
30 ..................................................................................................................... 2.042 1.274 1.168 1.488
31 ..................................................................................................................... 2.040 1.268 1.168 1.481
32 ..................................................................................................................... 2.037 1.263 1.167 1.474
33 ..................................................................................................................... 2.035 1.258 1.167 1.467
34 ..................................................................................................................... 2.032 1.253 1.166 1.461
35 ..................................................................................................................... 2.030 1.248 1.166 1.455
36 ..................................................................................................................... 2.028 1.244 1.165 1.450
37 ..................................................................................................................... 2.026 1.240 1.165 1.444
38 ..................................................................................................................... 2.024 1.236 1.165 1.439
39 ..................................................................................................................... 2.023 1.232 1.164 1.435
40 ..................................................................................................................... 2.021 1.228 1.164 1.430
TABLE 1.—FACTORS FOR CALCULATION OF CONFIDENCE AND TOLERANCE INTERVAL HALF RANGES—Continued
Tolerance interval with 75% coverage and 95%
Student’s t, confidence level
df t df
V df (95%) u n, (75%) kT
41 ..................................................................................................................... 2.020 1.225 1.164 1.425

42 ..................................................................................................................... 2.018 1.222 1.163 1.421
43 ..................................................................................................................... 2.017 1.218 1.163 1.417
44 ..................................................................................................................... 2.015 1.215 1.163 1.413
45 ..................................................................................................................... 2.014 1.212 1.163 1.410
46 ..................................................................................................................... 2.013 1.210 1.162 1.406
47 ..................................................................................................................... 2.012 1.207 1.162 1.403
48 ..................................................................................................................... 2.011 1.204 1.162 1.399
49 ..................................................................................................................... 2.010 1.202 1.162 1.396
50 ..................................................................................................................... 2.009 1.199 1.161 1.393
51 ..................................................................................................................... 2.008 1.197 1.161 1.390
52 ..................................................................................................................... 2.007 1.195 1.161 1.387
53 ..................................................................................................................... 2.006 1.192 1.161 1.384
54 ..................................................................................................................... 2.005 1.190 1.161 1.381
55 ..................................................................................................................... 2.004 1.188 1.160 1.379
56 ..................................................................................................................... 2.003 1.186 1.160 1.376
57 ..................................................................................................................... 2.002 1.184 1.160 1.374
58 ..................................................................................................................... 2.002 1.182 1.160 1.371
59 ..................................................................................................................... 2.001 1.180 1.160 1.369
60 ..................................................................................................................... 2.000 1.179 1.160 1.367
References 16.8 (t values) and 16.9 (v df and u n, (values).
* * * * * conducted to ensure the PEMS is maintenance must perform a minimum

