
                                                



Quality Assurance Requirements for
Continuous Opacity Monitoring Systems at
Stationary Sources

Response to Comments



                                       
                                       
                                       
                                       
                                       
                                       
                                       

                     (This page intentionally left blank.)


                                       
                                       
                      Quality Assurance Requirements for
                   Continuous Opacity Monitoring Systems at
                              Stationary Sources
                                       
                             Response to Comments



                        Air Quality Assessment Division
                 Office of Air Quality Planning and Standards
                     U.S. Environmental Protection Agency
                       Research Triangle Park, NC 27711
                                       
                                  March 2013
                                       
                                       
                                       
                                       
                                       













                     (This page intentionally left blank.)


                               Table of Contents
LIST OF ACRONYMS AND ABBREVIATIONS						iii
1.0 Introduction										1
      1.1 Background									1
      1.2 Commenters									1
      1.3 Organization of this Document						1	
2.0 Administrative Comments								3
      2.1 Legal Authority for Rule							3
      2.2 Procedure 3 vs. SIP Requirements						4
      2.3 Procedure 3 vs. Part 63 Rules							4
      2.4 Applicability									5
      2.5 Compliance Deadline								6
      2.6 Statutory and Executive Order Reviews					6	
            2.6.1 Paperwork Reduction Act						6
            2.6.2 National Technology Transfer and Advancement Act		6
      2.7 Terminology									7
            2.7.1 Consistently Achieve Quality Assured Data				7
            2.7.2 Out-of-Control Period							7
            2.7.3 Noncompliance vs. Failure to Meet					8
            2.7.4 Water Droplets							8
            2.7.5 Daily Calibrations							8
            2.7.6 Quarterly Calibrations							9
            2.7.7 Annual Zero Alignment Audit					9
            2.7.8 Zero-jig 								10
      2.8 ASTM Reference								11
3.0 Technical Comments									11
      3.1 Status Indicator Checks							11
      3.2 Units with Automatic Zero Compensation					11
      3.3 Calibration Attenuators (Neutral Density Filters) - Recalibration Frequency	12
      3.4 Quarterly Optical Alignment Checks						12
      3.5 COMS Removal for Annual Zero Alignment					13
      3.6 Quarterly Performance Audit Criteria  -  Misalignment Error			13
      3.7 Data Capture Criteria								14
      3.8 Corrective Action								14	
      3.9 30-Day Use of a Temporary Monitor						15
      3.10 QA/QC Reporting Requirements/Data Assessment Reports (DARs)	15
      3.11 Diagnostic Tests After Repairs (Table 17-1)					16
4.0 Miscellaneous Comments								17
      4.1 Alternative Compliance Option						17
      4.2 Employee Safety								17
      4.3 Equipment, Supplies, and Reagents						17
      4.4 QC Program									18
      4.5 Record Retention of QA/QC Procedures					18
      4.6 Consequences of Failing QC Audits						19
      4.7 Calibration and Standardization Procedures					19
      4.8 Quarterly Calculation of 95 Percent Confidence Interval			20
      4.9 Intervals to Record Data							20
      4.10 External Audit Device								20
      4.11 Clear Path Zero Value								20
      4.12 Limitations on the Use of COMS Data					21
      4.13 Data Retention								22
      4.14 Daily Calibration Drift Check Calculation					22
      4.15 Performance Specification 1 vs. Procedure 3					22





                      List of Acronyms and Abbreviations
                                       
§			Section (used in this document for sections of the CFR)
Act			Clean Air Act (42 U.S.C. 7401  -  7671q)
ASTM			American Society for Testing and Materials
CAA			Clean Air Act (42 U.S.C. 7401  -  7671q)
CD			Calibration Drift
CEMS			Continuous Emission Monitoring Systems
COMS			Continuous Opacity Monitoring Systems
DAR			Data Assessment Report
EPA			U.S. Environmental Protection Agency
FR			Federal Register
QA/QC		Quality Assurance/Quality Control
PS-1			Performance Specification 1
SIP			State Implementation Plan



1.0	Introduction
1.1	Background
	On February 14, 2012, the U.S. Environmental Protection Agency (EPA or "we") published a parallel proposed rule with a direct final rule titled Quality Assurance Requirements for Continuous Opacity Monitoring Systems at Stationary Sources. We proposed to establish quality assurance and quality control (QA/QC) procedures (referred to as Procedure 3) for continuous opacity monitoring systems (COMS) used to demonstrate continuous compliance with opacity standards in EPA issued or approved regulations. 
      This document presents EPA's responses to public comments received on the proposed rule. All significant issues raised in timely public comments have been addressed.
1.2	Commenters
	The public comment period initially ended on March 15, 2012 but was subsequently extended to April 30, 2012 via notices published in the Federal Register on March 8, 2012. Since we received adverse comments, we published a notice[4] to withdraw the direct final rule on March 28, 2012. We received comments from 27 commenters. This total counts each entity one time even if they submitted more than one comment letter. A list of the commenters and their affiliations is on page 2.  
1.3	Organization of this Document
	After this introductory section, this document is divided into three sections as follows:
   * Section 2  -  Administrative Comments
   * Section 3  -  Technical Comments
   * Section 4  -  Miscellaneous Comments


                        Commenters on COMS Procedure 3
                          Proposed February 14, 2012
                             EPA-HQ-OAR-2010-0873