8. In Appendix B, by adding operating properly. An ongoing sensor 9-run, 3-level (3 runs at each level) RA
Performance Specification 16 to read as evaluation procedure must be in place test (see Section 8.2).
follows: before the PEMS certification is 2.1.2 Compliance and Market
Appendix B—Performance complete. The amount of testing and Trading PEMS. PEMS that are used for
Specifications data validation we require depends continual compliance standards or in a
upon the regulatory needs, i.e., whether market trading program must perform a
* * * * * precise quantification of emissions will minimum 27-run, 3-level (9 runs at each
Performance Specification 16— be needed or whether indication of level) comparison test against the
Specifications and Test Procedures for exceedances of some regulatory reference method (RM) (see Section
Predictive Emission Monitoring threshold will suffice. Performance 8.1.6). The data are evaluated for bias
Systems in Stationary Sources criteria are more rigorous for PEMSs and by F-test and correlation analysis.
that are used in market-based programs 2.2 Periodic Quality Assurance (QA)
1.0 Scope and Application and for determining continual Assessments. All PEMSs are required to
1.1 Does this performance compliance with an emission limit than conduct quarterly relative accuracy
specification apply to me? If you, the those used to measure excess emissions audits (RAA) and yearly relative
source owner or operator, intend to use or indicate control device operation and accuracy test audits (RATA) to assess
a predictive emission monitoring system maintenance (O&M). You must perform ongoing PEMS operation.
(PEMS) to show compliance with your the initial certification test on your
3.0 Definitions
emission limitation(s), you must use the PEMS before reporting any PEMS data
procedures in this performance as quality-assured. The following definitions apply:
specification (PS) to determine whether 1.1.2 Is other testing required after 3.1 Centroidal Area means that area
your PEMS has acceptable performance. certification? After you initially certify in the center of the stack (or duct)
Use these procedures to certify your your PEMS, you must pass additional comprising no more than 1 percent of
PEMS after initial installation and periodic performance checks to ensure the stack cross-sectional area and having
periodically thereafter to ensure the the long-term quality of data. These the same geometric shape as the stack.
PEMS is operating properly. Additional periodic checks are listed in the table in 3.2 Data Recorder means the
tests may be required by an applicable Section 9. You are always responsible equipment that provides a permanent
regulation or by us, the reviewing for maintaining and operating your record of the PEMS output. The data
authority. If your PEMS contains a PEMS properly. recorder may include automatic data
diluent (O2 or CO2) measuring reduction capabilities and may include
component, this must be tested as well. 2.0 Summary of Performance electronic data records, paper records,
1.1.1 How do I certify my PEMS Specification or a combination of electronic data and
after it is installed? We require that a The following performance tests are paper records.
relative accuracy (RA) test and required in addition to equipment and 3.3 Defective sensor means a sensor
accompanying statistical tests be passed measurement location requirements. that is responsible for PEMS
in the initial certification test before 2.1 Initial PEMS Certification. malfunction or that operates outside the
your PEMS is acceptable for use in 2.1.1 Operation and Maintenance approved operating envelope.
demonstrating compliance with PEMS. PEMS that are used for excess 3.4 Diluent PEMS means the total
applicable requirements. Ongoing emission reporting and as indicators of equipment required to predict a diluent
quality assurance tests must be control device operation and gas concentration.
3.5 Operating envelope means the while the PEMS is operating at the inputs encountered during the
defined range of a parameter input that normal operating level. The RATA must certification test until the PEMS has a
is established during PEMS not be conducted in consecutive new operating envelope established.
development. Emission data generated quarters. 6.1.3 Source-Specific Operating
from parameter inputs that are outside 3.11 Reference Value means a PEMS Conditions. Identify any source-specific
the operating envelope are not baseline value established by RM testing operating conditions, such as fuel type,
considered quality assured and are under conditions when all sensors are that will affect the output of your PEMS.
therefore unacceptable. functioning properly. You may only use your PEMS under the
3.6 PEMS means all of the 3.12 Sensor Evaluation System
source-specific operating conditions it
equipment required to predict an means the equipment or procedure used
was certified for.
emission concentration or emission rate. to periodically assess the quality of
The system may consist of any of the sensor input data. This system may be 6.1.4 Ambient Conditions. You must
following major subsystems: sensors a sub-model that periodically cross- explain whether and how ambient
and sensor interfaces, emission model, checks sensor inputs against other conditions and seasonal changes affect
algorithm, or equation that uses process inputs or any other procedure that your PEMS. Some parameters such as
data to generate an output that is checks sensor integrity at least daily. absolute ambient humidity cannot be
proportional to the emission 3.13 Sensors and Sensor Interface manipulated during a test. The effect of
concentration or emission rate, diluent means the equipment that measures the ambient conditions such as humidity on
emission model, data recorder, and process input signals and transports the pollutant concentration must be
sensor evaluation system. Simple them to the emission prediction system. determined and this effect extrapolated
relationships that use fewer than 3 to include future anticipated conditions.
4.0 Interferences [Reserved] Seasonal changes and their effects on
variables may not be acceptable as
PEMS, and such systems must have the 5.0 Safety [Reserved] the PEMS must be evaluated unless you
Administrator’s approval before use. A 6.0 Equipment and Supplies can show that such effects are
PEMS may or may not predict emissions negligible.
data that are corrected for diluent. 6.1 PEMS Design. You must define
6.1.5 PEMS Principle of Operation.
3.7 Reconciled Process Data means and make available details on the design
If your PEMS is developed on the basis
substitute data that are generated by a of your PEMS. You must also establish
of known physical principles, you must
sensor evaluation system to replace that the following, as applicable:
6.1.1 Number of Input Parameters. identify the specific physical
of a failed sensor. assumptions or mathematical
3.8 Relative Accuracy means the An acceptable PEMS will normally use
three or more input parameters. You manipulations that support its
accuracy of the PEMS when compared operation. If your PEMS is developed on
to a RM at the source. The RA is the must obtain our permission on a case-
by-case basis to use a PEMS having the basis of linear or nonlinear
average difference between the pollutant
fewer than three input parameters. regression analysis, you must make
PEMS and RM data for a specified
6.1.2 Parameter Operating available the paired data (preferably in
number of runs plus a 2.5 percent
Envelopes. Before you evaluate your graphic form) used to develop or train
confidence coefficient, divided by the
PEMS through the certification test, you the model.
average of the RM tests or the emission
standard. For diluent PEMS, the RA must specify the input parameters your 6.1.6 Data Recorder Scale. If you are
may be expressed as a percentage PEMS uses, define their range of not using a digital recorder, you must
absolute difference between the PEMS minimum and maximum values choose a recorder scale that accurately
and RM. Alternative specifications may (operating envelope), and demonstrate captures the desired range of potential
be given for low-emitting units. the integrity of the parameter operating emissions. The lower limit of your data
3.9 Relative Accuracy Audit means a envelopes using graphs and data from recorder’s range must be no greater than
quarterly audit of the PEMS against a the PEMS development process. After 20 percent of the applicable emission
portable analyzer meeting the the certification test, the PEMS must be standard (if subject to an emission
requirements of ASTM D6522–00 or RM operated within these envelopes at all standard). The upper limit of your data
for a specified number of runs. times for the system to be acceptable. If recorder’s range must be determined
3.10 Relative Accuracy Test Audit these operating envelopes are not using the following table. If you obtain
means a RA test that is performed at clearly defined, the PEMS operation approval first, you may use other lower
least once every four calendar quarters will be limited to the range of parameter and upper recorder limits.
If PEMS is measuring . . . And if . . . Then your upper limit . . .
Uncontrolled emissions, such as NOX at the No regulation says otherwise .......................... Must be 1.25 to 2 times the average potential
stack of a natural gas-fired boiler. emission level.
Uncontrolled emissions, such as NOX at the A regulation says otherwise ............................. Must follow the other regulation.
stack of a natural gas-fired boiler.
Controlled emissions .......................................... .......................................................................... Must be 1.5 to 2.0 times the concentration of
the emission standard that applies to your
emission unit.
Continual compliance emissions for an applica- .......................................................................... Must be 1.1 to 1.5 times the concentration of
ble regulation. the emission standard that applies to your
emission unit.
6.1.7 Sensor Location and Repair. repairs and replacements. Permanently area which is not accessible, you may be
We recommend you install sensors in an installed platforms or ladders may not required to shut down the emissions
accessible location in order to perform be needed. If you install sensors in an unit to repair or replace a sensor. If
necessary after repairing or replacing a You may choose to perform more than two streams with different pollutant
sensor, correct the process data to match nine RM tests. If you perform more than concentrations are combined.
the data obtained from the originally nine tests, you may reject a maximum 4. You may select a different traverse
tested sensor, or conduct another RA of three tests as long as the total number point if you demonstrate and provide