ID
Commenter
Affiliation
0003
Stephenie Moyer
Nebraska Air Quality Division
0004
Donivan Porterfield
---
0006
Marsha Layman
Quality Assurance Support Group, LLC
0009
Andy Seeber
Wisconsin Department of Natural Resources
0010
Basil G. Constantelos
Midwest Generation LLC & Homer City EME Generation
0011
Douglas B. Price
Tesoro Companies, Inc.
0012
Monica Lopes
NAES Corporation
0013
Lauren Freeman
Utility Air Regulatory Group (UARG)  -  (pre-submittal - see 0029)
0014
Lisa Pongnon
Calpine Corporation
0015
Anthony C. Sullivan
Indianapolis Power & Light Co. (IPL)
0016
Cathe Kalisz
American Petroleum Institute (API)  -  (pre-submittal  -  see 0036)
0017
Robert A. Hirsch
Portland Cement Association (PCA)  -  (pre-submittal  -  see 0025)
0018
James S. Willis, III
Titan America
0019
Ross May
Searles Valley Minerals
0021
Jason Goodwin
CARBO Ceramics, Inc.
0022
Bob Machaver
---
0023
Jim Peeler
Emission Monitoring Inc.  -  (pre-submittal  -  see 0032)
0024
Jeffrey Poulton
Energy Recovery Operations, Inc.
0025
Robert A. Hirsch
Portland Cement Association (PCA)
0026
Christine Wyman
Class of `85 Regulatory Response Group
0027
Robert Bessette
Council of Industrial Boilers (CIBO)
0028
Mark Sedlacek
Los Angeles Department of Water and Power (LADWP)
0029
Lauren Freeman
Utility Air Regulatory Group (UARG)  -  w/ RMB comments attached
0030
Michael Bradley
The Clean Energy Group
0031
Kevin Dempsey
American Iron and Steel Institute (AISI)
0032
Jim Peeler
Emission Monitoring, Inc.
0033
Randal Oswald
Integrys Energy Group
0034
Gregory Wilkins
Marathon Petroleum Company 
0035
Lee Lemke
Georgia Mining Association
0036
Cathe Kalisz
American Petroleum Institute (API)
0037
Jay Hudson
Santee Cooper

	
	