test. All sensors must be calibrated as of test results used to determine the RA verification that it provides a
often as needed but in no event less is greater than or equal to nine and each representative sample.
often than recommended by the operating level has at least three tests. If you desire to test at only one
manufacturers be exceeded. You must report all data, including the traverse point, use the following
6.1.8 Sensor Evaluation System. rejected data. procedure, or provide supporting
Your PEMS must be designed to 8.2.3 Number of RM Tests for information for alternative procedures,
perform automatic or manual Continual Compliance and Market- to show that the single point yields
determination of defective sensors on at Trading PEMS. Conduct at least 27 RM representative results.
least a daily basis. This sensor tests at the following key parameter 1. Use Method 1 to establish the
evaluation system may consist of a operating levels: number and location of traverse points
sensor validation sub-model, a • Nine at a low level. that are normally used to sample the
comparison of redundant sensors, a spot • Nine at the normal operating level. stack or duct.
check of sensor input readings at a • Nine at a high level. 2. Following the RM procedures,
reference value, operation, or emission You may choose to perform more than measure emissions at each traverse
level, or other procedure that detects 9 RM runs at each operating level. If you point for a period of two minutes plus
faulty or failed sensors. Some sensor perform more than 9 runs, you may twice the response time of the RM.
evaluation systems generate substitute 3. Determine the average of the
reject a maximum of three runs as long
values (reconciled data) that are used emissions from all traverse points.
as the total number of runs used to
when a sensor is perceived to have 4. Choose the traverse point with
determine the RA at each operating
failed. You must have our prior emissions closest to the average
level is greater than or equal to 9.
approval before you use reconciled data. emissions from all points as the
8.2.4 Reference Method
6.1.9 Parameter Envelope sampling location for the RM tests.
Measurement Location. Select an 5. You may select a different traverse
Exceedances. Your PEMS must include accessible measurement point for the point if you can show that it provides
a plan to detect and notify the operator RM that will ensure that you measure a representative sample.
of parameter envelope exceedances. emissions representatively. Ensure the 8.2.6 Relative Accuracy Procedure.
Emission data collected outside any of location is at least two equivalent stack Perform the number of RA tests at the
the operating ranges will not be diameters downstream and a half levels required in Sections 8.2.2 and
considered quality assured. equivalent diameter upstream from the 8.2.3. For integrated samples, e.g.,
6.2 Recordkeeping. All valid data nearest flow disturbance such as the Method 3A or 7E, make a sample
recorded by the PEMS must be used to control device, point of pollutant traverse of at least 21 minutes, sampling
calculate the emission value. For a valid generation, or other place where the for 7 minutes at each traverse point. For
hourly average emission value, each 15- pollutant concentration or emission rate grab samples, e.g., Method 3 or 7, take
minute quadrant of the hour in which can change. You may use a half one sample at each traverse point,
the unit combusts any fuel must contain diameter downstream instead of the two scheduling the grab samples so that they
at least one valid emission value. diameters if you meet both of the are taken simultaneously (within a 3-
7.0 Reagents and Standards [Reserved] following conditions: minute period) or at an equal interval of
• Changes in the pollutant time apart over a 21-minute (or less)
8.0 Sample Collection, Preservation, concentration are caused solely by
Storage, and Transport period. A test run for grab samples must
diluent leakage, such as leaks from air be made up of at least three separate
8.1 Initial Certification. Use the heaters. measurements.
following procedure to certify your • You measure pollutants and Where multiple fuels are used in the
PEMS. Complete all PEMS training diluents simultaneously at the same monitored unit and the fuel type affects
before the certification. location. the predicted emissions, determine a RA
8.2 Relative Accuracy Test. 8.2.5 Traverse Points. Select traverse for each fuel unless the effects of the
8.2.1 Reference Methods. Unless points that ensure you obtain alternative fuel on predicted emissions
otherwise specified in the applicable representative samples. Conduct all RM or diluent were addressed in the model
regulations, you must use the test tests within 3 cm of each selected training process. You may only use fuels
methods in Appendix A of this part for traverse point but no closer than 3 cm in your unit that have been evaluated
the RM test. Conduct the RM tests at to the stack or duct wall. The minimum this way.
three operating levels of the key requirements for selecting traverse 8.2.4 Correlation of RM and PEMS
parameter that affects emissions, e.g., points are as follows: Data. Mark the beginning and end of
load level. Conduct the specified 1. Establish a measurement line across each RM test run (including the exact
number of RM tests at the low the stack that passes through the center time of day) on the permanent record of
(minimum to 50 percent of maximum), and in the direction of any expected PEMS output. Correlate the PEMS and
normal, and high (80 percent to stratification. the RM test data as to the time and
maximum) operating levels as 2. Locate a minimum of three traverse duration using the following steps:
practicable. points on the line at 16.7, 50.0, and 83.3 A. Determine the integrated pollutant
8.2.2 Number of RM Tests for O&M percent of the stack inside diameter. concentration for the PEMS for each
PEMS. Conduct at least nine RM tests at 3. If the stack inside diameter is corresponding RM test period.
the following key parameter operating greater than 2.4 meters, you may locate B. Consider system response time, if
levels: the three traverse points on the line at important, and confirm that the pair of
• Three at a low level. 0.4, 1.2, and 2.0 meters from the stack results are on a consistent moisture,
• Three at the normal level. or duct wall. You cannot use this option temperature, and diluent concentration
• Three at a high level. after wet scrubbers or at points where basis.
C. Compare each average PEMS value Use the following guidelines to make
to the corresponding average RM value. these comparisons.
If . . . Then . . . And then . . .
The RM has an instrumental or integrated non- Directly compare RM and PEMS results..
instrumental sampling technique.
The RM has a grab sampling technique ............ Average the results from all grab samples Compare this average RM result with the
taken during the test run. The test run must PEMS result obtained during the run.
include ≥ 3 separate grab measurements.
8.2.5 Relative Accuracy for O&M waived by the Administrator if the occur that result in a significant change
PEMS. Use the paired PEMS and RM emission concentration is less than 50 in the emission rate (e.g., turbine aging,
data and the equations in Section 12.2 percent of the applicable emission process modification, new process
to calculate the RA in the units of the standard. Requests for waiver must be operating modes, or changes to emission
applicable emission standard. For this accompanied by RM documentation of controls), your PEMS must be recertified
3-level RA test, calculate the RA at each the emission concentration. The waiver using the tests and procedures in
operation level. will be based on the measured value at Section 8.1. For example, if you initially
8.3 Statistical Tests for PEMS that the time of the waiver. Should a operated the emissions unit at 80–100
are Used for Continual Compliance or subsequent RATA identify a change in percent of its range, you would have
Market-Trading. In addition to the RA the RM measured value by more than 30 performed the initial test under these
determination, evaluate the paired RA percent, the correlation analysis test conditions. Later, if you wanted to
and PEMS data using the following must be repeated at the next RATA. operate the emission unit at 50–100
statistical tests. 8.3.4 Additional Statistical Tests. percent of its range, you must conduct
8.3.1 Bias Test. From the RA data Consult the reviewing authority with another RA test and statistical tests, as
taken at the normal operating level, jurisdiction over your emissions unit for applicable, under the new conditions of
determine if a bias exists between the additional requirements. 50–100 percent of range. These tests
RM and PEMS. Use the equations in 8.4 Reporting. Summarize in tabular must demonstrate that your PEMS
Section 12.3.1. form the results of the RA and statistical provides acceptable data when
8.3.2 F-test. Perform a separate F-test tests. Include all data sheets, operating in the new range or with the
for the RA paired data from each calculations, and charts (records of new critical PEMS parameter(s). The
operating level to determine if the RM PEMS responses) necessary to verify requirements of Section 8.1 must be
and PEMS variances differ by more than your PEMS’s meeting the performance completed by the earlier of 60 unit
might be expected from chance. Use the specifications. Include in the report the operating days or 180 calendar days
equations in Section 12.3.2. documentation used to establish your after the failed RATA or after the change
8.3.3 Correlation Analysis. Perform a PEMS parameter envelopes. Consult the that caused a significant change in
correlation analysis on all RA paired EPA regional office or permitting emission rate.
data from all operating levels, authority with jurisdiction over your 9.0 Quality Control.
combined, to determine how well the emissions unit for additional You must incorporate a QA plan
RM and PEMS correlate. Use the requirements. beyond the initial PEMS certification
equations in Section 12.3.3. 8.5 Reevaluating Your PEMS After a test to verify that your system is
If the process cannot be varied to Failed Test, Change in Operations, or generating quality-assured data. The QA
produce a concentration change Change in Critical PEMS Parameter. plan must include the components of
sufficient for a successful correlation After initial certification, if a quarterly this section.
test because of its technical design, the RAA or yearly RATA is failed due to a 9.1 QA/QC Summary. Conduct the
correlation analysis may be temporarily problem with the PEMS, or if changes applicable ongoing tests listed below.
ONGOING QUALITY ASSURANCE TESTS

PEMS Regulatory
Test Acceptability Frequency
Purpose
Sensor Evaluation Check ........................ All ....................... ................................................................ Daily.

RAA ......................................................... Compliance ........ 3-test average ≤ 10% of simultaneous Each quarter except quarter when
PEMS average. RATA performed.
RATA ....................................................... All ....................... Same as for RA in Sec. 13.1 ................ Yearly in quarter when RAA not per-
formed.
Bias Correction ........................................ All ....................... If davg > |cc| ............................................ Determine factor after each RATA.
PEMS Training ........................................ All ....................... If Fcritical ≥ F r ≥ 0.8 ................................ After initial and subsequent RATAs.
Sensor Evaluation Alert Test .................. All ....................... See Section 6.1.8 .................................. After each PEMS training.
9.2 Daily Sensor Evaluation Check. least three 30-minute portable analyzer basis in the quarter that the RAA is not
Your sensor evaluation system must determinations each quarter a RATA is preformed.
check the integrity of each PEMS input not performed.
at least daily. 9.4 Yearly Relative Accuracy Test
9.3 Quarterly Relative Accuracy Audit. Perform a minimum 9-run RATA
Audit. Perform a RAA consisting of at at the normal operating level on a yearly
10.0 Calibration and Standardization between paired RM and PEMS differences d̄ is greater than the absolute
[Reserved] observations using Equation 16–1. value of the confidence coefficient (cc),
11.0 Analytical Procedure [Reserved] your PEMS must incorporate a bias
1 n factor to adjust future PEMS values as
12.0 Calculations and Data Analysis d= ∑ di
n i =1
Eq. 16-1 in Equation 16–5.
12.1 Nomenclature.
B = PEMS bias adjustment factor. 12.2.2 Standard Deviation. Calculate PEMSiAdjusted = PEMSi × B Eq. 16-5
cc = Confidence coefficient. the standard deviation of the differences
di = Difference between each RM and Where:
using Equation 16–2 (positive square
PEMS run. root).
d̄ = Arithmetic mean of differences for d
all runs.  2
 B = 1+ Eq. 16-6a
 n  PEMS
 n ∑ di  
ei = Individual measurement provided
by the PEMS or RM at a particular level.   i =1   12.3.2 F-test. Conduct an F-test for
em = Mean of the PEMS or RM  ∑ di − n  each of the three RA data sets collected
at different parameter operating levels.
measurements at a particular level. s d =  i =1  Eq. 16-2
ep = Individual measurement provided  n −1  Calculate the variances of the PEMS and
by the PEMS.   the RM using Equation 16–6.
ev = Individual measurement provided   n
 