2.0	Administrative Comments
0.1 Legal Authority for Rule
Comment:
      Four commenters (0025, 0029, 0034, and 0036) commented on our legal authority to impose the Procedure 3 QA/QC requirements. One of these commenters (0025) stated that we did not provide the CAA provision that allows us to impose the QA/QC requirements, and another one of the commenters (0029) requested that we identify the circumstances and legal authority for imposing the QA/QC requirements. 
Response:
      Section 301(a) of the CAA authorizes the EPA Administrator to prescribe such regulations as are necessary to carry out his functions under the CAA and CAA section 114 authorizes the EPA Administrator to require any person who owns or operates an emission source to, among other things, install, use, and maintain such monitoring equipment and use such audit procedures, or methods as may be necessary to determine whether the owner/operator is in violation of a standard of performance under CAA section 111.  EPA has previously determined that the use of COMS is, in appropriate circumstances, necessary to ensure continuous compliance with opacity standards.  EPA has now determined that the Procedure 3 QA/QC requirements are necessary to ensure that COMS used to demonstrate continuous compliance with opacity standards are properly maintained and operated.. Procedure 3 applies to all COMS used to demonstrate continuous compliance with opacity standards specified in new source performance standards (NSPS) promulgated by EPA pursuant to section 111(b) of the Clean Air Act, 42 U.S.C. §4211(b).
Comment:
      The final two commenters (0034, 0036) stated that without amending individual subparts to reference Procedure 3, we don't have the authority to impose Procedure 3 on any COMS except those used for continuous compliance under Part 60 and that were constructed or modified after Procedure 3 was proposed.
Response:
      As stated above, we believe that we have the authority to impose COMS Procedure 3 QA/QC requirements to ensure that COMS used to demonstrate continuous compliance with opacity standards are properly maintained and operated.  We believe that this authority extends to requiring sources constructed, reconstructed or modified before COMS Procedure 3 was proposed to comply with the requirements to ensure that COMS used to demonstrate continuous compliance with opacity standards as required by a section 111(b) new source performance standards (NSPS) are properly maintained and operated. Part 60.13(a) of title 40 of the Code of Federal Regulations (CFR) specifically contemplates this We added a statement noting this authority in the paragraph titled "Background" in the preamble.
2.2	Procedure 3 vs. SIP Requirements
Comment:
      Six commenters (0012, 0025, 0027, 0028, 0029, 0030) commented on Procedure 3 requirements vs. SIP requirements. Commenter 0012 noted that Procedure 3 does not clarify if sources located in a state that has issued COMS QA guidance would be required to implement both state COMS QA guidance & Procedure 3. Another one of the commenters (0025) stated that source owners and operators should not have to comply with federal and state opacity monitoring and related QA requirements, and commenter 0027 stated that it is not clear as to whether a source will have to comply with state QA/QC procedures, EPA procedures, or both. Another one of the commenters (0028) stated that the applicability statement may not be specific enough to prevent sources subject to Title V Operating Permits or state regulations from being required to comply with Procedure 3. Commenter 0029 requested that we clarify that Procedure 3 would not apply to COMS required under a SIP unless the SIP specifically states that Procedure 3 applies. The final commenter (0030) stated that the Data Assessment Report (DAR) does not appear to be standardized and would not fit into what states require. The commenter recommended that for states with delegated air programs that source owners and operators not be required to submit duplicate DARs.	 
Response: 
	In the final regulatory text we clarified that COMS Procedure 3 is only required to be used when a COMS is used to demonstrate continuous compliance with an opacity standard specified in a new source performance standard (NSPS) promulgated by EPA pursuant to section 111(d) if the Clean Air Act, 42 U.S.C. §7411(b).  Source owners and operators must comply with both state and EPA requirements where both sets of requirements apply. A state can choose to adopt Procedure 3 in lieu of state-specific requirements or determine that compliance with Procedure 3 satisfies the state-specific requirements if it wishes to do so.  In addition, a source subject to state-specific QA/QC procedures applicable to COMS can request in writing, pursuant to 40 CFR 60.13(i), that the Administrator approve the use of those procedures in lieu of COMS Procedure 3. If the Administrator does approve the use of the state-specific procedures, those procedures become the applicable requirements under the relevant NSPS. We also believe that the information required by Procedure 3 to be included in the DAR is necessary and appropriate and is consistent with what most states require.
2.3 Procedure 3 vs. Part 63 Rules
Comment:
      One commenter (0036) noted the potential for overlaps and conflicts between existing rules and Procedure 3.
Response:
	As reflected in the revised applicability statement, Procedure 3 only applies to COMS used to demonstrate continuous compliance with opacity standards specified in new source performance standards (NSPS) promulgated by EPA pursuant to section 111(b) of the Clean Air Act, 42 U.S.C. §7411(b).  As a result, the commenter's examples of potential overlap with specific requirements in rules promulgated pursuant to section 112 of the Clean Air Act, 42 U.S.C. §7412, are not relevant.  Further, he commenter did not demonstrate that an actual conflict exists. Our intention is, nevertheless, to avoid duplicative and conflicting requirements, so if a source owner or operator determines that the requirements in Procedure 3 are actually in conflict with requirements in an existing rule, we will appreciate it if the owner or operator notifies us of this fact.
0.4 Applicability
Comment:
	Seven commenters (0015, 0025, 0027, 0028, 0029, 0034, 0036) expressed concerns about the applicability statement. Three of the commenters (0015, 0025, 0027) requested that we define or clarify "federally enforceable regulations". 
Response:
	We agree that the phrase "federally enforceable regulations" is ambiguous and clarified our intent by changing the language to "new source performance standards (NSPS) promulgated by EPA pursuant to section 111(b) of the Clean Air Act, 42 U.S.C. §7411(b)".  
Comment:
	Three of the commenters (0029, 0034, 0036) noted that Procedure 3 refers to Performance Specification 1 (PS-1), and stated that PS-1 applies only to COMS installed after February 6, 2001. One of the three commenters (0029) recommended that we develop separate requirements for pre-and post-February 6, 2001 COMS. 
Response:
	PS-1 was developed to address the initial installation of COMS, whereas Procedure 3 addresses on-going QA/QC requirements. When we developed PS-1, we realized that some existing COMS were not being used for compliance purposes, and it was not our intent to require these source owners and operators to purchase new COMS. To accommodate these source owners and operators, we established the cutoff date of February 6, 2001. In other words, all new COMS installed after February 6, 2001 are required to meet PS-1 requirements. 
	Furthermore, Procedure 3 applies to all COMS (regardless of installation date) that are used to demonstrate continuous compliance with opacity standards in federally enforceable regulations as defined above. Procedure 3 does not prohibit source owners and operators from using COMS installed prior to February 6, 2001.
Comment:
	The final commenter (0028) recommended that Procedure 3 allow different QA/QC requirements for sources depending on opacity limits. For example, sources subject to an opacity limit less than 10 percent could be required to perform QA/QC semi-annually, and sources subject to an opacity limit greater than 10 percent could be required to perform QA/QC annually.
Response:
	We believe that the requirements in Procedure 3 are comprehensive and appropriate for all COMS regardless of their opacity limit(s). Procedure 3 will result in national consistency in the application of QA/QC procedures by applicable sources using COMS.
0.5     Compliance Deadline 
Comment:
	Fifteen commenters (0010, 0012, 0019, 0022, 0025, 0026, 0027, 0028, 0029, 0031, 0032, 0033, 0034, 0036, 0037) urged us to increase the compliance deadline beyond 60 days after the date of publication of the rule. One commenter (0014) stated that for COMS installed and certified prior to the rule effective date, the rule should provide grandfathering such that the COMS would need to be removed no later than three years from the effective date of the rule and every three years thereafter. Most commenters requested that the compliance deadline be either four or six months after the rule publication date to allow sufficient time to develop QA/QC procedures, provide training, etc.  
Response:
	We agree that 60 days from the date Procedure 3 is published in the Federal Register may not allow adequate time for all sources to comply with the QA/QC requirements. Therefore, we extended the compliance deadline to 180 days from the date Procedure 3 is published.
0.6  	Statutory and Executive Order Reviews
         
0.6.1 Paperwork Reduction Act
Comment:
	Three commenters (0025, 0027, 0036) mentioned that Procedure 3 will result in implementation costs and budgetary impacts.
Response:
	We believe that the requirements in applicable regulations are broad enough to include  the information collection requirements specified in Procedure 3. In addition, the requirements in Procedure 3 are modeled after manufacturers' maintenance recommendations.  As a result, EPA believes that most, if not all, owners/operators are already following procedures similar to those specified in Procedure 3.  Therefore, there are no additional costs, or reporting burden, associated with implementing Procedure 3.  
2.6.2	National Technology Transfer and Advancement Act
Comment: 
      Two commenters (0004, 0031) commented on the National Technology Transfer and Advancement Act (NTTAA) language in the preamble of the Procedure 3 rule. The commenters noted that Procedure 3 involves technical standards.
Response:
      Upon further review, the Agency agrees that Procedure 3 involves technical standards and revised the NTTAA language in the final rule.
   