by the RM.
∑ (e i − e m )
2
F = Calculated F-value.
 
i =1
n = Number of RM runs. 12.2.3 Confidence Coefficient. S2 = Eq. 16-6
PEMSi = Individual measurement Calculate the confidence coefficient n −1
provided by the PEMS. using Equation 16–3 and Table 16–1. Determine if the variance of the PEMS
PEMSiAdjusted = Individual data is significantly different from that
measurement provided by the PEMS Sd of the RM data at each level by
cc = t 0.025 Eq. 16-3
adjusted for bias. n calculating the F-value using Equation
P̄ĒM̄S̄ = Mean of the values provided 12.2.4 Relative Accuracy. Calculate 16–7.
by the PEMS at the normal operating the RA of your data using Equation 16–
range during the bias test. 4. S2 PEMS
r = coefficient of correlation. F= Eq. 16- 7
S2 RM
EP08AU05.089</GPH>
RA = Relative accuracy.
R̄M̄ = Average RM value. In cases d + cc Compare the calculated F-value with
RA = ×100 Eq. 16-4 the critical value of F at the 95 percent
where the average emissions for the test RM
are less than 50 percent of the confidence level with n-1 degrees of
12.3 Compliance and Market-
applicable standard, substitute the freedom. The critical value is obtained
EP08AU05.088</GPH>
Trading PEMS Statistical Tests. If your
emission standard value here in place of from Table 16–2 or a similar table for F-
PEMS will be used for continual
the average RM value. compliance or market-trading purposes, distribution. If the calculated F-value is
Sd = Standard deviation of differences. conduct the following tests using the greater than the critical value at any
S2 = variance of your PEMS or RM. information obtained during the RA level, your proposed PEMS is
unacceptable.
EP08AU05.087</GPH>
t0.025 = t-value for a one-sided, 97.5 tests. For the pollutant measurements at
percent confidence interval (see Table any one test level, if the mean value of For pollutant PEMS measurements, if
16–1). the RM is less than either 10 ppm or 5 the standard deviation of the RM is less
12.2 Relative Accuracy Calculations. percent of the emission standard, all than either 3 percent of the span or 5
Calculate the mean of the RM values. statistical tests are waived at that ppm, use a RM standard deviation of
EP08AU05.086</GPH>
Calculate the differences between the specific test level. For diluent either 5 ppm or 3 percent of span. For
pairs of observations for the RM and the measurements at any one test level, if diluent PEMS measurements, if the
PEMS output sets. Finally, calculate the the mean value of the RM is less than standard deviation of the reference
mean of the differences, standard 3 percent of span, all statistical tests are method is less than 3 percent of span,
deviation, confidence coefficient, and use a RM standard deviation of 3
EP08AU05.085</GPH>
waived for that specific test level.
PEMS RA, using Equations 16–1, 16–2, 12.3.1 Bias Test. Conduct a bias test percent of span.
16–3, and 16–4, respectively. For to determine if your PEMS is biased 12.3.3 Correlation Analysis.
compliance and market-trading PEMS, relative to the RM. Determine the PEMS Calculate the correlation coefficient
calculate the RA at each operating level. bias by comparing the confidence either manually using Eq. 16–8, on a EP08AU05.084</GPH>
The PEMS must pass the RA criterion at coefficient obtained from Equation graph, or by computer using all of the
each operating level. 16–3 to the arithmetic mean of the paired data points from all operating
12.2.1 Arithmetic Mean. Calculate differences determined in Equation levels. Your PEMS correlation must be
the arithmetic mean of the differences 16–1. If the arithmetic mean of the 0.8 or greater to be acceptable.
EP08AU05.083</GPH>
∑ e p e v − ( ∑ e p )( ∑ e v ) / n
r= ( Eq. 16-8)

 ∑ e p −

2
( ∑ e p )2 / n  ∑ e2v − ( ∑ ev )2 / n 
EP08AU05.082</GPH>
EP08AU05.081</GPH>
13.0 Method Performance. 13.3 PEMS Variance. Your TABLE 16–1.—T-VALUES FOR ONE-
calculated F-value must not be greater SIDED, 97.5 PERCENT CONFIDENCE
13.1 PEMS Relative Accuracy. See than the critical F-value at the 95-
the relevant regulation for the INTERVALS FOR SELECTED SAMPLE
percent confidence level for your PEMS SIZES†—Continued
applicable RA criterion. For PEMS to be acceptable.
installed to meet New Source 13.4 PEMS Correlation. Your n-1 t0.025
Performance Standards, the RA of your calculated r-value must be greater than
PEMS must be no greater than 10 or equal to 0.8 for your PEMS to be 8 ................................................ 2.365
percent when based upon the average acceptable. 9 ................................................ 2.306
RM data (which must be measured in 10 .............................................. 2.262
the units of your emission standard). 14.0 Pollution Prevention. [Reserved] 11 .............................................. 2.228
For emissions below 25 percent of the 15.0 Waste Management. [Reserved] 12 .............................................. 2.201
emission standard, 20 percent RA based 13 .............................................. 2.179
16.0 References. [Reserved] 14 .............................................. 2.160
upon the emission standard may be
15 .............................................. 2.145
used. For emissions below 10 percent of 17.0 Tables, Diagrams, Flowcharts, 16 .............................................. 2.131
the emission standard, average PEMS and Validation Data 17 .............................................. 2.120
measurements within 2 ppm of the RM 18 .............................................. 2.110
mean value constitutes an acceptable TABLE 16–1.—T-VALUES FOR ONE- 19 .............................................. 2.101
RA test. For diluent PEMS, an SIDED, 97.5 PERCENT CONFIDENCE 20 .............................................. 2.093
alternative criterion of ±1 percent INTERVALS FOR SELECTED SAMPLE 21 .............................................. 2.086
absolute difference between the PEMS 22 .............................................. 2.080
SIZES† 23 .............................................. 2.074
and RM may be used if less stringent.
24 .............................................. 2.069
13.2 PEMS Bias. Your PEMS data is n-1 t0.025 25 .............................................. 2.064
considered biased and must be adjusted 26 .............................................. 2.060
2 ................................................ 12.706
if the arithmetic mean (d) is greater than 27 .............................................. 2.056
3 ................................................ 4.303
the absolute value of the confidence 28 .............................................. 2.052
4 ................................................ 3.182
coefficient (cc) in Equations 16.1 and >29 ............................................ t-Table
5 ................................................ 2.776
16.3. In such cases, a bias factor must be 6 ................................................ 2.571 †(Use n equal to the number of data points
used to correct your PEMS data. 7 ................................................ 2.447 (n-1 equals the degrees of freedom).
TABLE 16–2.—F-VALUES FOR CRITICAL VALUE OF F AT THE 95 PERCENT CONFIDENCE LEVEL