0.7 Terminology
         
2.7.1	Consistently Achieve Quality Assured Data
Comment:
      Six commenters (0011,0025, 0028, 0029, 0030, 0036) commented on the provision in §2.0(3) that allows owners and operators who consistently achieve quality assured data to request a semi-annual audit frequency by submitting the request in writing to the Administrator. Three of the commenters (0011, 0025, 0029) noted that the phrase "consistently achieves quality assured data" is vague and undefined, and the other three commenters (0028, 0030, 0036) requested that this provision be self-implementing or automatic (i.e., not require a petition to the Administrator).
Response:
      We deleted the ambiguity in this provision, and specified in §2.0(2) that owners and operators who achieve quality assured data for four consecutive quarters may reduce their auditing frequency to semi-annually. If a performance audit is failed, the source must resume quarterly testing for that audit requirement until such time that it has again demonstrated successful performance over four consecutive quarters. In addition, we agreed to allow this provision to be self-implementing. In other words, we deleted the requirement to submit the request in writing to the Administrator.  
0.7.2 Out-of-Control Period
Comment:
      One commenter (0028) requested that we redefine the out-of-control period to match the definition in Procedure 1 for gas CEMS.  
Response:
      The commenter provided no basis for this request, and we believe that our proposed definition is appropriate for COMS. Therefore, we did not make any changes to the proposed definition.
2.7.3	Noncompliance vs. Failure to Meet
Comment:
	One commenter (0029) requested that we change the term noncompliance to the phrases "fails to meet" and "failure to meet" for the out-of-control period for annual zero alignments and quarterly performance audits. The commenter noted that failure to meet a QA criterion indicates that the COMS data is not quality assured and does not indicate that a source is in a state of noncompliance.
Response:
	We agree and replaced the term "noncompliance" with "fails to meet" and "failure to meet". 
0.7.4 Water Droplets
Comment:
	Three commenters (0029, 0031, 0036) commented on §4.0 that discusses avoiding water droplet interferences. One of the commenters (0029) is concerned that the statement indicating that opacity compliance determinations cannot be made when water droplets are present is too narrow. The commenter noted that whether used for continuous compliance determinations or not, COMS are not accurate in the presence of water droplets. The other two commenters (0031, 0036) noted that the proposed language could be interpreted to require a reheater.
Response:
	To address the concerns of all three commenters and to be consistent with the language in PS-1, we replaced the term "water droplets" with "condensed water vapor" and added the statement "Therefore, COMS must be located where condensed water vapor is not present."
0.7.5 Daily Calibrations
Comment:
	Six commenters (0006, 0010, 0012, 0022, 0030, 0036) commented on the system checks required to be conducted daily. Three of the commenters (0006, 0010, 0030) requested that we define daily by specifying a minimum number of operating hours to qualify as a day, and one of the commenters (0036) requested that we clarify that daily means calendar days. Another one of the commenters (0012) requested that we consider allowing sources to discontinue daily calibration checks for units that don't operate for one week or more and require those sources to resume daily calibration checks the day before the unit is scheduled to resume operations.  
Response:
      We re-defined daily to mean any portion of a calendar day that a unit operates and believe that by doing so, we have addressed the commenters' concerns.
Comment:
       The final commenter (0022) stated that daily calibrations should only be required on days that the unit is operating on a fuel subject to opacity monitoring.  
Response:
      We believe this level of detail is inappropriate to include in Procedure 3.
0.7.6 Quarterly Calibrations
Comment:
	Eight commenters (0006, 0012, 0022, 0026, 0029, 0030, 0033, 0036) requested that we define quarterly (i.e., include the minimum number of operating hours to qualify as an operating quarter), and five of the commenters (0006, 0012, 0026, 0029, 0033) recommended that we  define quarter as one in which the unit operates at least 168 hours in a calendar quarter.
Response:
      We agreed to clarify the term "quarterly" and defined it in the final rule as a calendar quarter in which a unit operates at least 168 hours.
2.7.7	Annual Zero Alignment Audit
Comment:
      Six commenters (0012, 0021, 0026, 0028, 0030, 0036) commented on the annual zero alignment audit requirement. Two of the commenters (0030, 0036) mentioned that the term "annual" was unclear and requested that we define annual (e.g., specify minimum number of operating hours to qualify as a year), and one commenter (0012) inquired about whether the zero alignment audit should be performed annually or every 12 months.  
Response:
      We added clarity by defining "annually" as a period wherein the unit is operating at least 28 days in a calendar year.
Comment:
      One commenter (0026) requested that the rule provide a grace period for completion of audits and stated that the annual zero alignment must be conducted while the unit is offline.  
Response:
      However, the rule does not require that the annual zero alignment be conducted while the unit is offline.
Comment:
      Another commenter (0021) stated that the annual zero alignment audit should not be required unless the COMS is replaced.   
Response:
      We disagree. Based on input from COMS manufacturers, to ensure that the zero alignment is accurate, it is necessary to conduct this audit annually. 
Comment:
      The final commenter (0028) requested that the zero alignment be conducted every three years instead of annually and stated that testing annually will require a substantial amount of manpower that may not significantly improve the value of the data. The commenter recommended that we allow incentives based on system performance criteria instead of imposing a mandatory annual requirement.