d.f. for S2PEMS
d.f. for S2RM
1 2 3 4 5 6 7 8 9 10 11 12
1 ............................................................... 161.4 199.5 215.7 224.6 230.2 234.0 236.8 238.9 240.5 241.8 243.0 243.9
2 ............................................................... 18.51 19.00 19.16 19.25 19.30 19.33 19.35 19.37 19.38 19.50 19.40 19.41
3 ............................................................... 10.13 9.552 9.277 9.117 9.014 8.941 8.887 8.845 8.812 8.786 8.763 8.745
4 ............................................................... 7.709 6.944 6.591 6.388 6.256 6.163 6.094 6.041 5.999 5.964 5.935 5.912
5 ............................................................... 6.608 5.786 5.410 5.192 5.050 4.950 4.876 4.818 4.773 4.735 4.703 4.678
6 ............................................................... 5.987 5.143 4.757 4.534 4.387 4.284 4.207 4.147 4.099 4.060 4.027 4.000
7 ............................................................... 5.591 4.734 4.347 4.120 3.971 3.866 3.787 3.726 3.677 3.637 3.603 3.575
8 ............................................................... 5.318 4.459 4.066 3.838 3.688 3.581 3.501 3.438 3.388 3.347 3.312 3.284
9 ............................................................... 5.117 4.257 3.863 3.633 3.482 3.374 3.293 3.230 3.197 3.137 3.102 3.073
10 ............................................................. 4.965 4.103 3.709 3.478 3.326 3.217 3.136 3.072 3.020 2.978 2.942 2.913
11 ............................................................. 4.844 3.982 3.587 3.357 3.204 3.095 3.012 2.948 2.896 2.854 2.817 2.788
12 ............................................................. 4.747 3.885 3.490 3.259 3.106 2.996 2.913 2.849 2.796 2.753 2.717 2.687
* * * * * the most recent edition of EPA’s Reference Materials.’’ Joint publication

Traceability Protocol No. 1 (See Citation by NBS and EPA–600/7–81–010,
Appendix F—[Amended]
2). Procedures for preparation of CRM’s Revised 1989. Available from the U.S.
9. In Procedure 1 of Appendix F, by are described in Citation 1. Procedures Environmental Protection Agency.
revising paragraph (3) of Section 5.1.2 for preparation of ETPM’s are described Quality Assurance Division (MD–77).
and Section 8 as follows: in Citation 2. As an alternative to CRM’s Research Triangle Park, NC 27711.
Procedure 1. Quality Assurance or ETPM gases, Method 205 (See
2. ‘‘EPA Traceability Protocol For
Requirements for Gas Continuous Citation 3) may be used.
Assay And Certification Of Gaseous
Emission Monitoring Systems Used for The difference between the actual Calibration Standards.’’ EPA–600/R–97/
Compliance Determination concentration of the audit gas and the 121, September 1997. Available from
* * * * * concentration indicated by the monitor EPA’s Emission Measurement Center at
5.1.2 Cylinder Gas Audit (CGA). is used to assess the accuracy of the www.epa.gov/ttn/emc.
CEMS.
* * * * * 3. Method 205, ‘‘Verification of Gas
(3) Use Certified Reference Materials * * * * *
Dilution Systems for Field Instrument
(CRM’s) (See Citation 1) audit gases that 8. Bibliography Calibrations,’’ 40 CFR 51, Appendix M.
have been certified by comparison to
National Institute of Standards and 1. ‘‘A Procedure for Establishing * * * * *
Technology (NIST) or EPA Traceability Traceability of Gas Mixtures to Certain 10. In Procedure 2, by revising
Protocol Materials (ETPM’s) following National Bureau of Standards Standard Section 10.1, paragraph (3) of Section
10.4, paragraph (2) of Section 12.0 as the PM CEMS Knowledge Document 12.0 What calculations and data
follows: (see section 16.5 of PS–11). analysis must I perform for my PM
Procedure 2—Quality Assurance * * * * * CEMS?
Requirements for Particulate Matter 10.4 What are my limits for * * * * *
Continuous Emission Monitoring excessive audit inaccuracy? (2) How do I calculate ACA accuracy?
Systems at Stationary Sources * * * * * You must use either Equation 2–1a or 2–
* * * * * (3) What are the criteria for excessive 1b to calculate ACA accuracy for each
10.1 When should I use paired trains ACA error? Your PM CEMS is out of of the three audit points. However,
for reference method testing? Although control if the results of any ACA exceed when calculating ACA accuracy for the
not required, we recommend that you ±10 percent of the average audit value, first audit point (0 to 20 percent of
should use paired-train reference as calculated using Equation 2–1a, or measurement range), you must use
method testing to generate data used to 7.5 percent of the applicable standard,
Equation 2–1b to calculate ACA
develop your PM CEMS correlation and as calculated using Equation 2–1b,
accuracy if the reference standard value
for RCA testing. Guidance on the use of whichever is greater.
(Rv) equals zero.
paired sampling trains can be found in * * * * *
R CEM − R V
ACA Accuracy = × 100% ( Eq. 2-1a)
RV
Where: ACA Accuracy=The ACA accuracy at RCEM = Your PM CEMS response to the
each audit point, in percent, reference standard, and
RV = The reference standard value.
C CEM − C RV
ACA Accuracy = × 100% ( Eq. 2-1b)
Cs
Where: 1.0 Scope and Application pesticide product registrations (70 FR

ACA Accuracy = The ACA accuracy at 1.1 Applicability. * * * In order for 31401, June 1, 2005).
each audit point, in percent, the test method results to be indicative DATES: Comments, identified by the
CCEM = The PM concentration that of plant performance, the time of day of docket identification (ID) number OPP–
corresponds to your PM CEMS the run should vary.
response to the reference standard, 2005–0069, must be received on or
as calculated using the correlation * * * * *Q P=’03’≤ before August 31, 2005.
equation for your PM CEMS, [FR Doc. 05–15330 Filed 8–5–05; 8:45 am]
ADDRESSES: Follow the detailed
CRV = The PM concentration that BILLING CODE 6560–50–P
instructions as provided under
corresponds to the reference ADDRESSES in the Federal Register
standard value in units consistent document of June 1, 2005.
with CCEM, and ENVIRONMENTAL PROTECTION
Cs = The PM concentration that AGENCY FOR FURTHER INFORMATION CONTACT:
corresponds to the applicable Karen Angulo, Registration Division
emission limit in units consistent 40 CFR Part 180 (7505C), Office of Pesticide Programs,
with CCEM. Environmental Protection Agency, 1200
* * * * * [OPP–2005–0069; FRL–7729–4] Pennsylvania Ave., NW., Washington,
11. The authority citation for Part 63 DC 20460–0001; telephone number:
Inert Ingredients; Proposal to Revoke
continues to read as follows: (703) 306–0404; fax number: (703) 305–
34 Pesticide Tolerance Exemptions for
Authority: 42 U.S.C. 7401 et seq. 31 Chemicals; Reopening of Comment 0599; e-mail address:
Period angulo.karen@epa.gov.
PART 63—NATIONAL EMISSION
STANDARDS FOR HAZARDOUS AIR AGENCY: Environmental Protection SUPPLEMENTARY INFORMATION:
POLLUTANTS FOR SOURCE Agency (EPA). I. General Information
CATEGORIES
ACTION: Proposed rule; reopening of
A. Does this Action Apply to Me?
12. In Method 303 of Appendix A, by comment period.
adding the following sentence to The Agency included in the proposed
Section 1.1: SUMMARY: This document reopens the
rule a list of those who may be
public comment periodof EPA’s
Appendix A to Part 63—Test Methods potentially affected by this action. If you
proposal to revoke 34 exemptions from
the requirement of a tolerance that are have questions regarding the
* * * * *
EP08AU05.091</GPH>
associated with 31 inert ingredients applicability of this action to a

Method 303—Determination of Visible because, according to Agency records, particular entity, consult the person
Emissions From By-Product Coke Oven these substances are no longer listed under FOR FURTHER INFORMATION
Batteries contained in active Federal Insecticide, CONTACT.
EP08AU05.090</GPH>
* * * * * Fungicide, and Rodenticide Act (FIFRA)

Proposed Rule: Air Pollution Standards of Performance For New Stationary Sources: Predictive Emission Monitoring Systems Performance Specifications Testing and Monitoring Provisions Amendments

Hochgeladen von

Dokumentinformationen

Originaltitel

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Proposed Rule: Air Pollution Standards of Performance For New Stationary Sources: Predictive Emission Monitoring Systems Performance Specifications Testing and Monitoring Provisions Amendments

Hochgeladen von

Copyright:

Verfügbare Formate

45608 Federal Register / Vol. 70, No.

151 / Monday, August 8, 2005 / Proposed Rules

Examples of regulated entities SIC codes NAICS codes

Industrial, Commercial, Institutional Steam Generating Units ................................................................................. 3569 332410

Portland Cement Manufacturing .............................................................................................................................. 3559 333298

Hazardous Waste Incinerators ................................................................................................................................ 4953 562211

Fossil Fuel Steam Generators ................................................................................................................................. 3569 332410

Rubber Tire Manufacturing ...................................................................................................................................... 3011 326211

Where: n = the number of test runs. Equations 11–19 through 11–21,

b2 = (Eq. 11- 21)

(ii) Calculate the 95 percent first calculating the C coefficients (C0 to

Calculate D using Equation 11–25 for

EL = PM emission limit, as described resulting equation will have the form

in section 13.2. ŷ’ = the predicted log PM concentration

corresponds to Dmin, as indicated in Where: in paragraph (1)(i) of this section.

3 ....................................................................................................................... 3.182 2.920 1.266 3.697

41 ..................................................................................................................... 2.020 1.225 1.164 1.425

* * * * * conducted to ensure the PEMS is maintenance must perform a minimum

If PEMS is measuring . . . And if . . . Then your upper limit . . .

If . . . Then . . . And then . . .

ONGOING QUALITY ASSURANCE TESTS

Sensor Evaluation Check ........................ All ....................... ................................................................ Daily.

TABLE 16–2.—F-VALUES FOR CRITICAL VALUE OF F AT THE 95 PERCENT CONFIDENCE LEVEL

* * * * * the most recent edition of EPA’s Reference Materials.’’ Joint publication

Where: 1.0 Scope and Application pesticide product registrations (70 FR

associated with 31 inert ingredients applicability of this action to a

* * * * * Fungicide, and Rodenticide Act (FIFRA)

Das könnte Ihnen auch gefallen