Response:
      Section 10.3(2) allows owners and operators to use an external zero device to determine the zero alignment. In other words, source owners and operators are not required to remove the COMS annually if they use the external zero device alternative.
0.7.8 Zero-jig
Comment:
	Two commenters (0026, 0033) commented on the term zero-jig in the annual zero alignment alternative in §10.3(2). Both commenters indicated that the use of the term zero-jig was unclear and inadequately defined. They also requested that we consult with COMS manufacturers to identify appropriate methods for conducting the annual zero alignment audit.
Response:
      We agree with the commenters that the term zero-jig was not well-defined in the proposed rule and deleted the term from the final rule. The final rule uses the term external zero device only. We also discussed the annual zero alignment method with COMS manufacturers and confirmed that the method in the rule is appropriate.
7.7 ASTM Reference
Comment:
      One commenter (0032) requested that Procedure 3 reference the most recent version of the ASTM D 6216 only. Another commenter (0029) noted that the proposed rule suggests that source owners and operators must comply with all three versions of the ASTM and suggested that we specify that source owners and operators may comply with the version in effect when the COMS was certified or, using their opinion, either of the two later versions.  
Response:
      We believe that only the latest ASTM version should be referenced since the prior versions may contain obsolete information. Therefore, we referenced only the latest ASTM version (ASTM D6216-12) in the final rule.
       3.0 Technical Comments
3.1	Status Indicator Checks
Comment:
	Eight commenters (0006, 0010, 0022, 0027, 0029, 0034, 0036, 0037) noted that a fault status indicator does not necessarily mean that the data is invalid, and one commenter (0015) asked for clarification regarding whether corrective action, recordkeeping, and reporting are required when status indicator checks illuminate.   
      We agree that a status indicator is a warning that opacity readings are nearing the limit and that the data is not necessarily invalid. Therefore, we deleted the statement in §10.1(3), "All COMS data recorded during periods in which the fault status indicators are illuminated are to be considered invalid." You must continue to record data when the status indicator check illuminates.  
      As stated in §10.1(3), you must take appropriate corrective action based on the manufacturer's recommendations when the COMS is operating outside preset limits. Section 10.5 requires you to have a corrective action program to address the repair and/or maintenance of your COMS. Although it is a good practice to keep records, recordkeeping and reporting are not required for daily status indicators.
3.2	Units with Automatic Zero Compensation
Comment:
	One commenter (0014) asked that we clarify that §10.2(1) applies only to COMS with automatic zero compensation, and two commenters (0026, 0033) requested that we delete the requirement to record the value of the compensation applied to the effluent because some COMS are not equipped to record or display the zero compensation value.  
Response:
	The language in §10.2(1) states "For units with automatic zero compensation...".  In other words, §10.2(1) only applies to COMS equipped with automatic zero compensation.
3.3	Calibration Attenuators (Neutral Density Filters) - Recalibration Frequency
Comment:
      Five commenters (0021, 0027, 0029, 0034, 0036) commented on the recalibration frequency of calibration attenuators (or neutral density filters). Three of the commenters (0021, 0027, 0029) stated that the requirement to inspect and calibrate neutral density filters (calibration attenuators) every six months is overly burdensome. The other two commenters (0034, 0036) requested that Procedure 3 require calibration attenuators to be recalibrated annually in lieu of semi-annually. In addition, one of the commenters (0036) suggested that we replace the term "neutral density filters" with "calibration attenuators, either primary or secondary" to be consistent with language in Performance Specification 1.
Response:
      We agree that inspecting and calibrating neutral density filters (attenuators) every six months is overly burdensome and revised the language in §10.2(2) to require source owners and operators to inspect and calibrate attenuators annually as opposed to every six months. In addition, if two consecutive annual calibrations are within 0.5 percent opacity, the attenuators can be inspected and calibrated once every five years. We agree with the terminology suggestion and replaced the term "neutral density filters" with "calibration attenuators, either primary or secondary".
3.4	Quarterly Optical Alignment Checks
Comment:
	Five commenters (0026, 0027, 0028, 0030, 0033) noted that it is not feasible to check the optical alignment when the stack T is +/-50% of the typical operating T, and one commenter (0015) noted that §10.2(3) requires a unit to be online on a quarterly basis even if it would not be required otherwise.  
Response:
      We agree and revised the language in §10.2(3) as follows. "You must check the optical alignment of the COMS in accordance with the instrument manufacturer's recommendations. If the optical alignment varies with stack temperature, perform the optical alignment test when the unit is operating."
3.5	COMS Removal for Annual Zero Alignment
Comment:
      Nine commenters (0011, 0015, 0021, 0024, 0026, 0027, 0029, 0034, 0036) remarked that removing the COMS from the stack exposes the COMS to potential damage, presents a safety hazard, and is a monetary burden. Six of these commenters (0015, 0021, 0026, 0029, 0034, 0036) requested that we delete the requirement to remove the COMS to conduct zero alignment audits, and one (0027) of the commenters suggested that the rule require the zero alignment audit when the boiler is down.  
Response:
      The zero alignment test is necessary because the simulated zero device (an automated mechanism that produces a simulated clear path condition) is used daily to check the COMS's zero drift. If the simulated zero device is inaccurate, this will directly impact the instrument's daily zero drift readings. The accuracy of the simulated zero device is determined by comparing it to actual clear path conditions. The best way to establish this relationship between the simulated zero device and clear path conditions is by conducting the zero alignment test. The best way to ensure clear path conditions is by removing the COMS from it's monitoring location and setting it up at ground level. Based on input from COMS manufacturers, it was determined that the zero alignment test needs to be done off-stack annually unless you choose the alternative that allows the installation of an external zero device (which allows COMS removal from the stack every three years). In addition, based on our conversations with manufacturers, we believe that the risks for damage to occur when removing the COMS from the stack are minimal.
3.6	Quarterly Performance Audit Criteria  -  Misalignment Error
Comment:
	Four commenters (0027, 0029, 0034, 0036) noted that the quarterly performance audit criterion in §10.4(3)(i) is problematic. The proposed rule states that the COMS is out-of-control if the results of a quarterly performance audit indicate noncompliance with the criterion that the optical alignment misalignment error exceeds 3 percent opacity. The commenters noted that there is no reliable methodology that relates the amount of misalignment with opacity, and one of the commenters (0029) suggested that we change the excessive inaccuracy criterion to adjust alignment relative to a specific reference mark or condition.
Response:
      We agree with the commenters and changed the quarterly performance audit noncompliance criterion in §10.4(3)(i) to "the optical alignment indicator does not show proper alignment (i.e., does not fall within a specific reference mark or condition)."
3.7	Data Capture Criteria
Comment:
	Nine commenters (0006, 0009, 0010, 0012, 0022, 0028, 0029, 0034, 0036) expressed concerns regarding the criterion for data capture and requested clarification regarding the definition of "operating hours" or that we delete the data capture criterion from Procedure 3 and address it in the applicable subpart.
Response:
      We agree that data capture criterion should be addressed in the applicable subpart and revised §10.4(4) to reference the applicable subpart.
3.8	Corrective Action
Comment:
      Six commenters (0006, 0010, 0014, 0025, 0029, 0036) commented on language in §10.5. Two commenters (0029, 0036) requested that we delete the specific corrective action requirements because they are too restrictive.  
Response:
      We agree with these commenters and believe that it is appropriate to include corrective action maintenance and repair procedures in a guidance document. Therefore, we revised the language in §10.5 to reference a guidance document that discusses maintenance and repair procedures and diagnostic testing.
Comment:
      Two of theses commenters (0006, 0010) requested that we allow the use of a probationary calibration test to conditionally validate data. Another one of these commenters (0014) noted that there are five categories listed under §10.5, but the proposed rule mentioned four classes of maintenance and repair procedures. The final commenter (0025) suggested that we consider changing §10.5 (Corrective Action) to §11.0 since §10.0 is entitled "What calibration and standardization procedures must I perform for my COMs?"   
Response:
      Since we deleted the maintenance and repair categories and specific correction action maintenance and repair procedures, these comments are no longer applicable.
3.9	30-Day Use of a Temporary Monitor
Comment:
      Three commenters (0010, 0034, 0036) commented on the use of a temporary monitor in §10.6. Two of the commenters (0034, 0036) noted that §10.6 does not clearly define when a temporary COMS is collecting valid data.
Response:
      We agree that we did not clearly define when a temporary COMS is collecting valid data. Therefore, we added a statement to §10.6 to indicate that data generated by the temporary COMS is considered valid when paragraphs (i) through (ix) of §10.6 (1) have been met.
Comment:
      The other commenter (0010) stated that 30 days is not enough time to remove an existing monitor, ship it off-site for repairs, and return it to service.   
Response:
      In the proposed rule, we based our decision to allow the use of a temporary monitor for 30 days on discussions with COMS manufacturers. The manufacturers indicated that 30 days would be adequate time to troubleshoot, repair, and return COMS to owners and operators in most cases. However, to accommodate rare situations that may take longer, such as special orders, we have increased the timeframe for temporary monitor use from 30 days to 45 days.	
3.10	QA/QC Reporting Requirements/Data Assessment Reports (DARS)
Comment:
	Five commenters (0006, 0012, 0021, 0027, 0029) commented on the QA/QC reporting requirements referred to as Data Assessment Reports. One of the commenters (0006) asked for clarification regarding whether the reporting requirements in §10.9 includes daily status indicator checks, quarterly cumulative zero compensation checks, and annual primary zero alignment checks required in §10.0. Another commenter (0012) stated that it is unclear whether §10.9 requires results of quarterly audits only or all daily zero and upscale drift checks performed during the quarter and the results of the quarterly audits to be included in the DAR.
Response:
      The reporting requirements in §10.9 include daily status indicator checks, quarterly cumulative zero compensation checks, and annual primary zero alignment checks required in §10.0. To address both of these comments, we clarified the requirement by referencing §§10.1, 10.2, and 10.3 in §10.9.
Comment:
      Two commenters (0021, 0027) noted that Procedure 3 refers to an example DAR in Procedure 1 but noted that the DAR in Procedure 1 is inconsistent with the requirements specified in §10.0 of Procedure 3.  
Response:
      After reviewing §10.9, we realized that the DAR in Procedure 1 was not an appropriate example. The DAR in Procedure 1 addresses CEMS, whereas Procedure 3 deals with COMS.  Therefore, we deleted the reference to the Procedure 1 DAR.
Comment:
      One commenter (0029) noted that it is not appropriate to establish a reporting frequency in a QA procedure and requested that we defer to the applicable regulation.  
Response:
      We agree and deleted the reference to reporting results at the interval specified in this procedure.    
3.11	Diagnostic Tests After Repairs (Table 17-1)
Comment:
	Five commenters (0011, 0022, 0029, 0034, 0036) provided comments regarding Table 17-1 which included diagnostic tests after repairs. Four of these commenters (0011, 0029, 0034, 0036) recommended that we delete the diagnostic test requirements (Table 17-1) from Procedure 3, and three of these commenters (0011, 0029, 0036) further recommended that we specify diagnostic test requirements in a guidance document.  
Response:
      We decided to delete Table 17-1 because we realize that it is necessary to allow flexibility for repairs due to variations in COMS designs. Therefore, in lieu of the diagnostic tests required in Table 17-1, suggested diagnostic tests after repairs are given in a guidance document.   
Comment:
      In addition, one of the commenters (0034) stated that Procedure 3 does not address how to handle COMS data collected during a diagnostic test event. 
Response:
      Data collected during a diagnostic test event is considered invalid because the COMS is out-of-control during this time.
Comment:
      One of the commenters (0022) suggested that in lieu of conducting a COMS 7-day calibration drift (CD) check following certain maintenance activities that Procedure 3 should clarify that this audit check can be performed with the unit off-line. 
Response:
      This comment no longer applies since we deleted Table 17-1.
4.0 	Miscellaneous Comments
4.1	Alternative Compliance Option
Comment:
	One commenter (0026) requested that we adopt an additional alternative compliance option that would allow sources to implement QA/QC requirements based on manufacturer and vendor recommendations in lieu of implementing Procedure 3.  
Response:
      We believe that we allow the use of manufacturer's recommendations where appropriate in Procedure3. Procedure 3 establishes the minimum QA/QC procedures for the entire QA/QC program.
4.2	Employee Safety
Comment:
	One commenter (0025) stated that is is not appropriate to discuss employee safety in Procedure 3 and recommended that we delete the employee safety discussion in §5.0.  
Response:
However, we are not imposing safety regulations in Procedure 3. We are simply putting people on notice that there are safety concerns potentially associated with these QA/QC requirements. 
4.3	Equipment, Supplies, and Reagants
Comment:
	One commenter (0029) requested that we delete the statements in §6.0 and §7.0  tying equipment and standards to PS-1 because the statements could be misread to require source owners and operators with COMS that were installed prior to 2001 to unnecessarily replace equipment that is still performing well. The commenter suggested that we indicate that §6.0 and §7.0 are reserved and stated that these sections are unnecessary since the relevant terms & equipment are defined elsewhere in the rule.
Response:
      To clarify the Agency's intent, we added the statement, "You are not required to purchase a new COMS if your existing COMS meets the requirements of Procedure 3" to §6.0 and §7.0.
4.4	QC Program
Comment:
	Two commenters (0021, 0025) commented on the QC measures required in §9.0. One commenter (0021) stated that the requirement to develop a separate QC program is overly burdensome and redundant, and the other commenter (0025) requested that we change the requirement for the QC program to describe specific procedures to a recommendation to include the procedures.  
Response:
      We believe that the QC measures in §9.0 are necessary and appropriate to ensure proper operation and maintenance of COMS. Your QC program may reference documents, such as, the manufacturer's instructions for the COMS.
4.5	Record Retention of QA/QC Procedures
Comment:
	One commenter (0029) noted that it is not appropriate to establish the record retention period in QA/QC procedures. The commenter also stated that we cannot require maintenance of records for more than three years without providing justification.
Response:
      We agree with both of these comments and deleted the record retention period from §9.1. We retained the language in §10.11 that states that records must be kept for the period specified in the regulations requiring use of the COMS.
4.6	Consequences of Failing QC Audits
Comment:
	Two commenters (0027, 0029) commented on the language in §9.2 regarding the consequences of failing QC audits. One commenter (0027) stated that it is not rational to mandate specific corrective actions and mentioned that the second audit failure may be caused by something totally unrelated to the first one. The other commenter (0029) stated that the requirement to revise QC procedures or repair or replace the COMS to correct deficiencies causing audit failures is unreasonable and beyond the scope of Procedure 3.
Response:
      We revised the language in §9.2 as follows. "If you fail two consecutive annual audits, two consecutive quarterly audits, or five consecutive daily checks, you must either revise your QC procedures or determine if your COMS is malfunctioning.  If you determine that your COMS is malfunctioning, you must then take the necessary correction action as specified in §10.5.  If you determine that your COMS requires extensive repairs, you may use a substitute COMS provided the substitute meets the requirements in §10.6."	
4.7	Calibration and Standardization Procedures
Comment:
	One commenter (0011) asked if the requirements of §10.0(1) to perform routine system checks would be met by performing daily calibrations and optical checks. The commenter also inquired about the difference between paragraphs 10.0(1) and 10.0(2).
Response:
      The requirements of §10.0(1) to perform routine system checks would be met by performing daily calibrations and optical checks. In response to the commenters' second question, paragraph §10.0(1) discusses maintenance requirements that must occur more frequently than quarterly, such as, daily drift checks, and §10.0(2) discusses quarterly performance audit requirements.
Comment:
      Two commenters (0029, 0032) stated that the 90 to 110 percent criterion for the simulated clear path condition in the annual zero alignment as specified in §10.0(3) should be changed to 90 to 190 percent as specified in ASTM D 6216. However, one commenter (0032) retracted their request to change this criterion upon realizing that the intended range in ASTM D 6216 is 90 to 110 percent and noted that they plan to change the range to 90 to 110 percent in the ASTM D 6216-12.
Response:
      In response, the most recent version of the ASTM D 6216 (ASTM D 6216-12) specifies the criterion as 90 to 110 percent, and Procedure 3 should be consistent with this language. Therefore, we did not change the proposed criterion in the final rule. In other words, the final rule specifies 90 to 110 percent as did the proposed rule.
4.8	Quarterly Calculation of 95 Percent Confidence Interval
Comment:
	One commenter (0021) stated that it is overly burdensome to calculate a 95% confidence interval quarterly as required in §10.2(2). The commenter stated that the 95% confidence interval provides limited information and requested that the quarterly frequency be reduced to annually.  
Response:
      We disagree that this requirement is overly burdensome. Furthermore, the commenter did not provide a rationale to support reducing the requirement from quarterly to annually. Therefore, this requirement remains the same as proposed.
4.9	Intervals to Record Data
Comment:
	One commenter (0028) requested that we specify the interval for recording data and noted that they currently record data at six-minute intervals.  
Response:
      We believe most COMS record data at six-minute intervals, and we are not suggesting that owners/operators change that recording interval. Furthermore, we do not believe that it is necessary to specify intervals for recording data. 
4.10	External Audit Device
Comment:
	One commenter (0022) requested that we clarify the requirement in §10.2(2) to confirm that the external audit device produces a zero value within one percent opacity.   
Response:
      We agreed to clarify this language and re-worded the requirement in §10.2(2) to state "When the external audit device is installed, with no calibration filter inserted, the COMS measurement reading must be less than or equal to one percent opacity."
4.11	Clear Path Zero Value
Comment:
      Three commenters (0027, 0032, 0036) commented on the clear path zero value. One commenter (0032) requested that we delete the allowance to perform the primary clear path calibration for the COMS as installed if the process is not operating and the monitor path is free of particulate matter since this allowance conflicts with ASTM D 6216-07. 
Response:
      We agree and deleted this allowance from §10.3(1). Therefore, allowing the COMS to be removed every three years is no longer permissible using the primary zero alignment method; for this method, the COMS must be removed annually. As a result of deleting this allowance, it was necessary to change "COMS may be removed" in the second sentence of §10.3(1) to "COMS must be removed". Also, as a result of deleting the allowance in §10.3(1), it was necessary to delete §10.3(4) because the proposed language in §10.3(4) implied that the COMS would have to remain on the stack.	
Comment:
	Another commenter (0036) stated that the clear path zero determination should be based on an average value or some other statistically-derived value they believe is more representative and defensible in lieu of the minimum observed value. The commenter suggested that the zero value be based on the 5[th] percentile lowest value.
Response:
      However, we deleted the option to perform the zero alignment test with the COMS on stack, with the monitor path free of particulate matter from §10.3(1). Therefore, this comment is no longer applicable.
Comment:
      The final commenter (0027) noted that §10.3(1) allows an adjustment of the clear path zero as long as the two-hour readings don't vary more than 3% opacity, whereas §10.4(2) defines the COMS as out of control if the zero alignment exceeds 2% opacity. The commenter believes that clear stack zero alignment should be acceptable if readings vary by less than 3% opacity. 
Response:
      Since we deleted reference to 3% opacity in §10.3(1), this comment is no longer applicable. 
4.12	Limitations on the Use of COMS Data
Comment:
      One commenter (0025) requested that we delete §10.8 which addresses limitations on the use of COMS data collected during out-of-control periods.  
Response:
      We disagree. We believe that it is appropriate to address limitations on the use of COMS data in Procedure 3. We have consistently addressed these limitations in other QA/QC procedures (i.e., Procedures 1 and 2).  
4.13	Data Retention
Comment:
	One commenter (0025) noted that §10.11 of Procedure 3 states that QA data must be maintained for the period in the applicable regulation, whereas §10.6(9) references permanently maintained records.
Response:
      We agree that the language in §10.6(9) could be interpreted to imply that records for temporary monitors are to be kept on a permanent basis which was not our intent.  Therefore, we have deleted the phrase "or in other appropriate permanently maintained records" from §10.6(9).  All records generated as a result of meeting Procedure 3 requirements must be kept for the period specified in the applicable regulation as stated in §10.11.
4.14	Daily Calibration Drift Check Calculation
Comment:
	One commenter (0012) stated that the daily calibration drift (CD) check calculations are not in §12.0 of PS-1 as stated in the proposed rule.  
Response:
      We added the word "quarterly" to the question in §12.0 to make it clear that the calculations referred to in §12.0 are for quarterly performance audits.
4.15	PS-1 vs. Procedure 3
Comment:
      One commenter (0028) recommended that we revise PS-1 prior to issuing the final Procedure 3 requirements.
Response:
      We don't believe that it is necessary to revise PS-1 prior to publishing the final Procedure 3 because PS-1 deals with the initial installation of a COMS; Procedure 3 deals with the on-going operation of a COMS and ensures the instrument continues to produce quality data. Revisions to PS-1 will have little to no impact on Procedure 3.
